From: Lisp Newbie
Subject: Looking for Lisp compiler
Date: 
Message-ID: <4PKO9.164844$6k.3348354@news1.west.cox.net>
Is there such thing as free Lisp compiler, that can work under Linux and
Windows, develiver executables, have sockets implementation and support
multithreading?

Thank you

From: Coby Beck
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <aufv30$2uf5$1@otis.netspace.net.au>
"Lisp Newbie" <········@lisp.programmer> wrote in message
····························@news1.west.cox.net...
> Is there such thing as free Lisp compiler, that can work under Linux and
> Windows, develiver executables, have sockets implementation and support
> multithreading?

Start at www.lisp.org

Delivering executables is the only thing there I don't know if you can get
for free.  Usually it is not really what people need/want anyway.

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Dave Bakhash
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <c293cokb2iz.fsf@no-knife.mit.edu>
"Coby Beck" <·····@mercury.bc.ca> writes:

> Delivering executables is the only thing there I don't know if you can
> get for free.  Usually it is not really what people need/want anyway.

It's important for a lot of things.  To me, the ideal thing is to have a
self-contained executable that depends on as few things as possible
going right.  These days, I've found that having the right system
configuration, with the tons of OSs, libraries, application versions,
etc. is becomming harder and harder -- especially in the Unix world.  

In a way, Mac OS X + Fink was my attempt to have something as nice as MS
Windows, but at the same time, robust like FreeBSD, and uniform.  I
think it wasn't a bad call, but now the problem is that LispWorks
doesn't run on Mac OS X.  ACL does, except for the GUI stuff (which I
don't really care about anyway).  But getting back to the main story...

...delivering an executable is the end result of what you want to do
with a Lisp application in many cases.  In the case of LispWorks, for
example, you can export your application with some of the basic tools
you use during development (e.g. the editor, and a listener).  While the
delivered image isn't too tiny in these cases, it's nice to deliver such
a product, and make use of of it all.

My company, for example, delivers not only our core application, but a
built-in listener, editor, and event log console that make use of the LW
development environment.  I think there's a lot of value in that (at
least for companies that sell tools).

dave
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <863cokavu1.fsf@bogomips.optonline.net>
Dave Bakhash <·····@alum.mit.edu> writes:

> "Coby Beck" <·····@mercury.bc.ca> writes:
> 
> > Delivering executables is the only thing there I don't know if you can
> > get for free.  Usually it is not really what people need/want anyway.
> 
> It's important for a lot of things.  To me, the ideal thing is to have a
> self-contained executable that depends on as few things as possible
> going right.  These days, I've found that having the right system
> configuration, with the tons of OSs, libraries, application versions,
> etc. is becomming harder and harder -- especially in the Unix world.  

I agree the big honken static exe is a good thing, especially now
with disk and ram so cheap, iff you are distributing commercial 
software.  Otherwise it is not that important.

> 
> ...delivering an executable is the end result of what you want to do
> with a Lisp application in many cases.  In the case of LispWorks, for
> example, you can export your application with some of the basic tools
> you use during development (e.g. the editor, and a listener).  While the
> delivered image isn't too tiny in these cases, it's nice to deliver such
> a product, and make use of of it all.

Yes in almost all commercial, especially shrink rapped, software.  Now
if you want do a commercial thing you should be willing, happy even,
to pay vendors money to help you make money.  The creation of an exe
in one of the things that I would generally consider a necessary part
of commercial development.  Otherwise it is nice to have, but by no
means necessary for non commercial use.

Now the original poster wanted the ability to create exe file on
linux and windows and he can get it for 900 USD retail, lispworks pro. 
Now everything else he can get for free(cmu, ecl, clisp).  The ability to 
create exe's does not help you do anything but distribute applications 
that are no longer lisp code, they are object/byte code and that looks
like a commercial interest to me.  I think that commercial software is 
a good thing, how else can some one get the software that they need but
no one will enjoy making(among other things). 

marc
 
From: Dave Bakhash
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <c29y96c9e68.fsf@no-knife.mit.edu>
Marc Spitzer <········@optonline.net> writes:

> Now everything else he can get for free(cmu, ecl, clisp).  The ability
> to create exe's does not help you do anything but distribute
> applications that are no longer lisp code, they are object/byte code
> and that looks like a commercial interest to me.

I might be misunderstanding your post, but I know that at least ecl can
pretty much dump out an executable that actually _is_ compiled by
(g)cc.  You can write something in CL using ECL and without close
inspection one would never guess that it was done in CL (except insofar
as it didn't crash).

dave
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86of77qpu9.fsf@bogomips.optonline.net>
Dave Bakhash <·····@alum.mit.edu> writes:

> Marc Spitzer <········@optonline.net> writes:
> 
> > Now everything else he can get for free(cmu, ecl, clisp).  The ability
> > to create exe's does not help you do anything but distribute
> > applications that are no longer lisp code, they are object/byte code
> > and that looks like a commercial interest to me.
> 
> I might be misunderstanding your post, but I know that at least ecl can
> pretty much dump out an executable that actually _is_ compiled by
> (g)cc.  You can write something in CL using ECL and without close
> inspection one would never guess that it was done in CL (except insofar
> as it didn't crash).

Yes ecls does do that on posix systems( I never meant to imply it could
not) but not on windows and the original poster wanted both.  To go off
on a bit of a tangent I could see ecl used in several cool projects,
an apache module like mod_perl or a stored procedure language for
postgres for example or clxemacs which you are interested in.

My main point was that the win/linux exe thing was a
distribution/commercial request and when that happens you really want
a vendor and a support contract with said vendor.  You see I would
want to have someone who knows much more about the CL I am using be
bound under contract to care about my problems, for a fixed fee.
This way I get work arounds and patches and features that I need
but are not fun or interesting to do, a large amount of high quality 
documentation presented in a consistent format comes to mind.

marc
From: Pascal Bourguignon
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <877kduon2x.fsf@thalassa.informatimago.com>
Marc Spitzer <········@optonline.net> writes:

> Dave Bakhash <·····@alum.mit.edu> writes:
> 
> > Marc Spitzer <········@optonline.net> writes:
> > 
> > > Now everything else he can get for free(cmu, ecl, clisp).  The ability
> > > to create exe's does not help you do anything but distribute
> > > applications that are no longer lisp code, they are object/byte code
> > > and that looks like a commercial interest to me.
> > 
> > I might be misunderstanding your post, but I know that at least ecl can
> > pretty much dump out an executable that actually _is_ compiled by
> > (g)cc.  You can write something in CL using ECL and without close
> > inspection one would never guess that it was done in CL (except insofar
> > as it didn't crash).
> 
> Yes ecls does do that on posix systems( I never meant to imply it could
> not) but not on windows and the original poster wanted both.  To go off
> on a bit of a tangent I could see ecl used in several cool projects,
> an apache module like mod_perl or a stored procedure language for
> postgres for example or clxemacs which you are interested in.
> 
> My main point was that the win/linux exe thing was a
> distribution/commercial request and when that happens you really want
> a vendor and a support contract with said vendor.  You see I would
> want to have someone who knows much more about the CL I am using be
> bound under contract to care about my problems, for a fixed fee.
> This way I get work arounds and patches and features that I need
> but are not fun or interesting to do, a large amount of high quality 
> documentation presented in a consistent format comes to mind.
> 
> marc

The point is  that software with freedom is  not antinomic to contract
support. On the contrary.  How  could you offer contract support for a
commercial product you  don't have the sources? Or  you're not allowed
to modify the sources?

So, you  may shop at the  monopoly of contract support  offered by the
closed  proprietary  implementation,  or  at  the  various  competting
contract support offers for free software.

Moreover, guess who will have the  most pressure to help you resolve a
problem, the developper of commercial tools who have thousands or tens
of thousands customers and who is busy developping the next version to
be  able to  charge the  upgrade price,  or the  small  shop dedicated
support team  who will have  a couple dozen  of custmers and  for whom
you'll be relatively a much more important customer?


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. do not adjust your minds. -- Salman Rushdie
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <867kdu6680.fsf@bogomips.optonline.net>
Pascal Bourguignon <···@informatimago.com> writes:

> 
> The point is  that software with freedom is  not antinomic to contract
> support. On the contrary.  How  could you offer contract support for a
> commercial product you  don't have the sources? Or  you're not allowed
> to modify the sources?

I do not and that is pretty self evident.

> 
> So, you  may shop at the  monopoly of contract support  offered by the
> closed  proprietary  implementation,  or  at  the  various  competting
> contract support offers for free software.

Well the only place that I have seen this is with linux distributions
and the first thing they will tell you is you need to get into a
supported configuration.  If you contract with red hat you run red hat
or you get no help.  If you compile the newest glibc from source and
call them they will say put our supported version in place or we can
not help you under your contract.  If you want us to look at it, it
will be time & materials and can I please have a credit card number.

> 
> Moreover, guess who will have the  most pressure to help you resolve a
> problem, the developper of commercial tools who have thousands or tens
> of thousands customers and who is busy developping the next version to
> be  able to  charge the  upgrade price,  or the  small  shop dedicated
> support team  who will have  a couple dozen  of custmers and  for whom
> you'll be relatively a much more important customer?
> 

this is just ignorant, do the math.  Let's say a support contract for
open source product X is 5,000 USD/year and you have 5, not 2, dozen
customers.  This gives you 300,000 USD/year.  Now you need to pay for
office space, 800 phone service, commercial grade internet connection,
handle billing and collections, if you take credit cards they take
their cut off the top and so does paypal.  We have not even gotten to
the point of hiring the people needed to do the support work yet and
much of the money is gone, I will be nice 1/3 is gone.  I am counting
the commission on the sale that the sales person gets up front, lets
say 10% .  Now you need to hire people to actually do tech support.  to
have 16x7 coverage on the phone requires a min of 3 people, 2 work m-f
and 1 does 2 doubles on the weekend(40k/person gross cost not just
pay), you now have 80k USD left.  I forgot to take out the commission
on the sales, another 30k, 50k left.  So now you can get 1 more person
to answer the phones, you cannot afford a Sr./Guru type person with
the money and you would need at least 2 of them, think vacation/sick days
and people leave.  With 2 1/2 times your customer base it still makes
no sence.

Now on to the tech issues:
1: you get me a patch and it it not accepted in to the source code of
   the product I am using.
2: You only support the current version (or an older version) of the
   product that I cannot use for other reasons, I can only do apache 1.x 
   because the modules I use are not thread safe and you now only support
   apache 2.x
3: you go out of business, I have no support

Oh yea, I forgot *Taxes* and *Fees*.

marc

> 
> -- 
> __Pascal_Bourguignon__                   http://www.informatimago.com/
> ----------------------------------------------------------------------
> There is a fault in reality. do not adjust your minds. -- Salman Rushdie
From: Pascal Bourguignon
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87lm2amt8g.fsf@thalassa.informatimago.com>
Marc Spitzer <········@optonline.net> writes:

> Pascal Bourguignon <···@informatimago.com> writes:
> 
> > 
> > The point is  that software with freedom is  not antinomic to contract
> > support. On the contrary.  How  could you offer contract support for a
> > commercial product you  don't have the sources? Or  you're not allowed
> > to modify the sources?
> 
> I do not and that is pretty self evident.
> 
> > 
> > So, you  may shop at the  monopoly of contract support  offered by the
> > closed  proprietary  implementation,  or  at  the  various  competting
> > contract support offers for free software.
> 
> Well the only place that I have seen this is with linux distributions
> and the first thing they will tell you is you need to get into a
> supported configuration.  If you contract with red hat you run red hat
> or you get no help.  If you compile the newest glibc from source and
> call them they will say put our supported version in place or we can
> not help you under your contract.  If you want us to look at it, it
> will be time & materials and can I please have a credit card number.
 
The  point is that  any and  all unix/linux  experts could  maintain a
linux   installation  either  as   external  contractee   or  internal
employees, depending on your needs.  The point is that RedHat gave you
a quote, and  now you can shop and find other  linux experts who could
maintain your  patched installation, while this would  not be possible
at all with proprietary software.

> > Moreover, guess who will have the  most pressure to help you resolve a
> > problem, the developper of commercial tools who have thousands or tens
> > of thousands customers and who is busy developping the next version to
> > be  able to  charge the  upgrade price,  or the  small  shop dedicated
> > support team  who will have  a couple dozen  of custmers and  for whom
> > you'll be relatively a much more important customer?
> 
> this is just ignorant, do the math.  Let's say a support contract for
> open source product X is 5,000 USD/year and you have 5, not 2, dozen
> customers.  This gives you 300,000 USD/year.  Now you need to pay for
> office space, 800 phone service, commercial grade internet connection,
> handle billing and collections, if you take credit cards they take
> their cut off the top and so does paypal.  We have not even gotten to
> the point of hiring the people needed to do the support work yet and
> much of the money is gone, I will be nice 1/3 is gone.  I am counting
> the commission on the sale that the sales person gets up front, lets
> say 10% .  Now you need to hire people to actually do tech support.  to
> have 16x7 coverage on the phone requires a min of 3 people, 2 work m-f
> and 1 does 2 doubles on the weekend(40k/person gross cost not just
> pay), you now have 80k USD left.  I forgot to take out the commission
> on the sales, another 30k, 50k left.  So now you can get 1 more person
> to answer the phones, you cannot afford a Sr./Guru type person with
> the money and you would need at least 2 of them, think vacation/sick days
> and people leave.  With 2 1/2 times your customer base it still makes
> no sence.
 
> Now on to the tech issues:
> 1: you get me a patch and it it not accepted in to the source code of
>    the product I am using.

Because Microsoft or  the other big software editors  do offer support
for custom patched software?

> 2: You only support the current version (or an older version) of the
>    product that I cannot use for other reasons, I can only do apache 1.x 
>    because the modules I use are not thread safe and you now only support
>    apache 2.x
> 3: you go out of business, I have no support

Oops.   Hopefully, then  you'll be  able to  choose between  the 11000
"linux consulting" offers returned by:

http://www.google.com/search?as_q=&num=100&hl=en&ie=UTF-8&oe=UTF-8&btnG=Google+Search&as_epq=Linux+consulting&as_oq=&as_eq=&lr=lang_en&as_ft=i&as_filetype=&as_qdr=m3&as_occt=any&as_dt=i&as_sitesearch=.com&safe=images


Or the 42000  "linux support" offers returned by:

http://www.google.com/search?q=+%22Linux+support%22+site%3A.com&btnG=Google+Search&num=100&hl=en&lr=lang_en&ie=UTF-8&oe=UTF-8&as_qdr=m3&as_qdr=m3

I don't doubt  that amongst those 42000 hits you'll be  able to find a
corporation  more supple than  RedHat on  the patch  and customization
side of your problem.

> Oh yea, I forgot *Taxes* and *Fees*.
> 
> marc





-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86lm2am94m.fsf@bogomips.optonline.net>
Pascal Bourguignon <···@informatimago.com> writes:

> Marc Spitzer <········@optonline.net> writes:
> 
> > Pascal Bourguignon <···@informatimago.com> writes:
> 
> The  point is that  any and  all unix/linux  experts could  maintain a
> linux   installation  either  as   external  contractee   or  internal
> employees, depending on your needs.  The point is that RedHat gave you
> a quote, and  now you can shop and find other  linux experts who could
> maintain your  patched installation, while this would  not be possible
> at all with proprietary software.

We are back to time and materiels.  This means I do not have a fixed
cost for support.  And now I am now, as a business person, trying to
figure out who is a linux/unix expert, a recipee for success.  And 
what you are talking about is system administration not tech support.

> 
> > > Moreover, guess who will have the  most pressure to help you resolve a
> > > problem, the developper of commercial tools who have thousands or tens
> > > of thousands customers and who is busy developping the next version to
> > > be  able to  charge the  upgrade price,  or the  small  shop dedicated
> > > support team  who will have  a couple dozen  of custmers and  for whom
> > > you'll be relatively a much more important customer?
> > 
> > this is just ignorant, do the math.  Let's say a support contract for
> > open source product X is 5,000 USD/year and you have 5, not 2, dozen
> > customers.  This gives you 300,000 USD/year.  Now you need to pay for
> > office space, 800 phone service, commercial grade internet connection,
> > handle billing and collections, if you take credit cards they take
> > their cut off the top and so does paypal.  We have not even gotten to
> > the point of hiring the people needed to do the support work yet and
> > much of the money is gone, I will be nice 1/3 is gone.  I am counting
> > the commission on the sale that the sales person gets up front, lets
> > say 10% .  Now you need to hire people to actually do tech support.  to
> > have 16x7 coverage on the phone requires a min of 3 people, 2 work m-f
> > and 1 does 2 doubles on the weekend(40k/person gross cost not just
> > pay), you now have 80k USD left.  I forgot to take out the commission
> > on the sales, another 30k, 50k left.  So now you can get 1 more person
> > to answer the phones, you cannot afford a Sr./Guru type person with
> > the money and you would need at least 2 of them, think vacation/sick days
> > and people leave.  With 2 1/2 times your customer base it still makes
> > no sence.
>  
> > Now on to the tech issues:
> > 1: you get me a patch and it it not accepted in to the source code of
> >    the product I am using.
> 
> Because Microsoft or  the other big software editors  do offer support
> for custom patched software?
>

The reason this is a problem is now you have forked the code to
add/fix the feature you needed.  You now have your own source tree/set
of patches to maintain until you decide to get back with everyone
else.  This costs money and leads to more points of failure.
 
> > 2: You only support the current version (or an older version) of the
> >    product that I cannot use for other reasons, I can only do apache 1.x 
> >    because the modules I use are not thread safe and you now only support
> >    apache 2.x
> > 3: you go out of business, I have no support
> 
> Oops.   Hopefully, then  you'll be  able to  choose between  the 11000
> "linux consulting" offers returned by:
> 

The money I have for support is spent, to get these people to support
me I need to pay them right?  Now I go to my boss and say "Boss Bobo's
linux support is dead and I need more money to go with louies linux
support", boss says " I have to talk to my boss for the money".  So
now my boss looks stupid to his boss at the very least.  From all
levels above me the shit/question rolls down "why did you not use
red hat for support?  They are still in business."  Now this is a fair
question that I need to have a good answer for or I stand a good
chance of taking a hit on my review/contract renewal.  I just do not
have the problem if I go with the vendor.

> http://www.google.com/search?as_q=&num=100&hl=en&ie=UTF-8&oe=UTF-8&btnG=Google+Search&as_epq=Linux+consulting&as_oq=&as_eq=&lr=lang_en&as_ft=i&as_filetype=&as_qdr=m3&as_occt=any&as_dt=i&as_sitesearch=.com&safe=images
> 
> 
> Or the 42000  "linux support" offers returned by:
> 
> http://www.google.com/search?q=+%22Linux+support%22+site%3A.com&btnG=Google+Search&num=100&hl=en&lr=lang_en&ie=UTF-8&oe=UTF-8&as_qdr=m3&as_qdr=m3
> 
> I don't doubt  that amongst those 42000 hits you'll be  able to find a
> corporation  more supple than  RedHat on  the patch  and customization
> side of your problem.
> 

see above for why keeping your own patches around is not the best
idea.   If I was going to do that I would probably use IBM( even
bigger then Red Hat), but they now have moved on to their own
distribution for use with their rack mount pc's.

You see once I start using a custom, as in using source code not at
"where you get kernels" is not a good idea for small operations to
do.  With red hat and IBM economies of scale kick in.

marc

> > Oh yea, I forgot *Taxes* and *Fees*.
> > 
> > marc
> 
> 
> 
> 
> 
> -- 
> __Pascal_Bourguignon__                   http://www.informatimago.com/
> ----------------------------------------------------------------------
> There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Pascal Bourguignon
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87wultjjph.fsf@thalassa.informatimago.com>
Marc Spitzer <········@optonline.net> writes:
> 
> The money I have for support is spent, to get these people to support
> me I need to pay them right?  Now I go to my boss and say "Boss Bobo's
> linux support is dead and I need more money to go with louies linux
> support", boss says " I have to talk to my boss for the money".  So
> now my boss looks stupid to his boss at the very least.  From all
> levels above me the shit/question rolls down "why did you not use
> red hat for support?  They are still in business."  Now this is a fair
> question that I need to have a good answer for or I stand a good
> chance of taking a hit on my review/contract renewal.  I just do not
> have the problem if I go with the vendor.

The  fact  is  that   linux  consulting/support  corporation  are  not
necessarily  in the insurance  business, that  is the  risk management
business.  Big  corporations do everything including that.   You pay a
prime,  and the  probabilities are  you won't  get any  'benefit', you
won't need any  support activity.  

But you  could get an insurance  contract with a  respected and stable
insurance corporation and then be  able to pay for any intervention by
any linux consulting company available at the time the problem occurs.

But then  the prime asked by  your insurance corporation  may hint you
that to get  your money worth in support, you'd have  to pay much more
than what RedHat  and the other ask anyway.  I  speak of real service,
not shit umbrellas.

-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86u1gwzoke.fsf@bogomips.optonline.net>
Pascal Bourguignon <···@informatimago.com> writes:

> Marc Spitzer <········@optonline.net> writes:
> > 
> > The money I have for support is spent, to get these people to support
> > me I need to pay them right?  Now I go to my boss and say "Boss Bobo's
> > linux support is dead and I need more money to go with louies linux
> > support", boss says " I have to talk to my boss for the money".  So
> > now my boss looks stupid to his boss at the very least.  From all
> > levels above me the shit/question rolls down "why did you not use
> > red hat for support?  They are still in business."  Now this is a fair
> > question that I need to have a good answer for or I stand a good
> > chance of taking a hit on my review/contract renewal.  I just do not
> > have the problem if I go with the vendor.
> 
> The  fact  is  that   linux  consulting/support  corporation  are  not
> necessarily  in the insurance  business, that  is the  risk management
> business.  Big  corporations do everything including that.   You pay a
> prime,  and the  probabilities are  you won't  get any  'benefit', you
> won't need any  support activity.  

The reason you buy insurance is that when an event happens that can
destroy your business someone else is required to pay for it so you
can recover and continue doing business.  The benefit you get is that
you still have a business because you can meet your obligations, you
can reorder the stock that burned up in the fire because of the
insurance check.  The freedom from that threat is the benefit.

I would think it depends on what is being insured.  Most businesses do
not carry "business continuation" insurance, they carry things like
fire, theft, flood and liability insurance.  And for those you are
right the underwriter does not care about tech support for your
business, you are not paying him to cover it.  Now if they were
insuring the gross profit from your ecommerce site, ie site can not
transact business(doors are closed) insurer starts cutting checks,
they will look at your operation in great detail.  In fact they will
probably require changes in it to reduce down time or just charge you
a (much)higher premium.  One thing they might require is 24x7 admin
support, not just operators. Another one is getting Sr management involved
much earlier in the outage then is normal and he has to stay available
for the duration.  And real tech support, not Louie's Linux hut.  

> 
> But you  could get an insurance  contract with a  respected and stable
> insurance corporation and then be  able to pay for any intervention by
> any linux consulting company available at the time the problem occurs.
> 

And this would cost more or less then just going to redhat or ibm
for a normal or even premium service contract?

> But then  the prime asked by  your insurance corporation  may hint you
> that to get  your money worth in support, you'd have  to pay much more
> than what RedHat  and the other ask anyway.  I  speak of real service,
> not shit umbrellas.
> 

possibly not, remember the first thing asked for is your support contract.
So if you have no contract it is time and materials or to put it another
way 3-400 usd/hour/person also many service organizations have different 
levels of contract support and if you have the better contract you get more
help faster and access to more spares that are stashed closer to your site.
And if someone with a contract needs access to the same people that are 
working with you they get it, as per the contract and you can wait. 

Real support is not cheap, but when the business goes away when the
computer goes away then it is much cheaper then not having it.

marc
From: Frank A. Adrian
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <UYHP9.328$mg5.72131@news.uswest.net>
Marc Spitzer wrote:
> Real support is not cheap, but when the business goes away when the
> computer goes away then it is much cheaper then not having it.

Yes.  And when you really need uptime, you ain't gonna be looking at either 
crap Linux OR Windows and the boxes they run on.  You're going to be going 
with something that has both design for reliability AND support built in.  
And that means something old and tested by a company that's been in the 
business for more than a few years.  Probably something like OS/390 or 
OS/400.  Yeah, it costs more, but those things stay up.  And, if they do 
fall down, you got REAL backup.

faa
From: Damond Walker
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <BA37C047.8059%damosan@comcast.net>
On 12/29/02 2:23 PM, in article ···················@news.uswest.net, "Frank
A. Adrian" <·······@ancar.org> wrote:

[snip]

> Probably something like OS/390 or OS/400.

Sad but true.  The '400 is a stable box...how many Lisp implementations are
there for that machine?  :)

Damond
From: Frank A. Adrian
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <mHvQ9.322$yy5.185948@news.uswest.net>
Damond Walker wrote:

> Sad but true.  The '400 is a stable box...how many Lisp implementations
> are there for that machine?  :)

Unfortunately, none that I know of.  Even most of the Open Source ones would 
be hard to port due to some of the unusual properties of the machine.  To 
get the code to run efficiently would be a trick, too, given the 
limitations on pointer manipulation.  On the other hand, I think a Lisp 
with 64- or 128-bit pointers would be pretty cool (even if only 48 of them 
are physical on most boxes).

faa
From: Damond Walker
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <BA389251.8087%damosan@comcast.net>
On 1/1/03 1:14 AM, in article ····················@news.uswest.net, "Frank
A. Adrian" <·······@ancar.org> wrote:

> Unfortunately, none that I know of.  Even most of the Open Source ones would
> be hard to port due to some of the unusual properties of the machine.  To
> get the code to run efficiently would be a trick, too, given the
> limitations on pointer manipulation.  On the other hand, I think a Lisp
> with 64- or 128-bit pointers would be pretty cool (even if only 48 of them
> are physical on most boxes).

Probably right as, at least in my limited experience, the '400 is tuned for
running RPG/COBOL code.  The 390 doesn't share this property does it?
Doesn't it (or can it) run some version of Unix?

Damond
From: Doug McNaught
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <m3d6ngiv8m.fsf@abbadon.mcnaught.org>
Damond Walker <·······@comcast.net> writes:

> Probably right as, at least in my limited experience, the '400 is tuned for
> running RPG/COBOL code.  The 390 doesn't share this property does it?
> Doesn't it (or can it) run some version of Unix?

Modern S/390 ("zSeries") and AS/400 ("iSeries") can run Linux
alongside the native OS, under VM or in an LPAR (logical partition).
AFAIK they are PowerPC-based so there's no reason CMUCL, SBCL or even
maybe OpenMCL couldn't be ported there.  CLISP would likely be an easy
port if it's not done already...

As far as the native OS (S/390 and OS/400) I don't know, but I
certainly haven't heard of any of the free implementations running
there.

-Doug
From: Matthew Danish
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <20030101132132.C12928@lain.cheme.cmu.edu>
On Wed, Jan 01, 2003 at 01:24:25PM -0500, Doug McNaught wrote:
> CLISP would likely be an easy port if it's not done already...

There appears to be a clisp package for Debian/s390, though I haven't
tried it personally.

-- 
; Matthew Danish <·······@andrew.cmu.edu>
; OpenPGP public key: C24B6010 on keyring.debian.org
; Signed or encrypted mail welcome.
; "There is no dark side of the moon really; matter of fact, it's all dark."
From: Ng Pheng Siong
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <aujfd4$mmv$1@reader01.singnet.com.sg>
According to Pascal Bourguignon  <···@informatimago.com>:
> Moreover, guess who will have the  most pressure to help you resolve a
> problem, the developper of commercial tools who have thousands or tens
> of thousands customers and who is busy developping the next version to
> be  able to  charge the  upgrade price,  or the  small  shop dedicated
> support team  who will have  a couple dozen  of custmers and  for whom
> you'll be relatively a much more important customer?

Sir, you need to upgrade that black and white television of yours. The
upgrade price isn't much.


-- 
Ng Pheng Siong <····@netmemetic.com> * http://www.netmemetic.com
From: Pascal Bourguignon
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87ptrmmu11.fsf@thalassa.informatimago.com>
····@netmemetic.com (Ng Pheng Siong) writes:

> According to Pascal Bourguignon  <···@informatimago.com>:
> > Moreover, guess who will have the  most pressure to help you resolve a
> > problem, the developper of commercial tools who have thousands or tens
> > of thousands customers and who is busy developping the next version to
> > be  able to  charge the  upgrade price,  or the  small  shop dedicated
> > support team  who will have  a couple dozen  of custmers and  for whom
> > you'll be relatively a much more important customer?
> 
> Sir, you need to upgrade that black and white television of yours. The
> upgrade price isn't much.

Yes, sorry if I sound caricatural,  but I can only compare the support
I've never  been able to  get from Microsoft/Sony  for the Vaio  of my
niece (for a MS-Windows-Me problem) vs. the bugs in NeXTSTEP BSD layer
that  never have been  corrected in  any new  version (at  least Apple
freed  Darwin) vs.   the support  and  maintainance I  get (and  give)
everyday on usenet for GPL'ed or BSD'ed software.


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. do not adjust your minds. -- Salman Rushdie
From: Erik Naggum
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3250119501765203@naggum.no>
* Pascal Bourguignon
| Yes, sorry if I sound caricatural, but I can only compare the
| support I've never been able to get from Microsoft/Sony for the
| Vaio of my niece (for a MS-Windows-Me problem) vs. the bugs in
| NeXTSTEP BSD layer that never have been corrected in any new
| version (at least Apple freed Darwin) vs. the support and
| maintainance I get (and give) everyday on usenet for GPL'ed or
| BSD'ed software.

  So what you need is to become a customer of a commercial enterprise
  that provides much /better/ service than you can get with the Free
  Software and you would instantly change your mind?  Some of us have
  had that experience.  I strongly suggest that you accept that people 
  can actually have happy experiences with commercial software despite
  your unhappy experiences and therefore find your caricature damning
  for you, not for the commercial software you have yet to experience.

  What is so tragic about so many (rabid) Free Software adherents is
  that they remain Microsoft victims even when they decry its evil.
  It is not unlike freeing slaves whose slave psychology remains, and
  instead of experiencing the liberty to partake in the possibilities
  before them and enjoy the freedom to fail without punishment and
  reap the benefits of their success, they cling to the security of
  knowing the outcome of their actions as determined by their master,
  and so freedom presents them with problems, not opportunities.  The
  Microsoft victims truly believe that Free Software is the solution
  to their problems, but their problem was getting no support and now
  they think software must be free in order to get good support.  It
  is all hogwash!  This is not what Free Software was about!  This is
  not the reason people fight for Free Software!  Microsoft does not
  even /exist/ in the Free Software universe -- Microsoft is entirely
  irrelevant.  Whether Microsoft provides shitty or no service is of
  /zero/ consequence in the Free Software universe.  Free Software is
  not about support at all.  Using it as an argument is specious at
  best.  Free Software is about your ability qua user to look under
  the hood.  That (or if) you can hire somebody do this for you is
  entirely immaterial.  The freedom to ask somebody to do something
  for you in exchange for another service or money is an aspect of
  the environment in which Free Software exists, but the freedom to
  look under the hood is essentially disconnected from this aspect.

  Once bitten, twice shy, the saying goes, but it appears to me that
  many people are of the "bitten, thence shy" persuasion, and it is
  as impossible to change their mind as it is to unbite them.  The
  doctrine of forming all of one's opinions after the first encounter
  with anything is well known under a variety of names, all nasty and
  properly derisive of those who succumb to it, but until and unless
  many people understand that they do in fact subscribe to it, they
  appear to believe they have a right to an opinion about that which
  they have yet to experience based on what they have experienced so
  far.  Some even claim that this lunacy is "scientific" and that one
  cannot escape prejudice.  Again, hogwash!  People can /think/ and
  they can make new choices without any obligation to tell you about
  it, so no amount of "historic evidence" can predict the future of
  someone's choices, very much unlike physics.  To become satisfied
  that you have seen all there is to see from someone is the ultimate
  disrespect towards that person.  Most people acknowledge this when
  they face the naked facts, but still return to this loathsome and
  vile practice whenever they hope to get away with it, such as when
  they deal with "businesses" instead of "people".

  Please try to understand that the quality of the support you get is
  completely irrelevant to the issue of Free Software.  Would you
  abandon Free Software as a concept if you got lousy service from
  one or even many suppliers?  But you abandoned commercial software
  because of bad experience with one or a few suppliers of commercial
  software?  Pardon the passion when I ask you a heartfelt question:
  What the hell were you /thinking/?  Drawing that kind of conclusion
  from that kind of evidence strongly suggests that you are unable to
  think clearly and instead try to prop up your emotional conclusions
  with delusions of logic, with mere rationalization.  You are not
  alone in this mind-boggling stupidity, just as there are numerous
  people out in the real world who still believe that as soon as they
  know somebody's /race/ they can stop learning about them as people.

  You should be smart enough to realize that your personal experience
  is entirely immaterial as part of a critique of a system, and even
  more so to support your negative attitude to future experiences and
  those of other people you do not know.  You ought to be /ashamed/
  of yourself for trying to pull such a fantastically stupid stunt in
  a newsgroup of /really/ smart people.  The /arrogance/ of trying to
  fool people with such mind-numbing idiocy has marked /your/ past
  and has obliterated any trust in your thinking ability so far.  But
  according to the principles I laid down above, you, too, have the
  ability to think and make some intelligent choices.  Now, hear me
  when I tell you that nobody is interested in your past, in what you
  believe is your personality or your identity or your personal pride.
  You are only your arguments and your articles to this forum, and we
  know /nothing/ about you as a person, nor does anybody care.  Only
  a bigoted idiot would deny /you/ the chance to change your mind and
  fix your mistakes.  That means that your personal defense of your
  past is not suitable for this forum, however strongly you feel the
  urge to defend yourself.  In a debate, you are /expected/ to learn
  when you are mistaken.  Being a good debater means that you grasp
  the consequences of having your arguments shot down: You re-open
  them only if you have a solid argument to support them and refute
  the refutation, and attacking those who shot them down /does not/
  count in that regard.  And please realize this: This debate is not
  about what /you/ feel like doing, what /you/ have experienced that
  gave rise to your desires, but about something that could apply to
  other people with other experiences than yours.  Therefore, what
  you need to present as an argument is not what you experienced, but
  why anyone should care about it.

  But, since you have opened this argument in this newsgroup, where
  it has been opened hundreds of times before, even though it does
  not belong here, the sheer arrogance you display by expressing a
  belief that nobody has ever heard /your/ arguments before, I have
  also formed a set of expectations about your ability to reason and
  think clearly and grasp what you should do to present a coherent
  argument to us that is at least /somewhat/ novel in this forum if
  you have to discuss such an /off-topic/ issue here to begin with.
  That does not mean that you cannot start thinking without telling
  me or that you must feel obliged to continue down only one path.  I
  ask you, implore you, to reconsider whether you think anyone here
  is even remotely interested in your personal experiences or has not
  heard your pro-Free Software arguments before and whether you could
  hope to change anyone else's mind when yours is apparently made up
  for good.  People here are generally experienced programmers, not a
  cobble of newbies who need to be led onto the One True Path by
  another newcomer to the newsgroup.  When older and wiser people
  than you do not accept your self-evident truth, it is time for you
  to stop and think: at the very least, it is not self-evident, and
  it may not even be the truth.  Another thing you might want to
  consider is that those who want to discuss the programming language
  (family|Common) Lisp may simply ignore you and hope you go away,
  and that those who respond to you are just really pissed off that
  we have yet /another/ Free Software zealot waste everyone's time
  with this non-issue in this forum.  The desire to have everyone
  else agree to one's personal opinion may run strong in some people,
  but it has no place in a programming language forum, no matter how
  strongly felt the opinion or its gravity to the world in general.
  Those who want do discuss this issue are generally found over in
  gnu.misc.discuss.

  As a personal advice to you, I would very strongly suggest that you
  adopt a different attitude.  Try some humility towards people you
  do not know -- they may be the people who some day decide your very
  future, or they may just be the people who could help you, but may
  decide not to based on your perceived arrogance and aggressiveness
  about what you currently believe and your apparent lack of ability
  to listen to people unless they do exactly as you tell them to.  If
  you see other people as negative or harsh towards you, it is a good
  idea to examine what you have done to them first rather than go on
  and aggravate matters by doing it again -- people react negatively
  for a reason, and insisting that you are faultless tends to annoy
  people more than anything else.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: wni
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <kDwP9.489308$%m4.135999@rwcrnsc52.ops.asp.att.net>
Erik Naggum wrote:

>   You should be smart enough to realize that your personal experience
>   is entirely immaterial as part of a critique of a system, and even
>   more so to support your negative attitude to future experiences and
>   those of other people you do not know.  

So it's your experience that counts? What makes you think that you are
qualified to criticize while the other people can't?


>   You ought to be /ashamed/
>   of yourself for trying to pull such a fantastically stupid stunt in
>   a newsgroup of /really/ smart people.

I think you are talking about *you*, the only smart one.

>  The /arrogance/ of trying to
>   fool people with such mind-numbing idiocy has marked /your/ past
>   and has obliterated any trust in your thinking ability so far.

What's idotic about Bourguignon's argument? Don't you build your
arguments from your experience or you don't have any experience
on anything at all? The pretense of objectivity is a dirty trick
you play on your sycophants and horrify innocent people.

>  But  according to the principles I laid down above, you, too, have the
>   ability to think and make some intelligent choices.  Now, hear me
>   when I tell you that nobody is interested in your past, in what you
>   believe is your personality or your identity or your personal pride.
>   You are only your arguments and your articles to this forum, and we
>   know /nothing/ about you as a person, nor does anybody care.

We are not interested in your opinions, past, and whatever half-cooked,
densely packed dirty laundry of yours. Treating people this way has long
been your scare tactics and the laughing butt of comp.lang.lisp.

>  Only
>   a bigoted idiot would deny /you/ the chance to change your mind and
>   fix your mistakes.  That means that your personal defense of your
>   past is not suitable for this forum, however strongly you feel the
>   urge to defend yourself.  In a debate, you are /expected/ to learn
>   when you are mistaken.  

You are really being patronizing. You have nothing to show why people
should listen to this.


wni at attbi dot com
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86y969z40w.fsf@bogomips.optonline.net>
wni <···@nospam.attbi.com> writes:

> Erik Naggum wrote:
> 
> >   You should be smart enough to realize that your personal experience
> >   is entirely immaterial as part of a critique of a system, and even
> >   more so to support your negative attitude to future experiences and
> >   those of other people you do not know.
> 
> 
> So it's your experience that counts? What makes you think that you are
> qualified to criticize while the other people can't?
> 

No it is his well formulated arguments that count.  And I have never 
seen him say "do not do it", but I have seen him say "do not do such a
bad job of it"

> 
> >   You ought to be /ashamed/
> >   of yourself for trying to pull such a fantastically stupid stunt in
> >   a newsgroup of /really/ smart people.
> 
> I think you are talking about *you*, the only smart one.
> 

stuck in 3rd grade are we?

> >  The /arrogance/ of trying to
> >   fool people with such mind-numbing idiocy has marked /your/ past
> >   and has obliterated any trust in your thinking ability so far.
> 
> What's idotic about Bourguignon's argument? Don't you build your
> arguments from your experience or you don't have any experience
> on anything at all? The pretense of objectivity is a dirty trick
> you play on your sycophants and horrify innocent people.
> 

did you try google?  He has answered why XML sucks several times.

> >  But  according to the principles I laid down above, you, too, have the
> >   ability to think and make some intelligent choices.  Now, hear me
> >   when I tell you that nobody is interested in your past, in what you
> >   believe is your personality or your identity or your personal pride.
> >   You are only your arguments and your articles to this forum, and we
> >   know /nothing/ about you as a person, nor does anybody care.
> 
> We are not interested in your opinions, past, and whatever half-cooked,
> densely packed dirty laundry of yours. Treating people this way has long
> been your scare tactics and the laughing butt of comp.lang.lisp.

Actually some of us are. 

> 
> >  Only
> >   a bigoted idiot would deny /you/ the chance to change your mind and
> >   fix your mistakes.  That means that your personal defense of your
> >   past is not suitable for this forum, however strongly you feel the
> >   urge to defend yourself.  In a debate, you are /expected/ to learn
> >   when you are mistaken.
> 
> 
> You are really being patronizing. You have nothing to show why people
> should listen to this.
> 

He is right?  You are judged here by the work you produce here, posts
mostly but it also includes research.  If the original poster had
googled for "Erik xml" in the CLL archives he could have found out why
Erik dislikes xml.  And he could have gotten a different response from
Erik by posting a well researched article with counter information
supporting his view that xml does not suck.  And if he could not
refute Erik's arguments then perhaps xml does suck after all.

> 
> wni at attbi dot com

And since you have been here long enough to know how Erik will respond
to stupid dick head messages like this, you claim long lurker status
above.  You must want to start a stupid fuckking shit storm, for no
good reason, that will waste a lot of time and effort.

Please go fuck off, and please do it somewhere else.

marc

ps it is not a full moon out so why start this shit tonight?
From: wni
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <WKJP9.374895$GR5.112414@rwcrnsc51.ops.asp.att.net>
Marc Spitzer wrote:
> wni <···@nospam.attbi.com> writes:
> 
> 
>>
>>
>>So it's your experience that counts? What makes you think that you are
>>qualified to criticize while the other people can't?
>>
> 
> 
> No it is his well formulated arguments that count.  And I have never 
> seen him say "do not do it", but I have seen him say "do not do such a
> bad job of it"

I fail to see how he did the job better. Stuffing a lot of twisted
sentences with no logic foundation is basically what I see in his
articles. Yes, you can claim and claim and claim on the Usenet without
being critically judged by real experts (who might never read this
group), but that doesn't make you an authority. One particular
silly instance is his attack on Paul Graham's "Being Popular" (
which dismisses Java) but he himself dismissed a lot of languages.
This kind of double standard is what makes him unbearably harsh.


> 
>>> But  according to the principles I laid down above, you, too, have the
>>>  ability to think and make some intelligent choices.  Now, hear me
>>>  when I tell you that nobody is interested in your past, in what you
>>>  believe is your personality or your identity or your personal pride.
>>>  You are only your arguments and your articles to this forum, and we
>>>  know /nothing/ about you as a person, nor does anybody care.
>>
>>We are not interested in your opinions, past, and whatever half-cooked,
>>densely packed dirty laundry of yours. Treating people this way has long
>>been your scare tactics and the laughing butt of comp.lang.lisp.
> 
> 
> Actually some of us are. 

That's your freedom and nobody will condemn you for that.

> 
> 
>>> Only
>>>  a bigoted idiot would deny /you/ the chance to change your mind and
>>>  fix your mistakes.  That means that your personal defense of your
>>>  past is not suitable for this forum, however strongly you feel the
>>>  urge to defend yourself.  In a debate, you are /expected/ to learn
>>>  when you are mistaken.
>>
>>
>>You are really being patronizing. You have nothing to show why people
>>should listen to this.
>>
> 
> 
> He is right?  You are judged here by the work you produce here, posts
> mostly but it also includes research.  If the original poster had
> googled for "Erik xml" in the CLL archives he could have found out why
> Erik dislikes xml.  And he could have gotten a different response from
> Erik by posting a well researched article with counter information
> supporting his view that xml does not suck.  And if he could not
> refute Erik's arguments then perhaps xml does suck after all.
> 

So what makes Mr Naggum the King in opinions that people have to submit
"well researched article?" Last time I checked, I can't find one single
reference to any of his name in

           http://citeseer.nj.nec.com/cs

That means he has no acadmic credential for people to believe him. So
we have to dismiss people like Philip Wadler and believe in Erik
Naggum instead? The massive number of "articles" posted by Mr Naggum
are mostly unfounded claims and personal attacks. From my point of
view, some of them made sense but I don't think I have the "high
gound" to judge people working on XML as "stupid."

Yes, Knuth, Perlis, Dijkstra, Hoare, Reyolds, Landin, etc. can
call me "stupid" whenever they want, because they've got credentials.
Alas, they are too smart to do such a thing.

So, who are you guys to judge people? If Mr Naggum has any refereed
paper in either a reputed conference, workshop, (even better) journal
to show us how *well researched* his arguments are, then people might
have the obligation to show him that. Until then, this kind of one
way street is nonsense.

> And since you have been here long enough to know how Erik will respond
> to stupid dick head messages like this, you claim long lurker status
> above.  You must want to start a stupid fuckking shit storm, for no
> good reason, that will waste a lot of time and effort.
> 
> Please go fuck off, and please do it somewhere else.
> 
> marc
> 
> ps it is not a full moon out so why start this shit tonight?
> 

A lot of people described c.l.l as a private elite club where only
certain class of people are welcome. From the above message, it
seems this *club* has been penetrated by *foul-mouth street gangs*
that blocks the area with such kind of sabre rattling.

wni at attbi dot com
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86hecwzang.fsf@bogomips.optonline.net>
wni <···@nospam.attbi.com> writes:

> Marc Spitzer wrote:
> > wni <···@nospam.attbi.com> writes:
> >
> 
> >>
> >>
> >>So it's your experience that counts? What makes you think that you are
> >>qualified to criticize while the other people can't?
> >>
> > No it is his well formulated arguments that count.  And I have never
> > seen him say "do not do it", but I have seen him say "do not do such
> > a bad job of it"
> 
> I fail to see how he did the job better. Stuffing a lot of twisted
> sentences with no logic foundation is basically what I see in his
> articles. Yes, you can claim and claim and claim on the Usenet without
> being critically judged by real experts (who might never read this
> group), but that doesn't make you an authority. One particular
> silly instance is his attack on Paul Graham's "Being Popular" (
> which dismisses Java) but he himself dismissed a lot of languages.
> This kind of double standard is what makes him unbearably harsh.
> 

You started this with an out of the blue personnel attack on Erik,
that you chose to do for your own reasons.  This is a remarkably
stupid thing to do for anyone who has been here for any length
of time.  Unless having someone verbally slap you around and having
this archived now and forever is your idea of fun.  If it is please
find another hobby or practice it elsewhere.

And a little well reasoned verbal slapping around is not in any way
shape or form unendurable, it the grand scheme of things it dose not
even count on scale of humanities experience with harshness.  Think of
some of the things that were normal for people to endure just a few
hundreds years ago, half or more of your children would die before the
age of 5 for example, which was endured for a long time. 

> 
> >
> 
> >>> But  according to the principles I laid down above, you, too, have the
> >>>  ability to think and make some intelligent choices.  Now, hear me
> >>>  when I tell you that nobody is interested in your past, in what you
> >>>  believe is your personality or your identity or your personal pride.
> >>>  You are only your arguments and your articles to this forum, and we
> >>>  know /nothing/ about you as a person, nor does anybody care.
> >>
> >>We are not interested in your opinions, past, and whatever half-cooked,
> >>densely packed dirty laundry of yours. Treating people this way has long
> >>been your scare tactics and the laughing butt of comp.lang.lisp.
> > Actually some of us are.
> 
> 
> That's your freedom and nobody will condemn you for that.
> 

How nice of you to say so.


> >
> 
> >>> Only
> >>>  a bigoted idiot would deny /you/ the chance to change your mind and
> >>>  fix your mistakes.  That means that your personal defense of your
> >>>  past is not suitable for this forum, however strongly you feel the
> >>>  urge to defend yourself.  In a debate, you are /expected/ to learn
> >>>  when you are mistaken.
> >>
> >>
> >>You are really being patronizing. You have nothing to show why people
> >>should listen to this.
> >>
> > He is right?  You are judged here by the work you produce here, posts
> 
> > mostly but it also includes research.  If the original poster had
> > googled for "Erik xml" in the CLL archives he could have found out why
> > Erik dislikes xml.  And he could have gotten a different response from
> > Erik by posting a well researched article with counter information
> > supporting his view that xml does not suck.  And if he could not
> > refute Erik's arguments then perhaps xml does suck after all.
> >
> 
> 
> So what makes Mr Naggum the King in opinions that people have to submit
> "well researched article?" Last time I checked, I can't find one single
> reference to any of his name in
> 
>            http://citeseer.nj.nec.com/cs
> 

Erik has well thought out and logical reasons for disliking xml that is
why and they are posted.  The other guy was I like it so there.

that is why

> That means he has no acadmic credential for people to believe him. So
> we have to dismiss people like Philip Wadler and believe in Erik
> Naggum instead? The massive number of "articles" posted by Mr Naggum
> are mostly unfounded claims and personal attacks. From my point of
> view, some of them made sense but I don't think I have the "high
> gound" to judge people working on XML as "stupid."
> 
> Yes, Knuth, Perlis, Dijkstra, Hoare, Reyolds, Landin, etc. can
> call me "stupid" whenever they want, because they've got credentials.
> Alas, they are too smart to do such a thing.
> 

So they do not even acknowledge your existence, they are smarter then I.

> So, who are you guys to judge people? If Mr Naggum has any refereed
> paper in either a reputed conference, workshop, (even better) journal
> to show us how *well researched* his arguments are, then people might
> have the obligation to show him that. Until then, this kind of one
> way street is nonsense.
> 

Well he has posted them here and they are in google for your reading
pleasure, go read.

> > And since you have been here long enough to know how Erik will respond
> > to stupid dick head messages like this, you claim long lurker status
> > above.  You must want to start a stupid fuckking shit storm, for no
> > good reason, that will waste a lot of time and effort.
> > Please go fuck off, and please do it somewhere else.
> 
> > marc
> 
> > ps it is not a full moon out so why start this shit tonight?
> 
> >
> 
> 
> A lot of people described c.l.l as a private elite club where only
> certain class of people are welcome. From the above message, it
> seems this *club* has been penetrated by *foul-mouth street gangs*
> that blocks the area with such kind of sabre rattling.
> 

Well you come here and start throwing verbal shit around because you
felt like it.  You then get offended because you are asked to fuck off
and go away.  Well fuck off and go away, take the hint this time.

marc

> wni at attbi dot com
From: wni
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <XXMP9.376309$GR5.113303@rwcrnsc51.ops.asp.att.net>
Marc Spitzer wrote:

>>
>>A lot of people described c.l.l as a private elite club where only
>>certain class of people are welcome. From the above message, it
>>seems this *club* has been penetrated by *foul-mouth street gangs*
>>that blocks the area with such kind of sabre rattling.
>>
> 
> 
> Well you come here and start throwing verbal shit around because you
> felt like it.  You then get offended because you are asked to fuck off
> and go away.  Well fuck off and go away, take the hint this time.

I am really sorry that I wrote *foul-mouth street gangs*. It should
have been *foul-mouth attack dogs*.

You can have the last bark.


wni at attbi dot com
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <864r8wyxuy.fsf@bogomips.optonline.net>
wni <···@nospam.attbi.com> writes:

> Marc Spitzer wrote:
> 
> >>
> >>A lot of people described c.l.l as a private elite club where only
> >>certain class of people are welcome. From the above message, it
> >>seems this *club* has been penetrated by *foul-mouth street gangs*
> >>that blocks the area with such kind of sabre rattling.
> >>
> > Well you come here and start throwing verbal shit around because you
> 
> > felt like it.  You then get offended because you are asked to fuck off
> > and go away.  Well fuck off and go away, take the hint this time.
> 
> I am really sorry that I wrote *foul-mouth street gangs*. It should
> have been *foul-mouth attack dogs*.

A properly trained attack dog make no noise, they don't bark they rip
your throat out.  You may be thinking of a guard dog, the noise can be
a deterrent.  It depends how they are trained.

woof

marc



> 
> You can have the last bark.
> 
> 
> wni at attbi dot com
From: Tim Daly, Jr.
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <874r8p3nf7.fsf@bob.intern>
wni <···@nospam.attbi.com> writes:

> I am really sorry that I wrote *foul-mouth street gangs*. It should
> have been *foul-mouth attack dogs*.
> 
> You can have the last bark.
> 
> 
> wni at attbi dot com

Does wni really mean weenie?

-Tim
From: Kenny Tilton
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3E0FA423.5030000@nyc.rr.com>
> A lot of people described c.l.l as a private elite club where only
> certain class of people are welcome.

Really? I would love some details. People somewhere are sitting around 
talking about c.l.l? A /lot/ of people? Do they also talk about Lisp? If 
so, what do they say? If not, why are they even reading cll?

I realize you are not speaking strictly, because no un-moderated NG can 
be private or elite. Unwelcome would be possible, if the preponderance 
of opinion re some poster was so negative that they could not take the 
heat and went away. I've seen that happen, with people who were really 
off base. Nothing wrong with making bozos feel unwelcome, I hope.

Periodically someone crawls out of the woodlurk to upbraid Erik, but it 
is hard to do that without self-referentially despoiling the joint and 
they take heat for that. Then, since they had nothing else to offer cl, 
they usually just fade away.

As you will (fade away), unless you want to talk about Lisp.

But first, since you have bandied about hearsay accusations against cll, 
please now back it up by specifying a few (or one!) threads in which 
reasonable people were made to feel unwelcome by the private elite.

We'll wait while you get back to "a lot of people" for those references. 
I am really curious because the only folks I have ever seen get into 
protracted battles with any cll regular are people willing to engage in 
protracted battles, and that is always optional.

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Tim Bradshaw
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ey3of733845.fsf@cley.com>
* Kenny Tilton wrote:
> I realize you are not speaking strictly, because no un-moderated NG
> can be private or elite. Unwelcome would be possible, if the
> preponderance of opinion re some poster was so negative that they
> could not take the heat and went away. I've seen that happen, with
> people who were really off base. Nothing wrong with making bozos feel
> unwelcome, I hope.

The last person who could be thought of as being hounded out of cll
who I can think of was Ilias, although he kind of hounded himself out
- I think by the end that basically no one would talk to him and he
got bored and wandered off.  Can anyone - *anyone* - give examples of
someone who really had a lot (or anything) to contribute who was
hounded out?

I think that there are two relevant features of cll which make it
somewhat ruder than (some - by no means all) other groups. (1) There
are a relatively large number of people who turn up with extremely
wrong and very negative views, and who are not interested in learning
the truth. I don't read any other language groups, but I *really*
doubt that the C groups, for instance, get this. (2) Some regular
posters say you are an idiot when they think you are.  So what's wrong
with that?

--tim
From: Chris Gehlker
Subject: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA35C17C.24DDB%gehlker@fastq.com>
On 12/30/02 7:16 AM, in article ···············@cley.com, "Tim Bradshaw"
<···@cley.com> wrote:

> I think that there are two relevant features of cll which make it
> somewhat ruder than (some - by no means all) other groups.

I think there is really only one thing that seems to be particularly
relevant. The sense of some that Lisp is poorly regarded by the programming
community in general. It is, after all, a commonplace that the vehemence
with which people express their views is directly proportional to their
expectation that those views are unpopular.

> (1) There
> are a relatively large number of people who turn up with extremely
> wrong and very negative views, and who are not interested in learning
> the truth.

How do you know whether people are interested in learning the truth when
they "turn up"?

> I don't read any other language groups, but I *really*
> doubt that the C groups, for instance, get this.

I suspect that the number of people who think C is a very flawed language
dwarfs the number of people who think that about Lisp. The difference is
that the C coders simply don't feel threatened by people who don't like C.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <bs33l49n.fsf@ccs.neu.edu>
Chris Gehlker <·······@fastq.com> writes:

> I suspect that the number of people who think C is a very flawed language
> dwarfs the number of people who think that about Lisp. The difference is
> that the C coders simply don't feel threatened by people who don't like C.

I don't think any Lisp aficionado feels threatened by *anyone* who
doesn't like lisp.  I think `annoyed' is the word.
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA35F0C9.2504B%gehlker@fastq.com>
On 12/30/02 12:01 PM, in article ············@ccs.neu.edu, "Joe Marshall"
<···@ccs.neu.edu> wrote:

> Chris Gehlker <·······@fastq.com> writes:
> 
>> I suspect that the number of people who think C is a very flawed language
>> dwarfs the number of people who think that about Lisp. The difference is
>> that the C coders simply don't feel threatened by people who don't like C.
> 
> I don't think any Lisp aficionado feels threatened by *anyone* who
> doesn't like lisp.  I think `annoyed' is the word.

But fear/anger or concern/annoyance if you prefer milder terms, are at root
the same emotion. At least the physiologists & social psychologists think
so.

This is all speculation anyway. I did do one experiment to test it though. I
searched a couple of Visual Basic groups for any signs that the posters were
as angry as the typical poster here. They clearly are not.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Kaz Kylheku
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <cf333042.0212301637.723dfedb@posting.google.com>
Chris Gehlker <·······@fastq.com> wrote in message news:<······················@fastq.com>...
> On 12/30/02 12:01 PM, in article ············@ccs.neu.edu, "Joe Marshall"
> <···@ccs.neu.edu> wrote:
> 
> > Chris Gehlker <·······@fastq.com> writes:
> > 
> >> I suspect that the number of people who think C is a very flawed language
> >> dwarfs the number of people who think that about Lisp. The difference is
> >> that the C coders simply don't feel threatened by people who don't like C.
> > 
> > I don't think any Lisp aficionado feels threatened by *anyone* who
> > doesn't like lisp.  I think `annoyed' is the word.
> 
> But fear/anger or concern/annoyance if you prefer milder terms, are at root
> the same emotion. At least the physiologists & social psychologists think
> so.
> 
> This is all speculation anyway. I did do one experiment to test it though. I
> searched a couple of Visual Basic groups for any signs that the posters were
> as angry as the typical poster here. They clearly are not.

VB-pushing idiots have nothing to be arrogant about, and are too
stupid to realize when someone proposes some wrong way of doing
something. Everything seems like a neat idea when all you know is VB;
the whole world resembles a child's crayon drawing hanging on a
refridgerator, depicting sunshine, stick figures holding hands, bees
landing on flowers and the like.

Speaking of C, the comp.lang.c newsgroup is one of the most hostile
ones. It's probably worse than comp.lang.lisp. It is indeed is quite
harsh toward ill-considered criticisms of the C language. It's a lot
less tolerant of implementation-specific discussions than
comp.lang.lisp. Here you can ask something that pertains only to, say,
LispWorks, Corman Lisp or CLISP without being told off. The harshness
is an explicit culture in comp.lang.c; it's meta-discussed from time
to time and generally accepted as a virtue. The basic tenet is to be
kind to genuine newbies with topical questions, and merciless toward
dumb answers. Correctness is very important; it's hard to get away
with any nonportable coding example, or assertion that is not backed
by the ISO standard.

I think that psychonanalyzing people in Usenet forums is futile. What
appears as anger or annoyance is often just a deliberate writing
voice. When you know you are right, it's extremely entertaining to don
an air of complete arrogance, and watch the little psycho-analyzing
idiots fall for it and get their feathers all ruffled.

If you want to psychoanalyze, start with this principle: people do
voluntarily that which gives them some kind of pleasure (whether it is
healthy or destructive is a different matter---contrast listening to
Beethoven with injecting heroin).
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA366E0D.251B3%gehlker@fastq.com>
On 12/30/02 5:37 PM, in article
····························@posting.google.com, "Kaz Kylheku"
<···@ashi.footprints.net> wrote:

> VB-pushing idiots have nothing to be arrogant about, and are too
> stupid to realize when someone proposes some wrong way of doing
> something. Everything seems like a neat idea when all you know is VB;
> the whole world resembles a child's crayon drawing hanging on a
> refridgerator, depicting sunshine, stick figures holding hands, bees
> landing on flowers and the like.
> 
> Speaking of C, the comp.lang.c newsgroup is one of the most hostile
> ones. It's probably worse than comp.lang.lisp.

I simply can't find anything posted on comp.lang.c over the last three days
that sounds remotely as angry as your paragraph that I quoted above.

The VB programmer that I know personally is neither arrogant, stupid nor an
idiot. She is very aware that if she lost her job as a VB programmer, she
would lose her visa and, considering her nationality, most of her life
expectancy.

I suppose if she went on about the virtues of VB compared to other languages
I would become mildly annoyed but she doesn't do that. But she seems to
accept it for what it is, an entry level language for people to use while
they study something more expressive.

In fact, the only people who advocate for VB seem to be paid salespeople.
There is Bruce Tognazzini but he is so clueless that he is actually funny.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Nils Goesche
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87fzseu5nf.fsf@darkstar.cartan>
Chris Gehlker <·······@fastq.com> writes:

> On 12/30/02 5:37 PM, in article
> ····························@posting.google.com, "Kaz Kylheku"
> <···@ashi.footprints.net> wrote:
> 
> > VB-pushing idiots have nothing to be arrogant about, and are
> > too stupid to realize when someone proposes some wrong way of
> > doing something. Everything seems like a neat idea when all
> > you know is VB; the whole world resembles a child's crayon
> > drawing hanging on a refridgerator, depicting sunshine, stick
> > figures holding hands, bees landing on flowers and the like.
> > 
> > Speaking of C, the comp.lang.c newsgroup is one of the most
> > hostile ones. It's probably worse than comp.lang.lisp.
> 
> I simply can't find anything posted on comp.lang.c over the
> last three days that sounds remotely as angry as your paragraph
> that I quoted above.

Look harder, then.  Maybe google for comp.lang.c and ``Kaz
Kylheku�� for a start ;-) Also, look for ``Dan Pop��.  Our local
politeness freaks should also google for Richard Heathfield, an
extremely polite English Gentleman who never gets out of line but
for some reason is regularly part of extremely vicious flame wars
(also in comp.programming) and being attacked by idiots accusing
him of all kinds of bad things (like being a Nazi and whatnot,
even though he only talks about C!?).  Facing Erik's style, it is
not as easy to see that it is indeed the people attacking him and
not himself who is to blame for those flame wars, and seeing how
someone else who is always as polite as one can possibly be is
the target of the very same aggressions by very similar people
might be very instructive for our ``feel-good jerks (tm)��.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA3681B1.251BD%gehlker@fastq.com>
On 12/30/02 10:19 PM, in article ··············@darkstar.cartan, "Nils
Goesche" <···@cartan.de> wrote:

>> Chris Gehlker <·······@fastq.com> writes:
>> 
>> I simply can't find anything posted on comp.lang.c over the
>> last three days that sounds remotely as angry as your paragraph
>> that I quoted above.
> 
> Look harder, then.  Maybe google for comp.lang.c and ``Kaz
> Kylheku�� for a start ;-) Also, look for ``Dan Pop��.  Our local
> politeness freaks should also google for Richard Heathfield, an
> extremely polite English Gentleman who never gets out of line but
> for some reason is regularly part of extremely vicious flame wars
> (also in comp.programming) and being attacked by idiots accusing
> him of all kinds of bad things (like being a Nazi and whatnot,
> even though he only talks about C!?).

I tried this experiment to discover that Kaz was apparently less angry when
he was a regular contributor to the C lists.  Dan Pop doesn't appear to be
angry at all, and if anybody is attacking Richard Heathfield, Google knows
it not.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Matthew Danish
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <20021231110844.B12928@lain.cheme.cmu.edu>
On Mon, Dec 30, 2002 at 11:05:21PM -0700, Chris Gehlker wrote:
> On 12/30/02 10:19 PM, in article ··············@darkstar.cartan, "Nils
> Goesche" <···@cartan.de> wrote:
> > Look harder, then.  Maybe google for comp.lang.c and ``Kaz
> > Kylheku?? for a start ;-) Also, look for ``Dan Pop??.  Our local
> > politeness freaks should also google for Richard Heathfield, an
> > extremely polite English Gentleman who never gets out of line but
> > for some reason is regularly part of extremely vicious flame wars
> > (also in comp.programming) and being attacked by idiots accusing
> > him of all kinds of bad things (like being a Nazi and whatnot,
> > even though he only talks about C!?).
> I tried this experiment to discover that Kaz was apparently less angry when
> he was a regular contributor to the C lists.  Dan Pop doesn't appear to be
> angry at all, and if anybody is attacking Richard Heathfield, Google knows
> it not.

Don't know how you missed it.

But I did find this gem from Richard himself:

``Martin Ambuhl (the C expert than whom you claimed to know more,
remember?) used to sigblock a guy who said something like "A thick skin
is a gift from God". This is especially true in the comp.lang.c
newsgroup.''

-- 
; Matthew Danish <·······@andrew.cmu.edu>
; OpenPGP public key: C24B6010 on keyring.debian.org
; Signed or encrypted mail welcome.
; "There is no dark side of the moon really; matter of fact, it's all dark."
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey365taqb8j.fsf@cley.com>
* Matthew Danish wrote:

> ``Martin Ambuhl (the C expert than whom you claimed to know more,
> remember?) used to sigblock a guy who said something like "A thick skin
> is a gift from God". This is especially true in the comp.lang.c
> newsgroup.''

I've just had the most weird garden-path experience from this quote.
I read `sigblock' as being SIGBLOCK (which there isn't, but never
mind) then was trying to read the whole thing as some kind of
metaphorical uses of Unix signalling terminology to talk about kill
files or something.  Very, very weird.

--tim
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <d6nij7lp.fsf@ccs.neu.edu>
Tim Bradshaw <···@cley.com> writes:

> I've just had the most weird garden-path experience from this quote.
> I read `sigblock' as being SIGBLOCK (which there isn't, but never
> mind) then was trying to read the whole thing as some kind of
> metaphorical uses of Unix signalling terminology to talk about kill
> files or something.  Very, very weird.

kill -file ....
From: Nils Goesche
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <878yy6tiry.fsf@darkstar.cartan>
Chris Gehlker <·······@fastq.com> writes:

> On 12/30/02 10:19 PM, in article ··············@darkstar.cartan, "Nils
> Goesche" <···@cartan.de> wrote:
> 
> >> Chris Gehlker <·······@fastq.com> writes:
> >> 
> >> I simply can't find anything posted on comp.lang.c over the
> >> last three days that sounds remotely as angry as your paragraph
> >> that I quoted above.
> > 
> > Look harder, then.  Maybe google for comp.lang.c and ``Kaz
> > Kylheku�� for a start ;-) Also, look for ``Dan Pop��.  Our local
> > politeness freaks should also google for Richard Heathfield, an
> > extremely polite English Gentleman who never gets out of line but
> > for some reason is regularly part of extremely vicious flame wars
> > (also in comp.programming) and being attacked by idiots accusing
> > him of all kinds of bad things (like being a Nazi and whatnot,
> > even though he only talks about C!?).
> 
> I tried this experiment to discover that Kaz was apparently
> less angry when he was a regular contributor to the C lists.
> Dan Pop doesn't appear to be angry at all, and if anybody is
> attacking Richard Heathfield, Google knows it not.

Heh.  Absolutely wrong.  Your google skills are lacking to say
the least.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Thomas Stegen
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e11b7c1@nntphost.cis.strath.ac.uk>
Chris Gehlker wrote:

> I tried this experiment to discover that Kaz was apparently less angry when
> he was a regular contributor to the C lists.  Dan Pop doesn't appear to be
> angry at all, and if anybody is attacking Richard Heathfield, Google knows
> it not.

Search for richard heathfield eric gorr. Then look for the thread code
indentation tools. Also look for richard and edward g nilges in
comp.programming. Also try out Thomas Bushcloo and PaulG in conjunction
with Richard Heathfield.

-- 
Thomas.
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0212312145.5b949dd2@posting.google.com>
Chris Gehlker <·······@fastq.com> wrote in message news:<······················@fastq.com>...
> On 12/30/02 10:19 PM, in article ··············@darkstar.cartan, "Nils
> Goesche" <···@cartan.de> wrote:
> 
> >> Chris Gehlker <·······@fastq.com> writes:
> >> 
> >> I simply can't find anything posted on comp.lang.c over the
> >> last three days that sounds remotely as angry as your paragraph
> >> that I quoted above.
> > 
> > Look harder, then.  Maybe google for comp.lang.c and ``Kaz
> > Kylheku�� for a start ;-) Also, look for ``Dan Pop��.  Our local
> > politeness freaks should also google for Richard Heathfield, an
> > extremely polite English Gentleman who never gets out of line but
> > for some reason is regularly part of extremely vicious flame wars
> > (also in comp.programming) and being attacked by idiots accusing
> > him of all kinds of bad things (like being a Nazi and whatnot,
> > even though he only talks about C!?).

The facts are these.

Om May this year I suggested that comp.programming discuss the US Data
Quality Act.

I was repeatedly and repetitiously attacked by Richard Heathfield for
being "off topic" despite the topicality of an act which prevents
government entities from posting data and its software when
corporations use software with documented failure rates of 75%.

It is unlikely that a crude Lisp program could detect the pattern
Richard used, for in fact he used no rude words.

The rudeness was constituted in his repeatedly posting the same
objections even after I made a good faith and polite attempt to answer
them.

It was less like a student using "bad" language and more like a
person, in a conversational group or seminar, who repeatedly returns
to a question that has been responded to.

Richard, last summer, repeatedly and in a mechanical fashion which
failed to address the issues, returned to themes that I and others
addressed.  Such behavior in a classroom would be labeled disruptive. 
Of course, Internet discussions are not led by a "teacher."  But this
makes his behavior only more disruptive, for when a class conducts a
roundtable discussion in which the instructor leaves the room, it is
considered very bad form for one discussant to latch onto one other
discussant, and make the same objection, repeatedly.

The practice is called "crosstalk" and it disrupts group problem
solving.

Richard's behavior had the effect of making him and myself the
discussion "stars" when in fact I wanted to listen to what other
people experienced in the way corporations disregard software quality
and corporate governance while demanding that environmental groups and
governments apply standards to which the corporations fail to occur.

As it happened, no body was able to make a case in the corporation's
favor.  In the entire discussion, which was very large, nobody even
attempted to post facts showing me wrong about high failure rates in
private "enterprise" systems.  Because of Richard's disruptive
behavior, nobody every got around to doing their homework on the side
of the corporation.

Instead, the discussion began with repeated challenges to its
topicality, which were answered in good faith, and then devolved into
a series of meta-flames.  Richard's behavior was to blame for this,
and ultimately, I believe, he may have seen the error of his ways and
he left the discussion.

It is unlikely that a Lisp program could have detected this serious
violation of common courtesy.

As to the "Nazi" business.  I did compare Richard's use of electronic
media, to repeat the same concerns mechanically in order to foreground
them, to Fascist practice and not "Nazi" practice.  This comparision
was met with a complete misreading of Godwin's "law" which was
supposed by discussants to be a ban on comparisions to Hitler when in
fact Mike Godwin's law was "in Internet discussions, the probability
of comparision to Hitler approaches unity over time."

This misreading caused Heathfield to make a false claim, that I had
"called" him a "Nazi."

This business of using Lisp for courtesy indexing certainly sounds
like a bad idea, for no stylistic analysis can detect the presence or
absence of moral behavior.  However, I do suppose that aliterates can
use it for terrorist surveillance, and change its parameters more or
less at will until the usual suspects are rounded up.


    
> 
> I tried this experiment to discover that Kaz was apparently less angry when
> he was a regular contributor to the C lists.  Dan Pop doesn't appear to be
> angry at all, and if anybody is attacking Richard Heathfield, Google knows
> it not.
> 
> 
> 
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E1299AE.50903@nyc.rr.com>
Edward G. Nilges wrote:
> Richard's behavior had the effect of making him and myself the
> discussion "stars" when in fact I wanted to...

Did it ever occur to you to simply not respond to the guy? Or to respond 
with ten words max?

I think a huge problem in these deals is that people (quite reasonably!) 
read something attacking them or what they say and feel that unless they 
respond the community will judge them the loser. I wager what realy 
happens is that the community only does that when the attack is pretty 
solid. And on the flip side, ignoring crap wins the ignorer the 
admiration and gratitude of everyone for not perpetuating some silly, 
typical NG tit-for-tat flamewar.




-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301011138.24948d04@posting.google.com>
Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> Edward G. Nilges wrote:
> > Richard's behavior had the effect of making him and myself the
> > discussion "stars" when in fact I wanted to...
> 
> Did it ever occur to you to simply not respond to the guy? Or to respond 
> with ten words max?

You raise a good point.  The whole flaw with "rudeness indexing" is it
assumes that "rudeness" is exclusively individual, and this commits
what psychologists call the fundamental attribution error.

I bear some responsibility for Richard's rudeness because his
repetitious postings were CAUSED by my witty but at times over-the-top
responses.

But this means that it is absurd to think that by evaluating
INDIVIDUAL postings, one can come up with a meaningful index of
rudeness.  This is not to say that I don't find the venture
interesting.  I am afraid, however, that Google might use it to ban
posters.

I suppose that you could use Lisp for social research (has any Lisp
programmer done anything with Theodore Adorno's 1949 F-scale?) by
examining collective postings.  But this would require a political
re-education of the programmers and designers into an understanding
that Margaret Thatcher was wrong.

For example, I would be delighted if an AI program read every post in
the Data Quality Act and concluded that I was an angel of light, beset
by meanies and trolls.  Indeed, if I sat down and relearned Lisp, I
could write such a program and that, my friend, is the problem in the
first place.  As it is, I don't even have to use Lisp: I can use C:

     main { printf("Ed Nilges is quite a guy\n") }


> 
> I think a huge problem in these deals is that people (quite reasonably!) 
> read something attacking them or what they say and feel that unless they 
> respond the community will judge them the loser. I wager what realy 
> happens is that the community only does that when the attack is pretty 
> solid. And on the flip side, ignoring crap wins the ignorer the 
> admiration and gratitude of everyone for not perpetuating some silly, 
> typical NG tit-for-tat flamewar.

This is like telling the kid who is the victim of a bully that he
should take the long way to school.  Useful in a pragmatic sense,
socially, it has the effect of maintaining a bullying culture.

In this connection, I have posted, recently, a very similar, long,
article on the new Total Information Awareness program from our idiot
Administration in the same place as my Data Quality Act posting of May
2002.

In part because of what you say, I have decided NOT to reply to any
responses to this post.

This will avoid loading servers down with another monster thread.

I can do so in part because Peter Neumann accepted a much shorter, and
far mre toned down, precis of my views for comp.risks, and this is
also, probably, going to be published in a privacy journal.

I shal restrict my exchanges on the TIA to comp.risks where Peter is
the excellent moderator of same.

BTW, I note that this is a comp.lang.lisp discussion.  I do not
currently use Lisp, I have crossposted to comp.programming so that
readers of comp.programming know my decision, and I will discuss this
issue on comp.programming.
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E135476.1000701@nyc.rr.com>
Edward G. Nilges wrote:
> Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
>> And on the flip side, ignoring crap wins the ignorer the
>>admiration and gratitude of everyone for not perpetuating some silly, 
>>typical NG tit-for-tat flamewar.
> 
> 
> This is like telling the kid who is the victim of a bully that he
> should take the long way to school.  Useful in a pragmatic sense,
> socially, it has the effect of maintaining a bullying culture.

But if I make a post and three people respond, one with an article 
attacking me in ways I do not find interesting, I can ignore that one 
and respond to the other two. That becomes my response if anyone 
notices, if not at least the attack is not fueled by my actually responding.

This is more like turning on my Immateriality Cloak so the bully's punch 
goes right through me. I can refuse to fight without taking another 
route, I simply do not fight (respond) and no one can make me do so. So 
the bully analogy breaks down in what I think is an important way.

Importantly, I believe this does not encourage the attacker, it renders 
them impotent.

> In part because of what you say, I have decided NOT to reply to any
> responses to this post.

<g> well, i must admit it is hard to resist responding. good luck with 
your resolution.


-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301012033.544ef3a2@posting.google.com>
Kenny Tilton <·······@nyc.rr.com> wrote in message news:<················@nyc.rr.com>...
> Edward G. Nilges wrote:
> > Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> >> And on the flip side, ignoring crap wins the ignorer the
> >>admiration and gratitude of everyone for not perpetuating some silly, 
> >>typical NG tit-for-tat flamewar.
> > 
> > 
> > This is like telling the kid who is the victim of a bully that he
> > should take the long way to school.  Useful in a pragmatic sense,
> > socially, it has the effect of maintaining a bullying culture.
> 
> But if I make a post and three people respond, one with an article 
> attacking me in ways I do not find interesting, I can ignore that one 
> and respond to the other two. That becomes my response if anyone 
> notices, if not at least the attack is not fueled by my actually responding.
> 
> This is more like turning on my Immateriality Cloak so the bully's punch 
> goes right through me. I can refuse to fight without taking another 
> route, I simply do not fight (respond) and no one can make me do so. So 
> the bully analogy breaks down in what I think is an important way.
> 
> Importantly, I believe this does not encourage the attacker, it renders 
> them impotent.

The trouble was that Richard Heathfield raised an important issue. 
Large numbers of computer programmers do believe that their work and
its value is completely independent of society and are therefore
averse to any discussion of its wider import.  Richard felt that
comp.programming should not discuss wider issues of legislation.

My view was that programming won't be a profession until programmers
stop being mercenaries.  Therefore I went at it with Richard
Heathfield.  I probably did not convince him but I hope I made him
think.

> 
> > In part because of what you say, I have decided NOT to reply to any
> > responses to this post.
> 
> <g> well, i must admit it is hard to resist responding. good luck with 
> your resolution.

Not at all.  I keep promises.  Furthermore, I betya Al Dunbar has
organized a thread boycott behind my back because the guy doesn't want
to referee another Gangs of New York flame war.
From: Coby Beck
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <av0i4p$23p0$1@otis.netspace.net.au>
"Edward G. Nilges" <···········@yahoo.com> wrote in message
·································@posting.google.com...
> Kenny Tilton <·······@nyc.rr.com> wrote in message
news:<················@nyc.rr.com>...
> > Edward G. Nilges wrote:
> > > In part because of what you say, I have decided NOT to reply to any
> > > responses to this post.
> >
> > <g> well, i must admit it is hard to resist responding. good luck with
> > your resolution.
>
> Not at all.  I keep promises.

hmm... my irony meter just spun past the red zone!  ;)

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Strangely Placed
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <acec90c3.0301011225.51343e8b@posting.google.com>
Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> Edward G. Nilges wrote:
> > Richard's behavior had the effect of making him and myself the
> > discussion "stars" when in fact I wanted to...
> 
> Did it ever occur to you to simply not respond to the guy? Or to respond 
> with ten words max?

It occurred to /me/ (eventually). At the urging of comp.programming, I
killfiled Mr Nilges. There is only one unfortunate consequence of this
action, which is that he is now free to spew any old rubbish as if it
were fact. I urge you to check anything he tells you (even something
as simple as today's date!) before believing it. I've just been
reading some of his stuff in groups.google.com (a friend clued me in
to this latest idiocy of his via email), and the nonsense he is
spewing about me in this thread just beggars belief. The facts of any
exchange between myself and Mr Nilges (except for a very brief and
frosty email conversation which he will not allow me to disclose,
presumably because it embarrasses him) are available for you to check
via Google.

> 
> I think a huge problem in these deals is that people (quite reasonably!) 
> read something attacking them or what they say and feel that unless they 
> respond the community will judge them the loser.

Yes, I think that's fair. It takes courage to killfile someone like Mr
Nilges.

> I wager what realy 
> happens is that the community only does that when the
> attack is pretty  solid. And on the flip side,
> ignoring crap wins the ignorer the admiration
> and gratitude of everyone for not perpetuating
> some silly, typical NG tit-for-tat flamewar.

I hope so. I also hope I haven't started a whole new flame war by
responding in this thread. If so, my apologies.

-- 
Richard Heathfield
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E139310.6030507@nyc.rr.com>
Strangely Placed wrote:
> Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> 
>>Edward G. Nilges wrote:
>>
>>>Richard's behavior had the effect of making him and myself the
>>>discussion "stars" when in fact I wanted to...
>>
>>Did it ever occur to you to simply not respond to the guy? Or to respond 
>>with ten words max?
> 
> 
> It occurred to /me/ (eventually). At the urging of comp.programming, I
> killfiled Mr Nilges. There is only one unfortunate consequence of this
...
> 
> ....I also hope I haven't started a whole new flame war by
> responding in this thread. If so, my apologies.
> 

Who has their original Star Trek Concordance handy? I am looking for the 
episode in which some deep space low-life gets picked up by the 
enterprise and then starts wreaking havoc, followed by the good guy that 
has been chasing him through space trying to nam him. It ends with the 
good guy agreeing to some ploy which will have the unpleasant 
side-effect of leaving him locked eternally grappling with the bad guy, 
but hey, he does it to save the universe and all that.

That image of two folks locked in an eternal grappling match always 
comes to mind when I see a protracted flamewar (tho the good-bad thing 
does not apply, it's more like consensual S&M).

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Geoffrey Summerhayes
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <uJMQ9.4444$8n5.730054@news20.bellglobal.com>
"Kenny Tilton" <·······@nyc.rr.com> wrote in message ·····················@nyc.rr.com...
>
> Who has their original Star Trek Concordance handy? I am looking for the
> episode in which some deep space low-life gets picked up by the
> enterprise and then starts wreaking havoc, followed by the good guy that
> has been chasing him through space trying to nam him. It ends with the
> good guy agreeing to some ploy which will have the unpleasant
> side-effect of leaving him locked eternally grappling with the bad guy,
> but hey, he does it to save the universe and all that.
>
> That image of two folks locked in an eternal grappling match always
> comes to mind when I see a protracted flamewar (tho the good-bad thing
> does not apply, it's more like consensual S&M).
>

Sounds like this one:

The Alternative Factor
http://www.startrek.com/library/tos_episodes/episodes_tos_detail_68700.asp

--
Geoff
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E13A974.7050506@nyc.rr.com>
Geoffrey Summerhayes wrote:
> "Kenny Tilton" <·······@nyc.rr.com> wrote in message ·····················@nyc.rr.com...

>>That image of two folks locked in an eternal grappling match always
>>comes to mind when I see a protracted flamewar (tho the good-bad thing
>>does not apply, it's more like consensual S&M).
>>
> 
> 
> Sounds like this one:
> 
> The Alternative Factor
> http://www.startrek.com/library/tos_episodes/episodes_tos_detail_68700.asp

Bingo. Thx.



-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301031700.36ef0489@posting.google.com>
······@eton.powernet.co.uk (Strangely Placed) wrote in message news:<····························@posting.google.com>...
> Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> > Edward G. Nilges wrote:
> > > Richard's behavior had the effect of making him and myself the
> > > discussion "stars" when in fact I wanted to...
> > 
> > Did it ever occur to you to simply not respond to the guy? Or to respond 
> > with ten words max?
> 
> It occurred to /me/ (eventually). At the urging of comp.programming, I
> killfiled Mr Nilges. There is only one unfortunate consequence of this
> action, which is that he is now free to spew any old rubbish as if it
> were fact. I urge you to check anything he tells you (even something
> as simple as today's date!) before believing it. I've just been
> reading some of his stuff in groups.google.com (a friend clued me in
> to this latest idiocy of his via email), and the nonsense he is
> spewing about me in this thread just beggars belief. The facts of any
> exchange between myself and Mr Nilges (except for a very brief and
> frosty email conversation which he will not allow me to disclose,
> presumably because it embarrasses him) are available for you to check
> via Google.

Posters, be aware that the poster here is Richard Heathfield, the
instigator of a bullying campaign, who has changed his handle possibly
to conceal his identity, although it does appear below.

Richard, thanks most awfully for another document for my file
"Heathfield, Richard, pro se lawsuit, regarding, actionable libel,
under English and American, law."

You were urged to do so because we were collectively tired of your
repetitious and abusive responses to my concerns about the Data
Quality Act, which is of high technical and social revelance to
comp.programming.  In a subordinate fashion, the rest of the group was
also wearied by my detailed responses to all your points.

Your responses constituted a libelous attempt to damage my
professional, and therefore financial, standing, by making claims so
overbroad (as in the above, where you cast doubt on anything I say) as
to be provably false.

I have decided not to let you get away with this conduct this time. 
Therefore, I sent the post that appears in comp.programming about the
Total Information Awareness program to Peter Neumann, the moderator of
comp.risks.

He said it was overlong, and so broad in some political claims as to
lie outside the charter of comp.risks.  He asked me to write a smaller
article focusing on the technical objections and their immediate
political implications.  I did so and he released the article, which
is being considered for publication in a journal on these issues.

In consequence, Richard, I decided to not present you with another
opportunity to engage in libel and disruption, and won't be responding
to any comments on the comp.programming post.

You are instead encouraged to send any criticisms as to the revelance
or correctness of my views as expressed to Peter or to me by email. 
Contrary to what gswork implied, these posts, if acceptable and only
if, will appear as line comments in the same style as an unmoderated
group.

What is of general interest is the supposed liberal "bias" of
comp.risks that several neoconservative people have remarked upon.

What is said to be a liberal bias is merely the enforcement by a
virtuous human being, of common decency and fairness, which was absent
in your conduct last year.

Edward G. Nilges
304 Lakeside Drive
Valparaiso, IN  46383
···········@yahoo.com
http://members.screenz.com/edNilges


[On the subject of Lisp.  I have not used Lisp recently.  But ANY
language that memorializes the register structure of a vintage machine
certainly is suspect as a useful tool.]

[Professor McCarthy deserves credit for inventing the linked list. 
However, it is far less important than the concept of the stack, which
was independently invented by Dijsktra and by Turing.  Furthermore, OO
languages destroy most of the need for linked lists.]

[I do so need to be on topic, you chaps.]

> 
> > 
> > I think a huge problem in these deals is that people (quite reasonably!) 
> > read something attacking them or what they say and feel that unless they 
> > respond the community will judge them the loser.
> 
> Yes, I think that's fair. It takes courage to killfile someone like Mr
> Nilges.
> 
> > I wager what realy 
> > happens is that the community only does that when the
> > attack is pretty  solid. And on the flip side,
> > ignoring crap wins the ignorer the admiration
> > and gratitude of everyone for not perpetuating
> > some silly, typical NG tit-for-tat flamewar.
> 
> I hope so. I also hope I haven't started a whole new flame war by
> responding in this thread. If so, my apologies.
From: Rahul Jain
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <878yy0bbk0.fsf@localhost.localdomain>
···········@yahoo.com (Edward G. Nilges) writes:

> [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> language that memorializes the register structure of a vintage machine
> certainly is suspect as a useful tool.]
> 
> [Professor McCarthy deserves credit for inventing the linked list. 
> However, it is far less important than the concept of the stack, which
> was independently invented by Dijsktra and by Turing.  Furthermore, OO
> languages destroy most of the need for linked lists.]
> 
> [I do so need to be on topic, you chaps.]

Do you take us for such fools that we'll think that you're smarter
because of posting such ignorant drivel?

Hint: CAR and CDR are very simple to remember and have nothing to do
with memorizing register structures of anything. Also, I don't see how
Lisp destroys the need for linked lists or precludes the use of a stack.


--
Rahul Jain
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301050020.602302f8@posting.google.com>
Rahul Jain <·····@rice.edu> wrote in message news:<··············@localhost.localdomain>...
> ···········@yahoo.com (Edward G. Nilges) writes:
> 
> > [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> > language that memorializes the register structure of a vintage machine
> > certainly is suspect as a useful tool.]
> > 
> > [Professor McCarthy deserves credit for inventing the linked list. 
> > However, it is far less important than the concept of the stack, which
> > was independently invented by Dijsktra and by Turing.  Furthermore, OO
> > languages destroy most of the need for linked lists.]
> > 
> > [I do so need to be on topic, you chaps.]
> 
> Do you take us for such fools that we'll think that you're smarter
> because of posting such ignorant drivel?

Oh geez...
> 
> Hint: CAR and CDR are very simple to remember and have nothing to do
> with memorizing register structures of anything. Also, I don't see how
> Lisp destroys the need for linked lists or precludes the use of a stack.

Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
based on the IBM 701's register structure and CAR and CDR are based on
operations in its machine language.  As of now I stand by this claim.

You just misread me when you say "I don't see how...", because I said
OO, not Lisp "destroys the need for linked lists", and I did not say
that Lisp precludes the use of a stack.
From: Paul F. Dietz
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <q72dnUfkYORS3IWjXTWc2Q@dls.net>
Edward G. Nilges wrote:

> Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> based on the IBM 701's register structure and CAR and CDR are based on
> operations in its machine language.  As of now I stand by this claim.

'Architecture' == 'the name of two symbols'?

Incredibly lame.

	Paul
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301051311.67800184@posting.google.com>
"Paul F. Dietz" <·····@dls.net> wrote in message news:<······················@dls.net>...
> Edward G. Nilges wrote:
> 
> > Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> > based on the IBM 701's register structure and CAR and CDR are based on
> > operations in its machine language.  As of now I stand by this claim.
> 
> 'Architecture' == 'the name of two symbols'?
> 
> Incredibly lame.
> 
> 	Paul

'useless code' == 'one steenking bug'?

Lisp is useful and a lot of good work is done in Lisp.  But its
Original Sin will never be forgiven.  By me, that is.  For what it is
worth.
From: Marc Spitzer
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <864r8nrumt.fsf@bogomips.optonline.net>
···········@yahoo.com (Edward G. Nilges) writes:

> "Paul F. Dietz" <·····@dls.net> wrote in message news:<······················@dls.net>...
> > Edward G. Nilges wrote:
> > 
> > > Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> > > based on the IBM 701's register structure and CAR and CDR are based on
> > > operations in its machine language.  As of now I stand by this claim.
> > 
> > 'Architecture' == 'the name of two symbols'?
> > 
> > Incredibly lame.
> > 
> > 	Paul
> 
> 'useless code' == 'one steenking bug'?
> 
> Lisp is useful and a lot of good work is done in Lisp.  But its
> Original Sin will never be forgiven.  By me, that is.  For what it is
> worth.

Nothing.  

One of the real true metrics of a programming language is the scope of
problems that can be solved in it.  Lisp and CL have a long history of
solving problems that were thought to be too hard in other languages
or too hard period.  One of the more recent ones it the search engine
that runs travelocity.com airfare site.  Lisp & CL specifically have a
history of solving problems that were too hard for Algol style
languages(c, c++, pascal) to deal with.

marc
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3u1gnir54.fsf@cley.com>
* Edward G Nilges wrote:

> Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> based on the IBM 701's register structure and CAR and CDR are based on
> operations in its machine language.  As of now I stand by this
> claim.

Well, there are 978 symbols in the CL package.  2 of them come from
IBM 701 mnemonics, another two come from PDP-10 mnemonics.  So the
`architecture of Lisp' is based on the IBM 701's register
architecture?  Why not the DEC-10, which would certainly seem a more
plausible candidate given CL's maclisp heritage? Or, given that 99.6%
of the symbols in CL don't actually come from either machine, why not
just give up and think?  No, I know, it's easier to just not bother.

--tim
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301051309.78826847@posting.google.com>
Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> * Edward G Nilges wrote:
> 
> > Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> > based on the IBM 701's register structure and CAR and CDR are based on
> > operations in its machine language.  As of now I stand by this
> > claim.
> 
> Well, there are 978 symbols in the CL package.  2 of them come from
> IBM 701 mnemonics, another two come from PDP-10 mnemonics.  So the
> `architecture of Lisp' is based on the IBM 701's register
> architecture?  Why not the DEC-10, which would certainly seem a more
> plausible candidate given CL's maclisp heritage? Or, given that 99.6%
> of the symbols in CL don't actually come from either machine, why not
> just give up and think?  No, I know, it's easier to just not bother.

Thanks for the information and indeed the confirmation, Tim, but I am
not sure what you might mean by "thinking."  I fear it may be a sort
of slack-jawed, drool streaked refusal to interpret and in cybernetic
terms equal to watching Nul characters, or grass grow.

I do have a systematic tendency as a teacher and student formerly of
philosophy who entered the computer racket 30 years ago as a
draft-dodging scheme, to bring humanistic interpretation into the
field, and to me CAR and CDR are representative of a tendency in LISP
(please be advised that I have no special expertise in Lisp).

This is to tie the language, at a key point, to the vendor's hardware.
 It was also done in Fortran about the same time.

Whereas the Algol group pioneered what has turned out to be a better
approach, and that is for a quasi-civic agency to be technically
independent of computer vendors.  The Algol team devised the father of
most usable modern languages including C because as an independent
civic forum, they were immune to commercial pressures.

C does have a register declaration but note that it is as abstract and
not as specific as possible unlike CAR and CDR.

This immunity to marketing demands was interpreted as "failure to meet
deadlines" by the vendors whose language pioneered the use of
artificial cut-off dates to structure development for exclusively
private ends.

The split between civic technical society and commerce was evident as
early as 1968 in Saul Rosen's collection Programming Systems and
Languages, published that year by Addison-Wesley.  The late Edsger
Dijkstra to me was the Vaclav Havel of the split if we systematically
replace commerce and its implicit ideology with thug Communism.

OK, pal, you've confirmed that there is garbage in Lisp which does not
appear in Prolog and which in general does not belong in a programming
language, which should be machine and vendor independent.  But you
interpret from the slag heap that results from the small injection of
bias that no criticism of the result is possible.  With all do
respect, buddy, I demur.

99.6% is a meaningless number, for it is possible to exceed six sigma
in programming language design...it is possible to have NO reference
to unneeded garbage.


> 
> --tim
From: Matthew Danish
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <20030105171540.F12928@lain.cheme.cmu.edu>
What's all this about registers and architectures?

CONS creates a data structure of type CONS.  It is a function of two
arguments.  CAR is a function which accepts a CONS and returns the first
of the two values.  CDR is a function which accepts a CONS and returns
the second of the two values.  It is fairly straightforward to implement
said data structure without delving into any register or architectural
details.

That said conses may be used to create lists is incidental.  Thus when
dealing with conses merely as conses, the functions FIRST and REST only
serve to confuse; they appear to denote operations that might be
performed on lists.

I have no idea why you're making such a big deal about this rather
simple data structure.  Personally, I think that while conses can be
great, the true nature of Lisp lies in the importance of functions, and
the emphasis on programmability.  Few, if any, ALGOL-based languages
even strive to reach this goal as Common Lisp does [1].

Yes, I said Common Lisp.  This is the year 2003, not 1963.  The language
we discuss is here Common Lisp, not LISP, and it has far outgrown the
original acronym.  Yes, it has flaws, but no, this does not stop it from
being useful.  And there are far more important matters than whether the
names CAR and CDR should be used for standard functions, or not.

For your perusal:
http://www.lispworks.com/reference/HyperSpec/Front/index.htm


[1] Though one could say that Common Lisp is ALGOL-derived too, through
    Scheme, by the use of lexical scope.

-- 
; Matthew Danish <·······@andrew.cmu.edu>
; OpenPGP public key: C24B6010 on keyring.debian.org
; Signed or encrypted mail welcome.
; "There is no dark side of the moon really; matter of fact, it's all dark."
From: Andreas Eder
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m3u1gmh9en.fsf@elgin.eder.de>
Matthew Danish <·······@andrew.cmu.edu> writes:

> What's all this about registers and architectures?
> 
> CONS creates a data structure of type CONS.  It is a function of two
> arguments.  CAR is a function which accepts a CONS and returns the first
> of the two values.  CDR is a function which accepts a CONS and returns
> the second of the two values.  It is fairly straightforward to implement
> said data structure without delving into any register or architectural
> details.

What's all this about data structures? 
That's all way too abstract nonsense. Whenever I use a stack I am very
careful with my pushs and pops, not too break any plates!

'Andreas
-- 
Wherever I lay my .emacs, there�s my $HOME.
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ava9ue$383$1@newsreader2.netcologne.de>
Edward G. Nilges wrote:
> Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> 
>>* Edward G Nilges wrote:
>>
>>
>>>Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
>>>based on the IBM 701's register structure and CAR and CDR are based on
>>>operations in its machine language.  As of now I stand by this
>>>claim.
>>
>>Well, there are 978 symbols in the CL package.  2 of them come from
>>IBM 701 mnemonics, another two come from PDP-10 mnemonics.  So the
>>`architecture of Lisp' is based on the IBM 701's register
>>architecture?  Why not the DEC-10, which would certainly seem a more
>>plausible candidate given CL's maclisp heritage? Or, given that 99.6%
>>of the symbols in CL don't actually come from either machine, why not
>>just give up and think?  No, I know, it's easier to just not bother.
> 
> 
> Thanks for the information and indeed the confirmation, Tim, but I am
> not sure what you might mean by "thinking."

The names "car" and "cdr" are not an issue at all for Lisp programmers. 
On the contrary, some even appreciate them. If you are really interested 
why, just ask.

> I do have a systematic tendency as a teacher and student formerly of
> philosophy who entered the computer racket 30 years ago as a
> draft-dodging scheme, to bring humanistic interpretation into the
> field, and to me CAR and CDR are representative of a tendency in LISP

They are not.

> (please be advised that I have no special expertise in Lisp).

You are criticizing a very superficial aspect of Lisp. You should first 
study Lisp and find out what the essential aspects of Lisp are before 
you start to criticize it. You wouldn't criticize the Japanese language 
either just because of the "funny" characters.

> This is to tie the language, at a key point, to the vendor's hardware.
>  It was also done in Fortran about the same time.

car and cdr are not tied to any hardware. The semantics of car and cdr 
are clearly specified and can be implemented an any hardware. Lisp just 
happened to be implemented on a specific hardware first.

> Whereas the Algol group pioneered what has turned out to be a better
> approach, and that is for a quasi-civic agency to be technically
> independent of computer vendors.  The Algol team devised the father of
> most usable modern languages including C because as an independent
> civic forum, they were immune to commercial pressures.

It's not clear at all that the Algol approach is the better one. On the 
contrary, there are many signs that Algol-style languages have reached 
their limits. There is much research going on on things like 
aspect-oriented programming, generative programming, support for 
domain-specific languages etc. - things that are unnecessarily complex 
in Algol languages. Scripting languages like Python and Ruby are 
increasingly incorporating features originally found in Lisp. etc. pp.

Many things are continually being reinvented that are more or less 
already solved in Lisp.

> C does have a register declaration but note that it is as abstract and
> not as specific as possible unlike CAR and CDR.

car and cdr are not tied to registers at all.


Here are my favorite two papers about the essence of Lisp as far as I 
understand it.

* http://www.paulgraham.com/rootsoflisp.html
* ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-453.pdf

Further pointers can be found at 
http://www.pascalcostanza.de/lisp/guide.html and, of course, 
http://www.lisp.org



Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Paul F. Dietz
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <JmqdnQA0Uf4aMoWjXTWcog@dls.net>
Pascal Costanza wrote:

>> This is to tie the language, at a key point, to the vendor's hardware.
>>  It was also done in Fortran about the same time.
> 
> car and cdr are not tied to any hardware. The semantics of car and cdr 
> are clearly specified and can be implemented an any hardware. Lisp just 
> happened to be implemented on a specific hardware first.


I think what happens with these trolls is that they've try to learn Lisp,
failed to get it for some reason, then (as a defense mechanism) invented some
flaw in Lisp to rationalize their failure.  I had thought that 'too many
parens' was the stupidist of these excuses, but this one has
exceeded even that.

	Paul
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301052201.42d86d9c@posting.google.com>
Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
> Edward G. Nilges wrote:
> > Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> > 
> >>* Edward G Nilges wrote:
> >>
> >>
> >>>Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> >>>based on the IBM 701's register structure and CAR and CDR are based on
> >>>operations in its machine language.  As of now I stand by this
> >>>claim.
> >>
> >>Well, there are 978 symbols in the CL package.  2 of them come from
> >>IBM 701 mnemonics, another two come from PDP-10 mnemonics.  So the
> >>`architecture of Lisp' is based on the IBM 701's register
> >>architecture?  Why not the DEC-10, which would certainly seem a more
> >>plausible candidate given CL's maclisp heritage? Or, given that 99.6%
> >>of the symbols in CL don't actually come from either machine, why not
> >>just give up and think?  No, I know, it's easier to just not bother.
> > 
> > 
> > Thanks for the information and indeed the confirmation, Tim, but I am
> > not sure what you might mean by "thinking."
> 
> The names "car" and "cdr" are not an issue at all for Lisp programmers. 
> On the contrary, some even appreciate them. If you are really interested 
> why, just ask.

Programmers, to be professionals, need to abandon the need to be
entertained by their tools.  In my humble opinion, of course.
> 
> > I do have a systematic tendency as a teacher and student formerly of
> > philosophy who entered the computer racket 30 years ago as a
> > draft-dodging scheme, to bring humanistic interpretation into the
> > field, and to me CAR and CDR are representative of a tendency in LISP
> 
> They are not.
> 
> > (please be advised that I have no special expertise in Lisp).
> 
> You are criticizing a very superficial aspect of Lisp. You should first 
> study Lisp and find out what the essential aspects of Lisp are before 
> you start to criticize it. You wouldn't criticize the Japanese language 
> either just because of the "funny" characters.
> 
Lisp is a computer language, not a natural language, so there probably
is no question of ethnocentrism here, although I might be wrong.

As to studying it, I used it occasionally several years ago and find
no need for the artifact.  I have to judge it from the outside, but as
yet, nobunny has given me persuasive information as to its value.

> > This is to tie the language, at a key point, to the vendor's hardware.
> >  It was also done in Fortran about the same time.
> 
> car and cdr are not tied to any hardware. The semantics of car and cdr 
> are clearly specified and can be implemented an any hardware. Lisp just 
> happened to be implemented on a specific hardware first.
>
Jules Schwartz had the grace to be embarassed when Jovial received his
name, for he had not intended this immortality.  The best languages
are more impersonal and reflect the needs and values of a civic
society.
 
> > Whereas the Algol group pioneered what has turned out to be a better
> > approach, and that is for a quasi-civic agency to be technically
> > independent of computer vendors.  The Algol team devised the father of
> > most usable modern languages including C because as an independent
> > civic forum, they were immune to commercial pressures.
> 
> It's not clear at all that the Algol approach is the better one. On the 
> contrary, there are many signs that Algol-style languages have reached 
> their limits. There is much research going on on things like 
> aspect-oriented programming, generative programming, support for 
> domain-specific languages etc. - things that are unnecessarily complex 
> in Algol languages. Scripting languages like Python and Ruby are 
> increasingly incorporating features originally found in Lisp. etc. pp.

Well, I certainly agree that the procedural approach in Algol is out
of date and that concepts from Lisp are important.

Ruby incorporates objects.  But what, specifically, does it borrow
from Lisp?

> 
> Many things are continually being reinvented that are more or less 
> already solved in Lisp.

That is true.

The problem may be in commercial languages that the language cannot
treat its own text as an artifact, as does Lisp, because of something
called a market failure by economists.

I shall explain.  Aren't you glad?

What I mean is that often in a language intended for nonacademic use
you hunt in vain for a clear way to treat a program text as a thing in
its own right, able to be processed by statements in the language.

The "market failure" is the nonzero probability that some clown, who
has shelled out 400.00 for a single "seat", will then write one or two
lines of code that create...another seat, because the language is
fully capable of executing strings in its own language, to any level
needed.

For example, early editions of Basic, and the Rexx language, featured
an "eval" or a "system" command which took a string in the chosen
language, parsed it and executed it.  You could wrap this call in a
line of code and effectively give the store away on behalf of the
vendor, who does not receive the expected cash stream.

I believe that several Basic vendors came to grief because of this.

Even in a language like Basic, it is easy, but not as elegant as Lisp,
to implement eval as long as the compiler is about the shop at
execution time.  I feature an eval in my own quickBasicEngine, a .Net
compiler for the old Quick Basic language with extensions and
omissions at this time.

Now, in languages like Visual Basic, self-evaluation is possible but
you have to go through hoops.  Furthermore, there are explicit
prohibitions against licensing code that does this as shrink-wrapped
code.

The problem is to me unsolvable, for Microsoft has demonstrated that a
market for programming language compilers does not exist unless these
market failures are allowed.  Ethically, they deserve their ROI.

This is why we have to reinvent the wheel, my friend.


> 
> > C does have a register declaration but note that it is as abstract and
> > not as specific as possible unlike CAR and CDR.
> 
> car and cdr are not tied to registers at all.

CAR and CDR are only poetically referents to old architecture but my
(humanistic) point was the same as Dijkstra's: in applied mathematics
we need to be 100% vigilant even about the metaphors we use. 
Dijkstra, unlike nearly all other computer scientists, had an almost
Wittgensteinian respect for what Wittgenstein called "the bewitchment
of our intelligence by way of language."

Of course, youse guys may be above such bewitchment, being so smart
and everything.

> 
> 
> Here are my favorite two papers about the essence of Lisp as far as I 
> understand it.
> 
> * http://www.paulgraham.com/rootsoflisp.html
> * ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-453.pdf
> 
> Further pointers can be found at 
> http://www.pascalcostanza.de/lisp/guide.html and, of course, 
> http://www.lisp.org
>
Thank you I will check them out. 
> 
> 
> Pascal

What do you think of the language after which you were named?
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA3E7025.25560%gehlker@fastq.com>
On 1/5/03 11:01 PM, in article
····························@posting.google.com, "Edward G. Nilges"
<···········@yahoo.com> wrote:

> Programmers, to be professionals, need to abandon the need to be
> entertained by their tools.

This is among the saddest opinions that I've ever heard expressed.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Charlton Wilbur
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87adiexiyi.fsf@mithril.chromatico.net>
>>>>> "CG" == Chris Gehlker <·······@fastq.com> writes:

    CG> On 1/5/03 11:01 PM, in article
    CG> ····························@posting.google.com, "Edward
    CG> G. Nilges" <···········@yahoo.com> wrote:

    >> Programmers, to be professionals, need to abandon the need to
    >> be entertained by their tools.

    CG> This is among the saddest opinions that I've ever heard
    CG> expressed.

It's entirely consistent with a number of other opinions I've heard
bruited about, depressingly common:

-- that college (university) is the best four years of your life,
   because after college, people expect you to behave in a certain
   manner and take on certain responsibilities, which of course you
   can't avoid;

-- that Real Workers punch in at 9am and leave at 5pm (if not 8am to
   7pm), and that being interested in working 40 hours a week maximum
   means you are Not a Team Player;

-- that Real Programmers are the ones who use tools like Visual Basic
   and Java, because that's what Business demands, and nobody ever got
   fired for going with Java or VB;

-- that it's not a worthy job if you actually enjoy doing it;

-- and, for that matter, that Business is the 'real world', and that
   anything else is an interesting hobby with delusions of grandeur.

Charlton
(apparently not a Real Worker, Real Programmer, Team Player, or a
professional of any stripe)
From: Greg Menke
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m3smw61l6p.fsf@europa.pienet>
···········@yahoo.com (Edward G. Nilges) writes:

> Lisp is a computer language, not a natural language, so there probably
> is no question of ethnocentrism here, although I might be wrong.
> 
> As to studying it, I used it occasionally several years ago and find
> no need for the artifact.  I have to judge it from the outside, but as
> yet, nobunny has given me persuasive information as to its value.

If you admit you don't have deep experience with Lisp, then why do you
feel you are in a position where you can intelligently judge it?  If
you're too lazy to sit down and learn it and then make an informed
decision, please say so.


> 
> The "market failure" is the nonzero probability that some clown, who
> has shelled out 400.00 for a single "seat", will then write one or two
> lines of code that create...another seat, because the language is
> fully capable of executing strings in its own language, to any level
> needed.

If you think this is so easy in Lisp, I suggest you sit down and
actually try it and show us.  I'm sure the various Lisp vendors would
appreciate you demonstrating such an important issue.


> 
> For example, early editions of Basic, and the Rexx language, featured
> an "eval" or a "system" command which took a string in the chosen
> language, parsed it and executed it.  You could wrap this call in a
> line of code and effectively give the store away on behalf of the
> vendor, who does not receive the expected cash stream.

Clipper, the dbase clone, had an eval feature (as did dbase).  Both of
them were extremely successful for quite a few years.  Their failure
to succeed in the long run was due to other causes- principally Ashton
Tate abandoned dbase and Clipper was sold to CA who promptly let it
decay.  VB was just starting to be usable and capitalized on the lack
of competition.


> The problem is to me unsolvable, for Microsoft has demonstrated that a
> market for programming language compilers does not exist unless these
> market failures are allowed.  Ethically, they deserve their ROI.
> 
> This is why we have to reinvent the wheel, my friend.

This is why <you> feel it must be reinvented.  I don't think you have
established the causal link that goes from a primitive, limited
language to market success.

> 
> 
> > 
> > > C does have a register declaration but note that it is as abstract and
> > > not as specific as possible unlike CAR and CDR.
> > 
> > car and cdr are not tied to registers at all.
> 
> CAR and CDR are only poetically referents to old architecture but my
> (humanistic) point was the same as Dijkstra's: in applied mathematics
> we need to be 100% vigilant even about the metaphors we use. 
> Dijkstra, unlike nearly all other computer scientists, had an almost
> Wittgensteinian respect for what Wittgenstein called "the bewitchment
> of our intelligence by way of language."
> 
> Of course, youse guys may be above such bewitchment, being so smart
> and everything.

I think you should quit with the philosphers for a little while and
pick up a reasonable Lisp implementation for a few months.



(I apologize to c.l.l for wasting bandwidth on such a silly thread,
but I couldn't resist any more..)

Gregm
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avbtel$10v2$1@f1node01.rhrz.uni-bonn.de>
Edward G. Nilges wrote:
> Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
> 

>>The names "car" and "cdr" are not an issue at all for Lisp programmers. 
>>On the contrary, some even appreciate them. If you are really interested 
>>why, just ask.
> 
> 
> Programmers, to be professionals, need to abandon the need to be
> entertained by their tools.  In my humble opinion, of course.

...and I think that people work best when they are satisfied with the 
tools they use.

>>You are criticizing a very superficial aspect of Lisp. You should first 
>>study Lisp and find out what the essential aspects of Lisp are before 
>>you start to criticize it. You wouldn't criticize the Japanese language 
>>either just because of the "funny" characters.
> 
> Lisp is a computer language, not a natural language, so there probably
> is no question of ethnocentrism here, although I might be wrong.

Computer science was regarded as a kind of "applied mathematics" in the 
beginning, so there is indeed a tendency to take mathematics too 
seriously in CS. I would regard this as a variant of "ethnocentrism".

It's not clear at all that mathematics buys you a lot. In fact, there is 
evidence that this is counterproductve.

> Well, I certainly agree that the procedural approach in Algol is out
> of date and that concepts from Lisp are important.
> 
> Ruby incorporates objects.  But what, specifically, does it borrow
> from Lisp?

Ruby is strongly influenced by Smalltalk which in turn was strongly 
influenced by Lisp.

> The problem may be in commercial languages that the language cannot
> treat its own text as an artifact, as does Lisp, because of something
> called a market failure by economists.
> 
> I shall explain.
[...]

These are signs that certain forms of traditional capitalism are not 
appropriate in the context of software. There is a strong need for new 
economic models that reflect the boundary conditions imposed by 
software. Deliberate restrictions of software will not work in the long run.

>>Pascal
> 
> What do you think of the language after which you were named?

As an educational tool it's ok. Modula-2 and especially Oberon were 
important improvements. For system-level stuff I'd prefer to use Oberon 
rather than any language from the C family, but unfortunately that's not 
a viable option nowadays because unfortunately Oberon is a dead language.

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA3EE686.2557A%gehlker@fastq.com>
On 1/6/03 5:39 AM, in article ·············@f1node01.rhrz.uni-bonn.de,
"Pascal Costanza" <········@web.de> wrote:

> Ruby is strongly influenced by Smalltalk which in turn was strongly
> influenced by Lisp.

Ruby was also strongly influenced by CLU which could almost be considered a
dialect of Lisp. CLU became pretty popular in Japan while it seems to have
fallen off the radar in the rest of the world. This probably has to do with
the fact that NEC was one of the backers of the development of CLU and
served a role in promoting CLU analogous to Sun's role with Java.

In particular, the way Ruby syntax for blocks, the treatment of blocks as
closures and the use of proc and yield often seem strange to people with a
Common Lisp background because they are derived from CLU, not Common Lisp.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Thaddeus L Olczyk
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <7mmj1vkukvuha2vdo8c43kpssi3enb52ge@4ax.com>
On Mon, 06 Jan 2003 13:39:45 +0100, Pascal Costanza <········@web.de>
wrote:


>> What do you think of the language after which you were named?
>
>As an educational tool it's ok. Modula-2 and especially Oberon were 
>important improvements. For system-level stuff I'd prefer to use Oberon 
>rather than any language from the C family, but unfortunately that's not 
>a viable option nowadays because unfortunately Oberon is a dead language.
http://www.oberon.ethz.ch/
Looks more alive then Dylan to me.
From: Pascal Bourguignon
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87bs2uc6f1.fsf@thalassa.informatimago.com>
Pascal Costanza <········@web.de> writes:
> Computer science was regarded as a kind of "applied mathematics" in
> the beginning, so there is indeed a tendency to take mathematics too
> seriously in CS. I would regard this as a variant of "ethnocentrism".
> 
> It's not clear at all that mathematics buys you a lot. In fact, there
> is evidence that this is counterproductve.

Counterproductive, it depends for what purpose.  

If your purpose is to earn  googles of dollars, then yes, you'd rather
ship whatever code  that can be launched, the users will  do QA and be
happy paying for it.

If you purpose  is to get a correct and working  software in which you
may confide and put our life in  its bits, then maths is the way to go
(with a lot  of checking and some comprehensive  testing, but remember
that  a test  can only  show that  a  bug exist,  not that  no bug  is
present).


> These are signs that certain forms of traditional capitalism are not
> appropriate in the context of software. There is a strong need for new
> economic models that reflect the boundary conditions imposed by
> software. Deliberate restrictions of software will not work in the
> long run.


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avd5dm$343$1@newsreader2.netcologne.de>
Pascal Bourguignon wrote:

> If you purpose  is to get a correct and working  software in which you
> may confide and put our life in  its bits, then maths is the way to go
> (with a lot  of checking and some comprehensive  testing, but remember
> that  a test  can only  show that  a  bug exist,  not that  no bug  is
> present).

In fact, this isn't true. A test ultimately proves that the program is 
correct for the given test suite, and sometimes this is as good as it 
gets. A formal proof might give you new insights into the problem and/or 
solution domain, but doesn't ultimately prove that your program is 
correct because that very proof may contain bugs itself.

"Sorry, we have forgotten to cover that special case in our proof." is 
not better (nor worse) than "Sorry, we have forgotten to cover that 
special case in our test suite." - There are many examples in the 
history of mathematics where such bugs in proofs haven't been found for 
very long periods.

In fact, there is evidence that certain bugs are inherent in certain 
sufficiently complex problem domains and are seemingly not detectable 
until they occur. (I have got this tidbit from a talk about software 
development for power plants. Sorry, I am probably not able to give any 
references on this.)


Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Erann Gat
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0601032200400001@192.168.1.51>
In article <············@newsreader2.netcologne.de>, Pascal Costanza
<········@web.de> wrote:

> Pascal Bourguignon wrote:
> 
> > If you purpose  is to get a correct and working  software in which you
> > may confide and put our life in  its bits, then maths is the way to go
> > (with a lot  of checking and some comprehensive  testing, but remember
> > that  a test  can only  show that  a  bug exist,  not that  no bug  is
> > present).
> 
> In fact, this isn't true. A test ultimately proves that the program is 
> correct for the given test suite, and sometimes this is as good as it 
> gets.

In multithreaded or embedded systems the behavior of a program can be
different on multiple runs even when the initial conditions are as similar
as you can possibly make them.  See e.g. "On the Role of Stored Internal
State in the Control of Autonomous Mobile Robots", AI Magazine, Spring
1993, or the story of the Remote Agent bug (Google is your friend).  The
Remote Agent Bug story is also an example of how formal methods can find
bugs that even extensive testing and competent programmers miss.  So I'm
with Pascal B. on this one.  You can never be 100% sure a system is
bug-free, but the more different kinds if ways you express and cross-check
your intentions the better your chances of getting what you want.

E.
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avejfc$sd6$1@f1node01.rhrz.uni-bonn.de>
Erann Gat wrote:
> In article <············@newsreader2.netcologne.de>, Pascal Costanza
> <········@web.de> wrote:

>>A test ultimately proves that the program is 
>>correct for the given test suite, and sometimes this is as good as it 
>>gets.
> 
> In multithreaded or embedded systems the behavior of a program can be
> different on multiple runs even when the initial conditions are as similar
> as you can possibly make them.

OK, I stand corrected in this regard.

> You can never be 100% sure a system is
> bug-free, but the more different kinds if ways you express and cross-check
> your intentions the better your chances of getting what you want.

I agree. The important point is to gain different perspectives on the 
same domain.


Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Pascal Bourguignon
Subject: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87el7obnl0.fsf_-_@thalassa.informatimago.com>
IMO, the  interest of formal  proof of programs  is not so much  as an
alternative   to   test   cases   than   as  a   step   to   automatic
generation/verification of program.

That is,  an informal specification, an  informal program construction
and  some   informally  developped  test  cases,  as   done  by  human
programmers  can  show that  in  some  cases  the program  written  is
correct, and  the programmer can  even have confidence (by  his mental
process  while  constructing the  program)  that  he  wrote a  correct
program in all the cases specified.

But until we have a strong AI, this cannot be reproduced automatically
(and anyway,  check the current  software quality to see  what results
are attained this way).


If  instead we start  from formal  specifications and  formal software
development and verification, this can be implemented in software with
a  set of  tools that  don't amount  to strong  AI:  theorem proovers,
theorem checkers, program generators,  some goal directed search, etc,
that could be implemented today.


I know that  it's proven that no program can  prove the termination in
general.  But  if you check carefully,  you'll see that  to prove this
theorem, we apply  termination prover on itself.  So  we just found an
example, a  case where it's  not possible, but  it does not  mean that
this  program  cannot be  helpful  to  analyse  and verify  the  other
programs.  By  the way,  have you ever  found a programmer  capable of
proving that for  any specification he can himself  generate a correct
program in finite time?  (Since he has no formal specifications of his
own  brains, he would  have quite  some difficulties  to start  such a
proof).


Now, once we  have software programmers, I agree  that bug could still
be  generated.  Like in  our informal  world, where  to err  is human,
perhaps to err  will be software too (while I  hope they'll err less).
The  difference is  that  once a  bug  is detected  you  need to  stop
everything and  start a laborious  debugging/develop/test cycle before
installing   a   new   version.    A  software   formally   developped
automatically   could   be   monitored    by   AI   agents   and   the
debugging/correction  of the  formal  specification or  of the  formal
proof/regeneration   of  software/formal  verification   (and  perhaps
automatic generation and  run of test cases to  make everybody happy),
could be implemented and run automatically, reducing the down time.

-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Erik Naggum
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3250980130956676@erik.naggum.no>
* Pascal Bourguignon
| But until we have a strong AI, this cannot be reproduced automatically
| (and anyway, check the current software quality to see what results
| are attained this way).

  While the root cause of crime is obviously the laws, it really sounds
  like you believe that the system is to blame for individual failures.

| If instead we start from formal specifications and formal software
| development and verification, this can be implemented in software
| with a set of tools that don't amount to strong AI: theorem
| proovers, theorem checkers, program generators, some goal directed
| search, etc, that could be implemented today.

  Before you do this, you should to try give some program a program
  and wait for it to tell you what that program actually does.

| (Since he has no formal specifications of his own brains, he would
| have quite some difficulties to start such a proof).

  Do you really believe such hogwash or are you just pretending to be
  stupid for its effect?

| Like in our informal world, where to err is human, perhaps to err
| will be software too (while I hope they'll err less).

  You really have no idea what "to err is human" means, do you?

  I occasionally entertain a Microsoft "fan" with a discussion on the
  real problem behind viruses and malicious abuse of their software.
  I maintain that the problem is not that the software is not perfect
  or approved or proven correct or any such thing, but that it is not
  /defensive/ against harmful data.  There are basically two ways to
  build a society of autonomous units: You either design away all the
  anti-social flaws from the outset and let them go their happy way,
  or you design every single unit with an /immune system/ that copes
  with bad guys if and when they exist.  The problem is therefore not
  that some software gets viruses or rotten data, but that it keeps
  going as if it were completely unaffected by it.

  Drop the moronic "proof of correctness" and "verification" nonsense
  and design systems that can actually defend themself from bad input.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Thomas Stegen
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e1f56d1$1@nntphost.cis.strath.ac.uk>
Erik Naggum wrote:
>   You really have no idea what "to err is human" means, do you?

"To err is human" is suspect to interpretation and does not only have
one meaning. It should  be clear from context what is meant though.

>   Drop the moronic "proof of correctness" and "verification" nonsense
>   and design systems that can actually defend themself from bad input. 

Intuition is not always enough to "prove" or know that a system
can handle all types of bad input. Of course, a system which is well
designed in this regard might be easier to modify to deal with more
cases of bad input than a badly designed program.

That you call "proof of correctness" and verification" moronic nonsense
is perhaps evidence towards the fact that we do not know enough about
the software development process to make a judgement about the worth
of formal proofs.

-- 
Thomas.
From: Erik Naggum
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3251273208239716@naggum.no>
* Thomas Stegen
| That you call "proof of correctness" and verification" moronic
| nonsense is perhaps evidence towards the fact that we do not know
| enough about the software development process to make a judgement
| about the worth of formal proofs.

  But it is not the worth that I am worried about, it is the cost,
  which drives the /net/ worth deep into the red.  This is not
  because of the software /development/ process, but because of the
  /living/ conditions of the developed software -- what it means for
  software to coexist.  

  To create something new in a language that regards all unproven
  statements as if they were hostile, you would have to make each step
  of the expression so minute that the human mind would be incapable
  of accomplishing anything within budget.  We would no longer develop
  software in programming languages, we would spend all our time
  developing programming languages in which it would, someday, be
  possible to prove something useful.

  If we know enough about the problem to prove its specification and
  solution correct, there is no longer any reason to work on it, and
  the solution to the problem should simply be published and taught.
  Actum ne agas.  Move on to the next problem.

  It is not of the software we know too little, it is the real world.
  Human coping strategies are incommensurate with software coping
  strategies.  The belief that mathematical proofs will help software
  live long and prosper is bogus at its root.  Software will be just
  as much blindsided by the unpredictable real world as humans are, it
  will just happen in very different ways.  When humans are overloaded
  with tasks they are unable to complete, they break down, and so do
  computers, but we treat the same problem very differently.  When
  software or even hardware is given bad input, that is just as bad as
  poisoning or infecting humans, but the way humans respond to it is
  by means that have been optimized for survival of the base system
  while the bad input is being dealt with.  Software is allowed to die
  immediately upon encountering certain toxins, and evil software has
  no idea it is infected and go on to contaminate its environment.

  I think proofs of correctness and verification were really good
  ideas in the early days of computing.  Research on the optimal baby
  food for our fledgling computers was the right way to go then.  But
  today, computers have grown up and are extremely powerful, and they
  spend almost all their time waiting for something and have resources
  to spare.  Multics, if it had survived, would be far less complex
  than Unix of today despite the fact that the latter was born out of
  sheer frustration with the "wastefulness" of the former in the eyes
  of the myopic designers.  Today, we can let our computers do much
  more work at runtime than we could mere years ago.  Common Lisp has
  in many ways been eclipsed by the much slower interpreted languages
  that have become popular while it stressed compilation and hardware
  became fast enough to interpret other languages.  Building software
  immune systems would probably be more cost-effective than proving
  code correct.

  The complexity of software and systems that survive bad input is
  many orders of magnitude worse than that of software and systems
  that crash and burn on any bad input.  It necessarily means slower
  execution of (more system overhead for) the overall system, but not
  necessarily slower execution of critical sections.  We already see
  that the time spent satisfying computers that their input is correct
  is a major drag on the software development process.  Why have we
  created computers that are so picky when everything else we have
  created have been adaptive, dynamic systems in comparison?  Is it
  because we somehow think that humans should /not/ intervene when the
  computer makes a mistake or is given bad data, and therefore have
  not developed software that makes this process work well?

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Gary Klimowicz
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <uy95r2ngz.fsf@C1799538-A.attbi.com>
Erik Naggum <····@naggum.no> writes:

> * Thomas Stegen
> | That you call "proof of correctness" and verification" moronic
> | nonsense is perhaps evidence towards the fact that we do not know
> | enough about the software development process to make a judgement
> | about the worth of formal proofs.
> 
>   But it is not the worth that I am worried about, it is the cost,
>   which drives the /net/ worth deep into the red.  This is not
>   because of the software /development/ process, but because of the
>   /living/ conditions of the developed software -- what it means for
>   software to coexist.  

Drs. DeMillo, Lipton and Perlis wrote (in 1979) a great article about
this and other difficulties of formal proofs of correctness applied to
the programming world. As Erik points out, there are serious issues
with the living aspects of the program, the impossibility of
generating formal specifications for most "real world" applications.

The authors also make a good case that the believeability of
mathematical proofs actually has little to do with formal
methods. It's the severity of the review that proofs receive that give
them confidence in their veracity. Mathematicians appear to be more
skeptical about the application of formal methods to proofs than
computer scientists.

For automated proofs of programs, there is no such human review
involved, just the message "Verified!" at the end (or worse,
"Not verified. Here's the twenty-page dump of my reasoning....").

My experience has been that formal proofs of termination and
correctness helped me make sure that a small, particularly thorny
patch of code was right. But I still write the unit tests.  No matter
how good my proof was, the other developers believed the unit tests
more.

Here's a pointer to a scanned PDF of the article. I haven't seen
much in the advocacy of formal proof methods here in c.l.l. that
persuades me their arguments no longer apply.

http://www.demillo.com/Social_Processes.pdf


-- 
gak
From: Andreas Eder
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m34r8f3j68.fsf@elgin.eder.de>
Erik Naggum <····@erik.naggum.no> writes:

>   Drop the moronic "proof of correctness" and "verification" nonsense
>   and design systems that can actually defend themself from bad input.
> 
Why can't we do both? 
Try to make that parts of a program that have to cope with input - be
it from humans or other programs we can't control - as resilient to bogus
input as possible and try to prove correct the other parts deeper down
in the innards of our programs, that just have to handle known and
controlled input.
Why not have the best of both worlds?

Andreas
-- 
Wherever I lay my .emacs, there�s my $HOME.
From: Christopher Browne
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avpg3o$h9aok$2@ID-125932.news.dfncis.de>
Oops! Andreas Eder <············@t-online.de> was seen spray-painting on a wall:
> Erik Naggum <····@erik.naggum.no> writes:
>
>>   Drop the moronic "proof of correctness" and "verification" nonsense
>>   and design systems that can actually defend themself from bad input.
>> 
> Why can't we do both? 
> Try to make that parts of a program that have to cope with input - be
> it from humans or other programs we can't control - as resilient to bogus
> input as possible and try to prove correct the other parts deeper down
> in the innards of our programs, that just have to handle known and
> controlled input.
> Why not have the best of both worlds?

The problem is that by imagining that you can have a "proof of
correctness," you have leapt into error.

The only way that a "formal verification" is of any value is if you
know that the system specification is perfectly rigid and complete,
and it is rare for system specifications to be anywhere /near/ that
well-defined.

I have a project I'm working on where it is /known/ that the
specifications are incomplete.  We're working on software for an
accounting system, and we /know/ that the accountant is pretty
visually oriented, and that he won't be able to assess "right" or
"wrong" until the software actually gets implemented.

"Formal verification" might be useful in making sure that we have an
appropriate library for adding up values and making sure they are
rounded correctly, but aside from that, it's pretty futile because the
system specification isn't "written in stone;" it's "written in
Kleenex tissue."

And while this is a bit of a pathological case, it's not /way/ far
from the norm.  A lot of systems can only have vague specifications
because the people with purse strings aren't mathematicians and can't
even think about the notion of having a "provably correct" system
specification.
-- 
If this was helpful, <http://svcs.affero.net/rm.php?r=cbbrowne> rate me
http://cbbrowne.com/info/nonrdbms.html
Rules of  the Evil  Overlord #143.  "If  one of my  daughters actually
manages to win the hero and  openly defies me, I will congratulate her
on her  choice, declare a  national holiday to celebrate  the wedding,
and proclaim the  hero my heir. This will probably  be enough to break
up the relationship.  If not, at  least I am assured that no hero will
attack my  Legions of  Terror when  they are holding  a parade  in his
honor." <http://www.eviloverlord.com/>
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-1101031509170001@192.168.1.51>
In article <··············@ID-125932.news.dfncis.de>, Christopher Browne
<········@acm.org> wrote:

> The only way that a "formal verification" is of any value is if you
> know that the system specification is perfectly rigid and complete,

No, and I point once again to the Remote Agent experience as a counter
example.  The Remote Agent was about as poorly specified as any software
system has ever been.  Formal verification was nonetheless of great value.

E.
From: Andreas Eder
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m3iswv1szq.fsf@elgin.eder.de>
Christopher Browne <········@acm.org> writes:

> The problem is that by imagining that you can have a "proof of
> correctness," you have leapt into error.
> 
> The only way that a "formal verification" is of any value is if you
> know that the system specification is perfectly rigid and complete,
> and it is rare for system specifications to be anywhere /near/ that
> well-defined.

I would not consider that such a rare case. There are a lot of things
that have - or could have - a rather rigid soecification, e.g. network
protocols.

> "Formal verification" might be useful in making sure that we have an
> appropriate library for adding up values and making sure they are
> rounded correctly, but aside from that, it's pretty futile because the
> system specification isn't "written in stone;" it's "written in
> Kleenex tissue."

Another example is math libraries.

> 
> And while this is a bit of a pathological case, it's not /way/ far
> from the norm.  A lot of systems can only have vague specifications
> because the people with purse strings aren't mathematicians and can't
> even think about the notion of having a "provably correct" system
> specification.

Yes, but that is a situation that we should change. I'm all for giving
mathematicians a purse :-) maybe because I am a mathematician?

Andreas
-- 
Wherever I lay my .emacs, there�s my $HOME.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <smw3fui3.fsf@ccs.neu.edu>
Pascal Bourguignon <···@informatimago.com> writes:

> I know that  it's proven that no program can  prove the termination in
> general.  But  if you check carefully,  you'll see that  to prove this
> theorem, we apply  termination prover on itself.  So  we just found an
> example, a  case where it's  not possible, but  it does not  mean that
> this  program  cannot be  helpful  to  analyse  and verify  the  other
> programs. 

Many programmers are under the impression that although
non-termination is impossible to determine in general, that it is
probably determinable in practice, or at least in a large enough set
of cases as to be useful.  This isn't true.

It is quite easy to find very simple programs that cannot be proven to
terminate, and they come up in practice far more often than you'd
think.  Here is an artificial one I came up with:

(defun kernel (s i)
  (list (not (car s))
	(if (car s)
	    (cadr s)
	  (cons i (cadr s)))
	(cons 'y (cons i (cons 'z (caddr s))))))

(defconstant k0 '(t () (x)))

(defun mystery (list)
  (do ((result (list t list)
               (reduce #'kernel (if (car result)
                                    (cadr result)
                                    (caddr result))
                       :initial-value k0)))
      ((null (cadr result)))))

The kernel function trivially terminates, the call to reduce
terminates, the list over which mystery iterates is finite for all
iterations, yet it is not known whether mystery terminates for all
finite lists.  (Google on `Collatz Problem' for details.)

The reason I coded this is because the only obstacle to termination
analysis appears to be the iteration in `mystery'.  Yet iterations
such as this are not an uncommon feature in typical programs.

If you look at  

   http://www.math.niu.edu/~rusin/known-math/95/independence

you well see some examples where Godel's incompleteness theorem
(related to the halting problem) has cropped up in several `natural
independence phenomena' that don't require the trick of attempting to
apply a proof to itself.  One of the problems is `Kruskal's tree
theorem' that is involved in showing that certain orderings on trees
are well-founded.  

Again, tree traversal is a common programming technique.
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0801031114010001@k-137-79-50-101.jpl.nasa.gov>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> Pascal Bourguignon <···@informatimago.com> writes:
> 
> > I know that  it's proven that no program can  prove the termination in
> > general.  But  if you check carefully,  you'll see that  to prove this
> > theorem, we apply  termination prover on itself.  So  we just found an
> > example, a  case where it's  not possible, but  it does not  mean that
> > this  program  cannot be  helpful  to  analyse  and verify  the  other
> > programs. 
> 
> Many programmers are under the impression that although
> non-termination is impossible to determine in general, that it is
> probably determinable in practice, or at least in a large enough set
> of cases as to be useful.  This isn't true.
>
> It is quite easy to find very simple programs that cannot be proven to
> terminate, and they come up in practice far more often than you'd
> think.  Here is an artificial one I came up with:

This is a fallacious argument.  Your claim (slightly refactored) is that
it is not possible to determine whether or not a program halts in a
significant of useful cases.  But you support for this claim is that no
general proof is known for a particular (and rather artificial IMO) case. 
That seems like a non-sequitur to me at best.  That's like trying to
disprove that *most* (not all) crows are black by exhibiting a white crow.

Nonetheless, the references you point to make fascinating reading.

E.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <n0mbfklw.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
> 
> > Pascal Bourguignon <···@informatimago.com> writes:
> > 
> > > I know that  it's proven that no program can  prove the termination in
> > > general.  But  if you check carefully,  you'll see that  to prove this
> > > theorem, we apply  termination prover on itself.  So  we just found an
> > > example, a  case where it's  not possible, but  it does not  mean that
> > > this  program  cannot be  helpful  to  analyse  and verify  the  other
> > > programs. 
> > 
> > Many programmers are under the impression that although
> > non-termination is impossible to determine in general, that it is
> > probably determinable in practice, or at least in a large enough set
> > of cases as to be useful.  This isn't true.
> >
> > It is quite easy to find very simple programs that cannot be proven to
> > terminate, and they come up in practice far more often than you'd
> > think.  Here is an artificial one I came up with:
> 
> This is a fallacious argument.  Your claim (slightly refactored) is that
> it is not possible to determine whether or not a program halts in a
> significant of useful cases.  But you support for this claim is that no
> general proof is known for a particular (and rather artificial IMO) case. 

The reason I don't think it is completely fallacious is that I don't
think the example is too artificial.  The proof that a termination
prover fails upon itself seems extreme in that how often does one
reason about termination provers?  The example I gave was one that
involved very simple functions like CONS, CAR, CDR, EQ, DO, and
REDUCE, and used them in a way that (at least to me) doesn't seem
unusual.

> That seems like a non-sequitur to me at best.  

It may be a bit of leap, so let me fill in the gaps:

  Assume that it *is* possible to determine whether or not a program
  halts in virtually every useful case.

  Therefore, an undecidable program must be out of the ordinary.

  But the program exhibited *is* ordinary.

So my argument hinges on whether or not you think my example case is
ordinary.  (and as you indicated above, you don't)
Now, for the sake of pushing the point, exactly what is it about my
example case, aside from being undecidable, that pushes beyond the
realm of `ordinary' into `extraordinary'?

> That's like trying to disprove that *most* (not all) crows are black
> by exhibiting a white crow.

Well, the original argument was like:

    `The fact that all crows are black is disproved by this mutant
     crow, but that's a special case.  A lot of `normal' crows are 
     black.'

to which I responded:

    `This crow isn't black, and it doesn't seem to be mutant.  Looks a
     lot like any other crow other than the color.  I see no reason
     why crows like this should be rare.'

But there is a small problem with the analogy:  the color of a crow is
readily apparent.  The question of whether or not an interesting
set of programs terminate is not (and in fact, the question under
debate).
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0801031824340001@192.168.1.51>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> ···@jpl.nasa.gov (Erann Gat) writes:
> 
> > In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
> > 
> > > Pascal Bourguignon <···@informatimago.com> writes:
> > > 
> > > > I know that  it's proven that no program can  prove the termination in
> > > > general.  But  if you check carefully,  you'll see that  to prove this
> > > > theorem, we apply  termination prover on itself.  So  we just found an
> > > > example, a  case where it's  not possible, but  it does not  mean that
> > > > this  program  cannot be  helpful  to  analyse  and verify  the  other
> > > > programs. 
> > > 
> > > Many programmers are under the impression that although
> > > non-termination is impossible to determine in general, that it is
> > > probably determinable in practice, or at least in a large enough set
> > > of cases as to be useful.  This isn't true.
> > >
> > > It is quite easy to find very simple programs that cannot be proven to
> > > terminate, and they come up in practice far more often than you'd
> > > think.  Here is an artificial one I came up with:
> > 
> > This is a fallacious argument.  Your claim (slightly refactored) is that
> > it is not possible to determine whether or not a program halts in a
> > significant of useful cases.  But you support for this claim is that no
> > general proof is known for a particular (and rather artificial IMO) case. 
> 
> The reason I don't think it is completely fallacious is that I don't
> think the example is too artificial.

Even if the example is not artificial, one counterexample does not refute
a non-universal claim.

It may be that, say, half of all programs are amenable to some kind of
formal proof of correctness, and half are not.  I'd say that half of all
cases is a "large enough set to be useful."  But there would still be no
sport in finding counterexamples.

>  The example I gave was one that
> involved very simple functions like CONS, CAR, CDR, EQ, DO, and
> REDUCE, and used them in a way that (at least to me) doesn't seem
> unusual.

But it doesn't do anything useful.

> It may be a bit of leap, so let me fill in the gaps:
> 
>   Assume that it *is* possible to determine whether or not a program
>   halts in virtually every useful case.

That wasn't the claim.  The claim was "a large enough subset to be
useful", not "virtually every useful case."

E.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <7kdef5cm.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> It may be that, say, half of all programs are amenable to some kind of
> formal proof of correctness, and half are not.  I'd say that half of all
> cases is a "large enough set to be useful."  But there would still be no
> sport in finding counterexamples.

"Coby Beck" <·····@mercury.bc.ca> writes:
> 
> Sorry, but these arguments just don't fly.
> 

Ok.  I guess I'm unconvincing.  Allow me to appeal to algorithmic
information theory, then.  One of the results of algorithmic
information theory is that there is an upper bound on the complexity
of a program that a meta-program can reason about.  In other words, if
you have a program that you use to test terminating conditions, it
will be unable to prove termination of programs that are above the
algorithmic complexity of the termination prover.  (Or to seriously
oversimplify, a two page termination tester won't work on programs
longer than about two pages.)

See   http://www.cs.auckland.ac.nz/CDMTCS/chaitin/ 
for way too much information.
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0901031235090001@k-137-79-50-101.jpl.nasa.gov>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> ···@jpl.nasa.gov (Erann Gat) writes:
> 
> > It may be that, say, half of all programs are amenable to some kind of
> > formal proof of correctness, and half are not.  I'd say that half of all
> > cases is a "large enough set to be useful."  But there would still be no
> > sport in finding counterexamples.
> 
> "Coby Beck" <·····@mercury.bc.ca> writes:
> > 
> > Sorry, but these arguments just don't fly.
> > 
> 
> Ok.  I guess I'm unconvincing.  Allow me to appeal to algorithmic
> information theory, then.  One of the results of algorithmic
> information theory is that there is an upper bound on the complexity
> of a program that a meta-program can reason about.

There is an upper bound on the complexity of programs that meta-programs
can reason about *in general*.  But that leaves open a wide variety of
useful possibilities:

* a subset of programs that obey certain constraints that allow them to be
reasoned about

* a reasoner that is not 100% accurate, but which produces only false
negatives (i.e. it says "I can't prove this program is correct" when it is
in fact correct).

* a method of transforming programs that cannot be reasoned about into
ones that can.  (Wrapping a timeout around a program is one example of
such a transformation).

> In other words, if
> you have a program that you use to test terminating conditions, it
> will be unable to prove termination of programs that are above the
> algorithmic complexity of the termination prover.

No, it will be unable to prove the termination of *some* programs that are
above the algorithmic complexity of the prover.  The key question is: are
the programs that are left useful?  That is an open question, but
anecdotal evidence indicates that the answer is yes.

> See   http://www.cs.auckland.ac.nz/CDMTCS/chaitin/ 
> for way too much information.

Or http://www.flownet.com/gat/chaitin.html for way too little :-)

E.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <1y3mf0j7.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> > In other words, if you have a program that you use to test
> > terminating conditions, it will be unable to prove termination of
> > programs that are above the algorithmic complexity of the
> > termination prover.
> 
> No, it will be unable to prove the termination of *some* programs that are
> above the algorithmic complexity of the prover.

I think we must be a cross purposes here:  if the algorithmic
complexity of the prover is small, then it simply doesn't have enough
information capacity to make statements about the space of programs
larger than itself.  If it could, then you could encode the larger
programs in terms of the smaller, and contradict the supposition that
the program being examined is more complex than the prover. 

> The key question is: are the programs that are left useful?  

Well, another key question is `Are there very many programs left?'

> That is an open question, but anecdotal evidence indicates that the
> answer is yes.

What anecdotes?  The anecdotes I've heard say the opposite.

Consider this:  you certainly can't prove the termination of any
program capable of universal computation.  But it is surprisingly easy
to perform universal computation, and it often happens that systems
that were not designed for universal computation can be coaxed into
doing it.  
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0901031611350001@k-137-79-50-101.jpl.nasa.gov>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> ···@jpl.nasa.gov (Erann Gat) writes:
> 
> > > In other words, if you have a program that you use to test
> > > terminating conditions, it will be unable to prove termination of
> > > programs that are above the algorithmic complexity of the
> > > termination prover.
> > 
> > No, it will be unable to prove the termination of *some* programs that are
> > above the algorithmic complexity of the prover.
> 
> I think we must be a cross purposes here: if the algorithmic
> complexity of the prover is small, then it simply doesn't have enough
> information capacity to make statements about the space of programs
> larger than itself.

No, that's not true.  A prover cannot be guaranteed to produce a correct
answer on programs algorithmically more complex than itself.  But that is
different from what you seem to be saying, that a prover can be guaranteed
not to produce a correct answer in those cases.

Consider:

(defmacro with-timeout (time &body body)
   ... )

(defun analyze (code)
  (cond ( (and (eq (first code 'with-timeout))
               (numberp (second code)))
          (format t "This code is guaranteed to terminate in under ~A seconds"
                    (second code)) )
        (t (format t
 "I can't tell if this code will terminate or not.  Sorry."))))

This will produce the correct answer in some cases, and throw up its hands
in others.  Some (many, most) of the cases where it produces the correct
answer will be more algorithmically complex than the above code.  Whether
or not this is useful is an open question, but it is not prima facie
useless.

> If it could, then you could encode the larger
> programs in terms of the smaller, and contradict the supposition that
> the program being examined is more complex than the prover. 

No.  This is only true if the analyzer works perfectly.  If it returns the
correct result only some of the time (and returns "unknown" the rest of
the time) then the conditions of Chaitin's theorem are not violated.

> > The key question is: are the programs that are left useful?  
> 
> Well, another key question is `Are there very many programs left?'

There are in many (most) cases an infinite number of program left.  The
example above is one such case.

> > That is an open question, but anecdotal evidence indicates that the
> > answer is yes.
> 
> What anecdotes?  The anecdotes I've heard say the opposite.

Look up the Remote Agent Bug.

> Consider this:  you certainly can't prove the termination of any
> program capable of universal computation.  But it is surprisingly easy
> to perform universal computation, and it often happens that systems
> that were not designed for universal computation can be coaxed into
> doing it.

This is the white crows agument again.  I do not deny the existence of
white crows -- programs for which no meaningful proofs of correctness are
possible.  I do not even deny the existence of a great many white crows
(despite the fact that Chaitin's result only guarantees the existence of
one).  So no matter how many of them you exhibit you will not dissuade me
from my belief that, notwithstanding the white crows, there are
nonetheless a great many black crows in the world, having beheld a fair
number of them with my own eyes.

E.
From: Nikodemus Siivola
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avm9v8$2skos$1@midnight.cs.hut.fi>
Erann Gat <···@jpl.nasa.gov> wrote:

> This is the white crows argument again.  I do not deny the existence of
> white crows -- programs for which no meaningful proofs of correctness are
> possible.  I do not even deny the existence of a great many white crows

I'm confused: do you then choose to regard any program for which you
cannot obtain proof with suspicion, or what?

  -- Nikodemus
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-1001030858290001@192.168.1.51>
In article <··············@midnight.cs.hut.fi>, Nikodemus Siivola
<········@kekkonen.cs.hut.fi> wrote:

> Erann Gat <···@jpl.nasa.gov> wrote:
> 
> > This is the white crows argument again.  I do not deny the existence of
> > white crows -- programs for which no meaningful proofs of correctness are
> > possible.  I do not even deny the existence of a great many white crows
> 
> I'm confused: do you then choose to regard any program for which you
> cannot obtain proof with suspicion, or what?

Yes.  Or as an indication that the code ought to be rewritten.  Or thought
about very, very, very carefully.  Or placed in a computational
environment with backup safeguards so that if it doesn't work as expected
the results are not disastrous.

In other words, the fact that formal methods have limitations should not
be used as an excuse not to employ formal methods where they are useful. 
Conversely, the fact that formal methods are occasionally (often) useful
should not be used as an excuse to rely on them absolutely.  It's a simple
application of Gat's First Law: all extreme positions are wrong.

Some examples of useful results that could potentially be formally proven:

1) A Java program running on a particular implementation never generates
an illegal memory reference.

2) A user process running on a particular implementation of Unix cannot
acccess resource X.

3) Program P will either generate a correct result or indicate that it
cannot do so.

4) The probability that program P produces a correct result is >= X.

E.
From: Coby Beck
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avijs4$2b9i$1@otis.netspace.net.au>
"Joe Marshall" <···@ccs.neu.edu> wrote in message
·················@ccs.neu.edu...
> ···@jpl.nasa.gov (Erann Gat) writes:
>
> > That's like trying to disprove that *most* (not all) crows are black
> > by exhibiting a white crow.
>
> Well, the original argument was like:
>
>     `The fact that all crows are black is disproved by this mutant
>      crow, but that's a special case.  A lot of `normal' crows are
>      black.'
>
> to which I responded:
>
>     `This crow isn't black, and it doesn't seem to be mutant.  Looks a
>      lot like any other crow other than the color.  I see no reason
>      why crows like this should be rare.'
>

Sorry, but these arguments just don't fly.

[couldn't resist! :) ]
--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Pascal Bourguignon
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87n0ma76ou.fsf@thalassa.informatimago.com>
Joe Marshall <···@ccs.neu.edu> writes:

> Pascal Bourguignon <···@informatimago.com> writes:
> 
> > I know that  it's proven that no program can  prove the termination in
> > general.  But  if you check carefully,  you'll see that  to prove this
> > theorem, we apply  termination prover on itself.  So  we just found an
> > example, a  case where it's  not possible, but  it does not  mean that
> > this  program  cannot be  helpful  to  analyse  and verify  the  other
> > programs. 
> 
> Many programmers are under the impression that although
> non-termination is impossible to determine in general, that it is
> probably determinable in practice, or at least in a large enough set
> of cases as to be useful.  This isn't true.
> 
> It is quite easy to find very simple programs that cannot be proven to
> terminate, and they come up in practice far more often than you'd
> think.  Here is an artificial one I came up with:
> 
> (defun kernel (s i)
>   (list (not (car s))
> 	(if (car s)
> 	    (cadr s)
> 	  (cons i (cadr s)))
> 	(cons 'y (cons i (cons 'z (caddr s))))))
> 
> (defconstant k0 '(t () (x)))
> 
> (defun mystery (list)
>   (do ((result (list t list)
>                (reduce #'kernel (if (car result)
>                                     (cadr result)
>                                     (caddr result))
>                        :initial-value k0)))
>       ((null (cadr result)))))
> 
> The kernel function trivially terminates, the call to reduce
> terminates, the list over which mystery iterates is finite for all
> iterations, yet it is not known whether mystery terminates for all
> finite lists.  (Google on `Collatz Problem' for details.)
> 

Yes, it's similar to this numerical function:

(defun hotpo (n) 
    ;; hotpo stands for Half Or Triple Plus One
    (cond
		((= n 1)  t)
		((oddp n) (hotpo (+ 1 (* n 3))))
		(t        (hotpo (/ n 2)))))


Well, I  must say that  in 28 years  of programming, this is  the only
function  that I  wrote that  don't  obviously terminate.  (I think  I
proved  a few  years ago  that it  terminates though).   While  I must
confess that I didn't prove the termination of most of my functions, I
would be  very happy to  have an automatic  prover to check  my simple
algorithms.

> The reason I coded this is because the only obstacle to termination
> analysis appears to be the iteration in `mystery'.  Yet iterations
> such as this are not an uncommon feature in typical programs.

"typical programs" in what domain?  For 90% of the application domains
and  99.99% for  the programs,  such  algorithms don't  occur, or  are
avoided just because there  are alternatives whose termination is more
obvious. (percentages  invented here, but serriously,  think about all
these business or desktop applications  that could be much less bugged
if automatic tools was disponible to work on them).

> If you look at  
> 
>    http://www.math.niu.edu/~rusin/known-math/95/independence
> 
> you well see some examples where Godel's incompleteness theorem
> (related to the halting problem) has cropped up in several `natural
> independence phenomena' that don't require the trick of attempting to
> apply a proof to itself.  One of the problems is `Kruskal's tree
> theorem' that is involved in showing that certain orderings on trees
> are well-founded.  
> 
> Again, tree traversal is a common programming technique.

-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-0901030937520001@192.168.1.51>
In article <··············@thalassa.informatimago.com>, Pascal Bourguignon
<···@informatimago.com> wrote:

> Yes, it's similar to this numerical function:
> 
> (defun hotpo (n) 
>     ;; hotpo stands for Half Or Triple Plus One
>     (cond
>                 ((= n 1)  t)
>                 ((oddp n) (hotpo (+ 1 (* n 3))))
>                 (t        (hotpo (/ n 2)))))
> 
> 
> Well, I  must say that  in 28 years  of programming, this is  the only
> function  that I  wrote that  don't  obviously terminate.  (I think  I
> proved  a few  years ago  that it  terminates though).

I doubt it.  If you really proved it that would make you a candidate for
the Fields medal.  You'd probably remember that.

E.
From: Andreas Hinze
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp  compiler
Date: 
Message-ID: <3E1DCC42.7A8903A4@smi.de>
Erann Gat wrote:
> 
> In article <··············@thalassa.informatimago.com>, Pascal Bourguignon
> <···@informatimago.com> wrote:
> 
> > Yes, it's similar to this numerical function:
> >
> > (defun hotpo (n)
> >     ;; hotpo stands for Half Or Triple Plus One
> >     (cond
> >                 ((= n 1)  t)
> >                 ((oddp n) (hotpo (+ 1 (* n 3))))
> >                 (t        (hotpo (/ n 2)))))
> >
> >
> > Well, I  must say that  in 28 years  of programming, this is  the only
> > function  that I  wrote that  don't  obviously terminate.  (I think  I
> > proved  a few  years ago  that it  terminates though).
> 
> I doubt it.  If you really proved it that would make you a candidate for
> the Fields medal.  You'd probably remember that.
> 
If i remember correct it is called "Ulam's suggestion". He assumed that
this function terminates but AFAIK none was able to show a proof.
But maybe Pascal has solved it ;-)

Best
AHz
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <smw1dqeb.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
>
> > I think we must be a cross purposes here: if the algorithmic
> > complexity of the prover is small, then it simply doesn't have enough
> > information capacity to make statements about the space of programs
> > larger than itself.
> 
> No, that's not true.  A prover cannot be guaranteed to produce a correct
> answer on programs algorithmically more complex than itself.  But that is
> different from what you seem to be saying, that a prover can be guaranteed
> not to produce a correct answer in those cases.

I don't want to say that the prover can be guaranteed to not produce a
correct answer (i.e., that the prover *always* says `I dunno if this
halts').  

I think you will agree with me, however, that in a given space of
programs (say programs of N pages or less for some N), that there is
an upper limit on how many correct answers a prover of a certain
complexity can give.  I think what we are arguing is the nature or
value of this upper limit --- I think it is small, you seem to
disagree.

> > If it could, then you could encode the larger programs in terms of
> > the smaller, and contradict the supposition that the program being
> > examined is more complex than the prover.
> 
> No.  This is only true if the analyzer works perfectly.  If it
> returns the correct result only some of the time (and returns
> "unknown" the rest of the time) then the conditions of Chaitin's
> theorem are not violated.

Right.  But Chaitin's theorem *will* be violated if it returns a
correct result for *enough* cases.  Certainly if it returns `I dunno'
for virtually every case, there is no problem.  Certainly if it
returns a correct `yes' or `no' for virtually every case there *is* a
problem.  But the question is what is the maximum number of correct
answers it can give without violating Chaitin's theorem?

> Consider:
> 
> (defmacro with-timeout (time &body body)
>    ... )
> 
> (defun analyze (code)
>   (cond ( (and (eq (first code 'with-timeout))
>                (numberp (second code)))
>           (format t "This code is guaranteed to terminate in under ~A seconds"
>                     (second code)) )
>         (t (format t
>  "I can't tell if this code will terminate or not.  Sorry."))))
> 
> This will produce the correct answer in some cases, and throw up its hands
> in others.  Some (many, most) of the cases where it produces the correct
> answer will be more algorithmically complex than the above code.  Whether
> or not this is useful is an open question, but it is not prima facie
> useless.

Ok, I agree with you here.  But I want to point out something:  the
analyzer only produces answers for forms that begin `with-timeout' and
have a literal number for the time.  Of the space of lisp expressions
that are more algorithmically complex than the above code, the number
that begin this way is *very* small.

> > > The key question is: are the programs that are left useful?  
> > 
> > Well, another key question is `Are there very many programs left?'
> 
> There are in many (most) cases an infinite number of program left.  The
> example above is one such case.

Here I disagree.  While there are an infinite number of programs that
begin with `with-timeout', there are only a finite number of
computation steps that can be taken within the given timeout.
Programs that would differ in their computation if run *longer* than
the timeout are equivalent.

> > > That is an open question, but anecdotal evidence indicates that the
> > > answer is yes.
> > 
> > What anecdotes?  The anecdotes I've heard say the opposite.
> 
> Look up the Remote Agent Bug.

I have.  How does it apply to this argument?

> > Consider this:  you certainly can't prove the termination of any
> > program capable of universal computation.  But it is surprisingly easy
> > to perform universal computation, and it often happens that systems
> > that were not designed for universal computation can be coaxed into
> > doing it.
> 
> This is the white crows agument again.  I do not deny the existence of
> white crows -- programs for which no meaningful proofs of correctness are
> possible.  

I'm only talking about termination, not correctness.  In my `mystery
function', it is impossible to prove termination, but trivial to prove
that the answer is NIL.


> I do not even deny the existence of a great many white crows
> (despite the fact that Chaitin's result only guarantees the
> existence of one).

Chaitin's result guarantees a certain minimum ratio of white to black
crows.

> So no matter how many of them you exhibit you will not dissuade me
> from my belief that, notwithstanding the white crows, there are
> nonetheless a great many black crows in the world, having beheld a
> fair number of them with my own eyes.

No matter *how many*?  Sure, there are a lot of black crows, but the
question is:  if I were to shoot a crow at random, would it black?
(and are they good to eat?)  Godel and Turing showed the existence of
white crows by pointing out one, but exactly how rare an example is
it?  I claim that white crows *far* outnumber the black ones.  (And I
may end up eating crow on *that* claim.)
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-1001030842110001@192.168.1.51>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> I think you will agree with me, however, that in a given space of
> programs (say programs of N pages or less for some N), that there is
> an upper limit on how many correct answers a prover of a certain
> complexity can give.

No, I do not agree with that.  I gave an example of a prover that provides
a useful and correct result on an infinite number of programs.

> Right.  But Chaitin's theorem *will* be violated if it returns a
> correct result for *enough* cases.

I believe it is sufficient to return in incorrect result for a single
case, but I'm not 100% sure about that.

> Ok, I agree with you here.  But I want to point out something:  the
> analyzer only produces answers for forms that begin `with-timeout' and
> have a literal number for the time.  Of the space of lisp expressions
> that are more algorithmically complex than the above code, the number
> that begin this way is *very* small.

No, the number is infinite.  In fact, it is equal to the total number of
programs, both of which are countably infinite.

> Here I disagree.  While there are an infinite number of programs that
> begin with `with-timeout', there are only a finite number of
> computation steps that can be taken within the given timeout.
> Programs that would differ in their computation if run *longer* than
> the timeout are equivalent.

Yes, but you can't tell a priori which programs are equivalent in this
way.  (That's *your* argument!)  So two different-looking programs A and B
may in fact *be* computationally equivalent, but there's no way you can
ever know that for sure in general.  So it's a moot point.  Besides, every
program only has a finite time to run.  (It's an elementary excercise to
figure out an upper bound on how many computations the Universe is capable
of.  Take the number of elementary particles (~10^80), assume every one is
a processor running at the Planck frequency (10^43 Hz), pick a reasonable
number for the time until the heat death of the Universe, multiply and
voila!  It's not a particularly big number as numbers go.

> > > > That is an open question, but anecdotal evidence indicates that the
> > > > answer is yes.
> > > 
> > > What anecdotes?  The anecdotes I've heard say the opposite.
> > 
> > Look up the Remote Agent Bug.
> 
> I have.  How does it apply to this argument?

It's an anecdote about how formal methods found real bugs in real code
that could not have been (indeed *were* not) found any other way.  (Did
you read the paper?)

> > > Consider this:  you certainly can't prove the termination of any
> > > program capable of universal computation.  But it is surprisingly easy
> > > to perform universal computation, and it often happens that systems
> > > that were not designed for universal computation can be coaxed into
> > > doing it.
> > 
> > This is the white crows agument again.  I do not deny the existence of
> > white crows -- programs for which no meaningful proofs of correctness are
> > possible.  
> 
> I'm only talking about termination, not correctness.  In my `mystery
> function', it is impossible to prove termination, but trivial to prove
> that the answer is NIL.

Yes, but that proof is only possible because the code has a certain form. 
Any code of the form:

(defun foo ()
  (arbitrary-computation)
  X)

will either run forever or return X.  That's *my* point.  Code in certain
forms are amenable to proofs of certain useful properties.  My belief is
that there are enough practical programs in those forms to be useful (and
that indeed, the kinds of forms that are amenable to proof provides
guidance as to how good code should be written).

> > I do not even deny the existence of a great many white crows
> > (despite the fact that Chaitin's result only guarantees the
> > existence of one).
> 
> Chaitin's result guarantees a certain minimum ratio of white to black
> crows.

Then we're talking about two different results here.  Could you give me a
pointer to the minimum-ratio result?

> > So no matter how many of them you exhibit you will not dissuade me
> > from my belief that, notwithstanding the white crows, there are
> > nonetheless a great many black crows in the world, having beheld a
> > fair number of them with my own eyes.
> 
> No matter *how many*?  Sure, there are a lot of black crows, but the
> question is:  if I were to shoot a crow at random, would it black?
> (and are they good to eat?)  Godel and Turing showed the existence of
> white crows by pointing out one, but exactly how rare an example is
> it?  I claim that white crows *far* outnumber the black ones.  (And I
> may end up eating crow on *that* claim.)

Indeed, as there are in fact precisely equal numbers of white-crow and
black-crow programs, both numbers being precisely countably infinite. 
(This, and my passing familiarity with Chaitin's work, makes me skeptical
that minimum-ratio result you claim is true.  But I'm willing to be
convinced otherwise.)

E.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <n0m8esph.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
> 
> > I think you will agree with me, however, that in a given space of
> > programs (say programs of N pages or less for some N), that there is
> > an upper limit on how many correct answers a prover of a certain
> > complexity can give.
> 
> No, I do not agree with that.  I gave an example of a prover that provides
> a useful and correct result on an infinite number of programs.

Read my assertion again:  for a given space of programs.

There aren't an inifinite number of programs containing less than
10000 characters of source code, and your prover provides correct
answers for a subset of these.

> > But I want to point out something: the analyzer only produces
> > answers for forms that begin `with-timeout' and have a literal
> > number for the time.  Of the space of lisp expressions that are
> > more algorithmically complex than the above code, the number that
> > begin this way is *very* small.
> 
> No, the number is infinite.  In fact, it is equal to the total number of
> programs, both of which are countably infinite.

Cantor notwithstanding, for any finite bound on algorithmic
complexity, the number of programs that are not more complex is
finite, and the number of *those* that begin `with-timeout' is small. 

> > Here I disagree.  While there are an infinite number of programs that
> > begin with `with-timeout', there are only a finite number of
> > computation steps that can be taken within the given timeout.
> > Programs that would differ in their computation if run *longer* than
> > the timeout are equivalent.
> 
> Yes, but you can't tell a priori which programs are equivalent in this
> way.  (That's *your* argument!)  So two different-looking programs A and B
> may in fact *be* computationally equivalent, but there's no way you can
> ever know that for sure in general.  So it's a moot point.  Besides, every
> program only has a finite time to run.  (It's an elementary excercise to
> figure out an upper bound on how many computations the Universe is capable
> of.  Take the number of elementary particles (~10^80), assume every one is
> a processor running at the Planck frequency (10^43 Hz), pick a reasonable
> number for the time until the heat death of the Universe, multiply and
> voila!  It's not a particularly big number as numbers go.

Yes.  So we are talking about a finite set of programs (the ones that
differ in their computation steps before the heat death of the universe).

> > > > > That is an open question, but anecdotal evidence indicates that the
> > > > > answer is yes.
> > > > 
> > > > What anecdotes?  The anecdotes I've heard say the opposite.
> > > 
> > > Look up the Remote Agent Bug.
> > 
> > I have.  How does it apply to this argument?
> 
> It's an anecdote about how formal methods found real bugs in real code
> that could not have been (indeed *were* not) found any other way.  (Did
> you read the paper?)

Yes.  And I don't argue at all that formal methods are a bad idea.
Formal methods can be very useful.  

However, you can't use formal methods to reason about termination.
(and everyone accepts that assertion in the general case)  I'm
claiming that the specific cases where you *can* use formal methods to
reason about termination aren't that common in the space of typical
computer programs. 

> Yes, but that proof is only possible because the code has a certain form. 
> Any code of the form:
> 
> (defun foo ()
>   (arbitrary-computation)
>   X)
> 
> will either run forever or return X.  That's *my* point.  Code in certain
> forms are amenable to proofs of certain useful properties.  My belief is
> that there are enough practical programs in those forms to be useful (and
> that indeed, the kinds of forms that are amenable to proof provides
> guidance as to how good code should be written).

I agree with you here.  There are a *lot* of useful analyses that can
be done.  I just assert that termination isn't one of them. 

> > > I do not even deny the existence of a great many white crows
> > > (despite the fact that Chaitin's result only guarantees the
> > > existence of one).
> > 
> > Chaitin's result guarantees a certain minimum ratio of white to black
> > crows.
> 
> Then we're talking about two different results here.  Could you give me a
> pointer to the minimum-ratio result?

Chaitin's Omega is the `probability a program will halt'.  To compute
this, you would divide the number of programs that halt by the total
number of programs.  But that is infinity / infinity.  To give a more
meaningful result, you compute the ratio for programs of finite length
and divide by 2^^length.  

Another way to look at it is that for any space of programs of a
certain size or less, a certain number of the leading digits of Omega
provides the probability that a program chosen `randomly' from that
set halts.

Now, let us use our termination prover to estimate Omega.  We can get
a conservative upper and lower bound on Omega by counting the `i
dunno' response as either `yes' or `no' respectively.  For a given
program size, we simply enumerate all programs of that size or less
and feed them to our termination prover.

But the problem here is that Omega is algorithmically random:  every
bit in Omega is independent of every other.  So a program that
estimates n bits of Omega must itself be about n bits in size (as per
the definition of algorithmic complexity).  So our termination prover
won't give us any information about the bits of Omega beyond the
complexity of our termination prover.  i.e., when you try to use the
termination prover on programs much more complex than the termination
prover itself, you almost always get `i dunno' answers.
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-1001031205080001@k-137-79-50-101.jpl.nasa.gov>
In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:

> However, you can't use formal methods to reason about termination.
> (and everyone accepts that assertion in the general case)  I'm
> claiming that the specific cases where you *can* use formal methods to
> reason about termination aren't that common in the space of typical
> computer programs.

A couple of resonses to this:

1) In my line of business they are very common.  Spacecraft sequences are
essentially little computer programs (though the spacecraft folks tend not
to think about them this way, which is unfortunate).  The ontology for
sequences is constrained in such a way that all sequences are guaranteed
to terminate.  (I used to think they did this on purpose for precisely
this reason, but one of the latest developments around here is a "new
improved" sequencing system that is Turing-complete.  I'm having a hard
time convincing anyone that this is a bad idea.)

2) Which programs are common is a matter of choice.  We can choose to
engineer our code so that the analyzable cases are common.  BTW, when
thinking about this many people make the mistake of assuming that halting
is "good" and not halting is "bad".  This is not necessarily so.  An
operating system, for example, should never halt.  In fact, here's another
analyzable code pattern that is potentially useful:

(loop (ignore-errors ...))

> I agree with you here.  There are a *lot* of useful analyses that can
> be done.  I just assert that termination isn't one of them. 

And I dispute your assertion for the Nth time.

> > > > I do not even deny the existence of a great many white crows
> > > > (despite the fact that Chaitin's result only guarantees the
> > > > existence of one).
> > > 
> > > Chaitin's result guarantees a certain minimum ratio of white to black
> > > crows.
> > 
> > Then we're talking about two different results here.  Could you give me a
> > pointer to the minimum-ratio result?
> 
> Chaitin's Omega

[snip]

> when you try to use the
> termination prover on programs much more complex than the termination
> prover itself, you almost always get `i dunno' answers.

This tells us absolutely nothing about the performance of a termination
tester on programs that people might actually write, and even less about
its performance on programs that people might actually write in a context
where they actually care about whether the program terminates or not.

Look, we agree that termination testing is for all intents and purposes
impossible on arbitrary code.  Your conclusion is that termination testing
is impossible (or impractical), period, end of story.  My conclusion is:
if you care about termination, don't write arbitrary code, write code for
which proofs of termination are possible.  Yes, such programs are a
miniscule fraction of the space of possible programs, but so what? 
Programs that do *anything* interesting are a miniscule fraction of the
space of possible programs.  All the programs that will ever be written by
all the beings in the Universe capable of writing programs is a tiny, tiny
fraction of the space of all possible programs!  The game is not to fret
over what may or may not be possible for arbitrary programs.  The game is
to find programs that have properties that we care about.  There is no
reason why provable termination should not be one of those properties. 
Certainly neither Turing nor Chaitin provide any compelling reason.

E.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <hecgem8j.fsf@ccs.neu.edu>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
> 
> > However, you can't use formal methods to reason about termination.
> > (and everyone accepts that assertion in the general case)  I'm
> > claiming that the specific cases where you *can* use formal methods to
> > reason about termination aren't that common in the space of typical
> > computer programs.
> 
> A couple of resonses to this:
> 
> 1) In my line of business they are very common.  Spacecraft sequences are
> essentially little computer programs (though the spacecraft folks tend not
> to think about them this way, which is unfortunate).  The ontology for
> sequences is constrained in such a way that all sequences are guaranteed
> to terminate.  (I used to think they did this on purpose for precisely
> this reason, but one of the latest developments around here is a "new
> improved" sequencing system that is Turing-complete.  I'm having a hard
> time convincing anyone that this is a bad idea.)
> 
> 2) Which programs are common is a matter of choice.  We can choose to
> engineer our code so that the analyzable cases are common.  

Um.  Yeah.  If you don't program in a Turing complete language (or
restrict yourself to using a non-turning-complete subset of the
language) then you don't have a problem with termination.  I can see
why one would want to do this in a situation like a spacecraft.

> BTW, when thinking about this many people make the mistake of
> assuming that halting is "good" and not halting is "bad".  This is
> not necessarily so.  An operating system, for example, should never
> halt.  In fact, here's another analyzable code pattern that is
> potentially useful:
> 
> (loop (ignore-errors ...))

Yes, but it's probably the case that you want the form inside the
ignore-errors to halt.  You don't want your OS to halt, true, but you
*do* want most of your system calls to return control to the user
program. 

> Look, we agree that termination testing is for all intents and purposes
> impossible on arbitrary code.  Your conclusion is that termination testing
> is impossible (or impractical), period, end of story.  

Not quite.  My conclusion is that termination testing is impossible
(or impractical) on code that was developed (on a `standard'
turing-complete language like C, Java, or Lisp) without termination
testing in mind, and that *unless* you are careful it is rather easy
to write code that cannot be proven to terminate.  If you recall, my
original claim was that code that can stymie a termination analysis is
not restricted to termination analysis of termination analyzers, but
is rather far more common than people think.

> My conclusion is:  if you care about termination, don't write
> arbitrary code, write code for which proofs of termination are
> possible.

I can't argue with that.
From: Thomas F. Burdick
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <xcvof6okttk.fsf@famine.OCF.Berkeley.EDU>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <············@ccs.neu.edu>, Joe Marshall <···@ccs.neu.edu> wrote:
>
> > I agree with you here.  There are a *lot* of useful analyses that can
> > be done.  I just assert that termination isn't one of them. 
> 
> And I dispute your assertion for the Nth time.

There's another assertion here that I'd like to dispute: that there
are a lot of useful analyses that aren't equivalent to termination.
I'm working on a code-analysis/proof-generation system, and when
thinking about the various properties we might want to try to prove,
it turned out that most of them were equivalent to termination.  Of
course, maybe I'm using "useful" in a lot more optimistic of a sense
here.

(oblisp: the system analyses GCC object files on Mac OS X, but it's
actually implemented in Common Lisp, using OpenMCL)

-- 
           /|_     .-----------------------.                        
         ,'  .\  / | No to Imperialist war |                        
     ,--'    _,'   | Wage class war!       |                        
    /       /      `-----------------------'                        
   (   -.  |                               
   |     ) |                               
  (`-.  '--.)                              
   `. )----'                               
From: Rob Warnock
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <qaydnaclctLvx7yjXTWcoQ@giganews.com>
Erann Gat <···@jpl.nasa.gov> wrote:
+---------------
| Programs that do *anything* interesting are a miniscule fraction of the
| space of possible programs.  All the programs that will ever be written by
| all the beings in the Universe capable of writing programs is a tiny, tiny
| fraction of the space of all possible programs!  The game is not to fret
| over what may or may not be possible for arbitrary programs.  The game is
| to find programs that have properties that we care about.  There is no
| reason why provable termination should not be one of those properties. 
| Certainly neither Turing nor Chaitin provide any compelling reason.
+---------------

For one good reason: You can! That is, you *can* prove termination of a
*very large* class of "programs that have properties that [you] care about".
Just follow the discipline of writing your proof & program *together*, as
taught by Dijkstra in "A Discipline of Programming", or by David Gries in
"The Science of Programming" -- starting with the desired result at the end
and backing up step by step until you have proved that your program will
still work correctly (including handling "input errors") no matter what
initial conditions are (which include input data). At every single stage
in the development of your program, you use only forms which can be proven
correct (including termination) -- that old "precondition/postcondition"
stuff -- and you compose them to create larger forms using rules which
guarantee (more "preconditions/postconditions") that those proofs of
termination are not violated.  Voila! When you're done, you *know* your
program will terminate!! [Not only that, you know it will give the right
answers!]

As Dijkstra once said somewhere [*very* loosely paraphrased, I've forgotten
the exact quote], there are enough "interesting" programs that one can
write which can be proven to be correct to fill billions of lifetimes,
so why should one waste one's life writing *incorrect* ones?

[Partial answer: Writing provably-correct programs is hard work, and
requires thinking. Finding the proper preconditions for some loops
can be "hard mathematics" (though most are pretty easy). This group
knows better than most what "Joe Average Programmer" thinks about
hard math...]

-Rob

-----
Rob Warnock, PP-ASEL-IA		<····@rpw3.org>
627 26th Avenue			<URL:http://rpw3.org/>
San Mateo, CA 94403		(650)572-2607
From: Pascal Bourguignon
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87r8bi6mux.fsf@thalassa.informatimago.com>
····@rpw3.org (Rob Warnock) writes:

> Erann Gat <···@jpl.nasa.gov> wrote:
> +---------------
> | Programs that do *anything* interesting are a miniscule fraction of the
> | space of possible programs.  All the programs that will ever be written by
> | all the beings in the Universe capable of writing programs is a tiny, tiny
> | fraction of the space of all possible programs!  The game is not to fret
> | over what may or may not be possible for arbitrary programs.  The game is
> | to find programs that have properties that we care about.  There is no
> | reason why provable termination should not be one of those properties. 
> | Certainly neither Turing nor Chaitin provide any compelling reason.
> +---------------
> 
> For one good reason: You can! That is, you *can* prove termination of a
> *very large* class of "programs that have properties that [you] care about".
> Just follow the discipline of writing your proof & program *together*, as
> taught by Dijkstra in "A Discipline of Programming", or by David Gries in
> "The Science of Programming" -- starting with the desired result at the end
> and backing up step by step until you have proved that your program will
> still work correctly (including handling "input errors") no matter what
> initial conditions are (which include input data). At every single stage
> in the development of your program, you use only forms which can be proven
> correct (including termination) -- that old "precondition/postcondition"
> stuff -- and you compose them to create larger forms using rules which
> guarantee (more "preconditions/postconditions") that those proofs of
> termination are not violated.  Voila! When you're done, you *know* your
> program will terminate!! [Not only that, you know it will give the right
> answers!]
> 
> As Dijkstra once said somewhere [*very* loosely paraphrased, I've forgotten
> the exact quote], there are enough "interesting" programs that one can
> write which can be proven to be correct to fill billions of lifetimes,
> so why should one waste one's life writing *incorrect* ones?

Exactly. To get an illustration of this principle:
http://www.unitedmedia.com/comics/monty/archive/monty-20030112.html
:-)

-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Pekka P. Pirinen
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <u3cnxcckn.fsf@globalgraphics.com>
Joe Marshall <···@ccs.neu.edu> writes:
> Chaitin's Omega is the `probability a program will halt'. To compute
> this, you would divide the number of programs that halt by the total
> number of programs.  But that is infinity / infinity.  To give a more
> meaningful result, you compute the ratio for programs of finite length
> and divide by 2^^length.

This is theoretically interesting, but not a practical measure.  I
understand that you were not proposing it as such, but were
introducing it for the argument set out below, but I note that some
your other arguments were of the same form: estimating fractions of
(provably) terminating programs.  This is a mirage, there is no such
thing.  Arguing against you, Erann Gat was essentially proposing a
measure where the appromimating sets were programs of a given finite
length plus those programs wrapped in (with-timeout ...).  That's not
completely practical either, but would clearly result in a completely
different probability.  The point is that there is no value that is
the fraction of terminating programs - depending on how you choose to
approach infinity, you can get any value you want.
 
> Now, let us use our termination prover to estimate Omega.  We can get
> a conservative upper and lower bound on Omega by counting the `i
> dunno' response as either `yes' or `no' respectively.  For a given
> program size, we simply enumerate all programs of that size or less
> and feed them to our termination prover.
> 
> But the problem here is that Omega is algorithmically random:  every
> bit in Omega is independent of every other.  So a program that
> estimates n bits of Omega must itself be about n bits in size (as per
> the definition of algorithmic complexity).  So our termination prover
> won't give us any information about the bits of Omega beyond the
> complexity of our termination prover.  i.e., when you try to use the
> termination prover on programs much more complex than the termination
> prover itself, you almost always get `i dunno' answers.

No, that doesn't follow: only that you get 'I dunno' at least 1/2^n of
the time, to avoid being able to estimate bit n+1.  Since a
termination prover is a non-trivial piece of code, n is quite large,
so this limit is not much of an obstacle in practice.

Of course, this doesn't guarantee that any given termination prover
will reach the limit.  I completely agree that you have to write your
program with termination in mind for this to be practical.  I'd just
like to point out that I do that anyway.  Not with a rigor that would
work with a termination prover, but like any normal programmer, I
believe that code that I write will terminate - except in the rare
cases it's not supposed to - i.e., I have an informal proof in mind.
If I had to prove it formally, I'd have to write down a lot of detail
that I don't think about normally (and sometimes I'd find that my
proof doesn't actually work), but I certainly have the shape of it.
Programmers write with termination in mind, because they need their
programs to terminate.
-- 
Pekka P. Pirinen
The gap between theory and practice is bigger in practice than in theory.
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <fzrxc9kv.fsf@ccs.neu.edu>
···············@globalgraphics.com (Pekka P. Pirinen) writes:

> Joe Marshall <···@ccs.neu.edu> writes:
> > Chaitin's Omega is the `probability a program will halt'. To compute
> > this, you would divide the number of programs that halt by the total
> > number of programs.  But that is infinity / infinity.  To give a more
> > meaningful result, you compute the ratio for programs of finite length
> > and divide by 2^^length.
> 
> This is theoretically interesting, but not a practical measure.  I
> understand that you were not proposing it as such, but were
> introducing it for the argument set out below, but I note that some
> your other arguments were of the same form: estimating fractions of
> (provably) terminating programs.  This is a mirage, there is no such
> thing.

I didn't make up this measure, Gregory Chaitin did.  I disagree with
your assessment of it as being `a mirage'.  It certainly *seems* that
the notion of `probability of halting' is meaningful, and in any
finite set of programs, it is obvious that there are those that halt
and those that do not.

> Arguing against you, Erann Gat was essentially proposing a measure
> where the approximating sets were programs of a given finite length
> plus those programs wrapped in (with-timeout ...).  That's not
> completely practical either, but would clearly result in a
> completely different probability.  The point is that there is no
> value that is the fraction of terminating programs - depending on
> how you choose to approach infinity, you can get any value you want.

By introducing `with-timeout', Erann Gat created a time-complexity
constraint that wasn't in my original example.  As you recall, I was
arguing about programs of a certain algorithmic complexity.  Adding a
time-limit simply reduces the space of programs under consideration by
making all programs (that would have taken) longer than the time-limit
equivalent.

In any case, the original poster was wondering about applying a
termination prover to existing code.  No existing program is
infinitely long, and for the sake of argument, let us restrict our
analysis to something `reasonable', say a `small' application written
in Common Lisp (how about 50 pages of code or fewer).

Now there are a large, but finite number of 50-page common lisp
programs.  Some of them halt, some of them do not.  Simply divide the
number that halt by the number of 50-page programs and there's your
ratio.

> > Now, let us use our termination prover to estimate Omega.  We can get
> > a conservative upper and lower bound on Omega by counting the `i
> > dunno' response as either `yes' or `no' respectively.  For a given
> > program size, we simply enumerate all programs of that size or less
> > and feed them to our termination prover.
> > 
> > But the problem here is that Omega is algorithmically random:  every
> > bit in Omega is independent of every other.  So a program that
> > estimates n bits of Omega must itself be about n bits in size (as per
> > the definition of algorithmic complexity).  So our termination prover
> > won't give us any information about the bits of Omega beyond the
> > complexity of our termination prover.  i.e., when you try to use the
> > termination prover on programs much more complex than the termination
> > prover itself, you almost always get `i dunno' answers.
> 
> No, that doesn't follow: only that you get 'I dunno' at least 1/2^n of
> the time, to avoid being able to estimate bit n+1.  Since a
> termination prover is a non-trivial piece of code, n is quite large,
> so this limit is not much of an obstacle in practice.

Looks like I did the math wrong here.  I'm pretty confident of my
conclusion, but I can see that last statement is unconvincing.

> Of course, this doesn't guarantee that any given termination prover
> will reach the limit.  I completely agree that you have to write your
> program with termination in mind for this to be practical.  I'd just
> like to point out that I do that anyway.  Not with a rigor that would
> work with a termination prover, but like any normal programmer, I
> believe that code that I write will terminate - except in the rare
> cases it's not supposed to - i.e., I have an informal proof in mind.
> If I had to prove it formally, I'd have to write down a lot of detail
> that I don't think about normally (and sometimes I'd find that my
> proof doesn't actually work), but I certainly have the shape of it.
> Programmers write with termination in mind, because they need their
> programs to terminate.

Well, they need their programs to terminate for the `typical' input.
You probably don't write your programs with the goal of enforcing
termination no matter *what* the input is.
From: Pekka P. Pirinen
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <uznq3brty.fsf@globalgraphics.com>
Joe Marshall <···@ccs.neu.edu> writes:
> ···············@globalgraphics.com (Pekka P. Pirinen) writes:
> > Joe Marshall <···@ccs.neu.edu> writes:
> > > Chaitin's Omega is the `probability a program will halt'. [...]
> > [...] I note that some your other arguments were of the same form:
> > estimating fractions of (provably) terminating programs.  This is
> > a mirage, there is no such thing.
> 
> I didn't make up this measure, Gregory Chaitin did.  I disagree with
> your assessment of it as being `a mirage'.  It certainly *seems* that
> the notion of `probability of halting' is meaningful, and in any
> finite set of programs, it is obvious that there are those that halt
> and those that do not.

Chaitin's Omega is not mirage; the idea of a universal probability of
halting is.  In a finite set, the fraction of halting programs depends
on the set, and can be anything from 0 to 1.  If we try to extend this
idea to the set of all programs, the limit depends on the particular
approach to infinity, and can be anything from 0 to 1 (or undefined).
Sometimes in such situations, there is a particular way that is
_natural_, but for the purposes of this discussion, I don't see one
(because programmers can and do choose to create terminating
programs).

> > [...] a measure where the approximating sets were programs of a
> > given finite length plus those programs wrapped in (with-timeout
> > ...).  That's not completely practical either, but would clearly
> > result in a completely different probability.  [...]

This is an example of a different way of approaching infinity, which
would result in a probability > 0.5, rather than Omega (which, I
suspect, would have a similar value in all real computing systems; you
quoted ~2^-21 for Chaitin's UTM).

Also, it seems queer to talk about probability when were not making
random choises from these sets of programs; on the contrary, we're
expending considerable efforts to construct programs of very
particular kinds.  Since we were arguing the practicability of
termination proving, we should really be exploring what kinds of
provably-terminating programs are possible or impossible or
impractical.  I can't resist arguing theory, though :-)

> By introducing `with-timeout', Erann Gat created a time-complexity
> constraint that wasn't in my original example.  As you recall, I was
> arguing about programs of a certain algorithmic complexity.  Adding a
> time-limit simply reduces the space of programs under consideration by
> making all programs (that would have taken) longer than the time-limit
> equivalent.

Both spaces are finite sets, parameterized by complexity.  If you need
a larger set, choose a larger value for the parameter.  If you couple
the time limit to the same parameter, you get a family of sets that
covers all halting programs.  So I don't see any useful conclusion
that could be drawn from these considerations.

> In any case, the original poster was wondering about applying a
> termination prover to existing code.  No existing program is
> infinitely long, and for the sake of argument, let us restrict our
> analysis to something `reasonable', say a `small' application written
> in Common Lisp (how about 50 pages of code or fewer).
> 
> Now there are a large, but finite number of 50-page common lisp
> programs.  Some of them halt, some of them do not.  Simply divide the
> number that halt by the number of 50-page programs and there's your
> ratio.

Since the OP was wondering about existing code, let's talk about
existing 50-page programs, rather than all possible 50-page ones.
Avoid the issue that, due to environment issues, a lot of the bigger
ones are not conforming CL, by defining our space to include the set
of actual CL implementations.  Incorporate input into the limit of 50
pages (in some unspecified form).  Now we have a practical measure we
could calculate.  Some programs are designed not to halt (except when
told to), perhaps even a lot of the larger ones, but I'd still expect
the result to be nearer to 1 than 0 - certainly not 2^-21 = one in a
million.

> [Omega v. termination prover: 0 - 1]

> > [All programming involves informal termination proofs.]
> > Programmers write with termination in mind, because they need their
> > programs to terminate.
> 
> Well, they need their programs to terminate for the `typical' input.
> You probably don't write your programs with the goal of enforcing
> termination no matter *what* the input is.

Yes, that is true in practice, and it's one of the reasons why
provably-correct and provably-terminating programs are tedious to
construct: you're forced to consider all the rare and impossible-
in-practice cases.  Ideally, your programming system would help you
identify and handle such cases.  I was making it sound too easy.
-- 
Pekka P. Pirinen
A programming language is low level when its programs require attention to
the irrelevant.  - Alan Perlis
From: Kimmo T Takkunen
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <slrnb25pf6.1ok.ktakkune@sirppi.helsinki.fi>
In article <·············@globalgraphics.com>, Pekka P. Pirinen wrote:
> Joe Marshall <···@ccs.neu.edu> writes:
>> Chaitin's Omega is the `probability a program will halt'. To compute

> This is theoretically interesting, but not a practical measure.  
Agreed.

> The point is that there is no value that is
> the fraction of terminating programs - depending on how you choose to
> approach infinity, you can get any value you want.

That is wrong. Chaitin proved that there is (it's mathematically well
defined). Intresting thing is that he also proved that you can't know
this number. It's somewhere in ]0,1[ and irrational. Approximating it
is impossible. 

More can be read from http://www.cs.auckland.ac.nz/CDMTCS/chaitin

> The gap between theory and practice is bigger in practice than in theory.
Above is good summary of this discussion.

 Kimmo
-- 
((lambda (integer) ;; http://www.iki.fi/kt/ 
   (coerce (loop for i upfrom 0 by 8 below (integer-length integer)
                 collect (code-char (ldb (byte 8 i) integer))) 'string))
 100291759904362517251920937783274743691485481194069255743433035)
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <adi5c91a.fsf@ccs.neu.edu>
········@cc.helsinki.fi (Kimmo T Takkunen) writes:

> In article <·············@globalgraphics.com>, Pekka P. Pirinen wrote:
> > The point is that there is no value that is
> > the fraction of terminating programs - depending on how you choose to
> > approach infinity, you can get any value you want.
> 
> That is wrong. Chaitin proved that there is (it's mathematically well
> defined). Intresting thing is that he also proved that you can't know
> this number. It's somewhere in ]0,1[ and irrational. Approximating it
> is impossible. 

*Computing* it is impossible.  It is possible to know it and
approximate it.  For Chaitin's Universal Turing machine in

    G. J. Chaitin. Algorithmic Information Theory, Cambridge University Press,
    Cambridge, 1987. (third printing 1990)

The value of the first 80 bits of Omega is:

0.0000000000000000000010000001000000100000010000010000011100100111000101
0001010000


(See http://apru.nus.edu.sg/pdf%20files/student/Shu%20abstract.pdf)

Know you know.
From: Nikodemus Siivola
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avog11$2qbt6$1@midnight.cs.hut.fi>
Erann Gat <···@jpl.nasa.gov> wrote:

> It's an anecdote about how formal methods found real bugs in real code
> that could not have been (indeed *were* not) found any other way.  (Did
> you read the paper?)

I don't know about him, but I scanned it fairly well. It seems to me that
the "formal methods" that were used were in fact rather informal, and *could*
be claimed that a great deal of their success *could* be attributed
the the ad-hoc steps.

Also, seems that the paper equates testing with full system regression-
testing, and their formal methods had in some ways more to do with
unit testing than formal methods as you discuss them: they systematically
exercised bits of code with asserted pre- and post- conditions, while
you seem to be talking about pure code analysis. Have I misunderstood either
you or the paper?

Furthermore, since they carried the analysis out not on original code
but hand-translated code, I would argue that the process was more about
verification of algorithms that code.

A fascinating read, anyways! Gives a lot's of food for though, esp. regarding
unit-testing as a formal method.

  -- Nikodemus
From: Erann Gat
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <gat-1101031525230001@192.168.1.51>
In article <··············@midnight.cs.hut.fi>, Nikodemus Siivola
<········@kekkonen.cs.hut.fi> wrote:

> Erann Gat <···@jpl.nasa.gov> wrote:
> 
> > It's an anecdote about how formal methods found real bugs in real code
> > that could not have been (indeed *were* not) found any other way.  (Did
> > you read the paper?)
> 
> I don't know about him, but I scanned it fairly well. It seems to me that
> the "formal methods" that were used were in fact rather informal, and *could*
> be claimed that a great deal of their success *could* be attributed
> the the ad-hoc steps.

You'll have to take that up with the authors of the paper.

> Also, seems that the paper equates testing with full system regression-
> testing, and their formal methods had in some ways more to do with
> unit testing than formal methods as you discuss them: they systematically
> exercised bits of code with asserted pre- and post- conditions, while
> you seem to be talking about pure code analysis. Have I misunderstood either
> you or the paper?

I think probably yes, on both counts.  When you say they "systematically
excercised bits of code..." that sounds like you think they actually ran
the code.  They didn't.  The SPIN tool that they use does do an analysis
that is equivalent to "systematic excercise" of certain aspects of the
code, but the code doesn't actually run.  If it did, it would never
finish.  Also, I'm not so dogmatic about what to call "formal methods."  I
think the main point is to have redundant descriptions of one's
intentions, and some automated ways of checking consistency among those
descriptions.

> Furthermore, since they carried the analysis out not on original code
> but hand-translated code, I would argue that the process was more about
> verification of algorithms that code.

OK, fine.  Like I said, I don't want to be dogmatic about terminology. 
The point is that whatever you want to call it, what they did found bugs
that would not have been found by testing as shown by the fact that a bug
existed in the code that they didn't analyze that was in fact not found by
testing but nonetheless manifested itself in flight.

E.
From: Pascal Bourguignon
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <871y3jaaym.fsf@thalassa.informatimago.com>
···@jpl.nasa.gov (Erann Gat) writes:
> Yes, but that proof is only possible because the code has a certain form. 
> Any code of the form:
> 
> (defun foo ()
>   (arbitrary-computation)
>   X)
> 
> will either run forever or return X.  That's *my* point.  Code in certain
> forms are amenable to proofs of certain useful properties.  My belief is
> that there are enough practical programs in those forms to be useful (and
> that indeed, the kinds of forms that are amenable to proof provides
> guidance as to how good code should be written).

Yes.  For example, how would you proove that:

    (defun will-i-terminate ()
        (do nil ((member (read *query-io* t :eof) '(:quit :eof)) :done)))

will ever terminate?


Well, perhaps if we add formally the consiration that the program runs
on a concrete computer around a concrete start in a concrete universe,
giving  time limits  either for  a working  device, a  star  not going
super-nova, or a universe not  freezing, then we could proove that all
programs will end one way or another...


The  point is  that  the termination  of  some programs  (and of  most
practical  (interactive)   program)  cannot  be   determined  only  on
intrinsic properties.   It may be  sufficient to be able  to formalize
all (or perhaps  only some) of the needed  conditions for an expedient
termination.

So   even   if   it    cannot   be   automatically   determined   that
(arbitrary-computation) terminates,  perhaps some (trivial  but useful
nonetheless) terminating conditions could be automatically determined?


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <8yxpdt1d.fsf@ccs.neu.edu>
Pascal Bourguignon <···@informatimago.com> writes:

> So even if it cannot be automatically determined that
> (arbitrary-computation) terminates, perhaps some (trivial but useful
> nonetheless) terminating conditions could be automatically
> determined?

Check out 

@article{ lee01sizechange,
    author = "Chin Soon Lee and Neil D. Jones and Amir M. Ben-Amram",
    title = "The size-change principle for program termination",
    journal = "ACM SIG{\-}PLAN Notices",
    volume = "36",
    number = "3",
    pages = "81--92",
    year = "2001",
    url = "citeseer.nj.nec.com/lee01sizechange.html" }

ABSTRACT
  The `size-change termination' principle for a first-order functional
language with well-founded data is:  a program terminates on all
inputs if /every infinite call sequence/ (following program control
flow) would cause an infinite descent in some value.

  Size-change analysis is based only on local approximations to
parameter size changes derivable from program syntax.  The set of
infinite call sequences that follow program flow and can be recognized
as causing infinite descent is an omega-regular set, representable by
a Buchi automaton.  Algorithms for such automata can be used to decide
size-change termination.  We also give a direct algorithm operating on
`size-change graphs' (without the passage to automata).

  Compared to other results in the literature, termination analysis
based on the size-change principle is surprisingly simple and general:
lexical orders (also called lexcographic orders), indirect function
calls and permuted arguments (descent that is not /in-situ/) are all
handled /automatically and without special treatment/, with no need
for manually supplied argument orders, or theorem-proving methods not
certain to terminate at analysis time.

  We establish the problem's /intrinsic complexity/.  This turns out
to be surprisingly high, complete for PSPACE, in spite of the
simplicity of the principle.  PSPACE hardness is proved by a reduction
from Boolean program termination.  An interesting consequence:  the
same hardness result applies to many other analyses found in the
termination and quasi-termination literature.
From: Russell Wallace
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e282712.346295790@news.eircom.net>
On 10 Jan 2003 09:35:08 -0500, Joe Marshall <···@ccs.neu.edu> wrote:

>No matter *how many*?  Sure, there are a lot of black crows, but the
>question is:  if I were to shoot a crow at random, would it black?
>(and are they good to eat?)  Godel and Turing showed the existence of
>white crows by pointing out one, but exactly how rare an example is
>it?  I claim that white crows *far* outnumber the black ones.  (And I
>may end up eating crow on *that* claim.)

I suspect they don't.

Consider randomly generated programs of size, say, 1 million lines. I
will suggest that the majority of such programs will execute no more
than a few hundred lines of code before either halting or going into
an obviously infinite loop.

(Which of those two results is more common probably depends on the
details of the language, but I will suggest that for reasonably
designed languages - anything you'd want to make common use of for
either source or object code - one or the other will usually happen
quite quickly. Has anyone actually put this to the test?)

But even if white crows _were_ more common than black ones, it doesn't
greatly matter as long as the crow you can buy in the supermarket or a
restaurant is normally black, and I will claim this is the case. The
programs 90% of the world's programmers spend their time working on,
what do they do? Fundamentally simple transformations of data from one
format to another. Source -> object code, a sequence of user input ->
relational database entries, HTML -> bitmap etc. Programs like this
should halt when and only when they encounter an end of file or other
halt instruction in their input stream, and I will suggest that this
should usually be provable.

Bigger difficulties than generating proofs are:

- Getting the specification to match what the customer actually needs.

- Getting it close enough to what he'll need next month that you don't
have to redo half your work from scratch.

- Relying on all the layers of (usually buggy) runtime infrastructure.
I suspect runtime infrastructure/component problems nowadays account
for more failures than algorithmic errors in application code.

-- 
"Mercy to the guilty is treachery to the innocent."
Remove killer rodent from address to reply.
http://www.esatclear.ie/~rwallace
From: Joe Marshall
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <8yxjvhwy.fsf@ccs.neu.edu>
··@vorpalbunnyeircom.net (Russell Wallace) writes:

> On 10 Jan 2003 09:35:08 -0500, Joe Marshall <···@ccs.neu.edu> wrote:
> 
> >No matter *how many*?  Sure, there are a lot of black crows, but the
> >question is:  if I were to shoot a crow at random, would it black?
> >(and are they good to eat?)  Godel and Turing showed the existence of
> >white crows by pointing out one, but exactly how rare an example is
> >it?  I claim that white crows *far* outnumber the black ones.  (And I
> >may end up eating crow on *that* claim.)
> 
> I suspect they don't.
> 
> Consider randomly generated programs of size, say, 1 million lines. I
> will suggest that the majority of such programs will execute no more
> than a few hundred lines of code before either halting or going into
> an obviously infinite loop.

The ones that go into an infinite loop are among the ones that you
cannot prove will halt.

> But even if white crows _were_ more common than black ones, it doesn't
> greatly matter as long as the crow you can buy in the supermarket or a
> restaurant is normally black, and I will claim this is the case. The
> programs 90% of the world's programmers spend their time working on,
> what do they do? Fundamentally simple transformations of data from one
> format to another. Source -> object code, a sequence of user input ->
> relational database entries, HTML -> bitmap etc. Programs like this
> should halt when and only when they encounter an end of file or other
> halt instruction in their input stream, and I will suggest that this
> should usually be provable.

It is provable that a program that performs a bounded number of steps
on each item of data will terminate when the data stream is finite.
From: Russell Wallace
Subject: Re: Interest of formal proof  Was: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e284712.354488593@news.eircom.net>
On 17 Jan 2003 11:51:25 -0500, Joe Marshall <···@ccs.neu.edu> wrote:

>··@vorpalbunnyeircom.net (Russell Wallace) writes:
>>
>> Consider randomly generated programs of size, say, 1 million lines. I
>> will suggest that the majority of such programs will execute no more
>> than a few hundred lines of code before either halting or going into
>> an obviously infinite loop.
>
>The ones that go into an infinite loop are among the ones that you
>cannot prove will halt.

But if the infinite loop is simple - for example, if it returns
exactly to a previous state after a reasonable number of steps - you
can prove it _won't_ halt.

>It is provable that a program that performs a bounded number of steps
>on each item of data will terminate when the data stream is finite.

Yep.

-- 
"Mercy to the guilty is treachery to the innocent."
Remove killer rodent from address to reply.
http://www.esatclear.ie/~rwallace
From: Alain Picard
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87u1gmwq0o.fsf@ibook.optushome.com.au>
···········@yahoo.com (Edward G. Nilges) spouted:

> As to studying it, I used it occasionally several years ago and find
> no need for the artifact.  I have to judge it from the outside, but as
> yet, nobunny has given me persuasive information as to its value.

Now that we have your eloquent and considered opinion on Lisp,
would you please just _leave_?  

TIA.
From: Nils Goesche
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <lyy95yphkg.fsf@cartan.de>
Alain Picard <·······················@optushome.com.au> writes:

> ···········@yahoo.com (Edward G. Nilges) spouted:
> 
> > As to studying it, I used it occasionally several years ago and
> > find no need for the artifact.  I have to judge it from the
> > outside, but as yet, nobunny has given me persuasive information
> > as to its value.
> 
> Now that we have your eloquent and considered opinion on Lisp, would
> you please just _leave_?

He won't.  Not as long as anybody is responding or only mentioning
his name (as we have just seen).

            +-------------------+             .:\:\:/:/:.
            |   PLEASE DO NOT   |            :.:\:\:/:/:.:
            |  FEED THE TROLLS  |           :=.' -   - '.=:
            |                   |           '=(\ 9   9 /)='
            |   Thank you,      |              (  (_)  )
            |       Management  |              /`-vvv-'\
            +-------------------+             /         \
                    |  |        @@@          / /|,,,,,|\ \
                    |  |        @@@         /_//  /^\  \\_\
      @·@@·@        |  |         |/         WW(  (   )  )WW
      \||||/        |  |        \|           __\,,\ /,,/__
       \||/         |  |         |      jgs (______Y______)
   /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
======================================================================

Regards,
-- 
Nils G�sche
"Don't ask for whom the <CTRL-G> tolls."

PGP key ID 0x0655CFA0
From: Alain Picard
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87ptr9wzx0.fsf@ibook.optushome.com.au>
Nils Goesche <······@cartan.de> writes:

> He won't.  Not as long as anybody is responding or only mentioning
> his name (as we have just seen).

You're right.  Sort of a kibo from hell, I guess.
Lovely ASCII art, BTW.  Did you draw that?  Beautiful!  :-)
From: Nils Goesche
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <lyy95xnr6d.fsf@cartan.de>
Alain Picard <·······················@optushome.com.au> writes:

> Nils Goesche <······@cartan.de> writes:
> 
> > He won't.  Not as long as anybody is responding or only mentioning
> > his name (as we have just seen).
> 
> Lovely ASCII art, BTW.  Did you draw that?  Beautiful!  :-)

I wish :-) No, I stole it from the net, somewhere.  That was the best
one I could find.

Regards,
-- 
Nils G�sche
"Don't ask for whom the <CTRL-G> tolls."

PGP key ID 0x0655CFA0
From: Pascal Bourguignon
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87fzs6c6t7.fsf@thalassa.informatimago.com>
···········@yahoo.com (Edward G. Nilges) writes:
> > > C does have a register declaration but note that it is as abstract and
> > > not as specific as possible unlike CAR and CDR.
> > 
> > car and cdr are not tied to registers at all.
> 
> CAR and CDR are only poetically referents to old architecture but my
> (humanistic) point was the same as Dijkstra's: in applied mathematics
> we need to be 100% vigilant even about the metaphors we use. 
> Dijkstra, unlike nearly all other computer scientists, had an almost
> Wittgensteinian respect for what Wittgenstein called "the bewitchment
> of our intelligence by way of language."
> 
> Of course, youse guys may be above such bewitchment, being so smart
> and everything.

As mentioned  in another  post, CAR and  CDR are  only two slots  in a
small record.   Naming them  FIRST and REST  is giving them  a meaning
that  may be intersting  when you're  using lists,  but as  you write,
would be  a dangerous metaphor  when your data structure  is something
else,  be  it  a  binary  tree,  another kind  of  tree  or  something
completely different, like  a S-expr for example.  Why  would you want
to call  the + or  '(+ 1 2  3) the FIRST  and the arguments  the REST?
Would you not want then to call them OPERATOR and ARGUMENTS rather?

That's why CAR  and CDR are good: because  they are meaningless.  Then
you can assign them whatever meaning you want.


> > Pascal
> 
> What do you think of the language after which you were named?

I'm  not this  Pascal, but...   It  was one  of the  first language  I
learned and I liked it at the  time. Used to know the Report by heart.
But in the  category of Algol language, I then  moved to Modula-2, and
consider now Modula-3  the best of its category.   The fact that these
languages are designed keeping in  mind the simplicity of the compiler
(hence it's  correctness) is not alien  to my liking of  them.  When I
think of all these buffer  overrun bugs exploited by worms and viruses
that would not exist with them... (I do C only to bring the bread).


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301052223.3984e92e@posting.google.com>
Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
> Edward G. Nilges wrote:
> > Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> > 
> >>* Edward G Nilges wrote:
> >>
> >>
> >>>Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> >>>based on the IBM 701's register structure and CAR and CDR are based on
> >>>operations in its machine language.  As of now I stand by this
> >>>claim.
> >>
> >>Well, there are 978 symbols in the CL package.  2 of them come from
> >>IBM 701 mnemonics, another two come from PDP-10 mnemonics.  So the
> >>`architecture of Lisp' is based on the IBM 701's register
> >>architecture?  Why not the DEC-10, which would certainly seem a more
> >>plausible candidate given CL's maclisp heritage? Or, given that 99.6%
> >>of the symbols in CL don't actually come from either machine, why not
> >>just give up and think?  No, I know, it's easier to just not bother.
> > 
> > 
> > Thanks for the information and indeed the confirmation, Tim, but I am
> > not sure what you might mean by "thinking."
> 
> The names "car" and "cdr" are not an issue at all for Lisp programmers. 
> On the contrary, some even appreciate them. If you are really interested 
> why, just ask.
> 
> > I do have a systematic tendency as a teacher and student formerly of
> > philosophy who entered the computer racket 30 years ago as a
> > draft-dodging scheme, to bring humanistic interpretation into the
> > field, and to me CAR and CDR are representative of a tendency in LISP
> 
> They are not.
> 
> > (please be advised that I have no special expertise in Lisp).
> 
> You are criticizing a very superficial aspect of Lisp. You should first 
> study Lisp and find out what the essential aspects of Lisp are before 
> you start to criticize it. You wouldn't criticize the Japanese language 
> either just because of the "funny" characters.
> 
> > This is to tie the language, at a key point, to the vendor's hardware.
> >  It was also done in Fortran about the same time.
> 
> car and cdr are not tied to any hardware. The semantics of car and cdr 
> are clearly specified and can be implemented an any hardware. Lisp just 
> happened to be implemented on a specific hardware first.
> 
> > Whereas the Algol group pioneered what has turned out to be a better
> > approach, and that is for a quasi-civic agency to be technically
> > independent of computer vendors.  The Algol team devised the father of
> > most usable modern languages including C because as an independent
> > civic forum, they were immune to commercial pressures.
> 
> It's not clear at all that the Algol approach is the better one. On the 
> contrary, there are many signs that Algol-style languages have reached 
> their limits. There is much research going on on things like 
> aspect-oriented programming, generative programming, support for 
> domain-specific languages etc. - things that are unnecessarily complex 
> in Algol languages. Scripting languages like Python and Ruby are 
> increasingly incorporating features originally found in Lisp. etc. pp.
> 
> Many things are continually being reinvented that are more or less 
> already solved in Lisp.
> 
> > C does have a register declaration but note that it is as abstract and
> > not as specific as possible unlike CAR and CDR.
> 
> car and cdr are not tied to registers at all.
> 
> 
> Here are my favorite two papers about the essence of Lisp as far as I 
> understand it.
> 
> * http://www.paulgraham.com/rootsoflisp.html

I am afraid that this site contains some rather over the top claims,
starting with "In 1960, John McCarthy published a remarkable paper in
which he did for programming something like what Euclid did for
geometry."

Programming lacks a Euclid, unless that's Dijkstra, because it has not
solved the problem of deriving correct programs in full.

Furthermore, Lisp was introduced in 1960.  How could it have invented
recursion?  Work on what became Algol 60 was being released in the
late 1950s and this language contains recursion, does it not?

I agree that the full scale ability to treat the program as an
artifact was important, the question is, how important.  Ada Augusta
realised that the program itself was as physical as the data, and
prior to 1960, assembler programmers had long been amazing their
friends with self-modifying code.

While Lisp provides a more elegant way of operating than crude
self-modification, it allows a sort of Higher Obfuscation which in
practice creates confusing code.

Lisp genuinely invented the idea that a program can be a statement of
WHAT is required, as an expression, and not HOW to solve it.  The
question whether this is a good thing.  The claim at the site is that
the procedural orientation of Algol did damage before Lisp could save
the day.

The problem is that the HOW is clearer in some instances, and the WHAT
can be rather arrogant.  The "how" is Aristotle: the "what" is Plato. 
Many people prefer the livelier empiricism of Aristotle.

> * ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-453.pdf
> 
> Further pointers can be found at 
> http://www.pascalcostanza.de/lisp/guide.html and, of course, 
> http://www.lisp.org
> 
> 
> 
> Pascal
From: Paolo Amoroso
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <U48ZPqcdmU1sp6noqPwwfhU2wktR@4ax.com>
On 5 Jan 2003 22:23:48 -0800, ···········@yahoo.com (Edward G. Nilges)
wrote:

> Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
[...]
> > Here are my favorite two papers about the essence of Lisp as far as I 
> > understand it.
> > 
> > * http://www.paulgraham.com/rootsoflisp.html
> 
> I am afraid that this site contains some rather over the top claims,
> starting with "In 1960, John McCarthy published a remarkable paper in
> which he did for programming something like what Euclid did for
> geometry."

Paul Graham wrote two well known and highly respected Lisp books, and he
based a very successful business on the language. In an earlier article you
said "please be advised that I have no special expertise in Lisp". So whose
claims are over the top?


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3vg13tc0g.fsf@cley.com>
* Pascal Costanza wrote:

> car and cdr are not tied to any hardware. The semantics of car and cdr
> are clearly specified and can be implemented an any hardware. Lisp
> just happened to be implemented on a specific hardware first.

Indeed, the near-canonical example of closures is something like:

    (defun kons (a b)
      (lambda (k &optional (v nil vp))
        (ecase k
          ((kar)
           (if vp (setf a v) a))
          ((kdr)
           (if vp (setf b v) b)))))

    (defun kar (x)
      (funcall x 'kar))

    (defun (setf kar) (new x)
      (funcall x 'kar new))

    (defun kdr (x)
      (funcall x 'kdr))

    (defun (setf kdr) (new x)
      (funcall x 'kdr new))

Where are the registers here?

--tim

(I suspect that this is going to degenerate into a situation where
people will claim that this Nilges character was `hounded out' of cll,
so I'm going to avoid further `contributions' to this whole futility).

--tim
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301052145.307c339a@posting.google.com>
Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> * Pascal Costanza wrote:
> 
> > car and cdr are not tied to any hardware. The semantics of car and cdr
> > are clearly specified and can be implemented an any hardware. Lisp
> > just happened to be implemented on a specific hardware first.
> 
> Indeed, the near-canonical example of closures is something like:
> 
>     (defun kons (a b)
>       (lambda (k &optional (v nil vp))
>         (ecase k
>           ((kar)
>            (if vp (setf a v) a))
>           ((kdr)
>            (if vp (setf b v) b)))))
> 
>     (defun kar (x)
>       (funcall x 'kar))
> 
>     (defun (setf kar) (new x)
>       (funcall x 'kar new))
> 
>     (defun kdr (x)
>       (funcall x 'kdr))
> 
>     (defun (setf kdr) (new x)
>       (funcall x 'kdr new))
> 
> Where are the registers here?

I agree that Lisp is far more elegant.  David Gerlenter showed why in
Programming Linguistics.  But, you can write self-interpreters in any
language.  Furthermore, successors to Lisp exist which preserve this
elegance and discard its antiquarian features.
> 
> --tim
> 
> (I suspect that this is going to degenerate into a situation where
> people will claim that this Nilges character was `hounded out' of cll,
> so I'm going to avoid further `contributions' to this whole futility).
> 
Yeah, it's strange.  In unmoderated groups I am like Shakespeare's Tom
O'Bedlam, who is whipped from town to town (yet we must remember is
actually Gloucester's loyal son.)

But when I send articles about what I actually know to editors they
sometimes send me money, which is very cool to have.  Of course, these
do not contain irresponsible speculation and opinion, for the most
part.

I have a theory that entirely too many groups have in recent years
been so dominated by supertrolls (of the sort who are perhaps
over-eager to misread and prejudge people as trolls) and
superpredators.

As a result, deviance is normalized.  In this context, the non-deviant
becomes Tom O'Bedlam.

Spinoza said it best.  Enlightened men seek kindred spirits and then,
in Emily Dickinson's words, shut the Valves of attention like stone. 
I should wise up.

The only reason I persist is that I learn to deal, at arms length,
with bozos, of which there is a rich supply in the business world.

Thanks for updating me about Lisp.  
> --tim
From: wni
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ZP2S9.254347$qF3.23224@sccrnsc04>
This is probably some rare occasion that I realize where Mr Naggum's
anger came from. But I will try to address some with my limited
knowlege and ultimate ignorance.

Edward G. Nilges wrote:
> Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...

> I do have a systematic tendency as a teacher and student formerly of
> philosophy who entered the computer racket 30 years ago as a
> draft-dodging scheme, to bring humanistic interpretation into the
> field, and to me CAR and CDR are representative of a tendency in LISP
> (please be advised that I have no special expertise in Lisp).

So you want to read computational artifacts and language designs using
the "humanistic* glasses of yours. This is why you don't like car/cdr?
To my limited brain, the knowledge of the tie between car/cdr and
IBM 704 came very late when I read SICP. Before that, I was never
bothered by things in MACLISP except some mistakes I made because
of the scoping rules. What's your humanistic interpretation of this?

Choosing the bad names is a common mistake that happen all the time.
I don't like the word "closure" because I happen to learn algebraic
closure first in math. One mistake of typesetter gave us lambda
instead of what Church originally wrote. Most of time, I think these
"mistakes" add a lot of colorful human touch in there to make the
story extremely romantic and interesting.

> 
> This is to tie the language, at a key point, to the vendor's hardware.
>  It was also done in Fortran about the same time.

I think you are picking up a very small historic negligible "mistake"
(if you can call it that) and magnify it into something out of
proportion.

> 
> Whereas the Algol group pioneered what has turned out to be a better
> approach, and that is for a quasi-civic agency to be technically
> independent of computer vendors.  The Algol team devised the father of
> most usable modern languages including C because as an independent
> civic forum, they were immune to commercial pressures.

So the choice of car/cdr can be translated to "commercial pressures"
and render the Lisp language "uncivilized?" This is a stretch.

I don't want to criticize Algol-based languages here because of your
admirations from a false angle. Just understand why Dana Scott had
to resort to topological space and lattice theory to get the
semantics right for a toy Algol-like language.

I also object to the notion that C is strongly related to the Algol
family. My multi-year experiences in programming C convinced me that
C is related to assembly language. One big feature of Algol is the
block structure; a regular C programmer doesn't even know that.

> 
> C does have a register declaration but note that it is as abstract and
> not as specific as possible unlike CAR and CDR.

If you read John Lion's (possibly illegal then) Unix v7 source code,
you would have known that the C language was heavily tied to the
PDP-11 architecture.

I also fail to see why a register declaration is *more abstract*.

> 
[irrelevant paragraphs deleted]

> 
> 99.6% is a meaningless number, for it is possible to exceed six sigma
> in programming language design...it is possible to have NO reference
> to unneeded garbage.
> 

What kind of "garbage" are you talking about? I am beginning to doubt
my own reading capabilities because I just can't extract meaningful
message from the above paragraph.

One thing I'd like to "assert" is that many modern languages borrow
much much more from Lisp family. The only thing Common Lisp obvious
borrowed was the block structure introduced by Scheme. However, only
one thing they can't steal: the powerful idea of data == program,
which is why Lisp macros are so so powerful just the way it is instead
of the theory-laden *functor* approaches (which is still functional
instead of Algol-like). The original "sin" of using linked lists,
despite its simplicity, probably catapults that kind of thinking.




wni at attbi dot com
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301052259.dcced56@posting.google.com>
wni <···@nospam.attbi.com> wrote in message news:<······················@sccrnsc04>...
> This is probably some rare occasion that I realize where Mr Naggum's
> anger came from. But I will try to address some with my limited
> knowlege and ultimate ignorance.
> 
> Edward G. Nilges wrote:
> > Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
>  
> > I do have a systematic tendency as a teacher and student formerly of
> > philosophy who entered the computer racket 30 years ago as a
> > draft-dodging scheme, to bring humanistic interpretation into the
> > field, and to me CAR and CDR are representative of a tendency in LISP
> > (please be advised that I have no special expertise in Lisp).
> 
> So you want to read computational artifacts and language designs using
> the "humanistic* glasses of yours. This is why you don't like car/cdr?
> To my limited brain, the knowledge of the tie between car/cdr and
> IBM 704 came very late when I read SICP. Before that, I was never
> bothered by things in MACLISP except some mistakes I made because
> of the scoping rules. What's your humanistic interpretation of this?

I learned symbolic logic before having access to a computer, yet when
I started programming, I realized that it was not a good notation for
programming, at least at the time.

My need today is not to find a single magic language, it's to write
correct code, and I tend to look at languages like Lisp with a
jaundiced eye as a result.  Procedural code can be sometimes easier to
informally prove correct, and informal proofs of correctness are, to
me, what's important.

Indeed, an entire philosophy of mathematics, that of Brouwer's
"intuitionism" is critical of mathematical procedures that rely
excessively on axioms not spelled out in a "finitary" style.  That's
because of Russell's discovery of a proto-bug in his grand system of
Principia Mathematica, the existence of "sets that do not include
themselves."

This was a proto-bug in his mentor Frege's initial attempt, discovered
with the same dismay by Frege and Russell because, like modern
programmers, they found they had been misled by their choice of
axioms.

In a purely psychological sense, the Intuitionists relied on the
greater safety of finitary mathematics which abandoned the excluded
middle and existence assumptions.

Unlike Algol, Lisp more or less assumed that deduction and mathematics
were fully reliable, and that in 1960 we could proceed to use pure
deduction.  From the important perspective of checking results, this
missed the fact that empirically, ordinary slobs like me sometimes
need to have things spelled out to a (Kantian) intuition, as a
procedure.


> 
> Choosing the bad names is a common mistake that happen all the time.
> I don't like the word "closure" because I happen to learn algebraic
> closure first in math. One mistake of typesetter gave us lambda
> instead of what Church originally wrote. Most of time, I think these
> "mistakes" add a lot of colorful human touch in there to make the
> story extremely romantic and interesting.
> 
> > 
> > This is to tie the language, at a key point, to the vendor's hardware.
> >  It was also done in Fortran about the same time.
> 
> I think you are picking up a very small historic negligible "mistake"
> (if you can call it that) and magnify it into something out of
> proportion.
> 
> > 
> > Whereas the Algol group pioneered what has turned out to be a better
> > approach, and that is for a quasi-civic agency to be technically
> > independent of computer vendors.  The Algol team devised the father of
> > most usable modern languages including C because as an independent
> > civic forum, they were immune to commercial pressures.
> 
> So the choice of car/cdr can be translated to "commercial pressures"
> and render the Lisp language "uncivilized?" This is a stretch.
> 
> I don't want to criticize Algol-based languages here because of your
> admirations from a false angle. Just understand why Dana Scott had
> to resort to topological space and lattice theory to get the
> semantics right for a toy Algol-like language.
>
That is not necessarily a defect.  I used informal group theory to
optimize string and math expressions in a recent compiler.

In my book, there is nothing wrong with being as mathematically "deep"
as necessary in Getting It Right, but something wrong with havering
about 36 bit words in the modern era.

 
> I also object to the notion that C is strongly related to the Algol
> family. My multi-year experiences in programming C convinced me that
> C is related to assembly language. One big feature of Algol is the
> block structure; a regular C programmer doesn't even know that.
>

This is supposed to be a good thing??  
 
> > 
> > C does have a register declaration but note that it is as abstract and
> > not as specific as possible unlike CAR and CDR.
> 
> If you read John Lion's (possibly illegal then) Unix v7 source code,
> you would have known that the C language was heavily tied to the
> PDP-11 architecture.
> 
No, his code was.  Brian and Dennis excluded computer architecture
from the design of C except in the minimalist sense of assuming a Von
Neumann machine, whence the address operator.

> I also fail to see why a register declaration is *more abstract*.
>
The elegance is that it says nothing more than "yo, put this inna
register if you have a choice and you want to."  I realize that CAR
and CDR are not commands to do nasty things but in a stylistic sense
they are anomalies.

 
> > 
> [irrelevant paragraphs deleted]
> 
> > 
> > 99.6% is a meaningless number, for it is possible to exceed six sigma
> > in programming language design...it is possible to have NO reference
> > to unneeded garbage.
> > 
> 
> What kind of "garbage" are you talking about? I am beginning to doubt
> my own reading capabilities because I just can't extract meaningful
> message from the above paragraph.
>
I'm afraid that the unneeded garbage was IBM's drive to dominate the
US and western European computer market by giving the (false)
impression that in a Turing sense, IBM computers had "artificial
intelligence" whereas other makes did not.
 
> One thing I'd like to "assert" is that many modern languages borrow
> much much more from Lisp family. The only thing Common Lisp obvious
> borrowed was the block structure introduced by Scheme. However, only
> one thing they can't steal: the powerful idea of data == program,

The problem is that this is not a new idea, was not invented by
McCarthy, not specially a powerful idea, nor even, an absolutely true
one.

John von Neumann invented the only usable sense in which data is
approximately identical to program, that of the stored program.  Ergo,
it had existed for about 15 years in 1960.

It is not powerful because in most cases, programs SHOULD NOT
self-manipulate.  The idea that it is powerful creates what to me is
the fundamental mistake of AI, and that is that we can dimly
understand expertise and intelligence, code through a glass darkly,
using self-modification, and then impose the results on labor and the
consumer whether they like it or not.

It is true that you can do useful things with let us say natural
language text, using software.  But it seems that the reliable and
above all explainable results are based on a theory (such as the power
and the limits of actual formal languages) which can in principle be
expounded to the average person.

You see, Arthur C. Clarke was wrong.  Sufficiently advanced technology
had better not be magic, especially let us say if your onboard
computer fails at 12000 feet.

In 99% of computations you need data != program.

> which is why Lisp macros are so so powerful just the way it is instead
> of the theory-laden *functor* approaches (which is still functional
> instead of Algol-like). The original "sin" of using linked lists,
> despite its simplicity, probably catapults that kind of thinking.

I do not want to make the stupid mistake of saying, you can do all dis
stuff in Visual Basic.  But the question is whether Lisp provides the
right sort of expressiveness.

It would be equivalently stupid of me to write some cool code in Lisp
and then say, why this here is the Philosopher's Stone, by gum.  Maud!
 I done found de Philosopher's Stone!

That's the alienation in which we confuse our G-d given abilities with
a product.

I think overall the claims for Lisp are overblown.  The only Euclid
whose ghost walks the battlements these days is the late Dijkstra, who
showed us that we did not need a "powerful tool", but to think harder.
> 
> 
> 
> 
> wni at attbi dot com
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avbuq5$rok$1@f1node01.rhrz.uni-bonn.de>
Edward G. Nilges wrote:

> My need today is not to find a single magic language, it's to write
> correct code, and I tend to look at languages like Lisp with a
> jaundiced eye as a result.  Procedural code can be sometimes easier to
> informally prove correct, and informal proofs of correctness are, to
> me, what's important.

Mathematically inclined people always make this fundamental mistake: A 
proof of correctness, regardless whether formal or informal, doesn't buy 
you anything with respect to correctness. A formal description of a 
program needs to be produced by human beings, and therefore it can 
contain bugs. Anything produced by a human being can contain bugs, so a 
proof of correctness just adds another source of potential bugs, but 
doesn't prove correctness. It does so only for toy examples.

The only way to show that a program does what you expect is by actually 
running it. I think that that's the main reason why hackers don't use 
formal techniques.

(I don't say that formal techniques are totally useless - they're useful 
if they provide a new perspective on a domain. But when they merely 
repeat what's already stated, they're useless - and that's what proofs 
of correctness usually do.)

>>One thing I'd like to "assert" is that many modern languages borrow
>>much much more from Lisp family. The only thing Common Lisp obvious
>>borrowed was the block structure introduced by Scheme. However, only
>>one thing they can't steal: the powerful idea of data == program,
> 
> 
> The problem is that this is not a new idea, was not invented by
> McCarthy, not specially a powerful idea, nor even, an absolutely true
> one.
> 
> John von Neumann invented the only usable sense in which data is
> approximately identical to program, that of the stored program.  Ergo,
> it had existed for about 15 years in 1960.

...but McCarthy made it accessible in a high-level language.

> It is not powerful because in most cases, programs SHOULD NOT
> self-manipulate.  The idea that it is powerful creates what to me is
> the fundamental mistake of AI, and that is that we can dimly
> understand expertise and intelligence, code through a glass darkly,
> using self-modification, and then impose the results on labor and the
> consumer whether they like it or not.

Please update yourself about topics like aspect-oriented programming, 
generative programming, template meta-programming, logic 
meta-programming, intentional programming, and so on. You will see that 
there is a need for program transformations in areas apart from AI. 
Please also notice that there is a difference program transformation and 
self-modifying code. Common Lisp doesn't allow for self-modifying code.


Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: sv0f
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <none-0601030914390001@129.59.212.53>
In article <··············@cartan.de>, Nils Goesche <······@cartan.de> wrote:

>Pascal Costanza <········@web.de> writes:
>
>> Mathematically inclined people always make this fundamental mistake:
>> A proof of correctness, regardless whether formal or informal,
>> doesn't buy you anything with respect to correctness. A formal
>> description of a program needs to be produced by human beings, and
>> therefore it can contain bugs. Anything produced by a human being
>> can contain bugs, so a proof of correctness just adds another source
>> of potential bugs, but doesn't prove correctness. It does so only
>> for toy examples.
>
>I happen to disagree.  It is a common fallacy of /non-mathematically/
>inclined people to derive from the fact that /they/ cannot produce
>correct proofs the incorrect conclusion that mathematicians can't,
>either.  But they can.  You can actually learn to produce correct
>proofs.  And you have to, if you want to become a mathematician.  A
>``mathematician�� who produces incorrect proofs is only a crank.

"Beware of bugs in the above code; I have only proved it correct,
not tried it." -- Donald Knuth
From: Tim Haynes
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <867kdiwfpc.fsf@potato.vegetable.org.uk>
Nils Goesche <······@cartan.de> writes:

> Pascal Costanza <········@web.de> writes:
>
>> Mathematically inclined people always make this fundamental mistake:
>> A proof of correctness, regardless whether formal or informal,
>> doesn't buy you anything with respect to correctness. A formal
>> description of a program needs to be produced by human beings, and
>> therefore it can contain bugs. Anything produced by a human being
>> can contain bugs, so a proof of correctness just adds another source
>> of potential bugs, but doesn't prove correctness. It does so only
>> for toy examples.
>
> I happen to disagree. It is a common fallacy of /non-mathematically/
> inclined people to derive from the fact that /they/ cannot produce
> correct proofs the incorrect conclusion that mathematicians can't,
> either. But they can. You can actually learn to produce correct proofs.
> And you have to, if you want to become a mathematician. A
> ``mathematician�� who produces incorrect proofs is only a crank.

I suspect you miss the point; proof of correctness is not proof that you
wanted it to do the Right Thing.

>> The only way to show that a program does what you expect is by actually
>> running it. I think that that's the main reason why hackers don't use
>> formal techniques.
>
> The only reason we don't prove ``correctness�� of real-world programs is
> that doing so would be much too complicated and cost too much time in
> practice. (And maybe even impossible because of the lack of complete,
> formal specifications)

Taken as a whole, yes it would become unwieldy. 
Pragmatically, it's important to test that the most-relied-upon functions /
modules do what they're supposed to. Mathematical correctness probably
should feature most there, then rely on a certain amount of induction for
the rest. I think.

~Tim
-- 
And we feel those flickering moments        ·······@stirfried.vegetable.org.uk
Like silk, the flags of our days            |http://spodzone.org.uk/
From: Casper H.S. Dik
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e198b40$0$49100$e4fe514c@news.xs4all.nl>
Nils Goesche <······@cartan.de> writes:

>I happen to disagree.  It is a common fallacy of /non-mathematically/
>inclined people to derive from the fact that /they/ cannot produce
>correct proofs the incorrect conclusion that mathematicians can't,
>either.  But they can.  You can actually learn to produce correct
>proofs.  And you have to, if you want to become a mathematician.  A
>``mathematician�� who produces incorrect proofs is only a crank.

Ah, but you can produce a correct proof and proof a that a program
matches the specification perfectly.

I think, however, that the point is that you cannot proof the
specification to be correct, only that the program is a correct
implementation of the specification.

>The only reason we don't prove ``correctness�� of real-world programs
>is that doing so would be much too complicated and cost too much time
>in practice.  (And maybe even impossible because of the lack of
>complete, formal specifications)

Indeed, since we don't even have a formal specification of most programs,
the exercise is futile.

And if the size of the specification approaches the size of the program,
there's little reason to assume that the specification will not also
have bugs.

Casper
-- 
Expressed in this posting are my opinions.  They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.
From: Nikodemus Siivola
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avcaou$2po3n$2@midnight.cs.hut.fi>
Casper H.S. Dik <··········@sun.com> wrote:

> I think, however, that the point is that you cannot proof the
> specification to be correct, only that the program is a correct
> implementation of the specification.

Also, a typo in a mathematical proof is just a typo -- a typo in a 
program is a bug. Furthermore, a *full* proof of a program requires 
that you prove everything from the chip up: OS, the implementation 
of the language you are using, etc.

But: proofs of *algorithmic* correctness are quite another thing: your 
implementation may still be bug ridden, but at least you can be certain
in the knowledge that the theory is sound. ;)

Strange how often the two become confused in CS: theory/algorithm
vs. practice/implementation. Perfectly sensible people are willing to 
believe that proving the correctness of programs is a magic bullet that
make for better software, but none of them entertain the idea that proof 
of a theorem's correctness ensures that student are able to apply the 
theorem. 8)=

Cheers,

  -- Nikodemus
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avc1cf$sbc$1@f1node01.rhrz.uni-bonn.de>
Nils Goesche wrote:

> You can actually learn to produce correct
> proofs.  And you have to, if you want to become a mathematician.  A
> ``mathematician�� who produces incorrect proofs is only a crank.

...but why don't you just learn to produce correct programs? You can do 
that without resorting to mathematical proofs. ;)

I am only half-joking: Complex mathematical proofs require a lot of 
scrutiny and time until you can be certain that they are correct, while 
programs can always be tested, also in intermediate states, regardless 
of their complexity.

Of course, the choice of the "right" approach depends on many factors, 
like what risks are involved when the deployed program fails to work, or 
what technique you feel comfortable with.

But I stand by my claim: A formal proof of correctness adds another 
source of potential bugs. (If you feel comfortable enough with formal 
techniques that might be just fine, but there is no magic involved!)


Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Nils Kassube
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <813co6dvct.fsf@darwin.lan.kassube.de>
Pascal Costanza <········@web.de> writes:

> But I stand by my claim: A formal proof of correctness adds another
> source of potential bugs. (If you feel comfortable enough with formal

It's obviously true that a proof can be wrong. However, it's like
writing a second version in another programming language -- if you
notice any problems (different results), you take a hard look at both:
the proof and the source code.
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avd65r$4c4$1@newsreader2.netcologne.de>
Nils Kassube wrote:
> Pascal Costanza <········@web.de> writes:
> 
> 
>>But I stand by my claim: A formal proof of correctness adds another
>>source of potential bugs. (If you feel comfortable enough with formal
> 
> 
> It's obviously true that a proof can be wrong. However, it's like
> writing a second version in another programming language -- if you
> notice any problems (different results), you take a hard look at both:
> the proof and the source code.

There are problems that are inherent in a problem domain and can remain 
undetected regardless of the language. At least that's true even for 
switching between different programming paradigms. I suspect that's also 
true for switching to formal methods.

I think that it's a big mistake to trust any kind of method more than to 
trust human beings. If someone says that they have applied rigorous 
mathematical methods and are certain by now that their system works, 
then you show trust in them (or their judgement) and not in those 
methods. That's the way it should be IMHO.

One of the fundamental problems of computer science and the IT industry 
is that the majority of people thinks that good software stems from the 
methods and tools programmers use, while in fact it's foremostly the 
programmers that count. If they are good then they know what the 
appropriate tools are for a given task; if they are not good no tool 
whatsoever will save them.


Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <adichcq0.fsf@ccs.neu.edu>
Pascal Costanza <········@web.de> writes:

> I think that it's a big mistake to trust any kind of method more than
> to trust human beings. If someone says that they have applied rigorous
> mathematical methods and are certain by now that their system works,
> then you show trust in them (or their judgement) and not in those
> methods. That's the way it should be IMHO.

But what if the human is at odds with the mathematical result?  Then
what/who do you trust?
From: Darius
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <a40a15cb.0301091630.62bdaaa4@posting.google.com>
Joe Marshall <···@ccs.neu.edu> wrote in message news:<············@ccs.neu.edu>...
> Pascal Costanza <········@web.de> writes:
> 
> > I think that it's a big mistake to trust any kind of method more than
> > to trust human beings. If someone says that they have applied rigorous
> > mathematical methods and are certain by now that their system works,
> > then you show trust in them (or their judgement) and not in those
> > methods. That's the way it should be IMHO.
> 
> But what if the human is at odds with the mathematical result?  Then
> what/who do you trust?

To succinctly quote Nils Kassube: Neither.
From: Pascal Bourguignon
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <877kdic4dj.fsf@thalassa.informatimago.com>
Nils Goesche <······@cartan.de> writes:

> Pascal Costanza <········@web.de> writes:
> 
> > Mathematically inclined people always make this fundamental mistake:
> > A proof of correctness, regardless whether formal or informal,
> > doesn't buy you anything with respect to correctness. A formal
> > description of a program needs to be produced by human beings, and
> > therefore it can contain bugs. Anything produced by a human being
> > can contain bugs, so a proof of correctness just adds another source
> > of potential bugs, but doesn't prove correctness. It does so only
> > for toy examples.
> 
> I happen to disagree.  It is a common fallacy of /non-mathematically/
> inclined people to derive from the fact that /they/ cannot produce
> correct proofs the incorrect conclusion that mathematicians can't,
> either.  But they can.  You can actually learn to produce correct
> proofs.  And you have to, if you want to become a mathematician.  A
> ``mathematician�� who produces incorrect proofs is only a crank.
> 
> > The only way to show that a program does what you expect is by
> > actually running it. I think that that's the main reason why hackers
> > don't use formal techniques.
> 
> The only reason we don't prove ``correctness�� of real-world programs
> is that doing so would be much too complicated and cost too much time
> in practice.  (And maybe even impossible because of the lack of
> complete, formal specifications)

I find the principal  pratical obstacle of proving real-world programs
is the lack of proof  and formal specification of underlying software,
starting from  the OS.   (Of course  to prove the  OS, you  would need
first to have  a proof and formal specification  of the processor, but
these exists for some processors at least).

Linux is  3e6 LOC  IIRC, that  means that we'd  need about  30 million
lines of  formal specifications and  proof.  Hopefully, the  user mode
API  is more  restricted  and  a formal  specification  should not  be
sensibly bigger than the sum of (2) and (3) man pages.

> Regards,
> -- 
> Nils G�sche

-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: Paolo Amoroso
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <6I4ZPvnV=F1aCXYN1H16CtiqHgu4@4ax.com>
On 5 Jan 2003 22:59:55 -0800, ···········@yahoo.com (Edward G. Nilges)
wrote:

> In 99% of computations you need data != program.

What was your point about 99.6% being meaningless, and the possibility of
exceeding six sigma? Never mind.


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3znqftcjb.fsf@cley.com>
* Edward G Nilges wrote:
> (please be advised that I have no special expertise in Lisp).

That is very, very clear.

> This is to tie the language, at a key point, to the vendor's hardware.
>  It was also done in Fortran about the same time.

> [twaddle removed]

> OK, pal, you've confirmed that there is garbage in Lisp which does not
> appear in Prolog and which in general does not belong in a programming
> language, which should be machine and vendor independent.  
> [further twaddle excised]

Perhaps you should explain how the names of a couple of abstract
operations on pairs make the language machine and vendor dependent?

--tim
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301052226.511dc5dd@posting.google.com>
Tim Bradshaw <···@cley.com> wrote in message news:<···············@cley.com>...
> * Edward G Nilges wrote:
> > (please be advised that I have no special expertise in Lisp).
> 
> That is very, very clear.
> 
> > This is to tie the language, at a key point, to the vendor's hardware.
> >  It was also done in Fortran about the same time.
>  
> > [twaddle removed]
>  
> > OK, pal, you've confirmed that there is garbage in Lisp which does not
> > appear in Prolog and which in general does not belong in a programming
> > language, which should be machine and vendor independent.  
> > [further twaddle excised]
> 
> Perhaps you should explain how the names of a couple of abstract
> operations on pairs make the language machine and vendor dependent?

Did not make that claim, claimed instead that the spirit of the
language is toward obfuscation as a result.  There is for example,
entirely too much discussion of how brilliant it was to divide up a 36
bit word, discussion that has no place in a programming language
design at all.
> 
> --tim
From: Paolo Amoroso
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <qI4ZPqUPyk7a5sgONLh9t5bRVs9s@4ax.com>
On 5 Jan 2003 13:09:50 -0800, ···········@yahoo.com (Edward G. Nilges)
wrote:

> field, and to me CAR and CDR are representative of a tendency in LISP
> (please be advised that I have no special expertise in Lisp).
> 
> This is to tie the language, at a key point, to the vendor's hardware.
>  It was also done in Fortran about the same time.
> 
> Whereas the Algol group pioneered what has turned out to be a better
> approach, and that is for a quasi-civic agency to be technically
> independent of computer vendors.  The Algol team devised the father of
> most usable modern languages including C because as an independent
> civic forum, they were immune to commercial pressures.

Can you provide any references to substantiate your claim that early Lisp
implementations, due to commercial pressure, tied the language to the
vendor's hardware?


> language, which should be machine and vendor independent.  But you
> interpret from the slag heap that results from the small injection of
> bias that no criticism of the result is possible.  With all do

Well, it would certainly help if the one who does the criticism had at
least some Lisp expertise.


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Harley
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ofKdnT4jQ7Go54ejXTWcpQ@giganews.com>
"Edward G. Nilges" <···········@yahoo.com> wrote in message
·································@posting.google.com...
> I do have a systematic tendency as a teacher and student formerly of
> philosophy who entered the computer racket 30 years ago as a
> draft-dodging scheme, to bring humanistic interpretation into the
> field, and to me CAR and CDR are representative of a tendency in LISP
> (please be advised that I have no special expertise in Lisp).
>
> This is to tie the language, at a key point, to the vendor's hardware.
>  It was also done in Fortran about the same time.
>
> Whereas the Algol group pioneered what has turned out to be a better
> approach, and that is for a quasi-civic agency to be technically
> independent of computer vendors.  The Algol team devised the father of
> most usable modern languages including C because as an independent
> civic forum, they were immune to commercial pressures.

Now here is a gem - someone who doesn't like Lisp because it is too
commercial.

-- Harley
From: Rahul Jain
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87r8br8ksk.fsf@localhost.localdomain>
···········@yahoo.com (Edward G. Nilges) writes:

> Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> based on the IBM 701's register structure and CAR and CDR are based on
> operations in its machine language.  As of now I stand by this claim.

Then you have no idea WHY they were chosen to be the names of the
operators. It was because McCarthy couldn't think of a better term at
the time and it stuck. ACCIDENT, not intent. We have FIRST and REST for
those who are squeamish about learning TLAs and not caring about their
expansion.

> You just misread me when you say "I don't see how...", because I said
> OO, not Lisp "destroys the need for linked lists", and I did not say
> that Lisp precludes the use of a stack.

Then you misread the CL standard and confused Lisp with some language
that isn't OO. Also, your statement is horribly strange, since you seem
to imply that stacks are somehow better than linked lists or that stacks
are better than lisp.

If you're trolling, and you give no indication that you're not, please
don't post anything on-topic in c.l.l ever again.

--
Rahul Jain
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301051746.f905d15@posting.google.com>
Rahul Jain <·····@rice.edu> wrote in message news:<··············@localhost.localdomain>...
> ···········@yahoo.com (Edward G. Nilges) writes:
> 
> > Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> > based on the IBM 701's register structure and CAR and CDR are based on
> > operations in its machine language.  As of now I stand by this claim.
> 
> Then you have no idea WHY they were chosen to be the names of the
> operators. It was because McCarthy couldn't think of a better term at
> the time and it stuck. ACCIDENT, not intent. We have FIRST and REST for
> those who are squeamish about learning TLAs and not caring about their
> expansion.

The fact that John McCarthy could not think of a meaningful name for
the operators is a *prima facie* indicator of a problem.

It is my opinion, based on experience, that whenever you are coding,
or designing a language, and you CANNOT find names (perhaps compound)
and complete sentences that describe, in a simple or complex fashion,
what you are doing, then at worst you are creating a design flaw, and
at worst, a bug.

I have this view because I have what in technical philosophical terms
would be called a "realist" ontology of programming in which the
referent of any program is the objects and concepts outside of
cyberspace to which the program MUST correspond.

I am therefore troubled about the very basis of Lisp for the same
reason Fortran's continued survival is to me problematic.

> 
> > You just misread me when you say "I don't see how...", because I said
> > OO, not Lisp "destroys the need for linked lists", and I did not say
> > that Lisp precludes the use of a stack.
> 
> Then you misread the CL standard and confused Lisp with some language
> that isn't OO. Also, your statement is horribly strange, since you seem
> to imply that stacks are somehow better than linked lists or that stacks
> are better than lisp.

None of this has anything to do with anything, champ.  You failed to
comprehend what I said, and I am gonna leave it at that.

> 
> If you're trolling, and you give no indication that you're not, please
> don't post anything on-topic in c.l.l ever again.

Have a nice day.
From: Paul F. Dietz
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <__6cnX5QxY26foWjXTWc3A@dls.net>
Edward G. Nilges wrote:

> I have this view because I have what in technical philosophical terms
> would be called a "realist" ontology of programming in which the
> referent of any program is the objects and concepts outside of
> cyberspace to which the program MUST correspond.

Well, there's the problem.  You're simply fucked up.

	Paul
From: Matthew Danish
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <20030105212922.G12928@lain.cheme.cmu.edu>
On Sun, Jan 05, 2003 at 05:46:35PM -0800, Edward G. Nilges wrote:
> The fact that John McCarthy could not think of a meaningful name for
> the operators is a *prima facie* indicator of a problem.

What John McCarthy could or could not think of is not meaningful to the
current state of Lisp, since there have been many more people involved
over the past 45 years.

> It is my opinion, based on experience, that whenever you are coding,
> or designing a language, and you CANNOT find names (perhaps compound)
> and complete sentences that describe, in a simple or complex fashion,
> what you are doing, then at worst you are creating a design flaw, and
> at worst, a bug.

I have asked this very question on this newsgroup before: What would be
a better name for CAR and CDR?  The best proposal I've heard is LEFT and
RIGHT (or LHS and RHS) but even that still implies some kind of physical
orientation to a cons (the printed form does have this), which sounds
ridiculous if you consider a cons to be an abstract idea (and who knows
how it may be implemented).  If you can think of better names, I would
like to hear it.

(CAR and CDR also have the advantage that abbreviations such as CADDR
for nested applications can be easily understood, though LHS and RHS have
similar.)

If you think that a data structure which simply contains two objects
(and has been accepted by many as a basis for constructing singly-linked
lists, among other things) is a design flaw, then your short-sightedness
is absolutely astonishing.

> I have this view because I have what in technical philosophical terms
> would be called a "realist" ontology of programming in which the
> referent of any program is the objects and concepts outside of
> cyberspace to which the program MUST correspond.

My apologies, but you're going to have to clarify this sentence for me.

> I am therefore troubled about the very basis of Lisp for the same
> reason Fortran's continued survival is to me problematic.

If you think that conses are the ``basis of Lisp'' then you seriously
need to start reading.

    Edward:
> > > You just misread me when you say "I don't see how...", because I said
> > > OO, not Lisp "destroys the need for linked lists", and I did not say
> > > that Lisp precludes the use of a stack.

  Rahul:
> > Then you misread the CL standard and confused Lisp with some language
> > that isn't OO. Also, your statement is horribly strange, since you seem
> > to imply that stacks are somehow better than linked lists or that stacks
> > are better than lisp.
> 
> None of this has anything to do with anything, champ.  You failed to
> comprehend what I said, and I am gonna leave it at that.

No, you fell in Rahul's little trap.  He said Lisp, in place of OO, for
a reason.

-- 
; Matthew Danish <·······@andrew.cmu.edu>
; OpenPGP public key: C24B6010 on keyring.debian.org
; Signed or encrypted mail welcome.
; "There is no dark side of the moon really; matter of fact, it's all dark."
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301060023.92f5f5d@posting.google.com>
Matthew Danish <·······@andrew.cmu.edu> wrote in message news:<·····················@lain.cheme.cmu.edu>...

I am going to bail out here, since Lisp is not one of my strong
languages (although by definition, this is not a requirement for
evaluating a language: you have to make these judgements prior to
use.)  Thanks to all for the information on Lisp.  I still believe
that it is not quite the level of intellectual achievement represented
in the Web pages that were linked.

Any poster is welcome to send me comments by email.
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <k7hihyi7.fsf@ccs.neu.edu>
Matthew Danish <·······@andrew.cmu.edu> writes:

> I have asked this very question on this newsgroup before: What would be
> a better name for CAR and CDR?  

MOM and DAD?
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA3F808F.255A0%gehlker@fastq.com>
On 1/6/03 12:25 PM, in article ············@ccs.neu.edu, "Joe Marshall"
<···@ccs.neu.edu> wrote:

> Matthew Danish <·······@andrew.cmu.edu> writes:
> 
>> I have asked this very question on this newsgroup before: What would be
>> a better name for CAR and CDR?
> 
> MOM and DAD?

Yin and Yang?



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Fred Gilham
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <u7znqbfsyl.fsf@snapdragon.csl.sri.com>
Joe Marshall <···@ccs.neu.edu> writes:

> Matthew Danish <·······@andrew.cmu.edu> writes:
> 
> > I have asked this very question on this newsgroup before: What would be
> > a better name for CAR and CDR?  
> 
> MOM and DAD?

The neat thing about CAR and CDR is that they start and end with the
same letter, AND one has a vowel and the other has a consonant.  That
lets you do the CADAR trick and have a good chance that it will be
pronounceable.  I think Lisp got lucky with CAR and CDR.

Just because something's old and takes a little getting used to
doesn't mean it's bad.

-- 
Fred Gilham                                        ······@csl.sri.com
And then [Clinton] turned to Hunter Thompson, of all people, and said
with wholehearted fervor, "We're going to put one hundred thousand new 
police officers on the street."
I was up all night persuading Hunter that this was not a personal
threat.                                              -- P. J. O'Rourke
From: faust
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <pt8r1v00ge0oe16s1ar8bjla7ubpdgkl0v@4ax.com>
>> > I have asked this very question on this newsgroup before: What would be
>> > a better name for CAR and CDR?  
>> 
>> MOM and DAD?
>
>The neat thing about CAR and CDR is that they start and end with the
>same letter, AND one has a vowel and the other has a consonant.  

FAT and FART ?

giving :
FARFATFAR or FATTRTRTRTR

-- 
natsu-gusa ya     / tsuwamono-domo-ga / yume no ato
summer grasses / strong ones          / dreams site
 
Summer grasses,
All that remains
Of soldier's dreams
(Basho trans. Stryk)
From: Fred Gilham
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <u7r8bmfe7i.fsf@snapdragon.csl.sri.com>
faust <·······@optushome.com.au> writes:

> >> > I have asked this very question on this newsgroup before: What would be
> >> > a better name for CAR and CDR?  
> >> 
> >> MOM and DAD?
> >
> >The neat thing about CAR and CDR is that they start and end with the
> >same letter, AND one has a vowel and the other has a consonant.  
> 
> FAT and FART ?
> 
> giving :
> FARFATFAR or FATTRTRTRTR

I don't think so.

(CAR (CDR (CAR X))) = (CADAR X)

By analogy your version would be

(FAT (FART (FAT X))) = (FAARAT X)


It would be pronounced Fa-ar-at.  It would have kind of a
middle-eastern ring to it....

:-)

-- 
Fred Gilham                                     ······@csl.sri.com
The vicissitudes of history, however, have not dissuaded them from
their earnest search for a "third way" between socialism and
capitalism, namely socialism.   --- Fr. Richard John Neuhaus
From: Erik Naggum
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3250902827314549@erik.naggum.no>
* Matthew Danish
| I have asked this very question on this newsgroup before: What would
| be a better name for CAR and CDR? [...]  If you can think of better
| names, I would like to hear it.

  All kinds of dual opposites would do.  Some well-known examples
  include up, down; top, bottom; charm, strange; west, east; north,
  south; in, out; blue, red; mit, psu; pgp, gpg; cbs, fox.

| (CAR and CDR also have the advantage that abbreviations such as
| CADDR for nested applications can be easily understood, though LHS
| and RHS have similar.)

  This is an anachronistic argument.  If the names were not similar in
  this way, the way to refer to other elements would also change.  For
  instance, you might indicate a position in a tree by the number of
  branches to skip on alternating left and right sides before you walk
  down a branch in the other direction.

| If you think that a data structure which simply contains two objects
| (and has been accepted by many as a basis for constructing singly-
| linked lists, among other things) is a design flaw, then your short-
| sightedness is absolutely astonishing.

  What of it?  Who cares?  He expressed his opinion and he appears
  extremely unlikely to change it.  Nobody would dream of taking it
  seriously if they had their wits about them.  Why do you?

| If you think that conses are the "basis of Lisp" then you seriously
| need to start reading.

  Why should he?  Geez, the guy is obviously so full of hot air and
  has such an inflated ego that it interferes with weather balloons
  and supersonic aircraft.

  Anyone who needs to figure out the comp.lang.lisp community should
  take notes.  The treatment Edward G. Nilges has received here shows
  a very deep and serious /concern/ for what people believe, not just
  about our (Common) Lisp, but in general.  Of course, telling people
  what they should believe is about as welcome as missionaries, but
  our conviction that We know what is true and They do not probably
  makes Us inherently oblivious to how They feel about it.

  I thought if I had inspired anything of importance, it was not to
  have and show concern for all things very strongly, but to detect
  idiots early.  If all this deeply felt concern for what other people
  believe is a product of my attempts to make people aware of what is
  /really/ true about Common Lisp and other things I know about, I am
  sorry.  You should have adopted my Moron Early Warning System, MEWS
  (so I am an ailurophile /and/ a punster; sue me), instead.

  But let us have another row over rudeness, again.  They're /fun/!

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Harley
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <s0SdnaKMlOt84oejXTWcpQ@giganews.com>
"Erik Naggum" <····@erik.naggum.no> wrote in message
·····················@erik.naggum.no...
> You should have adopted my Moron Early Warning System, MEWS
>   (so I am an ailurophile /and/ a punster; sue me), instead.

I'm certain that most people on the newsgroup are glad that nobody else here
has adopted Erik's MUSE...

-- Harley
From: Erik Naggum
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3250923230883759@erik.naggum.no>
* "Harley" <······@speakeasy.net>
| I'm certain that most people on the newsgroup [whatever]

  Wow!  Impressive!  Care to share how you gain such certainty?

  Or did you not know that such statements trigger the MEWS?

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Harley
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <9SWdnTfmYagHWYajXTWcqw@giganews.com>
"Erik Naggum" <····@erik.naggum.no> wrote in message
·····················@erik.naggum.no...
> * "Harley" <······@speakeasy.net>
> | I'm certain that most people on the newsgroup are amewsed by all this.
>
>   Wow!  Impressive!  Care to share how you gain such certainty?

Not really, no.

>   Or did you not know that such statements trigger the MEWS?

Mais bien sur, mon cher ami.

-- Harley
From: Charlton Wilbur
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87wulgwsnr.fsf@mithril.chromatico.net>
>>>>> "EN" == Erik Naggum <····@erik.naggum.no> writes:

    EN>   But let us have another row over rudeness, again.  They're
    EN> /fun/!

Actually, for the record, I like rows over rudeness, because some of
the vitriol is worth rereading several times.  Interesting discussion
of LISP would be preferable, of course, but rows over rudeness are far
more entertaining than yet another round of 'cn u h3lp m3 wif my
h0mew0rk?'

Charlton
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3r8brt44b.fsf@cley.com>
* Edward G Nilges wrote:
> The fact that John McCarthy could not think of a meaningful name for
> the operators is a *prima facie* indicator of a problem.

> It is my opinion, based on experience, that whenever you are coding,
> or designing a language, and you CANNOT find names (perhaps compound)
> and complete sentences that describe, in a simple or complex fashion,
> what you are doing, then at worst you are creating a design flaw, and
> at worst, a bug.

> I have this view because I have what in technical philosophical terms
> would be called a "realist" ontology of programming in which the
> referent of any program is the objects and concepts outside of
> cyberspace to which the program MUST correspond.

> I am therefore troubled about the very basis of Lisp for the same
> reason Fortran's continued survival is to me problematic.

And presumably you have real problems with things like, say, physics
and (much worse) maths.  Every time you see these awful terms like
`Hamiltonian', `Lagrangian', `Dirac delta function' and `Einstein
tensor' you must be really worried...  With flaws like these, no
wonder these fields have been so sterile.

--tim
From: Paolo Amoroso
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <Vo8ZPvUPwZiZCZtircaTnhPHVc43@4ax.com>
On 5 Jan 2003 17:46:35 -0800, ···········@yahoo.com (Edward G. Nilges)
wrote:

> I am therefore troubled about the very basis of Lisp for the same
> reason Fortran's continued survival is to me problematic.

If you think Lisp is so bad, why do you continue bothering with it? Just
stop dealing with it, and life will be happy again out there in Plato's
iperuranio[*]

Or are you going to start a crusade against Lisp? If so, comp.lang.lisp may
not be a good starting point.


Paolo

[*] Apologies for providing the Italian version, but I don't know the
English equivalent.
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Bruce Hoult
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <bruce-48256A.18311206012003@copper.ipg.tsnz.net>
In article <···························@posting.google.com>,
 ···········@yahoo.com (Edward G. Nilges) wrote:

> Rahul Jain <·····@rice.edu> wrote in message 
> news:<··············@localhost.localdomain>...
> > ···········@yahoo.com (Edward G. Nilges) writes:
> > 
> > > Chill, Rahul!  As far as I know, the architecture of Lisp is indeed
> > > based on the IBM 701's register structure and CAR and CDR are based on
> > > operations in its machine language.  As of now I stand by this claim.
> > 
> > Then you have no idea WHY they were chosen to be the names of the
> > operators. It was because McCarthy couldn't think of a better term at
> > the time and it stuck. ACCIDENT, not intent. We have FIRST and REST for
> > those who are squeamish about learning TLAs and not caring about their
> > expansion.
> 
> The fact that John McCarthy could not think of a meaningful name for
> the operators is a *prima facie* indicator of a problem.
> 
> It is my opinion, based on experience, that whenever you are coding,
> or designing a language, and you CANNOT find names (perhaps compound)
> and complete sentences that describe, in a simple or complex fashion,
> what you are doing, then at worst you are creating a design flaw, and
> at worst, a bug.

So do you like C++?  It calls the same things foo.first and foo.second.

-- Bruce
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ptrahylg.fsf@ccs.neu.edu>
···········@yahoo.com (Edward G. Nilges) writes:

> As far as I know, the architecture of Lisp is indeed based on the
> IBM 701's register structure and CAR and CDR are based on operations
> in its machine language.  As of now I stand by this claim.

Then you must not be familiar with the architecture of Lisp.

I suggest that you read the `History of Lisp' by John McCarthy.
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301050022.4d9aeae1@posting.google.com>
Rahul Jain <·····@rice.edu> wrote in message news:<··············@localhost.localdomain>...
> ···········@yahoo.com (Edward G. Nilges) writes:
> 
> > [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> > language that memorializes the register structure of a vintage machine
> > certainly is suspect as a useful tool.]
> > 
> > [Professor McCarthy deserves credit for inventing the linked list. 
> > However, it is far less important than the concept of the stack, which
> > was independently invented by Dijsktra and by Turing.  Furthermore, OO
> > languages destroy most of the need for linked lists.]
> > 
> > [I do so need to be on topic, you chaps.]
> 
> Do you take us for such fools that we'll think that you're smarter
> because of posting such ignorant drivel?
> 
> Hint: CAR and CDR are very simple to remember and have nothing to do
> with memorizing register structures of anything. Also, I don't see how
> Lisp destroys the need for linked lists or precludes the use of a stack.

This shows the effect of Richard Heathfield's libelous conduct and is
a documentation of same, for Rahul apparently is a follower of
Heathfield and for this reason did not even read my claims with any
diligence.
From: Geoffrey Summerhayes
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <fg0S9.214$SQ5.78361@news20.bellglobal.com>
"Edward G. Nilges" <···········@yahoo.com> wrote in message ·································@posting.google.com...
>
> This shows the effect of Richard Heathfield's libelous conduct and is
> a documentation of same, for Rahul apparently is a follower of
> Heathfield and for this reason did not even read my claims with any
> diligence.

Pfui. He probably just read your post. Trivializing evidence that fails
to support your claim, ad hominem, appeal to authority, followed by a
silly argument against the Lisp language.
Then you follow it up with this piece of nonsense. You can stop now, you
have my vote for 'troll of the month'.

--
Geoff
From: Rahul Jain
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87vg148cmx.fsf@localhost.localdomain>
Rahul Jain <·····@rice.edu> writes:

> ···········@yahoo.com (Edward G. Nilges) writes:
> 
> > [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> > language that memorializes the register structure of a vintage machine
> > certainly is suspect as a useful tool.]
[...]
> Hint: CAR and CDR are very simple to remember and have nothing to do
> with memorizing register structures of anything.

Oops, misread/interpreted that. In any case, how is using two terms from
some cpu's terminology "memorializing" it? Especially when there are
alternative operators with the same funcitonality and more descriptive
names (for that same linked-list data structure you said that Lisp makes
unnecessary)

--
Rahul Jain
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E1871EF.4070905@nyc.rr.com>
Edward G. Nilges wrote:

> [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> language that memorializes the register structure of a vintage machine
> certainly is suspect as a useful tool.]

Have you no appreciation for tradition? I love this stuff, this sense of 
connection to primordial Lisp. This image I get of some hacker 
off-handedly overloading car and cdr because nothing else sprung to 
mind, or because it was more self documenting, or because it would make 
some ASM code easier to port!

besides, what's in a name?(tm)

and lastlymost, men's jackets still have the buttons one the other side 
so we can remove them with one hand while weilding a sword with the 
other. Does that make them suspect as useful garments? And what did 
left-handed swordsmen do, or right-handed swordswomen? Hmm?

:)

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301051755.452d75c8@posting.google.com>
Kenny Tilton <·······@nyc.rr.com> wrote in message news:<················@nyc.rr.com>...
> Edward G. Nilges wrote:
> 
> > [On the subject of Lisp.  I have not used Lisp recently.  But ANY
> > language that memorializes the register structure of a vintage machine
> > certainly is suspect as a useful tool.]
> 
> Have you no appreciation for tradition? I love this stuff, this sense of 
> connection to primordial Lisp. This image I get of some hacker 
> off-handedly overloading car and cdr because nothing else sprung to 
> mind, or because it was more self documenting, or because it would make 
> some ASM code easier to port!

I appreciate tradition and am the proud owner of an IBM 1401 simulator
written in Delphi.  But I do not try to solve problems with it.
> 
> besides, what's in a name?(tm)
> 
> and lastlymost, men's jackets still have the buttons one the other side 
> so we can remove them with one hand while weilding a sword with the 
> other. Does that make them suspect as useful garments? And what did 
> left-handed swordsmen do, or right-handed swordswomen? Hmm?
> 
The problem with these common metaphors is that in them, a programmer
is imagined to be someone trying to solve a problem in his own
self-interest.

There are two problems with this picture.

First of all, an alternative view, for which I am in debt to Larry
Lessig, is that programming is less problem-solving than legislating
and structuring electronic law.

A programmer, whose style of work and personal hygiene presents a
problem set to himself and to others may well think of himself as a
problem solver.

Whereas a truly great programmer who writes The Law in the form of
provably bug free code is not so much a problem solver as an artiste.

Secondly, programmers are acting as agents of corporations and other
institutions.

Both of these facts would probably indicate that a programming
language is not a cozy smoking jacket in which we are supposed to feel
all comfy.

Instead, it should be more like Dosteoevsky's Grand Inquisitor and
help us to search in our hearts for where we fail in terms of bugs.
> :)
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <av9qdh$shf$1@newsreader2.netcologne.de>
Edward G. Nilges wrote:

> [Professor McCarthy deserves credit for inventing the linked list. 
> However, it is far less important than the concept of the stack, which
> was independently invented by Dijsktra and by Turing.  Furthermore, OO
> languages destroy most of the need for linked lists.]

The fact that Lisp can be defined by a meta-circular interpreter is the 
far more important contribution.


Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Edward G. Nilges
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <f5dda427.0301051750.9e76873@posting.google.com>
Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
> Edward G. Nilges wrote:
> 
> > [Professor McCarthy deserves credit for inventing the linked list. 
> > However, it is far less important than the concept of the stack, which
> > was independently invented by Dijsktra and by Turing.  Furthermore, OO
> > languages destroy most of the need for linked lists.]
> 
> The fact that Lisp can be defined by a meta-circular interpreter is the 
> far more important contribution.

I agree that this is very cool as seen in Structure and Interpretation
of Computer Programs.  Of course, Visual Basic can also be defined in
this way but with considerably less elegance.  If I understand your
point, that is.  Being.  Sein. Zeit.

You can "define" any Turing-complete language by writing an
interpreter for it in its own language.  Perhaps there is something I
do not understand about being meta-circular.  It does sound
impressive, almost Hegelian.  Can you explain?
> 
> 
> Pascal
From: Pascal Costanza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <avbv30$rom$1@f1node01.rhrz.uni-bonn.de>
Edward G. Nilges wrote:

> You can "define" any Turing-complete language by writing an
> interpreter for it in its own language.  Perhaps there is something I
> do not understand about being meta-circular.  It does sound
> impressive, almost Hegelian.  Can you explain?

I don't know what you do and don't understand about meta-circular 
interpreters. Please be more specific. Of course, you can write an 
interpreter for any language in almost any language, but that's not the 
point. (And I don't want to see the meta-circular interpreter for C++ ;-)

Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Pascal Bourguignon
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87k7hicabd.fsf@thalassa.informatimago.com>
···········@yahoo.com (Edward G. Nilges) writes:
> Pascal Costanza <········@web.de> wrote in message news:<············@newsreader2.netcologne.de>...
> > Edward G. Nilges wrote:
> > 
> > > [Professor McCarthy deserves credit for inventing the linked list. 
> > > However, it is far less important than the concept of the stack, which
> > > was independently invented by Dijsktra and by Turing.  Furthermore, OO
> > > languages destroy most of the need for linked lists.]
> > 
> > The fact that Lisp can be defined by a meta-circular interpreter is the 
> > far more important contribution.
> 
> I agree that this is very cool as seen in Structure and Interpretation
> of Computer Programs.  Of course, Visual Basic can also be defined in
> this way but with considerably less elegance.  If I understand your
> point, that is.  Being.  Sein. Zeit.
> 
> You can "define" any Turing-complete language by writing an
> interpreter for it in its own language.  Perhaps there is something I
> do not understand about being meta-circular.  It does sound
> impressive, almost Hegelian.  Can you explain?

The difference is that it's quite easy to do in LISP, therefore it can
be  used everytime  it's needed.   That  is, a  program generating  or
modifying its own functions.

Of course, languages like  C, Pascal, SmallTalk, Modula-[23], etc have
all  compilers/interperters written  in themselves.   But  designing a
program generating some of its functions  is hard (in all but LISP and
SmallTalk) because  you must generate  text, launch a compiler  on it,
and dynamically  load the compiled functions  to be able  to run them.
To do this  you need to generate and use  external files, use external
tools  such   a  the  compiler  and  dynamic   loader,  with  changing
conventions and usage depending on  the platform, so this procedure is
brittle and hard to implement.

The fact that the LISP reader  builds lists (s-expr) and that the LISP
eval  function can  execute directly  these  lists is  what makes  the
difference.   You could  as  well design  a  pascal or  C function  to
compile a  string into  a function (since  these language have  half a
function type (pascal) and a function type (C)), but parameter passing
would be awkward  because the types are processed  at compilation time
(it would be difficult to  match types compiled in the running program
with types  compiled at run-time)  and the access to  global variables
other functions would be harder too.

To see the  difference in complexity, just compare the  size of a lisp
interpreter  and  that of  another  compiler  with  its libraries  and
linker. 


-- 
__Pascal_Bourguignon__                   http://www.informatimago.com/
----------------------------------------------------------------------
There is a fault in reality. Do not adjust your minds. -- Salman Rushdie
From: gswork
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <81f33a98.0301060041.442bbea4@posting.google.com>
···········@yahoo.com (Edward G. Nilges) wrote in message news:<····························@posting.google.com>...
> ······@eton.powernet.co.uk (Strangely Placed) wrote in message news:<····························@posting.google.com>...
> > Kenny Tilton <·······@nyc.rr.com> wrote in message news:<··············@nyc.rr.com>...
> > > Edward G. Nilges wrote:
> > > > Richard's behavior had the effect of making him and myself the
> > > > discussion "stars" when in fact I wanted to...
> > > 
> > > Did it ever occur to you to simply not respond to the guy? Or to respond 
> > > with ten words max?
> > 
> > It occurred to /me/ (eventually). At the urging of comp.programming, I
> > killfiled Mr Nilges. There is only one unfortunate consequence of this
> > action, which is that he is now free to spew any old rubbish as if it
> > were fact. I urge you to check anything he tells you (even something
> > as simple as today's date!) before believing it. I've just been
> > reading some of his stuff in groups.google.com (a friend clued me in
> > to this latest idiocy of his via email), and the nonsense he is
> > spewing about me in this thread just beggars belief. The facts of any
> > exchange between myself and Mr Nilges (except for a very brief and
> > frosty email conversation which he will not allow me to disclose,
> > presumably because it embarrasses him) are available for you to check
> > via Google.
> 
> Posters, be aware that the poster here is Richard Heathfield, the
> instigator of a bullying campaign, who has changed his handle possibly
> to conceal his identity, although it does appear below.

I wondered what happened to balloon this thread recently.    As to ID,
I think it's just an 'on the road' handle for RH, or something, if it
were subterfuge I'm sure it wouldn't be quite such a give away!

> He said it was overlong, and so broad in some political claims as to
> lie outside the charter of comp.risks.  He asked me to write a smaller
> article focusing on the technical objections and their immediate
> political implications.  I did so and he released the article, which
> is being considered for publication in a journal on these issues.

I read the abbreviated one too, it reads fine - infact I'm tempted to
believe you'd have stimulated some discussion on c.p. too with that
one.

> In consequence, Richard, I decided to not present you with another
> opportunity to engage in libel and disruption, and won't be responding
> to any comments on the comp.programming post.
> 
> You are instead encouraged to send any criticisms as to the revelance
> or correctness of my views as expressed to Peter or to me by email. 
> Contrary to what gswork implied, these posts, if acceptable and only
> if, will appear as line comments in the same style as an unmoderated
> group.

What I meant to imply was that a similarly styled [to these groups]
discussion was unlikely to occur, rather than would certainly not
occur.   The history of that group and it's traffic level would
indicate this.  Still, if your post causes a break with that pattern
then all credit to you.

I note you responded in a couple of other comp.* threads too recently,
joining in discussions other than those related to your own posts is
indeed a contributive and meritorious thing to do, will we see a
'semi-regular' EGN contributing on assorted threads on algorithms and
the like in 2003?  Perhaps we can hope for less distraction in order
to facilitate this.

> 
> What is of general interest is the supposed liberal "bias" of
> comp.risks that several neoconservative people have remarked upon.

I must have missed some posts!  comp.risks seems relatively unbiased
to me.
From: Wade Humeniuk
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BUHQ9.73996$xp4.2910811@news1.telusplanet.net>
"Kenny Tilton" <·······@nyc.rr.com> wrote in message ···················@nyc.rr.com...
> 
> 
> Edward G. Nilges wrote:
> > Richard's behavior had the effect of making him and myself the
> > discussion "stars" when in fact I wanted to...
> 
> Did it ever occur to you to simply not respond to the guy? Or to respond 
> with ten words max?
> 

It does not matter whether it occurred to him or not.  He did
what he did, he had his reasons.  One cannot determine what
the end result will be of any particular course of action.  What Edward did 
might have been the "best" course of action.  I think what is 
important is to give others the freedom  to follow a course of action 
(based on some absolute code of conduct, like, you cannot kill anyone).  
Even if what Edward did was not the "best" course, he did take action, 
and he has TAKEN responsibility for his behaviour.  Just like Erik says, 
some people blame others for their own personal behaviour.

> I think a huge problem in these deals is that people (quite reasonably!) 
> read something attacking them or what they say and feel that unless they 
> respond the community will judge them the loser. I wager what realy 
> happens is that the community only does that when the attack is pretty 
> solid. And on the flip side, ignoring crap wins the ignorer the 
> admiration and gratitude of everyone for not perpetuating some silly, 
> typical NG tit-for-tat flamewar.
> 

You think the problem comes down to caring what others think?
Loser vs. Winner?

Wade 
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E1356B9.8040701@nyc.rr.com>
Wade Humeniuk wrote:
> "Kenny Tilton" <·······@nyc.rr.com> wrote in message ···················@nyc.rr.com...
> 
>>
>>Edward G. Nilges wrote:
>>
>>>Richard's behavior had the effect of making him and myself the
>>>discussion "stars" when in fact I wanted to...
>>
>>Did it ever occur to you to simply not respond to the guy? Or to respond 
>>with ten words max?
>>
> 
> 
> It does not matter whether it occurred to him or not.  He did
> what he did, he had his reasons.

<heh-heh> I know the phrase "did it ever occur to you" usually means 
"You should have", so I should have picked a different phrase because I 
was honestly wondering whether that course had been considered and rejected.

>>I think a huge problem in these deals is that people (quite reasonably!) 
>>read something attacking them or what they say and feel that unless they 
>>respond the community will judge them the loser. ....
> 
> 
> You think the problem comes down to caring what others think?
> Loser vs. Winner?

Absolutely. These folks often have each other e-mail addresses. And 
sometimes are exchanging private mail while continuing the public joust. 
And will respond to messages like "take it outside" by explicitly 
stating that their articles are meant for everyone, not just their 
antagonist. Another bit of circumstantial evidence: there is no 
discussion going on after a very short time, rather minute (unfavorable) 
dissection of each other's articles looking for points to score.

We're the judges. :)

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Kaz Kylheku
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <cf333042.0301010804.600b6a0b@posting.google.com>
Chris Gehlker <·······@fastq.com> wrote in message news:<······················@fastq.com>...

Claim:

> The VB programmer that I know personally is neither arrogant, stupid nor an
> idiot. 

Supporting evidence, supposedly:

> She is very aware that if she lost her job as a VB programmer, she
> would lose her visa and, considering her nationality, most of her life
> expectancy.

The awareness that one must cling to a job for security is not a huge
intellectual leap. You have written nothing to reveal that it's an
important job that is being done well by someone competent.

But rest assured that if I had a heart, it would bleed for her
predicament.
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA387CC9.251FF%gehlker@fastq.com>
On 1/1/03 9:04 AM, in article
····························@posting.google.com, "Kaz Kylheku"
<···@ashi.footprints.net> wrote:

> Chris Gehlker <·······@fastq.com> wrote in message
> news:<······················@fastq.com>...
> 
> Claim:
> 
>> The VB programmer that I know personally is neither arrogant, stupid nor an
>> idiot. 
> 
> Supporting evidence, supposedly:
> 
>> She is very aware that if she lost her job as a VB programmer, she
>> would lose her visa and, considering her nationality, most of her life
>> expectancy.

Not supporting evidence but rather
1) A crude appeal to the emotions.
2) Some explanation for why I bothered to reply.
3) An example showing that not everyone who uses VB does so by choice.

> The awareness that one must cling to a job for security is not a huge
> intellectual leap. You have written nothing to reveal that it's an
> important job that is being done well by someone competent.

I certainly wasn't trying to demonstrate that. In fact, she assures me that
her job is menial, and that she is studying in order to *become* competent.
My point was simply that not all VB programmers so arrogant as to argue that
their language is superior to Lisp or anything else.

My problem is that I still don't know whether we disagree here or not and
I'll tell you why. You said:

> VB-pushing idiots have nothing to be arrogant about, and are too
> stupid to realize when someone proposes some wrong way of doing
> something. Everything seems like a neat idea when all you know is VB

There are two ways to take that. If you meant something like:

'People who assert that a language they know is superior in any respect to a
language that they don't know are being foolish.'

Then I agree completely. I tried to indicate my agreement by asserting that
I would become annoyed with my VB programmer friend if she advocated for VB.
Maybe that wasn't a clear as I'd hoped.

If you meant something like:

'Most VB programmers are fools.'

Then I disagree. *I* would be foolish to denigrate a language or it's users
when I know very little about it. What little I do know suggests that that
it was designed as an automation language for the MS Office suite and maybe
a few other MS products. I have no reason to doubt that it does that well. I
wouldn't even go so far as to assert that automating MS Office is
unimportant. I do believe that it is deadly boring and intellectually
undemanding but only because my friend who does it all day tells me so.


> But rest assured that if I had a heart, it would bleed for her
> predicament.

But I suspect that you do and it does.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Frank A. Adrian
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <t%GQ9.433$V05.31984@news.uswest.net>
Chris Gehlker wrote:

> What little I do know suggests that that
> it was designed as an automation language for the MS Office suite and
> maybe a few other MS products. I have no reason to doubt that it does that
> well.

Having used it I can attest that it doesn't do it that well.  All-in-all, 
it's one of the most unpleasant programming languages and environments I've 
used.  I'm glad I don't have to use it much anymore.

faa
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA389665.2520A%gehlker@fastq.com>
On 1/1/03 12:07 PM, in article ···················@news.uswest.net, "Frank
A. Adrian" <·······@ancar.org> wrote:

> Chris Gehlker wrote:
> 
>> What little I do know suggests that that
>> it was designed as an automation language for the MS Office suite and
>> maybe a few other MS products. I have no reason to doubt that it does that
>> well.
> 
> Having used it I can attest that it doesn't do it that well.  All-in-all,
> it's one of the most unpleasant programming languages and environments I've
> used.  I'm glad I don't have to use it much anymore.

Now I do have reason to doubt. Are there any alternatives that you prefer. I
know there is a thing derived from Ruby that's popular in Japan and that a
few people outside Japan seem to like but there doesn't seem to be any
documentation for it in English.

Would it make sense to write Office automation extension to Lisp? There do
seem to be jobs in the office automation area, even if they aren't well
paid.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Marc Spitzer
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <86smwctwp8.fsf@bogomips.optonline.net>
Chris Gehlker <·······@fastq.com> writes:

> On 1/1/03 12:07 PM, in article ···················@news.uswest.net, "Frank
> A. Adrian" <·······@ancar.org> wrote:
> 
> > Chris Gehlker wrote:
> > 
> >> What little I do know suggests that that
> >> it was designed as an automation language for the MS Office suite and
> >> maybe a few other MS products. I have no reason to doubt that it does that
> >> well.
> > 
> > Having used it I can attest that it doesn't do it that well.  All-in-all,
> > it's one of the most unpleasant programming languages and environments I've
> > used.  I'm glad I don't have to use it much anymore.
> 
> Now I do have reason to doubt. Are there any alternatives that you prefer. I
> know there is a thing derived from Ruby that's popular in Japan and that a
> few people outside Japan seem to like but there doesn't seem to be any
> documentation for it in English.
> 
> Would it make sense to write Office automation extension to Lisp? There do
> seem to be jobs in the office automation area, even if they aren't well
> paid.
> 

Well since MS claims that everything is driven by a com/activex api you
should be able to use anything that can access com/activex.  ACL, Corman 
and Lispworks all can do this, as far as I know.  Other languages would
also work, TLC, Python, Ruby and Perl for example and of course MS stuff.

One of the issues about office automation is that, from what I have seen,
it is generally very simple work and CL would probably be over kill for it.
You would probably have a better shot with one of the more popular scripting
languages.  If you were looking for work in that area.   

marc

ps I have a friend who did a lot of work with MS products and VB.  He said
that the real problem was 2 fold, first learning all the apis for the 
products and then learning which ones did not work as documented/how they
did behave.
From: Frank A. Adrian
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <YYQQ9.65$7I5.220717@news.uswest.net>
Chris Gehlker wrote:

> Are there any alternatives that you prefer.

Sadly, no.  If you want tight integration with the application, VB is the 
only thing that's out there that comes close.  I am not saying that it is 
unusable, what I am saying is that it is an unpleasant and underpowered 
language and that the bugs within the environment drive one crazy.

> Would it make sense to write Office automation extension to Lisp?

Probably not.  Most people wouldn't use it and the work to integrate the 
language wouldn't be worth it.  This is a shame, fo it would be a much more 
pleasant language to program in.

faa
 
From: Christopher Browne
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <auvr8u$an76q$1@ID-125932.news.dfncis.de>
Chris Gehlker <·······@fastq.com> wrote:
> On 1/1/03 12:07 PM, in article ···················@news.uswest.net, "Frank
> A. Adrian" <·······@ancar.org> wrote:
>
>> Chris Gehlker wrote:
>> 
>>> What little I do know suggests that that
>>> it was designed as an automation language for the MS Office suite and
>>> maybe a few other MS products. I have no reason to doubt that it does that
>>> well.
>> 
>> Having used it I can attest that it doesn't do it that well.  All-in-all,
>> it's one of the most unpleasant programming languages and environments I've
>> used.  I'm glad I don't have to use it much anymore.
>
> Now I do have reason to doubt. Are there any alternatives that you prefer. I
> know there is a thing derived from Ruby that's popular in Japan and that a
> few people outside Japan seem to like but there doesn't seem to be any
> documentation for it in English.
>
> Would it make sense to write Office automation extension to Lisp? There do
> seem to be jobs in the office automation area, even if they aren't well
> paid.

I doubt that the problems in this area are driven primarily by
deficiencies in the programming language.

Back when I played with Lotus Command Language (one of the most
unpleasant things I have ever had the displeasure of using), the
serious problems weren't even with the language, but rather with the
overall model of "system control."

LCL sucked quite incredibly badly, but replacing it with Lisp wouldn't
help appreciably.

In recent threads, the has assortedly been bashing and defending of
TeX; let me point out that there has been one would-be alternative
come along, called Lout.  It used a rather more Algol-like programming
scheme, with a complete absence of macros.  It never caught on,
because the problems people were having with TeX weren't actually
being solved by Lout.  It turns out that it's better to fight with the
pains of the TeX macro system and get benefit of the decent rendering
and the existing macro packages than it is to jump to Lout and
discover that you have to translate over your whole bibliography
database, and maybe have to rewrite quite a lot of code to get
references to work the way you want.

Heading back to Office, supposing you replace "VBA" with something
more palatable, you're still left with the ugliness of:
 a) Praying that there might be acceptable APIs to let you get at the
    actual application features, and
 b) Figuring out where to stick the code that won't be just as ugly as
    it was with VBA.

If the COM functions are ugly, and their use error-prone, it may be
that VBA makes that /worse/, but it's hardly likely that Lisp could
make it better.
-- 
If this was helpful, <http://svcs.affero.net/rm.php?r=cbbrowne> rate me
http://cbbrowne.com/info/lisp.html
"A program invented (sic) by a Finnish computer hacker and handed out free
in 1991 cost investors in Microsoft $11 billion (#6.75 billion) this week."
-- Andrew Butcher in the UK's Sunday Times, Feb 20th, 1999
From: Basile STARYNKEVITCH
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <q5r1y3w5ut1.fsf@hector.lesours>
>>>>> "Christopher" == Christopher Browne <········@acm.org> writes:


    Christopher> In recent threads, the has assortedly been bashing
    Christopher> and defending of TeX; let me point out that there has
    Christopher> been one would-be alternative come along, called
    Christopher> Lout.  It used a rather more Algol-like programming
    Christopher> scheme, with a complete absence of macros.  It never
    Christopher> caught on, because the problems people were having
    Christopher> with TeX weren't actually being solved by Lout.
    Christopher> [...]

Sorry for this off-topic correction...

Actually Lout (see http://snark.ptc.spbu.ru/~uwe/lout/ for details)
does have macros, and use them. Lout macros is basically a textual
replacement (without arguments).

To cite the Lout Expert Guide by Jeff Kingston:

``
               Macros provide a means of defining symbols which stand
          for a sequence of textual units rather than an object.  For
          example, the macro definition

               macro  @PP  {  //1.3vx  2.0f @Wide  &0i }

          makes Lout replace the symbol @PP by the given textual
          units before assembling its input into objects.  A similar
          macro to this one is used to separate the paragraphs of
          the present document.  The enclosing braces and any spaces
          adjacent to them are dropped, which can be a problem:  @PP2i
          has result //1.3vx 2.0f @Wide &0i2i which is erroneous.


''

So basically Lout macros are verbatim replacement without arguments
and they are extensively used!

I am not sure I fully follow Christopher's argument on why Lout never
caught on. I would say that it did came too late after TeX, and also
that its typography is not always as good as TeX's or LaTeX's
one. Perhaps the fact that Lout use PostScript native fonts and not
its own ones is also a reason. I still happen to occasionnaly use
Lout. The main feature I liked in it is that it is much smaller in
size that LaTeX!

Jeff Kingston is starting to work on Lout's successor [NonPareil]
which should have lots of features LaTeX (and Lout) don't have.


A Happy New Year to all


-- 

Basile STARYNKEVITCH         http://starynkevitch.net/Basile/ 
email: basile<at>starynkevitch<dot>net 
alias: basile<at>tunes<dot>org 
8, rue de la Fa�encerie, 92340 Bourg La Reine, France
From: faust
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <nq4e1vc0moh0djkte2dpqsh81u3a7nsb2c@4ax.com>
 Basile STARYNKEVITCH <·········@SPAM+starynkevitch.net.invalid> ,
emitted these fragments:

>I am not sure I fully follow Christopher's argument on why Lout never
>caught on. I would say that it did came too late after TeX, and also
>that its typography is not always as good as TeX's or LaTeX's
>one. 

It's typography left much to be desired.
In any case, Jeff's page says that it is in maintenance mode now.


>Jeff Kingston is starting to work on Lout's successor [NonPareil]
>which should have lots of features LaTeX (and Lout) don't have.

NonPareil is still vapourware.
There is one thesis on it.
But no software.

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Coby Beck
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <auqqdq$3j$1@otis.netspace.net.au>
"Kaz Kylheku" <···@ashi.footprints.net> wrote in message
·································@posting.google.com...
> voice. When you know you are right, it's extremely entertaining to don
> an air of complete arrogance, and watch the little psycho-analyzing
> idiots fall for it and get their feathers all ruffled.

Well, that pretty much defines "troll" and has little utility wether your
goal is to teach or to learn.

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Marc Spitzer
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <86vg1bx8ue.fsf@bogomips.optonline.net>
"Coby Beck" <·····@mercury.bc.ca> writes:

> "Kaz Kylheku" <···@ashi.footprints.net> wrote in message
> ·································@posting.google.com...
> > voice. When you know you are right, it's extremely entertaining to don
> > an air of complete arrogance, and watch the little psycho-analyzing
> > idiots fall for it and get their feathers all ruffled.
> 
> Well, that pretty much defines "troll" and has little utility wether your
> goal is to teach or to learn.

As far as I know part of the definition of troll is that they provide
nothing of value.  A correct answer is something of value, even if you 
are behaving as a complete asshole when providing it.  Ergo not a troll

marc

> 
> --
> Coby Beck
> (remove #\Space "coby 101 @ bigpond . com")
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <u1gvjhs6.fsf@ccs.neu.edu>
Chris Gehlker <·······@fastq.com> writes:

> But fear/anger or concern/annoyance if you prefer milder terms, are at root
> the same emotion. At least the physiologists & social psychologists think
> so.

Hmmm.  I'm annoyed at email spam.  I get angry at the people who send
it.  I don't think I fear it.

I fear that my kids may be injured in a car accident caused by a drunk
driver.  It would be more than an annoyance if such were to happen.

Perhaps at some `root level' the emotions are the `the same', but
I'd be annoyed at anyone that says the emotions are indistinguishable
beyond that.  And I don't fear physiologists and social psychologists
who may contradict me.
From: Gabe Garza
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87ptrjw8bd.fsf@ix.netcom.com>
Chris Gehlker <·······@fastq.com> writes:

> This is all speculation anyway. I did do one experiment to test it though. I
> searched a couple of Visual Basic groups for any signs that the posters were
> as angry as the typical poster here. They clearly are not.

If Common Lisp was as widely accepted as Visual Basic, maybe the
posters here would be holding hands and dancing under the rainbow
too. :)  

I think that a lot of the "anger" isn't so much because some people
dismiss and misunderstand Lisp, but because the majority of the
industry (and even a lot of academic realms) dismiss it--if not
explicitly then at least implicitly with indifference.  Someone coming
here and refusing to either listen or bite their tongue is just adding
insult to injury.  I think Lisp's lack of popularity is a major sore
point--and strangers walking up and poking at sores isn't welcome.
What makes Lisp's lack of popularity even more painful is that it is
symptomatic of major flaws with the computer world as a whole...

This is (of course!) based on my own opinion.  I don't mean to speak
for anyone else...

Gabe Garza
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA365F5A.25081%gehlker@fastq.com>
On 12/30/02 1:32 PM, in article ··············@ix.netcom.com, "Gabe Garza"
<·······@ix.netcom.com> wrote:

> Chris Gehlker <·······@fastq.com> writes:
> 
>> This is all speculation anyway. I did do one experiment to test it though. I
>> searched a couple of Visual Basic groups for any signs that the posters were
>> as angry as the typical poster here. They clearly are not.
> 
> If Common Lisp was as widely accepted as Visual Basic, maybe the
> posters here would be holding hands and dancing under the rainbow
> too. :)  

It's weird though. I don't think that very many people think VB is an
especially good language. Around here it's what gets taught, along with Java
and C#, at the trade schools and junior colleges. Lisp is still the language
at the university. I see below that you think that changing in some academic
areas. When MS first announced .Net there was a perception among the VB
programmers that they were abandoning VB altogether. Then there was real
fear and anger.

> I think that a lot of the "anger" isn't so much because some people
> dismiss and misunderstand Lisp, but because the majority of the
> industry (and even a lot of academic realms) dismiss it--if not
> explicitly then at least implicitly with indifference.

This seems pretty insightful. A lot of people seem to think that Lisp is a
great language but it's not the right language for what they are trying to
do. They *may* be right in some cases. My whole experience with VB was
adding some additional functions to an Excel spreadsheet. It was easier to
figure out how to do that in VB from scratch than to figure out how to do
that in a real language. As I recall, I was paging through the manual
looking for a way to extend the spreadsheet language and discovered 2 things
in about 10 minutes.

1) If I wanted to use anything but VB I had to send MS serious $ for a
developers kit.

2) There was an example close enough to what I was trying to do that I could
copy and edit it w/o really understanding it.

I don't know where I'm going with this except to suggest half way seriously
that a lot of the "programming" done is VB is pretty repetitive stuff and
that maybe we should all study it with a view to writing a VB programmer in
Lisp.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Ingvar Mattsson
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87n0mmo57b.fsf@gruk.tech.ensign.ftech.net>
Chris Gehlker <·······@fastq.com> writes:

> On 12/30/02 1:32 PM, in article ··············@ix.netcom.com, "Gabe Garza"
> <·······@ix.netcom.com> wrote:
> 
> > Chris Gehlker <·······@fastq.com> writes:
> > 
> >> This is all speculation anyway. I did do one experiment to test it though. I
> >> searched a couple of Visual Basic groups for any signs that the posters were
> >> as angry as the typical poster here. They clearly are not.
> > 
> > If Common Lisp was as widely accepted as Visual Basic, maybe the
> > posters here would be holding hands and dancing under the rainbow
> > too. :)  
> 
> It's weird though. I don't think that very many people think VB is an
> especially good language. Around here it's what gets taught, along with Java
> and C#, at the trade schools and junior colleges. Lisp is still the language
> at the university. I see below that you think that changing in some academic
> areas. When MS first announced .Net there was a perception among the VB
> programmers that they were abandoning VB altogether. Then there was real
> fear and anger.

Do the people in the VB group(s) get random people starting to post to
the group about how fundamentally flawed their language is and how
they would be much better off using something *vastly* better instead
and how *oibviously* the VB specification/reference is incorrect in
that it states <fact>?

That is, to my eye, what is currently the root cause of quite a few of
the harsh feelings. The rest seems to be harsh feelings about harsh
feelings.

//Ingvar
-- 
(defun m (a b) (cond ((or a b) (cons (car a) (m b (cdr a)))) (t ())))
From: Nils Kassube
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <81u1gt0wro.fsf@darwin.lan.kassube.de>
Chris Gehlker <·······@fastq.com> writes:

> do. They *may* be right in some cases. My whole experience with VB was
> adding some additional functions to an Excel spreadsheet. It was easier to
[...]
> 2) There was an example close enough to what I was trying to do that I could
> copy and edit it w/o really understanding it.

Sometimes (like after a few beers) I believe that this attitude is one of
the reasons of the current recession. BTW: Did you work for Enron?

Shit, I've seen too many economy classes where most students took a
pattern-matching approach to the exams, even worse: the teachers
advocated this behaviour, too. 
From: Erik Naggum
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3250272666307367@naggum.no>
* Chris Gehlker
| This is all speculation anyway. I did do one experiment to test it
| though. I searched a couple of Visual Basic groups for any signs
| that the posters were as angry as the typical poster here. They
| clearly are not.

  Did that include the newbies who rush in to make bold claims that
  the standard is broken and they have to tell everyone what it says
  because before they came along, all Common Lisp programmers and
  language lawyers were living in the dark?

  I have never seen the kind of pestering idiots we suffer here in
  comp.lang.lisp regularly anywhere else on this planet.  When I
  spent a lot more time trying to explain SGML to people, we had
  people who came rushing in to proclaim that so and so was broken
  and badly designed, but they actually /listened/ when they were
  told how and why it became that way and how it worked and how they
  could get what they wanted.  Peculiar to the comp.lang.lisp newbie
  is that he does not listen and does not stop talking.  And every
  one of them have evidently read previous posts where they find fuel
  for their massive confusion as well as reason to resist correction.

  Fighting off the lunatic fringe that invades comp.lang.lisp is like
  debunking physics lunatics like Santilli and his "magnecules" or
  discussing medicine in the presence of "alternative" medicinemen.
  The core problem is that some people will insist that whatever they
  heard first must be truer than anything they hear later, and they
  are just as obnoxious if they first hear some bogosity as if they
  think they grok Common Lisp from tinkering with Emacs Lisp.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3r8bz8a39.fsf@cley.com>
* Chris Gehlker wrote:

> But fear/anger or concern/annoyance if you prefer milder terms, are at root
> the same emotion. At least the physiologists & social psychologists think
> so.

Erm.  Then they should get some new brains, because theirs obviously
aren't working very well.  When I'm angry at my cat does that mean I'm
afraid of her?  When I'm annoyed by some fool on cll does that mean
I'm afraid of them?  Come on, *think*!  Don't listen to what some fool
social `scientist' says, think what these states actually *feel* like.

--tim
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA361416.2506A%gehlker@fastq.com>
On 12/30/02 2:34 PM, in article ···············@cley.com, "Tim Bradshaw"
<···@cley.com> wrote:

> * Chris Gehlker wrote:
> 
>> But fear/anger or concern/annoyance if you prefer milder terms, are at root
>> the same emotion. At least the physiologists & social psychologists think
>> so.
> 
> Erm.  Then they should get some new brains, because theirs obviously
> aren't working very well.  When I'm angry at my cat does that mean I'm
> afraid of her?  When I'm annoyed by some fool on cll does that mean
> I'm afraid of them?  Come on, *think*!  Don't listen to what some fool
> social `scientist' says, think what these states actually *feel* like.
> 
> --tim

I think they do actually feel very similar. When I get angry at my wife's
cat (my cat is perfect, I never get angry at him) it's because I see her
starting to squat outside the litter box or I find evidence that she missed
it. In either case, I *fear* that I am about to have an unpleasant
experience. Fear is a very strong word but it seems to me that it differs in
degree from such emotions as concern or worry, not in essence. Similarly
anger is a very strong word but it seems to me that annoyance isn't an
entirely different emotion, just a weaker form.

Obviously, when you get annoyed at somebody on the list you are in literal
fear of them as an individual. But isn't it possible that part of your
annoyance comes from the concern that the views they are expressing are
widely held?  



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3adindrqf.fsf@cley.com>
* Chris Gehlker wrote:

> I think they do actually feel very similar. 

Oh well, maybe for some people (perhaps people who live in some
permanent terror).  Not for me: I can be angry, afraid, or both, in
any degree.

> Obviously, when you get annoyed at somebody on the list you are in literal
> fear of them as an individual. But isn't it possible that part of your
> annoyance comes from the concern that the views they are expressing are
> widely held?  

I presume s/are/aren't/ above.  But no, when I get angry it's usually
because they are wasting my time (yes, this is selfish, but I find it
hard to get angry because someone else is wasting their time...).  I
don't, really, care whether bogus views are widely held: I *know* they
are, about all sorts of stuff, so I have no room for fear about
that...

I guess this is going nowhere - obviously psychowotsits can make all
sorts of random claims and be believed, but for me, I don't beleive
fear and anger have much in common.  I've been *very* frightened
indeed and *very* angry and they were really remarkably different
experiences.

--tim
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA36603E.25083%gehlker@fastq.com>
On 12/30/02 4:14 PM, in article ···············@cley.com, "Tim Bradshaw"
<···@cley.com> wrote:

> 
> I guess this is going nowhere - obviously psychowotsits can make all
> sorts of random claims and be believed, but for me, I don't beleive
> fear and anger have much in common.  I've been *very* frightened
> indeed and *very* angry and they were really remarkably different
> experiences.

OK. You are obviously the best expert on your own feelings.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Thomas Stegen
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3e119aba$1@nntphost.cis.strath.ac.uk>
Chris Gehlker wrote:
> You are obviously the best expert on your own feelings.

As a general claim I think this is very false. What is actually
cause and effect is rarely obvious at a low level. Take a drop of
water and a drop of oil. Then take a hammer and slam each drop
hard. What happens? The end result is more or less the same, but
the road there is very different.

Feelings and emotions are defense mechanisms and increases your
chances for survival. At least that was how they evolved when humans
were still living in the wild.

For some people the above claim might be true. For most, I don't think
the word expert rings true in any sense of the word.

-- 
Thomas.
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E110D4E.5000403@nyc.rr.com>
Chris Gehlker wrote:
> On 12/30/02 12:01 PM, in article ············@ccs.neu.edu, "Joe Marshall"
> <···@ccs.neu.edu> wrote:
> 
> 
>>Chris Gehlker <·······@fastq.com> writes:
>>
>>
>>>I suspect that the number of people who think C is a very flawed language
>>>dwarfs the number of people who think that about Lisp. The difference is
>>>that the C coders simply don't feel threatened by people who don't like C.
>>
>>I don't think any Lisp aficionado feels threatened by *anyone* who
>>doesn't like lisp.  I think `annoyed' is the word.
> 
> 
> But fear/anger or concern/annoyance if you prefer milder terms, are at root
> the same emotion. At least the physiologists & social psychologists think
> so.

Says you <g>:

from http://www.brain.riken.go.jp/bsinews/bsinews3/no3/speciale.html

"It has also been made clear that the regions related to fear and anger 
are clearly segregated each other within the amygdala. The same is the 
case in the hypothalamus and in the midbrain central gray."

and http://www.apa.org/science/ed-freyd.html

"Research in psychology provides some insight into the relationship 
between fear, anger, and hatred. Most animals naturally respond to 
threat with either flight or fight. When flight is not an option we are 
likely, at a very basic psychological and physiological level, to feel 
the need to fight. Anger and hatred of the enemy is one way our minds 
might try to help our bodies prepare for a frightening survival situation."

ie, they ain't the same but fear can lead to anger, which is a horse of 
a different color.

Actually I do recall a science times article about german shepherds in 
which it was said researchers concluded they were feeling the same 
emotion when barking ferociously as when afraid, but that was many years 
ago and may have been superseded.

I too looked askance at "threatened", but I figured I knew what you 
meant and did not want to get into hairsplitting. And to a degree there 
is a threat implicit in the drumbeat of misinformation. Repeat a lie 
often enough, etc.


-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA367277.251B5%gehlker@fastq.com>
On 12/30/02 8:18 PM, in article ················@nyc.rr.com, "Kenny Tilton"
<·······@nyc.rr.com> wrote:

> Says you <g>:
> 
> from http://www.brain.riken.go.jp/bsinews/bsinews3/no3/speciale.html
> 
> "It has also been made clear that the regions related to fear and anger
> are clearly segregated each other within the amygdala. The same is the
> case in the hypothalamus and in the midbrain central gray."

I'll be damned. There goes the rug from under my hypothesis.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey38yy6yyeg.fsf@cley.com>
* Kenny Tilton wrote:


> Actually I do recall a science times article about german shepherds in
> which it was said researchers concluded they were feeling the same
> emotion when barking ferociously as when afraid, but that was many
> years ago and may have been superseded.

I don't think you need articles in journals to know this: it's pretty
well known (at least it is among dog people I know) that dogs bark
when afraid: a dog that isn't frightened of you won't make a lot of
noise if it attacks you (but almost all dogs are frightened of people
in that kind of situation).

--tim
From: Rolf Marvin B�e Lindgren
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <spi11vkp8ogrudp2d0df8ttfaccli4btaj@4ax.com>
[Chris Gehlker]

| But fear/anger or concern/annoyance if you prefer milder terms, are at root
| the same emotion. At least the physiologists & social psychologists think
| so.

I would *love* to see the sources of this assertion.  I'm a
psychologist, by the way.
From: Matthias Heiler
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <av1fvs$jj$1@trumpet.uni-mannheim.de>
Hello,

Chris Gehlker wrote:

> This is all speculation anyway. I did do one experiment to test it though.
> I searched a couple of Visual Basic groups for any signs that the posters
> were as angry as the typical poster here. They clearly are not.

I can second this impression:  I'm reading comp.lang.c++.moderated, 
comp.std.c++, comp.lang.python, comp.lang.functional, and comp.lang.scheme 
besides cll.  In none of the other newsgroups people complain frequently 
about being treated 'rude' or 'harsh'. 

This is a good thing because you can point newcomers (once they mastered 
language basics) confidently to these newsgroups for new information.  So a 
friendly newsgroup actually helps spreading the use of a language.

One way to get the 'percieved hostility level' of this newsgroup down would 
be to ignore people who are percieved as hostile or unwilling to learn.  
This also would increase the 
useful-information/metadiscussion-how-to-behave-here ratio.

Matthias
From: Ng Pheng Siong
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <av1hq5$odp$1@reader01.singnet.com.sg>
According to Matthias Heiler  <······@nospm-uni-mannheim.de>:
> This is a good thing because you can point newcomers (once they mastered 
> language basics) confidently to these newsgroups for new information.  So a 
> friendly newsgroup actually helps spreading the use of a language.

And then the newbies go reinvent the same things over and over again. And
the level of discourse in these newsgroups don't seem to be able to rise
above these endless rehashes.



-- 
Ng Pheng Siong <····@netmemetic.com> * http://www.netmemetic.com
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E145D91.8020600@nyc.rr.com>
Matthias Heiler wrote:
> I can second this impression:  I'm reading comp.lang.c++.moderated, 
> comp.std.c++, comp.lang.python, comp.lang.functional, and comp.lang.scheme 
> besides cll.  In none of the other newsgroups people complain frequently 
> about being treated 'rude' or 'harsh'. 

Not here, either! What you do get are /bystanders/ who say cll is 
hostile based on how they see others being treated, but those others are 
too busy flaming away to complain (if they even feel put upon, which I 
doubt).

Overall, you are generalizing from certain threads/posters to the NG as 
a whole, and ya just can't do that, every Usenetter is their own keeper.

> 
> This is a good thing because you can point newcomers (once they mastered 
> language basics) confidently to these newsgroups for new information.  So a 
> friendly newsgroup actually helps spreading the use of a language.

I defy you to find an NG where newbies get better help, more graciously 
offered. Even the trolls get taken seriously longer than they deserve; 
you really have to prove yer a bozo to strike out here.

> 
> One way to get the 'percieved hostility level' of this newsgroup down would 
> be to ignore people who are percieved as hostile or unwilling to learn.  

And my point is that the inordinate amount of articles it takes for cll 
to collectively give up on those people proves that the conventional 
wisdom is wrong, we are in fact a bunch of feel-good, newbie-hugger jerks.

:)


-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Erik Naggum
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3250519673662774@naggum.no>
* Matthias Heiler
| In none of the other newsgroups people complain frequently about
| being treated 'rude' or 'harsh'.

  That is because there is no emotional gain from making complaints
  like that in those fora.  Watch this forum closely: Whenever some
  loser whines that he is being treated harshly because he acts like
  an immature child with a severe learning disability (but probably
  /is/ a grown man), loads of well-meaning idiots come to his rescue
  and validate his pain, regardless of the cause or even the accuracy
  of the complaint.

  One might wonder why people keep at this when it obviously does not
  work, but that is because it /does/ work: Someone who needs other
  people to validate his emotional problems will find them here.  It
  is not about newsgroup climate at all.  It is not about how rude or
  harsh people are to one another /at all/.  It is about how sulking
  and whining is /rewarded/ by people who are either taken for a ride
  or are quite as stupid as they appear when they keep acknowledging
  the irrelevant personal problems of the whining losers.

  In this forum, the exchange goes like this:

A: Please pay attention and try to understand what you are told.
B: You are so mean!  This is so unfair!  Boo hoo!  *whine* *cry*
A: Pull yourself together, you little fuck.
C: Awww, poor you!  A is so mean.  Here, let me cuddle and hug you.
D: Yeah, A is a goddamn son of a bitch!  This is a hostile place.
E: I *hate* A!  It is all A's fault.  I can't use Lisp because of A.
F: Tell me about it, A once told me to go stuff myself.
G: We should all be nicer to newbies.  Newbies are the future.
H: I'm a newbie and I'm scared of posting here because of A.
I: I'm with B.  Nobody should be told they are little fucks.
J: I'm outta here.  Python people are much nicer.
K: Look at what A made me do!
  etc

  In any other forum, the exchange goes like this:

A: Please pay attention and try to understand what you are told.
B: Well, I do try, but I don't understand X.
A: Ah, I see your problem.  X works like this, which is different.
B: Great, I got it.  Thanks!

  The problem here is not A or B, it is C through Z.  Absent C through
  Z, B would stand out like the moron he has acted like, and he would
  (and should) feel really stupid if nobody came to his rescue.  For
  some reason which I have yet to figure out, a cobble of /babies/
  play in comp.lang.lisp and appear to be in dire need of constant
  emotional support.  Fail to give them their support, and they cry.
  Fail to make them feel included and accepted, and they do not act in
  such a way that others would /want/ to include and accept them, but
  /demand/ to be included and accepted, as if it were their birthright
  and it was somehow denied them.

  Most grown adults look at a harsh reaction to something they have
  done with a desire to avoid future harsh reactions, even if they do
  not understand precisely what caused the reaction, so they want to
  know what they did wrong if they get a harsh reaction.  But not in
  comp.lang.lisp.  In comp.lang.lisp, you get to be a hatemongerer and
  a disruptive shithead if you get anything but the answer you have a
  God-given right to /demand/ from other people.  If something is
  explained to you and you do not understand it, most people realize
  that they may be lacking in background and that it would be a good
  idea to acquire that background on their own.  They may therefore
  seek advice on how to acquire it.  But not in comp.lang.lisp.  An
  answer that you do not immediately understand is probably an insult,
  and your best course of action is to elaborate on your ideas about
  the other person's personality or heritage or carreer or whatever.
  And this is /accepted/ by a cobble of crying babies who have just
  been /waiting/ for another chance to cry the forum another river.

  So to address the exact wording of your message, people /complain/
  about being treated rudely and harshly here, but they are actually
  not -- if anything, the treatment they get is civil compared to
  their own.  The complaints and the constant blame-throwing are all
  perfectly OK in the minds of a few people who spur them on and feel
  no regrets for going after individuals they believe are to blame
  like mad dogs.  But if they were really wanted proper behavior, how
  come they do not do something that could have /helped/?  Well, it is
  because they have no desire for proper behavior.  The lynch mob that
  persists in comp.lang.lisp has no such noble goals as civility or
  justice or fairness; their goal is simply to blame someone for their
  own emotional problems, and if they can find somebody that other
  people can agree is to blame for their failure to grow up and accept
  responsibility for their own emotional well-being, they gang up on
  that person (which just happens to be me here -- it is other people
  in a few other fora with the same kind of baby population) with a
  vengeance that I regard as prima facie evidence of rabid insanity,
  but which is probably arrested development at about 2 years of age.

  I suspect that given half a chance, a mob of insane babies would be
  able to destroy any forum, but the problem is that not enough people
  crush them at the outset.  Look at that whining weeny "wni" and how
  disruptive he managed to be because so many fucking morons had to
  chime in and validate his out-of-place emotional problems.  Where
  else would you find such a disgraceful display and have lots of
  people flock to support it?  The problem is here in comp.lang.lisp
  is that of /accepting/ the personal problems of lunatics who feel no
  reservations about "sharing" their personal hatred for other people
  with the whole forum, and those who obsess about civility /never/
  rise to tell /those/ people to shut the fuck up, do they?  In a
  truly /civilized/ society, which comp.lang.lisp is not, /personal/
  animosity has no place in public, and he who puts his personal
  problems up for public display has disgraced /himself/ regardless of
  his supposed message.  In comp.lang.lisp, a disgusting kindergarten
  of whining losers, other immature people rush to support the crying
  idiot instead of trying to do something that could actually help.

  What has in so many ways killed this forum is that so many egoists
  and egotists think this forum is about them and their feelings.  It
  is not.  This is, probably to the amazement of a lot of people, a
  forum created to help people use the programming language (family|
  Common) Lisp.  Discussions about individual people or about how you
  "feel" about other articles are out of order and out of place.

| One way to get the 'percieved hostility level' of this newsgroup
| down would be to ignore people who are percieved as hostile or
| unwilling to learn.

  One good way to get rid of the problem would be to stop talking
  about it and instead talk about something that could interest people
  who are, probably shocking to many of the current crop of readers,
  reading comp.lang.lisp to read about the programming language
  (family|Common) Lisp.  I occasionally receive mail from people who
  have given up on comp.lang.lisp because of the metadiscussions and
  the goddamn shitheads who think this forum exists for the express
  purpose of attacking me.

| This also would increase the useful-information/
| metadiscussion-how-to-behave-here ratio.

  The best thing that help reduce the meta-discussion to signal ratio
  here would be to stop engaging in meta-discussions.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: faust
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <115e1v49alu2vhvt3g38nq5vgub3lf3akp@4ax.com>
 Erik Naggum <····@naggum.no> ,  emitted these fragments:

>That is because there is no emotional gain from making complaints
>  like that in those fora.  Watch this forum closely: Whenever some
>  loser whines that he is being treated harshly because he acts like
>  an immature child with a severe learning disability (but probably
>  /is/ a grown man), loads of well-meaning idiots come to his rescue
>  and validate his pain

Or one of the regulars goes ballistic and suggests that he kills
himself.

Statistically, this is more likely.

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Wade Humeniuk
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <1noR9.100060$k13.3568333@news0.telusplanet.net>
"faust" <·······@optushome.com.au> wrote in message
·······································@4ax.com...
> Erik Naggum <····@naggum.no> ,  emitted these fragments:
>
> >That is because there is no emotional gain from making complaints
> >  like that in those fora.  Watch this forum closely: Whenever some
> >  loser whines that he is being treated harshly because he acts like
> >  an immature child with a severe learning disability (but probably
> >  /is/ a grown man), loads of well-meaning idiots come to his rescue
> >  and validate his pain
>
> Or one of the regulars goes ballistic and suggests that he kills
> himself.
>
> Statistically, this is more likely.
>
> --------------------------------------------------------
> Come see,
> real flowers
> of this pain-filled world.
>
> (from Basho)

Beat it faust (or more accurately, die)

WARNING! Completely off-topic.

A post for those who have the wrong idea of what it is to be nice.  This
drama is not only played out in c.l.l.

Excerpt from http://www.humankindness.org/winter98.html

Many people go to church on Sunday and then pass by a beggar on their
way home without stopping to help in any way. They choose a church
which will comfort them in a self-centered lifestyle rather than
challenge them to be true Christians. Being a true Christian is a
terrifying prospect. Being a true Buddhist, Jew or anything else is a
terrifying prospect. All the religions stress that we must die as
self-centered little idiots in order to discover new life as selfless,
loving, generous, fearless souls.

Die to my own plans and dreams? Die to my countless preferences and
aversions? Die to my pride and greed? Yes, yes, and yes. Die.

Father Murray Rogers, the beloved elder on our board of directors, had
a powerful experience along these lines many years ago. Father Murray
is a Christian who has been involved in the interfaith dialogue for
over fifty years. As part of his interfaith experience, he went to
Japan to spend time in a Zen Buddhist temple for a few months. The
Temple was a serenely beautiful place, extremely neat and orderly,
extremely quiet, like most Zen temples.

The abbot, a small, courteous man of few words, showed Murray around
for a half-hour or more, whispering "This is where you will eat,"
"This is where you will be meditating," "This is your room," and so
forth. In his room, just before turning to leave, the abbot leaned
forward toward Murray and whispered, "There is just one more thing."
And then he thrust his face directly into Murray's with a wild look
and screamed at the top of his lungs, "YOU MUST DIE!!!!!"
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E163612.8020000@nyc.rr.com>
Wade Humeniuk wrote:
> ... In his room, just before turning to leave, the abbot leaned
> forward toward Murray and whispered, "There is just one more thing."
> And then he thrust his face directly into Murray's with a wild look
> and screamed at the top of his lungs, "YOU MUST DIE!!!!!"

I once stayed in a motel like that, just outside the Lincoln Tunnel in NJ.

:)

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: faust
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <16oe1vc2nk9lkue5n5sgp8k8bvfiit9rno@4ax.com>
 "Wade Humeniuk" <····@nospam.nowhere> ,  emitted these fragments:

>>>  loads of well-meaning idiots come to his rescue  and validate his pain

>> Or one of the regulars goes ballistic and suggests that he kills
>> himself.

>And then he thrust his face directly into Murray's with a wild look
>and screamed at the top of his lungs, "YOU MUST DIE!!!!!"
>Beat it faust (or more accurately, die)

Aah ! Now I have attained enlightenment !

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Joe Marshall
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <vg12hza3.fsf@ccs.neu.edu>
Erik Naggum <····@naggum.no> writes:

>   In this forum, the exchange goes like this:
> 
> A: Please pay attention and try to understand what you are told.
> B: You are so mean!  This is so unfair!  Boo hoo!  *whine* *cry*
> A: Pull yourself together, you little fuck.
> C: Awww, poor you!  A is so mean.  Here, let me cuddle and hug you.
> D: Yeah, A is a goddamn son of a bitch!  This is a hostile place.
> E: I *hate* A!  It is all A's fault.  I can't use Lisp because of A.
> F: Tell me about it, A once told me to go stuff myself.
> G: We should all be nicer to newbies.  Newbies are the future.
> H: I'm a newbie and I'm scared of posting here because of A.
> I: I'm with B.  Nobody should be told they are little fucks.
> J: I'm outta here.  Python people are much nicer.
> K: Look at what A made me do!
>   etc...

There are more than a few of these exchanges that *start* with D!
From: Matt Curtin
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <864r8vtizb.fsf@rowlf.interhack.net>
Chris Gehlker <·······@fastq.com> writes:

> I suspect that the number of people who think C is a very flawed
> language dwarfs the number of people who think that about Lisp. The
> difference is that the C coders simply don't feel threatened by
> people who don't like C.

I think the real issue here is a bit more general.  How politicized
the language is will affect how much noise there is, largely as a
function of advocacy (for and against) the language in question.

Java newsgroups also have quite a lot of noise, including people from
Microsoft posting under Hotmail pseudonyms about how flawed it is, why
C# is so much better, blah blah blah blah... only to be exposed, and
to disappear ... often emerging with a new pseudonym as one of the
next new idiots to appear a few weeks later.

Even beyond programming languages, the same rule generally holds true,
e.g., Linux.

Creation of *.advocacy newsgroups can sometimes help to direct the
tripe, but certainly not all.  And the advocacy groups easily turn
into little more than threads of flamefests.

I cannot help but wonder how many personae that would no doubt exist
on a comp.lang.lisp.advocacy would actually be Lisp programs.  I also
cannot help but wonder (with some amusement at the very idea) how many
anti-Lisp zealots would lose their arguments to such programs.

-- 
Matt Curtin, CISSP, IAM, INTP.  Keywords: Lisp, Unix, Internet, INFOSEC.
Founder, Interhack Corporation +1 614 545 HACK http://web.interhack.com/
Author of /Developing Trust: Online Privacy and Security/ (Apress, 2001)
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3isxb1gsx.fsf@cley.com>
* Chris Gehlker wrote:

> How do you know whether people are interested in learning the truth when
> they "turn up"?

You don't.  But when you tell them , but when you explain things, and
they don't listen but repeat back the same stuff they said before, you
begin to realise something.

> I suspect that the number of people who think C is a very flawed language
> dwarfs the number of people who think that about Lisp. The difference is
> that the C coders simply don't feel threatened by people who don't like C.

As a proportion of users of the languages, I wonder if that is so.
However I don't feel threatened by people who don't like Lisp.  I do
feel annoyed by people who refuse to listen when things are explained
to them.

--tim
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA35F626.25057%gehlker@fastq.com>
On 12/30/02 11:51 AM, in article ···············@cley.com, "Tim Bradshaw"
<···@cley.com> wrote:

> * Chris Gehlker wrote:
> 
>> How do you know whether people are interested in learning the truth when
>> they "turn up"?
> 
> You don't.  But when you tell them , but when you explain things, and
> they don't listen but repeat back the same stuff they said before, you
> begin to realise something.

Clearly. I think I read something into "turn up" that you didn't intend.
> 
>> I suspect that the number of people who think C is a very flawed language
>> dwarfs the number of people who think that about Lisp. The difference is
>> that the C coders simply don't feel threatened by people who don't like C.
> 
> As a proportion of users of the languages, I wonder if that is so.

I don't have any hard evidence but I suspect it's so. I haven't met an
enthusiastic C user in years, and I know lots of C users. I do know many who
are enthusiastic about what they are doing in C, but that's not the same
thing. 



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3n0mn89tr.fsf@cley.com>
* Chris Gehlker wrote:
> I don't have any hard evidence but I suspect it's so. I haven't met an
> enthusiastic C user in years, and I know lots of C users. I do know many who
> are enthusiastic about what they are doing in C, but that's not the same
> thing. 

You've met one now.  I think C is cool: almost as cool as CL.  It's a
*completely* different approach suitable to very different problems,
and it has some annoying misfeatures (like CL), but it's just a great
language.  (Note: I said C, not C++, which I do not think is cool.)  I
think Java is cool too, though not in the same way, more as a
beautifully engineered solution to a number of technical and `social'
problems.  I even think Perl is cool, in a mutant sort of way.  So now
you can shoot me.

--tim
From: Lars Magne Ingebrigtsen
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m3k7hrglie.fsf@quimbies.gnus.org>
Tim Bradshaw <···@cley.com> writes:

> I even think Perl is cool, in a mutant sort of way.

So, what's your opinion on Tcl?  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
   ·····@gnus.org * Lars Magne Ingebrigtsen
From: Ingvar Mattsson
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87isxao2q4.fsf@gruk.tech.ensign.ftech.net>
Lars Magne Ingebrigtsen <·····@gnus.org> writes:

> Tim Bradshaw <···@cley.com> writes:
> 
> > I even think Perl is cool, in a mutant sort of way.
> 
> So, what's your opinion on Tcl?  :-)

If you replace [] with () and {} with '() it sort-of looks like a
line-oriented lisp (where, of course, each line in and of itself
should have ( at the start and ) at the end).

//Ingvar
-- 
((lambda (x) `(,x ',x)) '(lambda (x) `(,x ',x)))
	Probably KMP
From: Tim Bradshaw
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <ey3d6niyypc.fsf@cley.com>
* Lars Magne Ingebrigtsen wrote:

> So, what's your opinion on Tcl?  :-)

horrible.

--tim
From: Gareth McCaughan
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <slrnb11vcd.11v7.Gareth.McCaughan@g.local>
Chris Gehlker wrote:

>  I don't have any hard evidence but I suspect it's so. I haven't met an
>  enthusiastic C user in years, and I know lots of C users. I do know many who
>  are enthusiastic about what they are doing in C, but that's not the same
>  thing. 

I see someone else already owned up, so I will too. I'm
(moderately) enthusiastic about C. I think it's a nice
piece of design; it does what it does really rather well.

I even quite like C++, though it suffers badly from trying
(as it were) to compete with both C and CL.

I suspect that considerably more CL programmers than, say,
C programmers like a wide variety of other languages.
This may just be because by the time someone discovers CL
they've usually had time to find several other decent
languages first, but I don't think that's the whole story.

-- 
Gareth McCaughan  ················@pobox.com
.sig under construc
From: Paul Foley
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <m2vg1a2xx9.fsf@mycroft.actrix.gen.nz>
On Tue, 31 Dec 2002 02:14:37 +0000, Gareth McCaughan wrote:

> Chris Gehlker wrote:
>> I don't have any hard evidence but I suspect it's so. I haven't met an
>> enthusiastic C user in years, and I know lots of C users. I do know many who
>> are enthusiastic about what they are doing in C, but that's not the same
>> thing. 

> I see someone else already owned up, so I will too. I'm
> (moderately) enthusiastic about C. I think it's a nice
> piece of design; it does what it does really rather well.

AOL...and HNY (Happy New Year, of course)

-- 
Bonis nocet quisquis pepercit malis
                                                       -- Publilius Syrus
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(··@) "actrix.gen.nz>"))
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E10A228.4050800@nyc.rr.com>
Chris Gehlker wrote:
> On 12/30/02 7:16 AM, in article ···············@cley.com, "Tim Bradshaw"
> <···@cley.com> wrote:

> How do you know whether people are interested in learning the truth when
> they "turn up"?

They make that manifest in their responses to the helpful responses they 
get to their OPs. One thing that strikes me abou cl.l. is how wrong are 
those who call this NG unfriendly. Newcomers are always given the 
benefit of the doubt, and then one by one the regulars come up to speed 
onthe cluelessness we are dealing with. Some regulars get there /real 
fast/ <g>, but to their credit they almost always turn out to have been 
right in their quick assessments.

I am pretty sure about this, because I have been informally studying 
these close encounters of an nth kind since someone I know to be highly 
cool observed that cll was a hostile place.

> 
> The difference is
> that the C coders simply don't feel threatened by people who don't like C.

Right. To a large degree I have a thin skin re Lisp because the drumbeat 
of nonsense is so unrelenting, the ignorance so pervasive. In a few 
years when OReilly is accepting /only/ Lisp books for publication we 
won't even respond to naysayers.

:)



-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA35FBD0.25059%gehlker@fastq.com>
On 12/30/02 12:41 PM, in article ················@nyc.rr.com, "Kenny Tilton"
<·······@nyc.rr.com> wrote:

> Right. To a large degree I have a thin skin re Lisp because the drumbeat
> of nonsense is so unrelenting, the ignorance so pervasive. In a few
> years when OReilly is accepting /only/ Lisp books for publication we
> won't even respond to naysayers.
> 

I just can't find the naysayers. I don't know any personally and a Google
for "lisp sucks" turned up "Well, in this disciples point of view, Lisp has
a very high impedance mismatch with the human brain. This makes it very hard
to grok at first. But, that's about all that makes Lisp suck. So go read
Lisp Rocks again!" as the *most* negative comment in the first page of hits.

The O'Reilly thing is disturbing but I suspect that a lot of the motivation
there was commercial. There are high quality, free PDFs of Lisp books
floating around. O'Reilly notoriously cancelled Ruby books and lost their
enthusiasm for Ruby when the "PickAxe Book" showed up on the web.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Patrick W
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <87isxbuhhq.fsf@map.localdomain>
Chris Gehlker <·······@fastq.com> writes:

> I just can't find the naysayers. I don't know any personally and a
> Google for "lisp sucks" turned up "Well, in this disciples point of
> view, Lisp has a very high impedance mismatch with the human
> brain. This makes it very hard to grok at first. But, that's about
> all that makes Lisp suck. So go read Lisp Rocks again!" as the
> *most* negative comment in the first page of hits.

There are a few naysayers in other newsgroups.  But are they naysayers
because they believe Lisp sucks?  I doubt it.

People make a small number of reasoned decisions amidst much
uncertainty, and do the best they can.  It is a terrible waste of time
to regularly revise past decisions, backtrack, change course, move
forward, change mind, backtrack, change course, move on, etc.  People
make a few vital decisions early in an endeavour, and defend them
vigorously.  But not necessarily honestly.

Publicly justifying or privately rationalising a choice often involves
turning a deliberately blind or jaundiced eye to alternatives.  It's
one of the prices we pay for the capacity to focus and persevere.  (If
anyone out there is both (a) immune to all forms of wilful blindness;
(b) productive; I'd like to know how you do it).

It amazes me that so many people assume that the main purpose of
communicating is to share and to learn.  In many *many* different
contexts (some of them really scary), much communication depends on a
shared intention *not* to see, but to create increasingly robust
rationalisations for what is already 'known', or has already been
decided, or for what 'must' be the case even if it isn't, or what must
*not* be the case, even if it is.

This is pervasive, and can often be very subtle.  But spotting it in
usenet discussions is ridiculously easy.  Discussions involving
language comparisons are usually only fake preludes to conclusions
already reached, or rationalisations for decisions already made.  When
somebody mentions Lisp on Slashdot or comp.lang.flavour-of-the-month,
it becomes very obvious that some people would *rather* find evidence
that Lisp sucks than ... (well, you can guess the rest).

Even if we limit this to narrow technical issues, Lisp is far from
alone.  For each of the many justifiable criticisms of C++, there are
at least a dozen stupid ones that are bandied about by people who know
nothing, and don't *need* to know, because their only purpose is to
prop up a decision they made some time ago to not learn it, or to use
something else instead.  (Right decision, wrong reasons) ;-)

If Lisp people are more sensitive to illogical criticisms, it's partly
because they have better technical reasons to be frustrated, and
partly because they share at least one blind eye with the rest of the
population.
From: Chris Gehlker
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <BA367227.251B4%gehlker@fastq.com>
On 12/30/02 6:03 PM, in article ··············@map.localdomain, "Patrick W"
<···········@yahoo.com.au> wrote:

Such a good post that I feel like abandoning this thread. There is simply
nothing more to say.



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----
From: Paolo Amoroso
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <6cARPgjy=7Zs2F6nHlhxDLlECLXW@4ax.com>
On Mon, 30 Dec 2002 21:59:03 -0700, Chris Gehlker <·······@fastq.com>
wrote:

> On 12/30/02 6:03 PM, in article ··············@map.localdomain, "Patrick W"
> <···········@yahoo.com.au> wrote:
> 
> Such a good post that I feel like abandoning this thread. There is simply
> nothing more to say.

MP:THIS-THREAD-INTENTIONALLY-LEFT-BLANK


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Kenny Tilton
Subject: Re: Rudeness index was Looking for Lisp compiler
Date: 
Message-ID: <3E110361.8060203@nyc.rr.com>
Patrick W wrote:
> It is a terrible waste of time
> to regularly revise past decisions, backtrack, change course, move
> forward, change mind, backtrack, change course, move on, etc. 

Good lord! That's how I program!! So that's why it takes me so long.


> Publicly justifying or privately rationalising a choice often involves
> turning a deliberately blind or jaundiced eye to alternatives.

I thought you were going to say "involves lying like a rug", because 
none of us gets the introspection thing right. When someone asks me 
"why?", I warn them "I'll tell you, but I'll be making it up as I go along."

>  It's
> one of the prices we pay for the capacity to focus and persevere.  (If
> anyone out there is both (a) immune to all forms of wilful blindness;
> (b) productive; I'd like to know how you do it).

Yeah, but strike willful. To me it is like reading. We can do it quickly 
only if we go by partial information visually in re identifying words 
and on the other hand add information in the form of our expectations of 
what we are about to read based on context and all the knowledge and 
experience we bring to the reading act. No one /intends/ that, it's just 
how it goes.

> 
> It amazes me that so many people assume that the main purpose of
> communicating is to share and to learn.

fat chance! It is said that whenever two sailboats are in sight of each 
other there is a race going on. We need a line like that for Usenet.

But in this case... good post!

:)



-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Lars Magne Ingebrigtsen
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <m3n0mnmotg.fsf@quimbies.gnus.org>
Tim Bradshaw <···@cley.com> writes:

> (2) Some regular posters say you are an idiot when they think you
> are.  So what's wrong with that?

The ensuing "discussions" are tedious in the extreme.

-- 
(domestic pets only, the antidote for overdose, milk.)
   ·····@gnus.org * Lars Magne Ingebrigtsen
From: Kenny Tilton
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3E10A469.5050204@nyc.rr.com>
Lars Magne Ingebrigtsen wrote:
> Tim Bradshaw <···@cley.com> writes:
> 
> 
>>(2) Some regular posters say you are an idiot when they think you
>>are.  So what's wrong with that?
> 
> 
> The ensuing "discussions" are tedious in the extreme.
> 

I feel exactly the same way and often start to tell the correspondents 
to pipe down, but then I always remember that I am reading through all 
the garbage voluntarily for its amusement value. Doh!

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Tim Bradshaw
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ey3bs335l93.fsf@cley.com>
* Lars Magne Ingebrigtsen wrote:

> The ensuing "discussions" are tedious in the extreme.

Yes, I agree, but whose fault is that?  The person calling someone an
idiot, the person so called, all the other people who get offended by
one side or the other and then wade in, or, finally, all the people
who carry on reading the thread.  I think it's mostly the latter two
groups.

--tim
From: Nils Kassube
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <813cof46n6.fsf@darwin.lan.kassube.de>
Lars Magne Ingebrigtsen <·····@gnus.org> writes:

>> (2) Some regular posters say you are an idiot when they think you
>> are.  So what's wrong with that?
> The ensuing "discussions" are tedious in the extreme.

You should get a real news reader software with support for scoring
and the ability to ignore threads you don't want to read. 

SCNR ;-) 
From: Jacek Generowicz
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <tyfsmwecs6f.fsf@lxplus029.cern.ch>
Nils Kassube <····@kassube.de> writes:

> You should get a real news reader software with support for scoring
> and the ability to ignore threads you don't want to read. 

Do you know one capable of filtering out the 6 pages of Erik's
vituperation from a post, and showing me only the 7 lines of technical
excellence ? :-)
From: Nils Kassube
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <81isxa2tnz.fsf@darwin.lan.kassube.de>
Jacek Generowicz <················@cern.ch> writes:

> Do you know one capable of filtering out the 6 pages of Erik's
> vituperation from a post, and showing me only the 7 lines of technical
> excellence ? :-)

Sure. Use Gnus and write a small script using well-known AI
algorithms:

        http://aima.cs.berkeley.edu/

Fortunately you can now use Python as user-friendly language to
enhance Emacs:

http://www.iro.umontreal.ca/~pinard/pymacs/

No need for this bizarre language with too many parantheses. 
From: Jacek Generowicz
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <tyfbs325d8p.fsf@lxplus011.cern.ch>
Nils Kassube <····@kassube.de> writes:

> Jacek Generowicz <················@cern.ch> writes:
> 
> > Do you know one capable of filtering out the 6 pages of Erik's
> > vituperation from a post, and showing me only the 7 lines of technical
> > excellence ? :-)
> 
> Sure. Use Gnus and write a small script using well-known AI
> algorithms:

Actually, I rather suspect that that is how most of Erik's posts are
generated in the first place.
From: Nils Kassube
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <818yy62igt.fsf@darwin.lan.kassube.de>
Jacek Generowicz <················@cern.ch> writes:

> Actually, I rather suspect that that is how most of Erik's posts are
> generated in the first place.

I didn't do it, nobody saw me do it, there's no way you can prove
anything!
From: faust
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <rlg61v4h91t8guu5sof5hkcl8q7i8cuntg@4ax.com>
 Jacek Generowicz <················@cern.ch> ,  emitted these
fragments:

 
>> Sure. Use Gnus and write a small script using well-known AI
>> algorithms:
>
>Actually, I rather suspect that that is how most of Erik's posts are
>generated in the first place.


AI is not needed for that.
Just a simple lookup-table populated with vituperation would do the
job.

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Gareth McCaughan
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <slrnb14opp.11v7.Gareth.McCaughan@g.local>
Something calling itself faust wrote:

>   Jacek Generowicz <················@cern.ch> ,  emitted these
>  fragments:
>  
>   
> >> Sure. Use Gnus and write a small script using well-known AI
> >> algorithms:
> >
> >Actually, I rather suspect that that is how most of Erik's posts are
> >generated in the first place.

Oh look, Mrs Naggumbaiter is out again. How nice.
More information in, e.g., <··············@raw.grenland.fast.no>.
(I think the "S Campion / Adam Tissa" outbreak was more
recent; same person, I believe.)

-- 
Gareth McCaughan  ················@pobox.com
.sig under construc
From: faust
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <lcg61vsp6kfnk13uc6p760mgls07t9jk3l@4ax.com>
 Nils Kassube <····@kassube.de> ,  emitted these fragments:


>Fortunately you can now use Python as user-friendly language to
>enhance Emacs:
>
>http://www.iro.umontreal.ca/~pinard/pymacs/
>
>No need for this bizarre language with too many parantheses. 

Didnt know that Python had parentheses

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <aupqeq$upc$1@f1node01.rhrz.uni-bonn.de>
Tim Bradshaw wrote:

> I think that there are two relevant features of cll which make it
> somewhat ruder than (some - by no means all) other groups. (1) There
> are a relatively large number of people who turn up with extremely
> wrong and very negative views, and who are not interested in learning
> the truth. I don't read any other language groups, but I *really*
> doubt that the C groups, for instance, get this. (2) Some regular
> posters say you are an idiot when they think you are.  So what's wrong
> with that?

wrt (1): This is also the case for some of the regular posters. Some of 
them try to defend Common Lisp by all means, and at least one of them 
has even stated this as an explicit goal. I fail to see what this has to 
do with any notion of "truth".

wrt (2): Some of those alleged idiots are possibly no idiots at all, but 
   may decide to have something better to do than to repeatedly expose 
themselves to abusive language. I'd regard this as a serious loss.

Just my 0.02�.


Pascal

P.S.: I wonder what Common Lisp vendors think about this situation. 
After all, anyone that shows some interest in Common Lisp is a potential 
customer in the long run, unless scared away prematurely. Or am I 
overestimating the impact of newsgroups?

(Some background information: This is my first real newsgroup 
experience, and I have only made my first posting in this year's August. 
Before that, I have only occasionally touched newsgroups. So I possibly 
have wrong expectations...)

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Joe Marshall
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <lm27mpn1.fsf@ccs.neu.edu>
Pascal Costanza <········@web.de> writes:

> P.S.: I wonder what Common Lisp vendors think about this
> situation. After all, anyone that shows some interest in Common Lisp
> is a potential customer in the long run, unless scared away
> prematurely. 

And stupid, idiotic, or moronic people are more easily separated from
their money.
From: Erik Naggum
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3250258817513930@naggum.no>
* Pascal Costanza
| wrt (1): This is also the case for some of the regular posters.
| Some of them try to defend Common Lisp by all means, and at least
| one of them has even stated this as an explicit goal.  I fail to
| see what this has to do with any notion of "truth".

  I have never seen that happen.  Could you point me to at least a few
  examples where you believed this to take place?  Note that you have
  a pretty lousy track record of getting people's intentions right and
  that my willingness to take your word for it when you make this kind
  of extremely rude claim is exactly zero (and in your inimical style
  remove every shred of the target of your slander).  Your impression
  is not necessarily the pure and unadulterated truth just because it
  is your personal experience.  You arrive at "conclusions" about how
  other people feel and what their intentions are, and you annoy the
  hell out of me, at least, by insisting that you are right about this
  and that the people you make these claims about shall not correct
  you.  As for the statement of explicit goal, dig it up and quote it.
  If you have to chicken out of actually naming the people you talk
  /about/, you indicate to the whole world that you no longer have (if
  you ever had) any interest in /dialog/ with them, and that is the
  kind of rudeness you have pretty clearly indicated that /you/ do not
  like when it happens to others.  One would have thought that you, of
  all people, would at least pay attention to such things, but no.

| P.S.: I wonder what Common Lisp vendors think about this situation.

  No, you don't.  You say this only to affect people how people feel.

| Or am I overestimating the impact of newsgroups?

  Yes, and you have still to figure out that this is a /technical/
  forum.  I doubt that you even know what a technical forum is.

| (Some background information: This is my first real newsgroup
| experience, and I have only made my first posting in this year's
| August.  Before that, I have only occasionally touched newsgroups.
| So I possibly have wrong expectations...)

  Not only expectations, but preconceptions that cannot be changed by
  facts, and you simply /cannot/ keep quiet about them.  What makes
  you think that your behavior is not massively annoying and so scares
  away much more people than any other cause you think is evil?  You
  go out of your way to post off-topic drivel and make claims about
  other people that you cannot possibly know, and you are adamant that
  you are not part of the problem.  This fact alone pisses people off.

  Some people actually /listen/ before they talk and try to grasp the
  ambiance in a group before they pronounce their judgment and seek to
  change it to their liking.  The other Pascal also fails to do that,
  As do several past whiners and feelers.  This is not a coincidence.

  Please tell us why you could not control yourself and just /had/ to
  chip in with your vague, yet specifically /negative/, bullshit.

  What happened to that philosophy of yours about spewing feel-good
  crap so people felt included?  You demonstrate that you suffer from
  exactly the same problem you attack others for so cowardly: the
  incredibly primitive us-vs-them tribal attitude.  Upgrade yourself!
  Become a /thinking/ human being, and grow the hell up so you can
  start to assume responsibility for your /own/ behavior.

  Consider whether this forum benefits from yet more drivel from you,
  or whether it would really be improved a lot more if you shut up
  about your feelings and acted according to your own rules, instead
  of demanding that only the people you dislike do that.
  
-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <auq26q$11p8$1@f1node01.rhrz.uni-bonn.de>
Erik Naggum wrote:

>   Your impression
>   is not necessarily the pure and unadulterated truth just because it
>   is your personal experience.

Correct.

>   As for the statement of explicit goal, dig it up and quote it.

http://groups.google.com/groups?selm=3208226254834485%40naggum.net&output=gplain

Here is an excerpt:

"There are some people who still show signs of enthusiasm in the Common 
Lisp community, who still express love for their language, who still 
want the language to be fully implemented. [...] In any _other_ 
community, those who love the language would be celebrated and the 
nay-sayers would be run out of town, asked to go create their own 
community.  Instead, we let these corpses stay with us and spread death 
and gloom and pestilence, [...] We need to throw out and bury the 
corpses. [...]"


-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Tim Bradshaw
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ey3fzsf5lh8.fsf@cley.com>
* Pascal Costanza wrote:

> http://groups.google.com/groups?selm=3208226254834485%40naggum.net&output=gplain

> Here is an excerpt:

> "There are some people who still show signs of enthusiasm in the
> Common Lisp community, who still express love for their language, who
> still want the language to be fully implemented. [...] In any _other_
> community, those who love the language would be celebrated and the
> nay-sayers would be run out of town, asked to go create their own
> community.  Instead, we let these corpses stay with us and spread
> death and gloom and pestilence, [...] We need to throw out and bury
> the corpses. [...]"

This excerpt (I haven't checked the original article) does not seem,
to me, to have anything to do with your statement:

    Some of them try to defend Common Lisp by all means, and at least
    one of them has even stated this as an explicit goal.

Rather it seems to me to be encouraging people who don't like Lisp to,
well, do something else, rather than hanging around denigrating the
language in a forum dedicated to it.  How is this `defending [Common]
Lisp by all means'?  I find that a very strange reading of this
excerpt.

--tim
From: Erann Gat
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <gat-3012021259370001@k-137-79-50-101.jpl.nasa.gov>
In article <···············@cley.com>, Tim Bradshaw <···@cley.com> wrote:

> * Pascal Costanza wrote:
> 
> >
http://groups.google.com/groups?selm=3208226254834485%40naggum.net&output=gplain
> 
> > Here is an excerpt:
> 
> > "There are some people who still show signs of enthusiasm in the
> > Common Lisp community, who still express love for their language, who
> > still want the language to be fully implemented. [...] In any _other_
> > community, those who love the language would be celebrated and the
> > nay-sayers would be run out of town, asked to go create their own
> > community.  Instead, we let these corpses stay with us and spread
> > death and gloom and pestilence, [...] We need to throw out and bury
> > the corpses. [...]"
> 
> This excerpt (I haven't checked the original article) does not seem,
> to me, to have anything to do with your statement:
> 
>     Some of them try to defend Common Lisp by all means, and at least
>     one of them has even stated this as an explicit goal.
> 
> Rather it seems to me to be encouraging people who don't like Lisp to,
> well, do something else, rather than hanging around denigrating the
> language in a forum dedicated to it.  How is this `defending [Common]
> Lisp by all means'?  I find that a very strange reading of this
> excerpt.


The call to action in the last sentence is first-person: "We need to..." 
Reading "we" to mean "people who don't like Lisp" seems untenable to me. 
Furthermore, the action called for is to "throw out and bury the
corposes."  I don't see any plausible way to read this as meaning "go away
and do something else."

So I find your reading of this excerpt to be "encouraging people who don't
like Lisp to, well, do something else" to be very strange indeed.

E.
From: Tim Bradshaw
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ey365tbdrha.fsf@cley.com>
> The call to action in the last sentence is first-person: "We need to..." 
> Reading "we" to mean "people who don't like Lisp" seems untenable to me. 
> Furthermore, the action called for is to "throw out and bury the
> corposes."  I don't see any plausible way to read this as meaning "go away
> and do something else."

Oh, I see it as meaning `we (people who use Lisp and want to discuss
it in a useful way) need to make people who only want to be derogatory
about it to not post in cll, so we can get some useful discussion done
rather than endlessly dealing with them'.  Or in other words it's
saying that we (same we) should encourage, strongly, these people who
don't like Lisp (and have nothing constructive to say) to, well, go
away and do something else (but post in cll).

--tim
From: Damond Walker
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <BA37C1EC.805A%damosan@comcast.net>
On 12/30/02 2:56 PM, in article ···············@cley.com, "Tim Bradshaw"
<···@cley.com> wrote:

[snip]

> Rather it seems to me to be encouraging people who don't like Lisp to,
> well, do something else, rather than hanging around denigrating the
> language in a forum dedicated to it.

...and to which I give a general <clap-clap>.

Damo
From: Erik Naggum
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3250270580629981@naggum.no>
* Pascal Costanza
| Here is an excerpt:

  Sigh.  As usual, the distance between the text you quote and the
  conclusion you want to support is measured in astronomical units.

  Could you, for /once/, try to back up some of your conclusions with a
  relevant set of /connections/ between the facts and your conclusion?

  It is by now evident beyond doubt that you believe you see things
  that are simply not there, probably because you are /unaware/ of
  your very own "contribution" to the process.  This is usually called
  psychosis, but if you could at least /try/ to explain how you got
  from observation to conclusion, there is a nonzero probability that
  you can be shown that you have only seen your own assumptions and
  preconceptions and have discarded all the evidence to the contrary.
  As long as you only look for confirmation of your preconceptions and
  are as happy as an astrologist or any other mystic when a prediction
  comes true despite thousands of predictions that did not, there is
  no way to provide you with counter-information, because only the
  information that supports something you already believe is important
  to you.  Just as you look away when you do not feel sufficiently
  warm and fuzzy and positive (or whatever), you evidently also look
  away when something could prove you wrong, because this, too, would
  make you feel less than stellar.  To most sane people, the very idea
  of turning a blind eye to everything that feels bad is tantamount to
  discarding one's sanity, but as you have confirmed previously, you
  do indeed check your feelings first.  The only possible result of
  this willful denial of the external world is that you defend ever
  more misguided notions and keep inventing things to support them.
  To put it very mildly, this bugs the hell out of me, but I would not
  bother to respond unless you had this uncontrollable urge to keep
  posting negative bullshit about me.

  You have not actually listened to anything I have written, but only
  "confirm" your mental image of me and behave as if I /am/ the image
  in /your/ mind.  You obviously do not think this is wrong, or you
  would have got a grip on yourself and repaired yourself, but the
  more you keep this fantastically annoying game going where the whole
  purpose of your increasingly /untrue/ claims about other people is
  to pretend that what you want to believe is true.  Normal human
  *decency* prevents most people from continuing to spout lies and so
  much hostile crap about somebody else when they have been arrested
  for their inability to accept corrections, but you seem to have made
  it your purpose here to keep pretending two things: (1) that you are
  a perfect, flawless saint without whom comp.lang.lisp will die a
  most horrible death, and (2) that I am some devil that you have to
  fight.  Since you are emotion-driven to the point of neurosis, you
  will not even understand what is wrong with your behavior, but you
  /have/ to notice the negative emotions you feel when I ask you to
  please stop lying so much.  You are not /actually/ braindead, are
  you?  There is still /something/ in you that accepts input from the
  external world that makes you feel bad and then shut down.  Or have
  you shut down so completely that you are now entirely of prejudice
  and feel bad even before you have read what I write to you?  I have
  my deepest doubts about your ability to read /anything/ that does
  not keep you emotionally on top of the world, but your unceasing
  need to keep talking about me is /pathological/.

  You /always/ have to "share" your lies about me with this forum when
  you feel sufficiently morally outraged to abandon your rationality.
  Why do you have to keep lying?  What on /earth/ compels you to keep
  making these idiotic "contributions" here?  You /must/ understand,
  even with your emotional needs, that when you keep lying and making
  all your hateful claims about me, you make things /worse/?  You are
  not /actually/ completely braindead, are you?

  You do not have to keep adding fuel to the flames whenever some of
  your kind rises to make his moronic and most uncivil "objections"
  because his puny brainpower does not grasp what is going on and he
  feels excluded.  You evidently strongly support such people and have
  made it your purpose here to make such people attack me, but that
  you have failed to understand how thoroughly /antisocial/ you are
  and that people tire of this forum because of scumbags like you and
  those you support, is impossible.  /Something/ in you must be able
  to see the damage you do.  Yet you keep doing it.  The alternative
  is that you are so morally outraged that whatever intelligence you
  might possess is rendered inoperative while you go on your holy
  mission, that your intelligence is a fair-weather intelligence that
  can only help you when the external world rubs you the right way.

  Part of being a /social/ animal, Pascal Costanza, is being able to
  deal with other people without having to /express/ your primitive
  emotions of tribal inclusion or exclusion.  I /know/ that you do not
  want me in this forum.  I /know/ that you are such a staggeringly
  stupid person as to be a functional illiterate when it comes to deal
  with counter-information to your beliefs.  I /know/ that you have so
  many negative emotions that your need to share them stems from self-
  preservation: You would /explode/ if you could not vent your spleen.
  And therefore, because you are such a hideous character, you believe
  that other people are like you, and then you do the stupidest thing
  that inferior people can do: They attack other people by proxy, by
  projecting their own evil onto them, by demonizing them and seeing
  monsters, and most importantly, by terminating their /dialog/ with
  the person they /hate/.  You are so clearly malfunctioning when it
  comes to dealing with such simple things as not making things worse
  by rearing your ugly head, that you cannot be expected to understand
  how destructive you are, but there are people who express a deep-
  rooted exasperation about /you/ and /your/ need to keep supporting
  every flaming moron in this newsgroup.  You drive people away from
  this forum, Pascal Costanza, by never letting go.  Like maniacal
  lunatic, you just /have/ to keep posting your insane hatred towards
  me and you /cannot/ stop.  I fear that you actually feel that you
  have to speek up lest you "support" me by your silence, but I urge
  you as strongly as I can, to give it a rest, let it go, get over it,
  move on, and accept that you do not /have/ to involve yourself each
  and every goddamn time someone says something negative about me.  We
  /all/ know how you feel.  It is /impossible/ not to know how much
  you hate me and how much you need to blame me for everything that
  you cannot deal with, so you do not have to keep repeating it.  As
  for me, I could not care less what you feel if you only shut up.

  Seeking validation of your personal feelings are /not/ the purpose
  of comp.lang.lisp, Pascal Costanza.  Please be so /decent/ as to
  keep most of them to yourself when they are as hateful and negative
  as they are.  They have no place here.  /Try/ to manage to think
  even though you feel a strong desire not to.  This newsgroup is
  /not/ about how Pascal Costanza feels about Erik Naggum.  It is
  about the programming language (family|Common) Lisp.  Can you please
  try to grasp this and stop telling your obnoxious lies about me?

  And do /not/ send another emotional message to this forum.  /You/
  opened this issue by blaming me, again, and /you/ can accept that
  you have been castigated for this idiocy and simply cease and
  desist.  That I think you are among the worst shitheads this planet
  has spawned should be pretty clear, but you do /not/ improve the
  matter by "defending" yourself.  To defend a mistake is to make
  another mistake.  Just accept that you made a mistake and resolve
  /not to do it again/.

  So, can we have 2003 without any more of your off-topic attacks on
  me and without any more of your off-topic support of every loser who
  feels bad and brandishes his emotional dysfunction for all to see?
  Send the goddamn loser /mail/ if you have to take on the burden that
  others cannot bear, but figure it out: /You/ destroy this forum when
  /you/ abuse it to stage your moronic wars against someone because
  /you/ think they are to blame.  /You/ are to blame for the damage
  that /you/ do.  Accept responsibility for yourself and be /nice/ if
  that is how you think others should behave.  Can you do all this?
  If no, respond with an emotional self-defense, but if yes, you will
  simply behave well and not talk about it.  Let your /actions/ speak
  for you in 2003.  Manage to establish positive feelings where you
  feel none yourself, and you will have proven your claims.  If you
  keep causing negative emotions like you have in the past when you
  lambast others for not considering other people emotions, you will
  admit that your theory is not only false, but destructive.  Do you
  believe that your theories actually work?  /Prove/ it in 2003!

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Kenny Tilton
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3E10A3E6.503@nyc.rr.com>
Pascal Costanza wrote:
> P.S.: I wonder what Common Lisp vendors think about this situation. 

That would be interesting. I wager "there is no such thing as bad 
publicity" is close. :)

> After all, anyone that shows some interest in Common Lisp is a potential 
> customer in the long run, unless scared away prematurely. Or am I 
> overestimating the impact of newsgroups?

Two things: One, no, I do not think anyone ever got interested in any 
language and then punted because of its NG.

Second, do what I have been doing. Watch objectively what happens when 
someone asks for support here (if I can put it that way). They get 
incredible support. Free. Within hours.

Sounds damn atractive to me.



-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Jacek Generowicz
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <tyfy966ctiv.fsf@lxplus029.cern.ch>
Kenny Tilton <·······@nyc.rr.com> writes:

> I do not think anyone ever got interested in any language and then
> punted because of its NG.

A number of people on c.l.python have claimed to have lost interest
in Lisp because of c.l.lisp (or Erik specifically).

Most of these people give the impression of being the sort of person
that Erik would be very glad to have driven away.

I guess you could claim that they were not really interested in Lisp
in the first place.

Which reminds me of something you posted a couple of days ago ...

> > A lot of people described c.l.l as a private elite club where only
> > certain class of people are welcome.
> 
> Really? I would love some details. People somewhere are sitting
> around talking about c.l.l?

c.l.py for a start.

> A /lot/ of people? Do they also talk about Lisp?  If so, what do
> they say?

You don't want to know.

> If not, why are they even reading cll?

Many of them have stopped.
From: Donald Fisk
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3E10CDD5.904AF297@enterprise.net>
Tim Bradshaw wrote:

> The last person who could be thought of as being hounded out of cll
> who I can think of was Ilias, although he kind of hounded himself out
> - I think by the end that basically no one would talk to him and he
> got bored and wandered off.  Can anyone - *anyone* - give examples of
> someone who really had a lot (or anything) to contribute who was
> hounded out?

Well, the very first person to post here (First message of new
newsgroup, 1982-03-27) felt unwelcome last year (I'm outta here,
2001-11-15) on and has not returned.

But I think he's a rare exception.   Most of the people rejected
by the newsgroup's immune system probably deserved it.

> --tim

Le Hibou
-- 
Dalinian: Lisp. Java. Which one sounds sexier?
RevAaron: Definitely Lisp. Lisp conjures up images of hippy coders,
drugs, sex, and rock & roll. Late nights at Berkeley, coding in
Lisp fueled by LSD.   Java evokes a vision of a stereotypical nerd,
with no life or social skills.
From: Coby Beck
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <auqr4n$6m$1@otis.netspace.net.au>
"Tim Bradshaw" <···@cley.com> wrote in message
····················@cley.com...
> The last person who could be thought of as being hounded out of cll
> who I can think of was Ilias, although he kind of hounded himself out
> - I think by the end that basically no one would talk to him and he
> got bored and wandered off.  Can anyone - *anyone* - give examples of
> someone who really had a lot (or anything) to contribute who was
> hounded out?

Well, for various definitions of "contribute" and "hound out" I always
thought that way about Erann Gat's long absence.

>
> I think that there are two relevant features of cll which make it
> somewhat ruder than (some - by no means all) other groups. (1) There
> are a relatively large number of people who turn up with extremely
> wrong and very negative views, and who are not interested in learning
> the truth.

I really do agree with this.  Ignorance when given the confidence that comes
with popular consensus is a formidable force.  Logic and reason often have
little effect.

> (2) Some regular
> posters say you are an idiot when they think you are.  So what's wrong
> with that?

I think we've been over all sides of that one before.  If you are asking the
question then I am sure any answer I can come up with would not sway you.  I
don't think that agreeing to disagree is the wrong thing to do about that.

It's interesting to note that lately the number of "leave me alone, you
feel-good jerks" posts seem (to me at least) to greatly outweigh the "take a
pill, you foul-mouth heathens" posts.  Is it true?  and What could it mean?
stay tuned..... ;)

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <aurt0s$s96$1@f1node01.rhrz.uni-bonn.de>
Coby Beck wrote:
> "Tim Bradshaw" <···@cley.com> wrote in message
> ····················@cley.com...
> 
>>The last person who could be thought of as being hounded out of cll
>>who I can think of was Ilias, although he kind of hounded himself out
>>- I think by the end that basically no one would talk to him and he
>>got bored and wandered off.  Can anyone - *anyone* - give examples of
>>someone who really had a lot (or anything) to contribute who was
>>hounded out?
> 
> 
> Well, for various definitions of "contribute" and "hound out" I always
> thought that way about Erann Gat's long absence.

I second that.

>>I think that there are two relevant features of cll which make it
>>somewhat ruder than (some - by no means all) other groups. (1) There
>>are a relatively large number of people who turn up with extremely
>>wrong and very negative views, and who are not interested in learning
>>the truth.
> 
> 
> I really do agree with this.  Ignorance when given the confidence that comes
> with popular consensus is a formidable force.  Logic and reason often have
> little effect.

I have made different experiences. It's hard for Lisp newbies to get the 
basic facts about Lisp right, but I have made the experience that if you 
explain those facts repeatedly then they get it. It's much more 
effective than telling them that they are stupid and should grow up and 
what not. I have made this experiment recently in comp.lang.python and 
have seen two things happening: (1) People weren't nearly as hostile as 
some are in cll. (2) Some people have understood the basic facts about 
Lisp, and I have gotten some wonderful feedback.

You can easily find the thread in google.


Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Erann Gat
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <gat-3112021011550001@192.168.1.51>
In article <············@f1node01.rhrz.uni-bonn.de>, Pascal Costanza
<········@web.de> wrote:

> Coby Beck wrote:
> > "Tim Bradshaw" <···@cley.com> wrote in message
> > ····················@cley.com...
> > 
> >>The last person who could be thought of as being hounded out of cll
> >>who I can think of was Ilias, although he kind of hounded himself out
> >>- I think by the end that basically no one would talk to him and he
> >>got bored and wandered off.  Can anyone - *anyone* - give examples of
> >>someone who really had a lot (or anything) to contribute who was
> >>hounded out?
> > 
> > 
> > Well, for various definitions of "contribute" and "hound out" I always
> > thought that way about Erann Gat's long absence.
> 
> I second that.

Heh, thanks for the compliment, but which "long absence" are you referring
to?  I've posted over 100 articles to c.l.l. this year.  (I was gone for a
year circa 2000-2001, but I wasn't "hounded out", I was working for Google
and didn't have any time for fun :-(

Nonetheless, I do believe that many potential contributors are "hounded
out" along with the trolls (and that the trolls would go away a lot faster
if they were ignored rather than hounded).  The problem is that once
they're gone no one remembers them.  John Foderaro comes to mind as a much
better example than me of a contributor who really did get hounded out. 
His case is pretty much unambiguous for any reasonable reading of
"contribution" and "hounding."  And I believe that John's experience,
while perhaps extreme, is not unique.  :-(

E.
From: Coby Beck
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <autb9b$11vm$1@otis.netspace.net.au>
"Erann Gat" <···@jpl.nasa.gov> wrote in message
·························@192.168.1.51...
> In article <············@f1node01.rhrz.uni-bonn.de>, Pascal Costanza
> <········@web.de> wrote:
>
> > Coby Beck wrote:
> > >
> > > Well, for various definitions of "contribute" and "hound out" I always
> > > thought that way about Erann Gat's long absence.
> >
> > I second that.
>
> Heh, thanks for the compliment, but which "long absence" are you referring
> to?  I've posted over 100 articles to c.l.l. this year.  (I was gone for a
> year circa 2000-2001, but I wasn't "hounded out", I was working for Google
> and didn't have any time for fun :-(

That is probably the time period I am thinking of, though I thought it was
longer than that.  I just remember many flames involving you (and not just
Erik) when I began lurking and then suddenly nothing until (as you point out
above) this year.

The rest is obviously my incorrect assumption for which I apologize.

> Nonetheless, I do believe that many potential contributors are "hounded
> out" along with the trolls (and that the trolls would go away a lot faster
> if they were ignored rather than hounded).  The problem is that once
> they're gone no one remembers them.

I really do concur with this, both points.

> John Foderaro comes to mind as a much
> better example than me of a contributor who really did get hounded out.
> His case is pretty much unambiguous for any reasonable reading of
> "contribution" and "hounding."

Well, having participated in that "hounding" myself, I am sorry for the end
result though continue to think he was in the wrong in many important ways.
However, I hold a much different yardstick when it comes to mature,
technically knowledgable and experienced lispers.  I think if someone like
that leaves, it is for much more compilicated and personal reasons than a
few harsh posts.

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: Erann Gat
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <gat-3112021912160001@192.168.1.51>
In article <·············@otis.netspace.net.au>, "Coby Beck"
<·····@mercury.bc.ca> wrote:

> > John Foderaro comes to mind as a much
> > better example than me of a contributor who really did get hounded out.
> > His case is pretty much unambiguous for any reasonable reading of
> > "contribution" and "hounding."
> 
> Well, having participated in that "hounding" myself, I am sorry for the end
> result though continue to think he was in the wrong in many important ways.

That sounds to me like, "Well, I'm sorry John is gone, but that is just
the price we pay for doing the necessary (or at least morally justified)
thing of hounding people who are wrong in many important ways."

If this is a fair rephrasing of your sentiment, then I think you are wrong
in many important ways.  Does that justify my hounding you the way you and
others hounded John?

Ironically, I also happened to disagree with John concerning the topic
that sparked that particular flamefest.  (I wonder how many people can
remember what the fuss was about without looking it up.)  Regardless, I
consider his departure a net loss.  If you chase away everyone you
disagree with pretty soon there will be no one left.

> However, I hold a much different yardstick when it comes to mature,
> technically knowledgable and experienced lispers.  I think if someone like
> that leaves, it is for much more compilicated and personal reasons than a
> few harsh posts.

So what?  The end result is the same.  He's gone, and we are the poorer
for it.  IMO the Lisp community is far too small to allow itself to
indulge in ideological cleansing.

E.
From: Kenny Tilton
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3E129FCB.9060503@nyc.rr.com>
Erann Gat wrote:
> In article <·············@otis.netspace.net.au>, "Coby Beck"
> <·····@mercury.bc.ca> wrote:
> 
> 
>>>John Foderaro comes to mind as a much
>>>better example than me of a contributor who really did get hounded out.
>>>His case is pretty much unambiguous for any reasonable reading of
>>>"contribution" and "hounding."
>>
>>Well, having participated in that "hounding" myself, I am sorry for the end
>>result though continue to think he was in the wrong in many important ways.
> 
> 
> That sounds to me like, "Well, I'm sorry John is gone, but that is just
> the price we pay for doing the necessary (or at least morally justified)
> thing of hounding people who are wrong in many important ways."
> 
> If this is a fair rephrasing of your sentiment, then I think you are wrong
> in many important ways.  Does that justify my hounding you the way you and
> others hounded John?

Hounded schmounded. This is a public forum, ferchissakes. If some 
butthead (in whosever opinion) is posting nonsense, don't read it, and 
for god's sake don't respond to it.

But you and John participate in the back and forth as if it were the 
most important thing in the world. If that's what you think, super, 
wallow on, why stop? If not, why perpetuate those train wrecks of dialogue?

Regardless, do not blame your antagonist for the choices you make. It is 
simply astonishing to think I have to stop posting to an NG because some 
denizen has it in for me. Tthis is a great fucking NG. Pardon the 
suck-up, but the regulars here are the bomb. Someone has to stop sharing 
with them because of one or two antagonists? Not.

I'll take it further. Even when A and B get into it for a few weeks, 
once it is over it is over. Only when one party gratuitously drags up 
something from the past does the other party again take up arms. That's 
pretty mature of everyone, IMO.

peace y'all, 2003.

-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Daniel Barlow
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87wulo6d9q.fsf@noetbook.telent.net>
···@jpl.nasa.gov (Erann Gat) writes:

> So what?  The end result is the same.  He's gone, and we are the poorer
> for it.  IMO the Lisp community is far too small to allow itself to
> indulge in ideological cleansing.

One of the things I had confirmed to me at ILC is that the Lisp
community (even the predominantly-Common Lisp community) is bigger and
has a wider range of opinions than those which appear to be Allowed in
comp.lang.lisp

You could claim that this indicates that something is wrong with
comp.lang.lisp, but to be honest I think it's more likely that
something is wrog with Usenet.


Happy New Year.  

-dan

-- 

   http://www.cliki.net/ - Link farm for free CL-on-Unix resources 
From: Coby Beck
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <autucm$15oe$1@otis.netspace.net.au>
http://groups.google.com/groups?selm=oQZi7.967%24Fu5.687099%40typhoon.tampab
ay.rr.com&rnum=58
http://groups.google.com/groups?selm=QN6j7.32%248c3.46651%40typhoon.tampabay
.rr.com&rnum=78



"Erann Gat" <···@jpl.nasa.gov> wrote in message
·························@192.168.1.51...
> In article <·············@otis.netspace.net.au>, "Coby Beck"
> <·····@mercury.bc.ca> wrote:
>
> > > John Foderaro comes to mind as a much
> > > better example than me of a contributor who really did get hounded
out.
> > > His case is pretty much unambiguous for any reasonable reading of
> > > "contribution" and "hounding."
> >
> > Well, having participated in that "hounding" myself, I am sorry for the
end
> > result though continue to think he was in the wrong in many important
ways.
>
> That sounds to me like, "Well, I'm sorry John is gone, but that is just
> the price we pay for doing the necessary (or at least morally justified)
> thing of hounding people who are wrong in many important ways."

The scare quotes above were meant to imply a non-literal meaning for
hounding.  To be perfectly clear:  I do not feel I did anything wrong and
still stand behind my points and the tone they were stated in.  I even went
back to the "the \"loop\" macro" thread to review.  My "harshest" posts are
here:
http://groups.google.com/groups?selm=oQZi7.967%24Fu5.687099%40typhoon.tampab
ay.rr.com&rnum=58
http://groups.google.com/groups?selm=QN6j7.32%248c3.46651%40typhoon.tampabay
.rr.com&rnum=78

John chose not to respond to any of those points, but I never berated him
for it.  I think most people disagreeing with him did so in civil and
logical fashions.

> > However, I hold a much different yardstick when it comes to mature,
> > technically knowledgable and experienced lispers.  I think if someone
like
> > that leaves, it is for much more compilicated and personal reasons than
a
> > few harsh posts.
>
> So what?  The end result is the same.  He's gone, and we are the poorer
> for it.

I agree.  But John left for his own personal reasons.  There was alot of
history and a lot of politics and alot of personal feuding at play.
"FoderaroGate" was not an idealogical cleansing.  But I really did not
intend to get into this particular unpleasantness so deeply.

The situations that bother me more are those involving newcomers.

> IMO the Lisp community is far too small to allow itself to
> indulge in ideological cleansing.

I agree with your point.

--
Coby Beck
(remove #\Space "coby 101 @ bigpond . com")
From: faust
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <fqg61v0qigdchgj6uv1sts8mvuv1trirsg@4ax.com>
 ···@jpl.nasa.gov (Erann Gat) ,  emitted these fragments:

>John Foderaro comes to mind as a much
>better example than me of a contributor who really did get hounded out. 

See:
http://groups.google.com/groups?q=John+Foderaro&hl=en&lr=&ie=UTF-8&oe=utf-8&sa=N&tab=wg

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <av0djp$rt1$1@newsreader2.netcologne.de>
Erann Gat wrote:
> In article <············@f1node01.rhrz.uni-bonn.de>, Pascal Costanza
> <········@web.de> wrote:
> 
> 
>>Coby Beck wrote:
>>
>>>"Tim Bradshaw" <···@cley.com> wrote in message
>>>····················@cley.com...
>>>
>>>
>>>>The last person who could be thought of as being hounded out of cll
>>>>who I can think of was Ilias, although he kind of hounded himself out
>>>>- I think by the end that basically no one would talk to him and he
>>>>got bored and wandered off.  Can anyone - *anyone* - give examples of
>>>>someone who really had a lot (or anything) to contribute who was
>>>>hounded out?
>>>
>>>
>>>Well, for various definitions of "contribute" and "hound out" I always
>>>thought that way about Erann Gat's long absence.
>>
>>I second that.
> 
> 
> Heh, thanks for the compliment, but which "long absence" are you referring
> to?  

The one you announced in the beginning of October, but fortunately you 
didn't keep your "promise". ;-)


Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Nils Goesche
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <874r8utfao.fsf@darkstar.cartan>
Pascal Costanza <········@web.de> writes:

> Coby Beck wrote:
> > "Tim Bradshaw" <···@cley.com> wrote in message
> > ····················@cley.com...
> >
> >>Can anyone - *anyone* - give examples of someone who really
> >>had a lot (or anything) to contribute who was hounded out?

> > Well, for various definitions of "contribute" and "hound out"
> > I always thought that way about Erann Gat's long absence.

> I second that.

Well, I have a very different theory about that.  But I won't say
which.  Because, in order to do that, I would have to participate
in the extremely disgusting activity of publicly accusing and
criticizing other people's personalities in the third person.  I
am puzzled over the question why anybody would regard that as
acceptable behavior.  I guess it is some remnant of old Stalinist
show trials and the regular ``self-criticizing sessions��
practicized in factories in the Eastern block, communist party
cells all over the world and Hippie communes in the West.  In
fact, those rituals were nothing but vicious denunciation
rituals, often resulting in death for the so accused (in the
West, too, sometimes).  There was a slogan: ``The private is
political�� or some such, used to justify this atrocity.  That PC
people today still enjoy this activity so much is probably an
expression of their desire to extend their dictatorial control
even to people's private lives and people's minds.  For they have
always known that the free mind, the thinking man is their most
dangerous enemy.  Their control isn't total until they can make
people actually believe that ``2 + 2 = 5�� whenever they want.

> >>I think that there are two relevant features of cll which
> >>make it somewhat ruder than (some - by no means all) other
> >>groups. (1) There are a relatively large number of people who
> >>turn up with extremely wrong and very negative views, and who
> >>are not interested in learning the truth.

> > I really do agree with this.  Ignorance when given the
> > confidence that comes with popular consensus is a formidable
> > force.  Logic and reason often have little effect.

> I have made different experiences. It's hard for Lisp newbies
> to get the basic facts about Lisp right, but I have made the
> experience that if you explain those facts repeatedly then they
> get it.

Yes, that's the thing to do when they simply don't get it the
first time.  Unless the reason they don't get it is that they are
operating under wrong preconceptions.  In that case, a second
explanation will not work, either.  You have to make them get rid
of their wrong preconceptions first, and when there is strong
resistance to that, as is often the case, you have to apply
force.  Sure, this might hurt them, but it is for their own
benefit.  As Coby said, their wrong preconceptions are often
backed up by popular consensus, which might be the reason that we
have to break this resistance more often in comp.lang.lisp than
anywhere else.

> It's much more effective than telling them that they are stupid
> and should grow up and what not.

Even that can be very useful.  When people are not willing to
consider the possibility that they are wrong on basic
assumptions, in short, when they are being stupid and immature,
the only thing that might /possibly/ work is to slap them and
shake them.  You have to show them that what they are saying is
actually /not/ just another ``opinion�� an intelligent man may
have, and that they might as well try to continue defending for
ages, but totally irrational, stupid and brain-damaged.
Sometimes, the shock resulting from this actually works by making
them consider, possibly for the first time in their life, whether
their most dearest preconceptions might in fact be wrong.
Although this is a painful experience for them, and it might take
some time until they recover and actually begin to think, if it
works you have done them the greatest possible favor.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ausddn$mj3$1@newsreader2.netcologne.de>
Nils Goesche wrote:
> Pascal Costanza <········@web.de> writes:
> 
> 
>>Coby Beck wrote:
>>
>>>"Tim Bradshaw" <···@cley.com> wrote in message
>>>····················@cley.com...
>>>
>>>
>>>>Can anyone - *anyone* - give examples of someone who really
>>>>had a lot (or anything) to contribute who was hounded out?
> 
> 
>>>Well, for various definitions of "contribute" and "hound out"
>>>I always thought that way about Erann Gat's long absence.
> 
> 
>>I second that.
> 
> 
> Well, I have a very different theory about that.  But I won't say
> which.  
[...]

So what's your point?

>>I have made different experiences. It's hard for Lisp newbies
>>to get the basic facts about Lisp right, but I have made the
>>experience that if you explain those facts repeatedly then they
>>get it.
> 
> 
> Yes, that's the thing to do when they simply don't get it the
> first time.  Unless the reason they don't get it is that they are
> operating under wrong preconceptions.  In that case, a second
> explanation will not work, either.  You have to make them get rid
> of their wrong preconceptions first, and when there is strong
> resistance to that, as is often the case, you have to apply
> force.  

Nonsense!

> Sure, this might hurt them, but it is for their own
> benefit.

Pure and utter nonsense!

>>It's much more effective than telling them that they are stupid
>>and should grow up and what not.
> 
> 
> Even that can be very useful.  When people are not willing to
> consider the possibility that they are wrong on basic
> assumptions, in short, when they are being stupid and immature,
> the only thing that might /possibly/ work is to slap them and
> shake them.  

Did it ever occur to you that you might be the one who's wrong on basic 
assumptions?


Pascal

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Nils Goesche
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87znqmry2z.fsf@darkstar.cartan>
Pascal Costanza <········@web.de> writes:

> So what's your point?
...
> Nonsense!
...
> Pure and utter nonsense!
...
> Did it ever occur to you that you might be the one who's wrong
> on basic assumptions?

Oh, yes.  And sometimes, I am.  I change them, when I find out.
It is quite clear now, that you have never done the same before.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ausjh5$83d$1@newsreader2.netcologne.de>
Nils Goesche wrote:
> Pascal Costanza <········@web.de> writes:
> 
> 
>>So what's your point?
> 
> ...
> 
>>Nonsense!
> 
> ...
> 
>>Pure and utter nonsense!
> 
> ...
> 
>>Did it ever occur to you that you might be the one who's wrong
>>on basic assumptions?
> 
> 
> Oh, yes.  And sometimes, I am.  I change them, when I find out.
> It is quite clear now, that you have never done the same before.

Oh dear. What's the basis for that conclusion of yours? (Never?!?)

Anyway, you completely missed my point. (Obviously slapping you wasn't 
sufficient - seems like your theory isn't that well-founded. 
<blink>SATIRE</blink>)

Now here is my point: If you take it for granted that you might be 
wrong, what gives you the right to slap other people in their faces?


Pascal


-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Nils Goesche
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87r8byrs0t.fsf@darkstar.cartan>
Pascal Costanza <········@web.de> writes:

> Nils Goesche wrote:
> > Pascal Costanza <········@web.de> writes:
> >
> >>So what's your point?
> > ...
> >
> >>Nonsense!
> > ...
> >
> >>Pure and utter nonsense!
> > ...
> >
> >>Did it ever occur to you that you might be the one who's
> >>wrong on basic assumptions?

> > Oh, yes.  And sometimes, I am.  I change them, when I find
> > out.  It is quite clear now, that you have never done the
> > same before.
> 
> Oh dear. What's the basis for that conclusion of yours?
> (Never?!?)

That you are having such a hard time understanding all this.

> Anyway, you completely missed my point. (Obviously slapping you wasn't
> sufficient - seems like your theory isn't that
> well-founded. <blink>SATIRE</blink>)

Oh, I knew what you meant.

> Now here is my point: If you take it for granted that you might
> be wrong, what gives you the right to slap other people in
> their faces?

That I might be wrong does not imply that I must not assume that
I am right.  Suppose somebody gave me a ``proof�� that his method
of trisecting an angle with compass and straightedge worked in
all cases.  I know why no such ``proof�� is possible.  So, I will
not read his proof and, when I have had a nice day, try to
convince him that he shouldn't look for one, either.  Now, I know
that I might be wrong on this: Maybe there is something wrong
with Galois' theory, or maybe there is something wrong with
mathematics in general, or whatnot, and I simply missed it.  But
I think it is much more likely that my theory of what's wrong
with the other guy's thinking is correct and I will try to
explain this to him.  If you have ever dealt with an angle
trisector or a circle squarer or a Fermat's Last Theorem prover
(except Wiles), you'll know that these people will feel very
insulted and hurt because I am not willing to look at their proof
and show where they go wrong.  Sometimes they have spent years
looking for a proof, and I am telling them that all those time
was wasted?  This will hurt them very much.  The alternative is
being nice: Look at their proof and show them where they go
wrong.  They'll walk off and continue wasting their time trying
to find a proof.  Dealing with lefties and greens is very
similar, actually.  They'll shout and spit and yell at me that I
feel like the Exorcist himself, but there is no way to cure them
without making them angry (but I don't actually chant ``The power
of Smith compels you!��).

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Erann Gat
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <gat-3112021923510001@192.168.1.51>
In article <··············@darkstar.cartan>, Nils Goesche <···@cartan.de> wrote:

> Suppose somebody gave me a ``proof�� that his method
> of trisecting an angle with compass and straightedge worked in
> all cases.  I know why no such ``proof�� is possible.  So, I will
> not read his proof and, when I have had a nice day, try to
> convince him that he shouldn't look for one, either.

Sometimes impossible "proofs" can be useful.  In
http://www.flownet.com/gat/QM.pdf there is a "proof" that quantum theory
leads to faster-than-light communication.  Many people (including a few
card-carrying physicists) have found this "proof" enlightening because it
highlights a pervasive but false assumption in popular presentations of
quantum theory.  So if you refuse to read any proof that you know to be
impossible you may in fact be missing something important.

E.
From: Nils Goesche
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87wulodyc7.fsf@darkstar.cartan>
···@jpl.nasa.gov (Erann Gat) writes:

> In article <··············@darkstar.cartan>, Nils Goesche <···@cartan.de> wrote:
> 
> > Suppose somebody gave me a ``proof�� that his method of
> > trisecting an angle with compass and straightedge worked in
> > all cases.  I know why no such ``proof�� is possible.  So, I
> > will not read his proof and, when I have had a nice day, try
> > to convince him that he shouldn't look for one, either.
> 
> Sometimes impossible "proofs" can be useful.  In
> http://www.flownet.com/gat/QM.pdf there is a "proof" that
> quantum theory leads to faster-than-light communication.  Many
> people (including a few card-carrying physicists) have found
> this "proof" enlightening because it highlights a pervasive but
> false assumption in popular presentations of quantum theory.
> So if you refuse to read any proof that you know to be
> impossible you may in fact be missing something important.

Heh.  Well, actually, I /do/ read them sometimes.  Usually,
though, you can see after a few seconds what particular kind of
crankiness you are dealing with and looking any further into it
is a total waste of time then, if not for the entertainment
value.  There is even a very good book about cranks:

``Mathematical Cranks�� by Underwood Dudley:

http://www.amazon.com/exec/obidos/tg/detail/-/0883855070/qid=1041478036/sr=8-1/ref=sr_8_1/002-4146620-4803200?v=glance&s=books&n=507846

Highly recommended.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <av0d1j$qoi$1@newsreader2.netcologne.de>
Nils Goesche wrote:
> Pascal Costanza <········@web.de> writes:
> 

>>Now here is my point: If you take it for granted that you might
>>be wrong, what gives you the right to slap other people in
>>their faces?
> 
> 
> That I might be wrong does not imply that I must not assume that
> I am right.  

I didn't say so.

> Suppose somebody gave me a ``proof�� that his method
> of trisecting an angle with compass and straightedge worked in
> all cases.  I know why no such ``proof�� is possible.  So, I will
> not read his proof and, when I have had a nice day, try to
> convince him that he shouldn't look for one, either. 
[...]

It's interesting that you recur to the area of mathematics in order to 
illustrate your point. It's relatively easy to prove or disprove 
mathematical theories because you only need to show that they are either 
true or false. (Note that I have said _relatively_.)

Some properties of programming languages are also of that kind - for 
example you can prove certain properties of type systems. However, most 
of the far more interesting properties cannot be reasoned about in such 
terms. You cannot "mathematically" prove, say, the utility of a 
programming language. You cannot formally describe the importance of 
following a standard and to what extent.

The reason is that these properties are "socially constructed".

For example, this is one of the most fundamental errors of some members 
of the type system community: They claim that (static) type systems are 
useful and seem to imply that formal proofs of particular type systems 
back this claim somehow.

So my point is that, when it comes to programming languages (and 
computer science in general), there are very few facts that can easily 
determined to be right or wrong.

I don't think it makes any sense at all to try to convince someone that 
a particular programming language is "the best" or "the right". It's 
much better to try to enable people to make up their own minds - and 
this means that you need to show them the whole spectrum of 
possibilities. (For example, if someone asks whether static type systems 
are useful or not, you shouldn't only talk about those statically typed 
languages that can most easily be torn apart.)


Pascal

P.S., disclaimer: To make this absolutely clear, I think that Common 
Lisp is the best programming language out there, and I don't think that 
static type systems are very useful.

-- 
Given any rule, however �fundamental� or �necessary� for science, there 
are always circumstances when it is advisable not only to ignore the 
rule, but to adopt its opposite. - Paul Feyerabend
From: Nils Goesche
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87smwcdvin.fsf@darkstar.cartan>
Pascal Costanza <········@web.de> writes:

> Nils Goesche wrote:
> > Pascal Costanza <········@web.de> writes:
> >

> >>Now here is my point: If you take it for granted that you
> >>might be wrong, what gives you the right to slap other people
> >>in their faces?

> > That I might be wrong does not imply that I must not assume that
> > I am right.

> I didn't say so.

You didn't?  I think this is precisely what you're saying :-)

> > Suppose somebody gave me a ``proof�� that his method of
> > trisecting an angle with compass and straightedge worked in
> > all cases.  I know why no such ``proof�� is possible.  So, I
> > will not read his proof and, when I have had a nice day, try
> > to convince him that he shouldn't look for one, either.

> It's interesting that you recur to the area of mathematics in
> order to illustrate your point. It's relatively easy to prove
> or disprove mathematical theories because you only need to show
> that they are either true or false. (Note that I have said
> _relatively_.)

One reason I often choose examples from mathematics is simply
that those examples are usually the first that come to my mind.
In this case, I deliberately didn't try to find a different one
from another field to emphasize the important part of the
argument.  By giving my hypothetical opponent a position that is
obviously and indisputably wrong, I am directing the focus on the
/interesting/ part, which is: How do we deal with people who are
wrong?  Who make false assumptions and are not willing to change
their assumptions?  I claim that all you can do is /either/ to
give up hope, not helping them and knowingly letting them
continue to live in ignorance and darkness, /or/ insist that they
change their wrong assumptions (or way of thinking), even though
we know that doing so will be a painful process for them.

> Some properties of programming languages are also of that kind
> - for example you can prove certain properties of type
> systems. However, most of the far more interesting properties
> cannot be reasoned about in such terms. You cannot
> "mathematically" prove, say, the utility of a programming
> language. You cannot formally describe the importance of
> following a standard and to what extent.

I know that, of course.  But that's not the point.  We are
/assuming/ for the moment, and above you claim that you think
that we are allowed to /assume/ (under certain restrictions and
conditions that are not interesting at the moment) that we are
right and seeing clearly what's wrong with the other guy's
thinking.  Moreover, we are /assuming/, for the moment, that the
other guy is /not/ willing to give up his wrong preconceptions
despite our explanations why they are wrong.  What shall we do
then?  This seems to be the point of divergence.

> So my point is that, when it comes to programming languages
> (and computer science in general), there are very few facts
> that can easily determined to be right or wrong.

Sure, but that is entirely not the point.

> I don't think it makes any sense at all to try to convince
> someone that a particular programming language is "the best" or
> "the right".

Why not?  That I cannot formally prove that I'm right doesn't
mean that I must not try to convince the other guy, anyway.

> It's much better to try to enable people to make up their own
> minds - and this means that you need to show them the whole
> spectrum of possibilities.

Sure, that is the first step.  Sometimes, however, you can tell
that the other guy is ``blinded�� by wrong preconceptions and is
not able to see what's clearly lying in front of him.  In that
case, you have to do something about his blindness (or give up on
him.  I will give up when I think that the other guy is simply
too dumb to ever get the point.  Otherwise, I'll continue
trying).

> (For example, if someone asks whether static type systems are
> useful or not, you shouldn't only talk about those statically
> typed languages that can most easily be torn apart.)

Um, of course.  Do you think I don't do that?  I often take
examples in SML, which I think is the best of the statically
typed languages.  If Lisp was magically removed from the world,
SML is what I would use instead.  No Python, no Java.  If you
prefer Java, fine, you make your examples in Java.  That doesn't
imply everybody else has to make examples in Pascal Costanza's
favorite statically typed language, too.

Unless, of course, we are specifically talking about another
subfamily of languages.  When everybody is talking about C, or
C++, or Java, it doesn't always make much sense to chime in with
SML or whatnot.

Regards,
-- 
Nils G�sche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xD26EF2A0
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <av1bvs$106i$1@f1node01.rhrz.uni-bonn.de>
Nils Goesche wrote:
> Pascal Costanza <········@web.de> writes:
> 
> 
>>Nils Goesche wrote:
>>
>>>Pascal Costanza <········@web.de> writes:
>>>
> 
> 
>>>>Now here is my point: If you take it for granted that you
>>>>might be wrong, what gives you the right to slap other people
>>>>in their faces?
> 
> 
>>>That I might be wrong does not imply that I must not assume that
>>>I am right.
> 
> 
>>I didn't say so.
> 
> 
> You didn't?  I think this is precisely what you're saying :-)
> 
>

Uhm, well, what can I say? It's not what I am saying. ;-)

That I am right does not imply that I must assume that you are wrong, 
even if our views are incompatible.

> One reason I often choose examples from mathematics is simply
> that those examples are usually the first that come to my mind.

...and I think this is a sign that we're talking about different levels.

 > How do we deal with people who are
> wrong?  Who make false assumptions and are not willing to change
> their assumptions?  I claim that all you can do is /either/ to
> give up hope, not helping them and knowingly letting them
> continue to live in ignorance and darkness, /or/ insist that they
> change their wrong assumptions (or way of thinking), even though
> we know that doing so will be a painful process for them.

Well, I basically think that this is a self-fulfilling prophecy. If you 
focus too much on the painful elements of the process, it will turn out 
to be a painful process. If you focus on the rewarding aspects it will 
be a rewarding process.

>>So my point is that, when it comes to programming languages
>>(and computer science in general), there are very few facts
>>that can easily determined to be right or wrong.
> 
> 
> Sure, but that is entirely not the point.

That's _exactly_ my point!

>>(For example, if someone asks whether static type systems are
>>useful or not, you shouldn't only talk about those statically
>>typed languages that can most easily be torn apart.)
> 
> 
> Um, of course.  Do you think I don't do that?  

No. This was just an example. I am sorry if this has given a wrong 
impression about you.

> If you
> prefer Java, fine, you make your examples in Java.  That doesn't
> imply everybody else has to make examples in Pascal Costanza's
> favorite statically typed language, too.

Huh?!? What makes you think that Java is my favorite statically typed 
language? (Or am I just misunderstanding you?)

Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: faust
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <qjg61vcng2qul7ik50g7a252lgorig0mtg@4ax.com>
 Nils Goesche <···@cartan.de> ,  emitted these fragments:

>> Did it ever occur to you that you might be the one who's wrong
>> on basic assumptions?
>
>Oh, yes.  And sometimes, I am.  I change them, when I find out.
>It is quite clear now, that you have never done the same before.

Let me guess, you model yourself on Stalin ?

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: faust
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ufg61vkfumpt4d9kr5mkt0mddl4ppcogq9@4ax.com>
 Nils Goesche <···@cartan.de> ,  emitted these fragments:

>and when there is strong
>resistance to that, as is often the case, you have to apply
>force.  Sure, this might hurt them, but it is for their own
>benefit. 

Of course, it is more likely that rather than feeling any benefit,
they will just walk away and tell others about the "kooks in cll" just
like they are doing on several newsgroups and mailing lists right now.

--------------------------------------------------------
Come see,
real flowers
of this pain-filled world.

(from Basho)
From: Adam Warner
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <pan.2003.01.02.00.41.39.227403@consulting.net.nz>
Hi Pascal Costanza,

> I have made different experiences. It's hard for Lisp newbies to get the
> basic facts about Lisp right, but I have made the experience that if you
> explain those facts repeatedly then they get it. It's much more
> effective than telling them that they are stupid and should grow up and
> what not. I have made this experiment recently in comp.lang.python and
> have seen two things happening: (1) People weren't nearly as hostile as
> some are in cll. (2) Some people have understood the basic facts about
> Lisp, and I have gotten some wonderful feedback.
> 
> You can easily find the thread in google.

Having now read through most of the extraordinarily long thread I thought
your many responses were superb. I consider this your finest response:

http://groups.google.com/groups?as_umsgid=asdt1u%249tn%241%40newsreader2.netcologne.de

Rivalled by your introductory course to macros:
http://groups.google.com/groups?as_umsgid=aqkdoq%24db6%241%40newsreader2.netcologne.de

All the Lispers handled themselves wonderfully (e.g. many great
contributions from Kenny).

This is advocacy at its finest. Distilled truth put in terms that
newcomers can begin to understand.

Regards,
Adam
From: Dorai Sitaram
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <av1u63$idc$1@news.gte.com>
In article <···············@cley.com>, Tim Bradshaw  <···@cley.com> wrote:
>* Kenny Tilton wrote:
>> I realize you are not speaking strictly, because no un-moderated NG
>> can be private or elite. Unwelcome would be possible, if the
>> preponderance of opinion re some poster was so negative that they
>> could not take the heat and went away. I've seen that happen, with
>> people who were really off base. Nothing wrong with making bozos feel
>> unwelcome, I hope.
>
>The last person who could be thought of as being hounded out of cll
>who I can think of was Ilias, although he kind of hounded himself out
>- I think by the end that basically no one would talk to him and he
>got bored and wandered off.  Can anyone - *anyone* - give examples of
>someone who really had a lot (or anything) to contribute who was
>hounded out?

John Foderaro. 
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86k7hntim3.fsf@bogomips.optonline.net>
····@goldshoe.gte.com (Dorai Sitaram) writes:

> In article <···············@cley.com>, Tim Bradshaw  <···@cley.com> wrote:
> >* Kenny Tilton wrote:
> >> I realize you are not speaking strictly, because no un-moderated NG
> >> can be private or elite. Unwelcome would be possible, if the
> >> preponderance of opinion re some poster was so negative that they
> >> could not take the heat and went away. I've seen that happen, with
> >> people who were really off base. Nothing wrong with making bozos feel
> >> unwelcome, I hope.
> >
> >The last person who could be thought of as being hounded out of cll
> >who I can think of was Ilias, although he kind of hounded himself out
> >- I think by the end that basically no one would talk to him and he
> >got bored and wandered off.  Can anyone - *anyone* - give examples of
> >someone who really had a lot (or anything) to contribute who was
> >hounded out?
> 
> John Foderaro. 

As one of the major hounds in the last loop/alegroserv/if* macro suck
thread I can honestly say I do not think I or the rest of the people 
involved had a direct influence on him leaving.  The reason for this 
is the timing of his first apology.  The loop thread flared up over
labor day weekend, first weekend in sept and you get Monday as a
holiday, and he apologized at around 11am his time.  This points to
someone or group who has real influence over him, management and/or
coworkers, had a chat with him when he showed up to work about what
he was doing over the weekend.  After a few weeks he starts reverting
to type.  In a reply to one of his posts I asked where it the manager
that was vetting your posts and if he is sick I hope he gets better
soon. And your unsupervised post were harming Franz.  Soon after that 
he posted the "I'm outa here" message.

I personally think it was a good thing for Franz that he stopped
posting.  He was posting things that were not in Franz's best interest
and he was/is a senior/executive person at Franz.  And he was doing
most of them without a disclaimer of any sort on them to even try to
protect Franz, he was using his work account to post.

marc
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86bs2ztdy8.fsf@bogomips.optonline.net>
Marc Spitzer <········@optonline.net> writes:

> ····@goldshoe.gte.com (Dorai Sitaram) writes:
> 
> > In article <···············@cley.com>, Tim Bradshaw  <···@cley.com> wrote:
> > >* Kenny Tilton wrote:
> > >> I realize you are not speaking strictly, because no un-moderated NG
> > >> can be private or elite. Unwelcome would be possible, if the
> > >> preponderance of opinion re some poster was so negative that they
> > >> could not take the heat and went away. I've seen that happen, with
> > >> people who were really off base. Nothing wrong with making bozos feel
> > >> unwelcome, I hope.
> > >
> > >The last person who could be thought of as being hounded out of cll
> > >who I can think of was Ilias, although he kind of hounded himself out
> > >- I think by the end that basically no one would talk to him and he
> > >got bored and wandered off.  Can anyone - *anyone* - give examples of
> > >someone who really had a lot (or anything) to contribute who was
> > >hounded out?
> > 
> > John Foderaro. 
> 
> As one of the major hounds in the last loop/alegroserv/if* macro suck
> thread I can honestly say I do not think I or the rest of the people 
> involved had a direct influence on him leaving.  The reason for this 
> is the timing of his first apology.  The loop thread flared up over
> labor day weekend, first weekend in sept and you get Monday as a
> holiday, and he apologized at around 11am his time.  This points to
                                                    ^ on Tuesday

Thats what I get for not proof reading.

marc

> someone or group who has real influence over him, management and/or
> coworkers, had a chat with him when he showed up to work about what
> he was doing over the weekend.  After a few weeks he starts reverting
> to type.  In a reply to one of his posts I asked where it the manager
> that was vetting your posts and if he is sick I hope he gets better
> soon. And your unsupervised post were harming Franz.  Soon after that 
> he posted the "I'm outa here" message.
> 
> I personally think it was a good thing for Franz that he stopped
> posting.  He was posting things that were not in Franz's best interest
> and he was/is a senior/executive person at Franz.  And he was doing
> most of them without a disclaimer of any sort on them to even try to
> protect Franz, he was using his work account to post.
> 
> marc
From: Tim Bradshaw
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <ey3isx6tkr5.fsf@cley.com>
* Dorai Sitaram wrote:
> John Foderaro. 

OK, can you give an example of how he was hounded out?  I don't have a
really good internet connection at the moment (I'm on holiday in a mud
hut[1]), and google searches are a bit of a pain as a result, but my
memory is that he said really a lot of fairly silly, and probably
actually technically wrong things, and was picked up on them by a
number of people, and eventually basically stalked off.  I remember at
the time being very impressed by Franz being brave enough to allow him
to post from a Franz account and run the risk of making the company
look bad (and searchably bad, as well): I definitely wouldn't have
been that brave...

--tim

Footnotes: 
[1]  You think this is a joke?  It's not.
From: Oleg
Subject: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <aup07r$qkk$1@newsmaster.cc.columbia.edu>
wni wrote:

[ snipped ]

You may find it amusing that a search for
"idiot OR moron OR cretin" in comp.lang.lisp limited to year 2002 returns 
102 results. OTOH, a similar search in comp.lang.fortran produces only 9 
results; fa.caml - 1; comp.lang.c - 163 (You need to ignore google's 
overestimated expectations) *.fortran is 1.5-2 times smaller than *.lisp, 
and *.c is about 3 times larger.

I agree with you and Richard Fateman, who started a thread about this a 
couple of months ago, that CLL is quite uncivilized. Why it is so is a 
mystery.

If you are, say, a caltech professor, and are being attacked by a bunch of 
trolls with no high school, your best strategy is to simply ignore them 
(there is only one of you and very many of them ;) 

Oleg
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86znqnyj3r.fsf@bogomips.optonline.net>
Oleg <············@myrealbox.com> writes:

> wni wrote:
> 
> [ snipped ]
> 
> You may find it amusing that a search for
> "idiot OR moron OR cretin" in comp.lang.lisp limited to year 2002 returns 
> 102 results. OTOH, a similar search in comp.lang.fortran produces only 9 
> results; fa.caml - 1; comp.lang.c - 163 (You need to ignore google's 
> overestimated expectations) *.fortran is 1.5-2 times smaller than *.lisp, 
> and *.c is about 3 times larger.
> 
> I agree with you and Richard Fateman, who started a thread about this a 
> couple of months ago, that CLL is quite uncivilized. Why it is so is a 
> mystery.
> 
> If you are, say, a caltech professor, and are being attacked by a bunch of 
> trolls with no high school, your best strategy is to simply ignore them 
> (there is only one of you and very many of them ;) 

As one if the people who is being accused of being a high school dropout,
could you please provide proof of this.  Or just shut the fuck up and
go away you idiot or moron or creitin.

marc
From: Tim Daly, Jr.
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <871y3vjpnf.fsf@bob.intern>
Marc Spitzer <········@optonline.net> writes:

> Oleg <············@myrealbox.com> writes:
...
> > 
> > If you are, say, a caltech professor, and are being attacked by a bunch of 
> > trolls with no high school, your best strategy is to simply ignore them 
> > (there is only one of you and very many of them ;) 
> 
> As one if the people who is being accused of being a high school dropout,
> could you please provide proof of this.  Or just shut the fuck up and
> go away you idiot or moron or creitin.


I actually /am/ a high school dropout, and I consider it the only
rational choice.

Of course, those of you who did not have the privilege of attending a
suburban American high school may have occasion to disagree.

Credentials are of extremely limited value.  Unless he's willing to
demonstrate, don't trust a man presenting papers which state that he
can do backflips.

And never, ever, trust a man who wants, at all costs, to see your
backflip degree instead of your backflip.

-Tim
From: Erik Naggum
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3250532115226900@naggum.no>
* Tim Daly, Jr.
| Credentials are of extremely limited value.  Unless he's willing to
| demonstrate, don't trust a man presenting papers which state that
| he can do backflips.

  This misses the point of the qualification completely.  Nobody is
  interested in what you did to achieve your degree -- whatever it was
  will have become irrelevant by the time the degree is useful, if it
  ever was relevant.  In most cases the only important properties of a
  degree are to prove beyond reasonable doubt that the person holding
  it is able and willing to make personal and financial sacrifices in
  order to obtain a higher goal and be a small part of something much
  greater than himself, to follow bizarre rituals and orders to do
  amazingly meaningless tasks for several years because he is told it
  will lead to the higher goal, to submit his entire image of self-
  worth to the judgment of an authority who not only gets paid for
  crushing it, but will keep the money regardless of the outcome, for
  which he accepts all blame, and, finally, to realize and come to
  terms with the realization that he is not the smartest person on the
  planet regardless of how smart he is.  If you /pay/ for following
  orders and for being judged harshly and placed among increasingly
  smart people until you are no longer smart enough to keep moving
  upwards, imagine what the person hiring you will be /relieved/ of!

  The flip side of a degree thus granted is of course that anyone who
  has one associates learning with pain, and not just ordinary pain,
  but prolonged, excruciating pain and humiliation and sacrifice, and
  considers the prospect of learning anything new about as inspiring
  as the full set of root canals in a row without anesthesia.  Most
  degree-holders are thus one-degree ponies.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Oleg
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <av3co2$aof$1@newsmaster.cc.columbia.edu>
Erik Naggum wrote:

> * Tim Daly, Jr.
> | Credentials are of extremely limited value.  Unless he's willing to
> | demonstrate, don't trust a man presenting papers which state that
> | he can do backflips.
> 
>   This misses the point of the qualification completely.  Nobody is
>   interested in what you did to achieve your degree -- whatever it was
>   will have become irrelevant by the time the degree is useful, if it
>   ever was relevant.  In most cases the only important properties of a
>   degree are to prove beyond reasonable doubt that the person holding
>   it is able and willing to make personal and financial sacrifices in
>   order to obtain a higher goal and be a small part of something much
>   greater than himself, to follow bizarre rituals and orders to do
>   amazingly meaningless tasks for several years because he is told it
>   will lead to the higher goal, to submit his entire image of self-
>   worth to the judgment of an authority who not only gets paid for
>   crushing it, but will keep the money regardless of the outcome, for
>   which he accepts all blame, and, finally, to realize and come to
>   terms with the realization that he is not the smartest person on the
>   planet regardless of how smart he is. [...]

I wish I could remind you of this meaningless rant next time you need a 
surgery and are asking to be treated by someone who has a degree.

Oleg
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86d6neip5d.fsf@bogomips.optonline.net>
Oleg <············@myrealbox.com> writes:

> Erik Naggum wrote:
> 
> > * Tim Daly, Jr.
> > | Credentials are of extremely limited value.  Unless he's willing to
> > | demonstrate, don't trust a man presenting papers which state that
> > | he can do backflips.
> > 
> >   This misses the point of the qualification completely.  Nobody is
> >   interested in what you did to achieve your degree -- whatever it was
> >   will have become irrelevant by the time the degree is useful, if it
> >   ever was relevant.  In most cases the only important properties of a
> >   degree are to prove beyond reasonable doubt that the person holding
> >   it is able and willing to make personal and financial sacrifices in
> >   order to obtain a higher goal and be a small part of something much
> >   greater than himself, to follow bizarre rituals and orders to do
> >   amazingly meaningless tasks for several years because he is told it
> >   will lead to the higher goal, to submit his entire image of self-
> >   worth to the judgment of an authority who not only gets paid for
> >   crushing it, but will keep the money regardless of the outcome, for
> >   which he accepts all blame, and, finally, to realize and come to
> >   terms with the realization that he is not the smartest person on the
> >   planet regardless of how smart he is. [...]
> 
> I wish I could remind you of this meaningless rant next time you need a 
> surgery and are asking to be treated by someone who has a degree.
> 
> Oleg

In the US to be a family doctor takes 4 years of post medical school
training and the passing of the medical board exams before it starts.
The degree allows you to sit for the exam that is it.  And for
surgeons, depending on your specialty, I think the apprentice time is
in the 8-12 year range before they are allowed to do surgery
unsupervised.

marc  
From: Kaz Kylheku
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <cf333042.0301031138.5f4948c1@posting.google.com>
Oleg <············@myrealbox.com> wrote in message news:<············@newsmaster.cc.columbia.edu>...
> Erik Naggum wrote:
> 
> > * Tim Daly, Jr.
> > | Credentials are of extremely limited value.  Unless he's willing to
> > | demonstrate, don't trust a man presenting papers which state that
> > | he can do backflips.
> > 
> >   This misses the point of the qualification completely.  Nobody is
> >   interested in what you did to achieve your degree -- whatever it was
> >   will have become irrelevant by the time the degree is useful, if it
> >   ever was relevant.  In most cases the only important properties of a
> >   degree are to prove beyond reasonable doubt that the person holding
> >   it is able and willing to make personal and financial sacrifices in
> >   order to obtain a higher goal and be a small part of something much
> >   greater than himself, to follow bizarre rituals and orders to do
> >   amazingly meaningless tasks for several years because he is told it
> >   will lead to the higher goal, to submit his entire image of self-
> >   worth to the judgment of an authority who not only gets paid for
> >   crushing it, but will keep the money regardless of the outcome, for
> >   which he accepts all blame, and, finally, to realize and come to
> >   terms with the realization that he is not the smartest person on the
> >   planet regardless of how smart he is. [...]
> 
> I wish I could remind you of this meaningless rant next time you need a 
> surgery and are asking to be treated by someone who has a degree.

That rant in fact describes quite precisely and objectively what it
actually means to hold a degree; that is, what are the weakest
assumptions that can be made about someone who holds one, knowing only
that he or she holds one and nothing more.

There is no reason to believe that Erik will forget his ideas about
what it means to hold a degree at some point in the future when he
requires surgery, and that, for instance, he will place more trust in
a degree than he does now.

The institution where the operation will be performed already performs
screening for degrees, so it would be superfluous and irrational to
demand to be treated by someone who has one. Whoever will perform the
surgery will represent himself or herself as having a degree, having
already represented himself or herself that way to the institution.

It will make more sense to ask for someone who has a demonstrated
track record in achieving excellent results in performing that type of
surgery.
From: Oleg
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <av4tv2$33s$1@newsmaster.cc.columbia.edu>
Kaz Kylheku wrote:

> Oleg <············@myrealbox.com> wrote in message
> news:<············@newsmaster.cc.columbia.edu>...
>> Erik Naggum wrote:
>> 
>> > * Tim Daly, Jr.
>> > | Credentials are of extremely limited value.  Unless he's willing to
>> > | demonstrate, don't trust a man presenting papers which state that
>> > | he can do backflips.
>> > 
>> >   This misses the point of the qualification completely.  Nobody is
>> >   interested in what you did to achieve your degree -- whatever it was
>> >   will have become irrelevant by the time the degree is useful, if it
>> >   ever was relevant.  In most cases the only important properties of a
>> >   degree are to prove beyond reasonable doubt that the person holding
>> >   it is able and willing to make personal and financial sacrifices in
>> >   order to obtain a higher goal and be a small part of something much
>> >   greater than himself, to follow bizarre rituals and orders to do
>> >   amazingly meaningless tasks for several years because he is told it
>> >   will lead to the higher goal, to submit his entire image of self-
>> >   worth to the judgment of an authority who not only gets paid for
>> >   crushing it, but will keep the money regardless of the outcome, for
>> >   which he accepts all blame, and, finally, to realize and come to
>> >   terms with the realization that he is not the smartest person on the
>> >   planet regardless of how smart he is. [...]
>> 
>> I wish I could remind you of this meaningless rant next time you need a
>> surgery and are asking to be treated by someone who has a degree.
> 
> That rant in fact describes quite precisely and objectively what it
> actually means to hold a degree; that is, what are the weakest
> assumptions that can be made about someone who holds one, knowing only
> that he or she holds one and nothing more.
> 
> There is no reason to believe that Erik will forget his ideas about
> what it means to hold a degree at some point in the future when he
> requires surgery, and that, for instance, he will place more trust in
> a degree than he does now.
> 
> The institution where the operation will be performed already performs
> screening for degrees, so it would be superfluous and irrational to
> demand to be treated by someone who has one. Whoever will perform the
> surgery will represent himself or herself as having a degree, having
> already represented himself or herself that way to the institution.
> 
> It will make more sense to ask for someone who has a demonstrated
> track record in achieving excellent results in performing that type of
> surgery.

Suppose you live in a land where one does not need to be licensed or hold a 
degree to practice medicine. Given no other information, would you pick a 
physician who holds a degree from a respectable university, or would you 
pick one who only claims to have as much experience?

The whole issue is so commonsense that I feel ridiculous writing about it.

BTW, in USA and probably other countries, you have the option of being 
treated by medical and dental students supervised by licensed doctors. It's 
very cheap and you usually get what you pay for (I had a dental student as 
a roommate a while back)

Oleg
From: Erik Naggum
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3250631718586444@naggum.no>
* Oleg <············@myrealbox.com>
| Suppose you live in a land where one does not need to be licensed
| or hold a degree to practice medicine. Given no other information,
| would you pick a physician who holds a degree from a respectable
| university, or would you pick one who only claims to have as much
| experience?

  But suppose you live in the real world, how would you argue then?

  If you have a degree, the experience did not teach you the three
  things that I believe any employer value most highly in graduates
  -- discipline, humility, and delayed gratification -- all of these
  make them suitable as part of something greater than themselves.
  If you hire people, you want them to do what you tell them and to
  help you reach /your/ goals, not theirs.  They will not work for
  you unless they can reach some of their goals, too, of course, so
  the balancing act that this entails needs some tuning.  Someone who
  has gone through a serious amount of effort in order to complete an
  education on somebody else's terms, has demonstrated that he is not
  only able to subjugate his personal desires to the demands from his
  superiors, he has been able to obtain good results while doing it.

  If you took time out of your life to read what I wrote before you
  succumbed to the urge to respond, you would see that I ended the
  paragraph you did not understand with the following words, which
  you deleted, probably thinking you had already understood the point
  and did not need the punch line:

    If you /pay/ for following orders and for being judged harshly and
    placed among increasingly smart people until you are no longer
    smart enough to keep moving upwards, imagine what the person
    hiring you will be /relieved/ of!

  Do you grasp what this means, "Oleg"?  Do you think that perhaps
  the point was to demonstrate the /values/ of a degree to those
  hiring degree-holders?  Perhaps you have now demonstrated that your
  degree, if any, is completely worthless to any employer because you
  failed to get the /real/ message?  You do not have a degree, if you
  have one, because of what you know.  That is immaterial.  You got
  your degree, if any, because you could follow a scheme much larger
  than yourself and still fit in.  /This/ is valuable to an employer;
  you will get on-the-job training for your real work, anyway.  There
  are other ways to acquire these basic skills and character traits,
  but nothing really beats having an institution with a solid track
  record for weeding out the mistakes approve it.

  It appears that you think what I wrote was a slight towards those
  who hold degrees.  Now, why the fuck would I do that?  Perhaps you
  are just a wee bit too sensitive on this topic and have shown the
  whole world that you get your exercise from jumping to conclusions?

| The whole issue is so commonsense that I feel ridiculous writing
| about it.

  Good, I feel ridiculous reading your stupid retorts to arguments I
  have never made, but are you sure you know why you feel ridiculous?
  The argument you make is so trivial as to be laughable, but it is
  quite suspicious that you have to invent an imaginary land devoid
  of relevance to the real world in order to make such a commonplace
  observation.  Clearly, it is not easy for you to demonstrate its
  validity in the real world, or you would have done that, right?

  That you manage to believe that somebody disagrees with your point
  speaks volumes about you, but says absolutely nothing about either
  the argument that nobody has made or the people you try to make it
  appear as though have made an argument you fight so hard to counter
  for your very own personal reasons.

  When you feel ridiculous countering some obviously wrong claim,
  perhaps it is time for you to dismount your high horse and try to
  read what people /actually/ write instead of being laughed off it
  because you failed to get the point.  You do not want to be a high-
  horse dropout, do you?

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Oleg
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <av6gsk$51j$1@newsmaster.cc.columbia.edu>
Proving once again that brevity is the soul of wit, and the lack of both is 
the soul of Naggum, someone wrote:

[snipped]

> It appears that you think what I wrote was a slight towards those
> who hold degrees.  Now, why the fuck would I do that?

[snipped]

You are obviously just trolling, as usual. You wrote:

> Nobody is interested in what you did to achieve your degree -- whatever 
> it was will have become irrelevant by the time the degree is useful, if it
> ever was relevant. 

I gave a pretty obvious counterexample (perhaps not very obvious to you, but 
I guess that was to be expected): medical degrees are valued by both 
employers and sane consumers, and what you do to obtain them is clearly 
relevant. \end{proof}

Oleg
From: Kaz Kylheku
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <cf333042.0301041055.2ce9c42e@posting.google.com>
Oleg <············@myrealbox.com> wrote in message news:<············@newsmaster.cc.columbia.edu>...
> Proving once again that brevity is the soul of wit, and the lack of both is 
> the soul of Naggum, someone wrote:
> 
> [snipped]
> 
> > It appears that you think what I wrote was a slight towards those
> > who hold degrees.  Now, why the fuck would I do that?
> 
> [snipped]
> 
> You are obviously just trolling, as usual. You wrote:
> 
> > Nobody is interested in what you did to achieve your degree -- whatever 
> > it was will have become irrelevant by the time the degree is useful, if it
> > ever was relevant. 
> 
> I gave a pretty obvious counterexample (perhaps not very obvious to you, but 
> I guess that was to be expected): medical degrees are valued by both 
> employers and sane consumers, and what you do to obtain them is clearly 
> relevant. \end{proof}

Nonsense. Over ten years ago, my physician was a fellow with a 1948
degree hanging on his wall. If all he knew was the state of the art of
medicine in 1948, I would not have gone to him. He could only still be
a practising doctor forty years later through continuous study,
without which he would not only fail to learn anything new, but forget
everything he knew in the 1950's. I certainly could not care less what
he did in the 1940's to get his degree. Maybe back then they snuck out
to graveyards at midnight to purloin cadavers, and used leeches to
draw blood, who knows! :)
From: JB
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <av7kp7$csf8e$1@ID-167393.news.dfncis.de>
Kaz Kylheku wrote:
> Nonsense. Over ten years ago, my physician was a fellow
> with a 1948 degree hanging on his wall. If all he knew was
> the state of the art of medicine in 1948, I would not have
> gone to him. He could only still be a practising doctor
> forty years later through continuous study, without which
> he would not only fail to learn anything new, but forget
> everything he knew in the 1950's. I certainly could not
> care less what he did in the 1940's to get his degree.

Though I think that your attitude is sensible, I am sure 
that most people do not think so.
In Germany we believe that our physicians are educated very 
thoroughly though we are not familiar with the details. You 
can become a physician without having your MD degree but 
you will have a hard time if you have none as most people 
expect it, however irrational this may be.

Also, if you were educated at a foreign university, you have 
to make this public. For example, you are not ellowed to 
say "John Smith, MD" but instead you have to say "John 
Smith, MD / Univ Budapest. If a physician was educated at 
the Univerity of Budapest then his education was presumably 
a bit different and this is considered so important that 
your customers have to be warned of him.

Additionally, What your phycian actually learnt 40 years 
ago, may be outdated. But he learnt how to work and think 
scientifically and this will influence his attitude toward 
his science during his lifetime.

-- 
JB
From: Erik Naggum
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3250707721616483@naggum.no>
* Oleg <············@myrealbox.com>
| Proving once again that brevity is the soul of wit, and the lack of
| both is the soul of Naggum, someone wrote:

  Excuse me?

| You are obviously just trolling, as usual.

  Excuse me?

  Could you please ask someone with a suitable degree to repair you
  and/or your malfunctioning brain before you keep polluting this
  forum?

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Kaz Kylheku
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <cf333042.0301041048.599e2f30@posting.google.com>
Oleg <············@myrealbox.com> wrote in message news:<············@newsmaster.cc.columbia.edu>...
> Kaz Kylheku wrote:
> 
> > It will make more sense to ask for someone who has a demonstrated
> > track record in achieving excellent results in performing that type of
> > surgery.
> 
> Suppose you live in a land where one does not need to be licensed or hold a 

I do not live in such a land, nor would I. So the supposition is
meaningless.

> degree to practice medicine. Given no other information, would you pick a 
> physician who holds a degree from a respectable university, or would you 
> pick one who only claims to have as much experience?

This pathetic land could have no such thing as a ``respectable
university'', so this second alternative would effectively mean going
to a normal country, or relying on foreign talent. In that normal
country, you would not have to face a ridiculous exclusive-or choice
between a degree and experience.
From: Erik Naggum
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3250612397276876@naggum.no>
* Oleg <············@myrealbox.com>
| I wish I could remind you of this meaningless rant next time you
| need a surgery and are asking to be treated by someone who has a
| degree.

  I would ask for an expert with a valid license to practice surgery,
  not someone who merely has a degree.

  I feel a profound pity for you and everyone else who attach their
  personal identity, self-worth, and pride to their academic degree.
  It means, more than anything else, that acquiring the degree was
  not a rite of passage into a group, but a rite of passage out of a
  group, and this means that your future contributions will not be
  made relative to your new group, but relative to your old.  This
  explains your ridiculously insecure attitude towards high school
  dropouts.  The secret to feeling great about yourself is not to be
  found in searching for people who are less than you and then show
  yourself superior to them, but in searching for people who are more
  than you and then show yourself worthy of their company.  What I,
  for one, vehemently detest, is people who make it their purpose to
  show that nobody is better than they are.  Some of them are here.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86fzsbte2n.fsf@bogomips.optonline.net>
···@tenkan.org (Tim Daly, Jr.) writes:

> Marc Spitzer <········@optonline.net> writes:
> 
> > Oleg <············@myrealbox.com> writes:
> ...
> > > 
> > > If you are, say, a caltech professor, and are being attacked by a bunch of 
> > > trolls with no high school, your best strategy is to simply ignore them 
> > > (there is only one of you and very many of them ;) 
> > 
> > As one if the people who is being accused of being a high school dropout,
> > could you please provide proof of this.  Or just shut the fuck up and
> > go away you idiot or moron or creitin.
> 
> 
> I actually /am/ a high school dropout, and I consider it the only
> rational choice.

Looking  back on highschool I can see your point.  

> 
> Of course, those of you who did not have the privilege of attending a
> suburban American high school may have occasion to disagree.
> 
> Credentials are of extremely limited value.  Unless he's willing to
> demonstrate, don't trust a man presenting papers which state that he
> can do backflips.

That is besides the point and true.  The issue was that Oleg implied
that the well educated person(Caltech professor) should not bother
listening to the ignorant burger flipper(HS dropout).  And since I was
in the group not to be listened to I asked him to present his evidence.

> 
> And never, ever, trust a man who wants, at all costs, to see your
> backflip degree instead of your backflip.

I do not know about that.  If I can predict your behavior I can
'trust' you.  If the person is a fool he will consistently be a fool
and I can 'trust' that.

marc
From: Alain Picard
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <87n0mj6jtz.fsf@ibook.nsw.optushome.com.au>
Marc Spitzer <········@optonline.net> writes:

> That is besides the point and true.  The issue was that Oleg implied
> that the well educated person(Caltech professor) should not bother
> listening to the ignorant burger flipper(HS dropout).  

And if you bothered reading carefully, you would have noted the
difference between "not bother listening to" and "respond to attacks
from a bunch of trolls".

Oleg's exact words were:

> If you are, say, a caltech professor, and are being attacked by a bunch of 
> trolls with no high school, your best strategy is to simply ignore them 
> (there is only one of you and very many of them ;) 

Have you ever heard the expression: "If the shoe fits..." ?
Oleg did not name YOU personally.  That you take this sort of offense
at his general, and obviously true, statement, speaks volumes.

You wrote:
> As one if the people who is being accused of being a high school dropout,
> could you please provide proof of this.  Or just shut the fuck up and
> go away you idiot or moron or creitin.

Nobody was being accused of being a high school dropout.  Nobody even
intimated a that HS dropouts were neccessarily worthless�.  What WAS said
is that there are a lot more HS dropouts than caltech professors, AND
if they troll and attack you, you're better off ignoring them.

Now, is that clearer?


� Heck, isn't Jamie Zawinski a HS dropout?  And he's a recognized,
  outstanding programmer.
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86hecqit77.fsf@bogomips.optonline.net>
Alain Picard <·······················@optushome.com.au> writes:

> Marc Spitzer <········@optonline.net> writes:
> 
> > That is besides the point and true.  The issue was that Oleg implied
> > that the well educated person(Caltech professor) should not bother
> > listening to the ignorant burger flipper(HS dropout).  
> 
> And if you bothered reading carefully, you would have noted the
> difference between "not bother listening to" and "respond to attacks
> from a bunch of trolls".
> 
> Oleg's exact words were:
> 
> > If you are, say, a caltech professor, and are being attacked by a bunch of 
> > trolls with no high school, your best strategy is to simply ignore them 
> > (there is only one of you and very many of them ;) 
> 
> Have you ever heard the expression: "If the shoe fits..." ?
> Oleg did not name YOU personally.  That you take this sort of offense
> at his general, and obviously true, statement, speaks volumes.

Fuck wit the thing I was objecting to is exactly that kind of snide and
cowardly remark.  It implied that everyone Oleg did not agree with was
an ignorant failure.  He did not have the balls to come out and say it,
he just implied it and he could say "you misunderstood" if it blew up in
his face.

I even asked Oleg to produce his evidence that I did not graduate from
high school or college as he implied.  I have not seen a damn thing
from him.

> 
> You wrote:
> > As one if the people who is being accused of being a high school dropout,
> > could you please provide proof of this.  Or just shut the fuck up and
> > go away you idiot or moron or creitin.
> 
> Nobody was being accused of being a high school dropout.  Nobody even
> intimated a that HS dropouts were neccessarily worthless�.  What WAS said
> is that there are a lot more HS dropouts than caltech professors, AND
> if they troll and attack you, you're better off ignoring them.
> 
> Now, is that clearer?

It was perfectly clear before also, Oleg implied that the people who
he disagreed with were ignorant failures and should not be paid attention 
to.  

Is that clear?

And now I am being called a troll by a putz. 

Is that clear?

> 
> 
> � Heck, isn't Jamie Zawinski a HS dropout?  And he's a recognized,
>   outstanding programmer.

That is completely besides the point.  Calling or implying that someone
is a high school dropout is a insult.  It implies a serious failure, ie
you are stupid and ignorant or to put it another way you are a looser.

Now it is entirely possible that individuals do not fit the profile,
but the insult is based on the profile not the exceptions.

Is this clear?

marc
From: Kenny Tilton
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3E154002.8060700@nyc.rr.com>
Alain Picard wrote:
 > Marc Spitzer <········@optonline.net> writes:
 >
 >
 >>That is besides the point and true.  The issue was that Oleg implied
 >>that the well educated person(Caltech professor) should not bother
 >>listening to the ignorant burger flipper(HS dropout).
 >
 >
 >
 > And if you bothered reading carefully, you would have noted the
 > difference between "not bother listening to" and "respond to attacks
 > from a bunch of trolls".
 >
 > Oleg's exact words were:
 >
 >
 >> If you are, say, a caltech professor, and are being attacked by a 
bunch of
 >> trolls with no high school, your best strategy is to simply ignore them
 >> (there is only one of you and very many of them ;)


That is not how natural language works. It is not how it is produced, it
is not how it is understood. It /is/ how a logic engine would understand
NL if it were capapble of parsing the rest of the information, but there
ya go, logic engines do not understand as people do (and any AI
researcher can explain that that is why logic engines cannot parse NL).

It's no good falling back on mathematical interpretation of a speech act
("he said 'if'!") when valid human interpretations prove uncomfortable.
Oleg damned all attackers as non-HS trolls with his words.

But plz count me among those who place little value in formal education.

:)

-- 

   kenny tilton
   clinisys, inc
   http://www.tilton-technology.com/
   ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
   the bath water is cold." -- Lorraine Lee Cudmore
From: Tom Lord
Subject: formal ed meditation (was ...blah blah blah...)
Date: 
Message-ID: <v1ajs8fpib9846@corp.supernews.com>
        Kenny Tilton:

	But plz count me among those who place little value in formal
	education.


One nice thing about formal education is that it sustains
universities.  Universities are great environments in which to gather,
exchange ideas, and learn, especially outside of the formal tracks of
getting creds.  They often have awesome libraries with open stacks.
Late at night, lisp hackers can be found.  Strange music can be heard,
both at performances and below rehearsal-room windows.  Good
bookstores and coffeeshops find their niche.  Sex flourishes.  Bands
form.  PDP-11s get preserved and recycled.  Classroom art studios
smell good.  Galleries happen.  Politics is discussed.  Liquid
nitrogen can be "procured".  Cycles can be bumbed.  Frisbee can be
played.  Repertory cinemas make money.  Immigrant communities open
great, affordable restaurants.  Oral history flourishes.  Flyers are
everywhere.  Children are honored members of the community.
Economically oppressed communities are reached out to -- and a few
people, at least, respond.  Everyone knows a little bit about
everything.  Unions do good work.  Not only is your informal erdos
number as low as you like (given your displacement in time), but your
informal Anyone number (for all Anyone's in intellectual history) is
as low as you like.  The history of your region is available in
extraordinary detail.  Power struggles are optional.  Some of the
graffiti is profound and lasts for decades.  There's always _someone_
who can use your latest Emacs hack.  Most of the people around you
are, in their own way, improving.  Sometimes there are parties.
Drugs, sobriety, and law achieve some approximation of a healthy
balance.  Anyone can sit in on just about any "invited talk".  Some
totally brilliant people have offices and if you visit them, they just
might give you copies of some good papers to read.  Pride!  My god, so
much justified pride in doing a good job.  Almost everybody is "poor"
(but getting by).  The dissidents have tenure and a mature calmness.
Anyone without severe learning disabilities can get a degree, given
enough time and freedom from economic oppression.  Over time,
transient members of the community give long-standing members an
incredible snapshot of human development.  Elevator's can be hacked
("Close the fucking door").  Cheese coops can be formed.  Samizdat can
find its audience.  Every city should have at least one great
university and the common criticism of esoteric disciplines that the
departments mostly (formally) serve to create their own next-gen
faculty is no criticism at all.  Of course, all the universities have
gone to hell since the mid-80s.

Cred fetishes: no.  Formal institutes of education: yes.

"Requirements:  BS in CS or equivalent experience" -- yeah right,
-t
From: Raffael Cavallaro
Subject: Re: formal ed meditation (was ...blah blah blah...)
Date: 
Message-ID: <aeb7ff58.0301042022.4ecadf89@posting.google.com>
····@emf.emf.net (Tom Lord) wrote in message news:<··············@corp.supernews.com>...

[a moving ode to university communities]


One shouldn't underestimate the desire to be part of such a communtiy
as a motivator in jumping through the hoops necessary to obtain a
degree.
From: Ray Blaak
Subject: Re: formal ed meditation (was ...blah blah blah...)
Date: 
Message-ID: <9183fc69.0301052243.402ea4d4@posting.google.com>
····@emf.emf.net (Tom Lord) wrote in message news:<··············@corp.supernews.com>...
[about the good old days of the utopic university life, where everyone just
 gets along]

It sounds good. I wish for that life myself.

I wonder though if the wonderfulness came about because it was a low stress
life (in the sense of staying alive), not of the real world. Many of these
people lose their idealism when they are eventually forced to grow up.

Society in general was basically funding it, after all, poor students aside
(those libraries and facilities they're using are actually signs of great
wealth). The idea is (was?) that it is a really really good idea to pay for
bunches of really smart people to just hang out. Everyone benefits in the long
run.

The problem is that people in general are catching on that it is a good life
to try for. But the system won't work if everyone wants a piece of the
action. Hence the ever more restrictive filtering mechanisms in the form of
increased tuition, publishing requirements, and funding scarcity.

BTW, when did you graduate? The mid-80's? People tend to remember their
defining experiences as the end of a golden age. Nevertheless I think I agree
with you here.

--
Cheers,                                        The Rhythm is around me,
                                               The Rhythm has control.
Ray Blaak                                      The Rhythm is inside me,
·····@telus.net                                The Rhythm has my soul.
From: Tom Lord
Subject: Re: formal ed meditation (was ...blah blah blah...)
Date: 
Message-ID: <v1iro2n527ag7f@corp.supernews.com>
        ····@emf.emf.net (Tom Lord) wrote in message news:<··············@corp.supernews.com>...
        [about the good old days of the utopic university life, where everyone just
         gets along]

         It sounds good. I wish for that life myself.

What I'd really like to do is to run a small commercial software
practical-R&D lab in a university town, trying in part to expand the
economic basis of a good "university culture" -- to create more seats
at the table and to help bridge between for-profit businesses and
university culture.  I have a little research agenda and everything
(and, yes, it involves lisp.)

After all, in addition to being lots of fun, universities produce huge
amounts of very high quality "value".  They function better than most
companies in almost every way.   The culture is a big part of why.




         I wonder though if the wonderfulness came about because it
         was a low stress life (in the sense of staying alive), not of
         the real world. Many of these people lose their idealism when
         they are eventually forced to grow up.

         Society in general was basically funding it, after all, poor
         students aside (those libraries and facilities they're using
         are actually signs of great wealth). The idea is (was?) that
         it is a really really good idea to pay for bunches of really
         smart people to just hang out. Everyone benefits in the long
         run.

That's too simplistic a view.  It boarders on puritanism.

Universities mostly recover their operating expenses by selling goods
and services to both the public and private sectors.  They aren't
the kind of fantasy-land economic leeches you imply -- far from it.



         The problem is that people in general are catching on that it
         is a good life to try for. But the system won't work if
         everyone wants a piece of the action. Hence the ever more
         restrictive filtering mechanisms in the form of increased
         tuition, publishing requirements, and funding scarcity.


Nah.  The disease isn't excessive supply of university workers or
excessive demand from prospective students.  It's decadent values in
industry and the near collapse of the surrounding civilization.  In
technical fields, it's the shortage of demand for research products
combined with the _incredibly_ poor judgement of those few who are
spending money on research.  In industry generally, it's the shortage
of demand for workers who aren't undereducated drones.



	 [universities have declined since the mid-80s]

         BTW, when did you graduate? The mid-80's? People tend to
         remember their defining experiences as the end of a golden
         age. Nevertheless I think I agree with you here.

Graduate?  No.  I was a bright hanger-on, bordering on "mascot".
Lot's of people showed me things and gave me good reading materials.
The table scraps may very well be much better than the meal, if you
ask me.  That's part of why I don't like degree fetishes -- there's
little worse in "the real world" of job-hunting than interviewing with
a prospective employer who can't hold an intelligent conversation but
then dismissively asks "questions" about your academic creds.  (Like,
*literally*, in an interview for a *senior engineering position*,
"What can you tell me about about linked lists?" and "Please sit at
this terminal and type in a function in C that compares two strings." 
The project manager, after we talked for a few minutes, asked me (not
rhetorically or for evaluation purposes -- more like "please help me
with my job") to read competing proposals from her subordinates and
help her _understand_ them.   And then, yeah, "You don't seem to have
a degree listed on your resume....").

But yeah -- I left being a direct employee of universities around
1988.  I worked for a little while for the FSF when they lived mostly
on the MIT campus.  I lived in Berkeley for a bit.

One piece of evidence that at least CS departments have gone to hell
is to track their web sites for the past *10 years*.   There's
precious little activity.  Few signs of life.

-t
From: Oleg
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <av51dk$56a$1@newsmaster.cc.columbia.edu>
Alain Picard wrote:

> Marc Spitzer <········@optonline.net> writes:
> 
>> That is besides the point and true.  The issue was that Oleg implied
>> that the well educated person(Caltech professor) should not bother
>> listening to the ignorant burger flipper(HS dropout).
> 
> And if you bothered reading carefully, you would have noted the
> difference between "not bother listening to" and "respond to attacks
> from a bunch of trolls".
> 
> Oleg's exact words were:
> 
>> If you are, say, a caltech professor, and are being attacked by a bunch
>> of trolls with no high school, your best strategy is to simply ignore
>> them (there is only one of you and very many of them ;)
> 
> Have you ever heard the expression: "If the shoe fits..." ?
> Oleg did not name YOU personally.  That you take this sort of offense
> at his general, and obviously true, statement, speaks volumes.
> 
> You wrote:
>> As one if the people who is being accused of being a high school dropout,
>> could you please provide proof of this.  Or just shut the fuck up and
>> go away you idiot or moron or creitin.
> 
> Nobody was being accused of being a high school dropout.  Nobody even
> intimated a that HS dropouts were neccessarily worthless�.  What WAS said
> is that there are a lot more HS dropouts than caltech professors, AND
> if they troll and attack you, you're better off ignoring them.
> 
> Now, is that clearer?
> 
> 
> � Heck, isn't Jamie Zawinski a HS dropout?  And he's a recognized,
>   outstanding programmer.

Alain,

Thanks for a great elaboration! Now even Mr. Spitzer should be able to 
understand [1]. I found it ironic that he made nearly half a dozen mistakes 
in the very first sentence of his angry outburst, including one that 
completely alters the meaning of the message. ;-)

Cheers,
Oleg

[1] who am I kidding?
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86wulksx7p.fsf@bogomips.optonline.net>
Oleg <············@myrealbox.com> writes:

> Alain Picard wrote:
> 
> > Marc Spitzer <········@optonline.net> writes:
> > 
> >> That is besides the point and true.  The issue was that Oleg implied
> >> that the well educated person(Caltech professor) should not bother
> >> listening to the ignorant burger flipper(HS dropout).
> > 
> > And if you bothered reading carefully, you would have noted the
> > difference between "not bother listening to" and "respond to attacks
> > from a bunch of trolls".
> > 
> > Oleg's exact words were:
> > 
> >> If you are, say, a caltech professor, and are being attacked by a bunch
> >> of trolls with no high school, your best strategy is to simply ignore
> >> them (there is only one of you and very many of them ;)
> > 
> > Have you ever heard the expression: "If the shoe fits..." ?
> > Oleg did not name YOU personally.  That you take this sort of offense
> > at his general, and obviously true, statement, speaks volumes.
> > 
> > You wrote:
> >> As one if the people who is being accused of being a high school dropout,
> >> could you please provide proof of this.  Or just shut the fuck up and
> >> go away you idiot or moron or creitin.
> > 
> > Nobody was being accused of being a high school dropout.  Nobody even
> > intimated a that HS dropouts were neccessarily worthless�.  What WAS said
> > is that there are a lot more HS dropouts than caltech professors, AND
> > if they troll and attack you, you're better off ignoring them.
> > 
> > Now, is that clearer?
> > 
> > 
> > � Heck, isn't Jamie Zawinski a HS dropout?  And he's a recognized,
> >   outstanding programmer.
> 
> Alain,
> 
> Thanks for a great elaboration! Now even Mr. Spitzer should be able to 
> understand [1]. I found it ironic that he made nearly half a dozen mistakes 
> in the very first sentence of his angry outburst, including one that 
> completely alters the meaning of the message. ;-)
> 
> Cheers,
> Oleg
> 
> [1] who am I kidding?

At least the insult is out in the open now.  Look what happens when one
fuck-wit [1] gets in a feed back loop with another.  Erik I feel some
of your annoyance.

As was pointed out by other people as the language is used by people 
you did insult people, deliberately.  Now if you want to talk about
people who have great accomplishments take a look at Audie Murphy[2]
and he had no high school, he did not go because he had to support
his family as a young man.  

marc

Footnotes: 
[1]  fucking moron by choice not ability.

[2]  not because he was a movie star, here is a short list why:
    * Medal of Honor
    * Distinguished Service Cross
    * Silver Star with First Oak Leaf Cluster
    * Legion of Merit
    * Bronze Star Medal with "V" Device and First Oak Leaf Cluster
    * Purple Heart with Second Oak Leaf Cluster
From: Christopher C. Stacy
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <uel7tdey7.fsf@dtpq.com>
I think it says something when I see a spam article entitled 
"Natural Penis Enlargement" in this newsgroup, and I stop to 
wonder whether it's just another meta-discussion thread.
From: Erik Naggum
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3250255753045856@naggum.no>
* Oleg <············@myrealbox.com>
| I agree with you and Richard Fateman, who started a thread about
| this a couple of months ago, that CLL is quite uncivilized. Why it
| is so is a mystery.

  Suppose people suddenly became friendly in comp.lang.lisp.  Would
  you be able to notice it or would you still project animosity and
  hostility onto others so you could have an excuse to keep hating?
  Do you trouble-starters have any concept of personal responsibility
  or any idea of your ability to /change/ the climate by shutting up?

  This forum is uncivilized because uncivilized people think it is OK
  to be uncivilized here.  If you really want to know, ask "wni" why
  he felt no compunctions about posting /his/ hostility.  Normal and
  well-adjusted people have an element of restraint that he and those
  like him lack, and which furthermore means that their indignation
  is not only everybody's business, but that telling them to curtail
  /their/ abuse of this forum is grounds for further indignation.  If
  anyone still objects to my characterization of these people by
  their rightful name: morons, please explain why what they do is not
  utterly devoid of every symptom of intelligence.

  Get rid of the people who believe this forum exists to share their
  opinions on how well the forum functions, and it /will/ function.
  The solution is pretty simple: Stay focused on what you believe the
  forum should be about; exhibit concentration and goal-directedness;
  exercise restraint; demonstrate your ability to act in accordance
  with your own convictions regarding behavior in public; apply your
  intelligence and stop doing things that exacerbate the situations
  you do not desire and do things that would ameliorate it instead;
  realize that if you cannot do it yourself, demanding it of others
  is not only morally bankrupt, it demonstrates that the purpose of
  the demands is to make someone the fall guy instead of accepting
  responsibility for one's own behavior, so it can be improved -- all
  they actually want is to blame someone for their own shortcomings,
  and they should not be surprised that the fall guy objects to this
  unjustness.  Clean up your own act and show the world that /you/
  can behave, and all the problems will simply go away on their own.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Kenny Tilton
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <3E109722.2060806@nyc.rr.com>
Oleg wrote:
> wni wrote:
> 
> [ snipped ]
> 
> You may find it amusing that a search for
> "idiot OR moron OR cretin" in comp.lang.lisp limited to year 2002 returns 
> 102 results. OTOH, a similar search in comp.lang.fortran produces only 9 
> results; fa.caml - 1; comp.lang.c - 163 

Sloppy research! Most of that was me describing my own coding, and I do 
not post on those NGs.


-- 

  kenny tilton
  clinisys, inc
  http://www.tilton-technology.com/
  ---------------------------------------------------------------
"Cells let us walk, talk, think, make love and realize
  the bath water is cold." -- Lorraine Lee Cudmore
From: Oleg
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <auvp41$p4m$1@newsmaster.cc.columbia.edu>
Kenny Tilton wrote:

> Oleg wrote:
>> wni wrote:
>> 
>> [ snipped ]
>> 
>> You may find it amusing that a search for
>> "idiot OR moron OR cretin" in comp.lang.lisp limited to year 2002 returns
>> 102 results. OTOH, a similar search in comp.lang.fortran produces only 9
>> results; fa.caml - 1; comp.lang.c - 163
> 
> Sloppy research! Most of that was me describing my own coding, and I do
> not post on those NGs.
> 

I think you are trying to take too much credit. Upon closer examination, 
adding "-tilton" to the search querry has little effect, while, on the 
other hand, adding "-naggum" cuts the number of hits approximately in half, 
thus bringing CLL to the level of civility of CLC in this artificial setup, 
which is of course still short of ideal.

Oleg
From: Marc Spitzer
Subject: Re: CLL statistics for 2002 (was: Looking for Lisp compiler)
Date: 
Message-ID: <86of70tpqf.fsf@bogomips.optonline.net>
Oleg <············@myrealbox.com> writes:

> Kenny Tilton wrote:
> 
> > Oleg wrote:
> >> wni wrote:
> >> 
> >> [ snipped ]
> >> 
> >> You may find it amusing that a search for
> >> "idiot OR moron OR cretin" in comp.lang.lisp limited to year 2002 returns
> >> 102 results. OTOH, a similar search in comp.lang.fortran produces only 9
> >> results; fa.caml - 1; comp.lang.c - 163
> > 
> > Sloppy research! Most of that was me describing my own coding, and I do
> > not post on those NGs.
> > 
> 
> I think you are trying to take too much credit. Upon closer examination, 
> adding "-tilton" to the search querry has little effect, while, on the 
> other hand, adding "-naggum" cuts the number of hits approximately in half, 
> thus bringing CLL to the level of civility of CLC in this artificial setup, 
> which is of course still short of ideal.
> 
> Oleg

I am not clear, is your problem that the "bad" words were use or that
they were used inaccurately?

I know when I want to say truly horrid things I do not used the "bad"
words because they take away from the insult I am going to deliver.
The words I use are wholly acceptable in polite conversation until I
put them together in the proper order.

marc
From: Erik Naggum
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <3250224324597292@naggum.no>
* wni <···@nospam.attbi.com>
| So it's your experience that counts?

  Was that a question?  If so, no, that is /so/ not the point.

| What makes you think that you are qualified to criticize while the
| other people can't?

  Well, to be honest, your rather frantic response virtually proves
  that I am qualified to criticize and you are not.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
From: Geoffrey Summerhayes
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <9VTP9.1306$rr1.226649@news20.bellglobal.com>
"Erik Naggum" <····@naggum.no> wrote in message ·····················@naggum.no...
>
>   Well, to be honest, your rather frantic response virtually proves
>   that I am qualified to criticize and you are not.
>

That's odd. I feel a sudden impulse to read a Joseph Heller novel...

--
Geoff
From: Will Deakin
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <cc0d708e.0212311303.14a11b46@posting.google.com>
"Geoffrey Summerhayes" <·············@hotmail.com> wrote in message 
> That's odd. I feel a sudden impulse to read a Joseph Heller novel...
I realise that a certain work by Mr Heller has reached almost
cannonical status but would like to recommend "Picture This" or
"Something Happened" as both are excellent.

HNY

:)w
From: Joe Marshall
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <r8bzmpwe.fsf@ccs.neu.edu>
wni <···@nospam.attbi.com> writes:

> We are not interested in your [Erik's] opinions ...

Speak for yourself.

If you have no interest in Erik's opinions, why do you read them?
From: JB
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <auq2de$8m7d4$1@ID-167393.news.dfncis.de>
Obviously, Microsoft cannot afford to offer much support as 
they sell many copies of their usually cheap products. They 
would probably give you good support if you paid for this 
kind of support. I am sure that they offer this kind of 
support.

I have had the experience that I got very good support from 
commercial Lisp vendors like Franz of even Corman and got 
the same kind of support from Trolltech (Qt), though I was 
not their customer and from what I had posted here they 
could infer that it was not very likely that I should 
become one.
Probably, if I were their customer, their support were even 
better (though what I got was completely satisfactory 
anyway).

I looked at several Lisp compilers in the past and I think 
that if you are serious about Lisp, you should buy a 
commercial license. And if you want to work on Linux /and/ 
on Windows then you cannot buy Corman Lisp. In this case, 
you should buy LispWorks or ACL from Franz (which I should 
prefer).
I do not think that it will be easy to find a free compiler 
that meets your demands.
-- 
JB
From: Marc Spitzer
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <86ptrmmtjo.fsf@bogomips.optonline.net>
Pascal Bourguignon <···@informatimago.com> writes:

> 
> Moreover, guess who will have the  most pressure to help you resolve a
> problem, the developper of commercial tools who have thousands or tens
> of thousands customers and who is busy developping the next version to
> be  able to  charge the  upgrade price,  or the  small  shop dedicated
> support team  who will have  a couple dozen  of custmers and  for whom
> you'll be relatively a much more important customer?

I forgot something here it is:

A company gains reputation by having good customer support, but good
support costs money so you do not want your customers to use it.  And
if they do you want to take care of it as quickly as possible and at
the lowest level in the organization as possible, web sites are where
they want you to go first.  So I would expect a company that has lots
of customers to want to fix bugs and in the interim come up with work
arounds for the issues until a patch can be released and then get the
info on the website or level 1 phone support, its cheaper.

Also I am not talking about MS shrink rap crap when I talk about
support.  When I think of a support/maintenance contract I am thinking
about things like hp-ux or openview or oracle or solaris etc.  Those
products when you pay for support you get all new releases that happen
during the course of the contract as a matter of course, oracle does
split upgrade license from support license though.

If a CL vendor starts to shit on his customers they can and will go
else where.  There are 3 commercial vendors for linux/unix(allegro,
lispworks, scl) the same is true for windows.  This is not true for
VB.  And this is not counting all the free CL's out there(cmu, sbcl,
ecl, openmcl, clisp)

And I do not care who feels the most pressure, I care who gives me the
best information in a timely manner.  If I call big lisp vendor and
the level 1 phone support person says " I will mail you a link that 
explains the problem and the work around, it in the website, and a
patch is making it through QA as we speak.  Would you like an email
alert when it is available for download? thanks and have a nice day"
I am happy and I consider the money I spent very well spent.  The
reason for this is it helps me to get my job done in a timely manner
and that is how I make money, not pressuring some other guy who is
trying to fix a problem he has never seen before and who's company has
no reason to invest in a website like "big CL vendor" does, because
there is no economy of scale for the little guy that presents a
compelling business to spend the money to develop and maintain such a
site. 

marc
From: Hannah Schroeter
Subject: ECLS porting (was Re: Looking for Lisp compiler)
Date: 
Message-ID: <aunlau$h0l$1@c3po.schlund.de>
Hello!

Marc Spitzer  <········@optonline.net> wrote:
>[...]

>Yes ecls does do that on posix systems( I never meant to imply it could
>[...]

I've tried to port a current ecls release (0.7b) to OpenBSD/x86 and failed.
Has someone else succeeded in doing so and what changes were needed?

Porting a very old version (0.0g) worked, but things seem to have
changed significantly.

Kind regards,

Hannah.
From: Juan Jose Garcia-Ripoll
Subject: RE: ECLS porting (was Re: Looking for Lisp compiler)
Date: 
Message-ID: <3e107fb3$1_4@news.arrakis.es>
Hannah Schroeter <······@schlund.de> escribi� en el mensaje de noticias
············@c3po.schlund.de...
> I've tried to port a current ecls release (0.7b) to OpenBSD/x86 and
failed.
> Has someone else succeeded in doing so and what changes were needed?

Can you be more explicit about the difficulties? Maybe through private mail?
(····@arrakis.es) OpenBSD is one of the platforms I lack for testing.
However porting should be straightforward in the x86 platform -- I myself
ported ECLS to NetBSD long ago, although I could not maintain that platform
because my old NetBSD box died long ago.

Best regards

Juanjo
From: Peter Seibel
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <m3vg1gq8in.fsf@localhost.localdomain>
"Coby Beck" <·····@mercury.bc.ca> writes:

> "Lisp Newbie" <········@lisp.programmer> wrote in message
> ····························@news1.west.cox.net...
> > Is there such thing as free Lisp compiler, that can work under Linux and
> > Windows, develiver executables, have sockets implementation and support
> > multithreading?
> 
> Start at www.lisp.org
> 
> Delivering executables is the only thing there I don't know if you
> can get for free. Usually it is not really what people need/want
> anyway.

Supposing for a moment that not everyone who *thinks* they need/want
to deliver a standalone executable is totally on crack (and it must be
of *some* use as all the commercial vendors seem to have thought it
was worth implementing), can anyone describe more or less what
strategies are used in implementations that do generate standalone
executables? Maybe someone (maybe me) will say, hmmm, that doesn't
sound too hard and will figure out how to hack it into their favorite
free Lisp. Then the standard answer to this question could change
from, "you don't really want that and if you do really want it you
have to buy it" to "you don't really want that but if you do, you can
buy it or get it for free in implementation XXX."

FWIW, I imagine that in an implementation that compiles to native
machine code (CMUCL and SBCL as I understand it) building an
executable would presumably involve generating some wrapper code that
compiles into something that looks like what you get with a simple C
program with a 'int main()' function which in turn calls into some
entry point that the programmer specifies somehow and which is linked
against a library containing the lisp runtime support. Obviously if
it's going to a absolutely standalone executable this all needs to be
statically linked but I can imagine that some folks would also like
having the lisp runtime dynamically linked if they're going to have a
bunch of lisp apps running separately.

In a bytecode based implementation (CLISP, again, if I understand
correctly) presumably a "standalone app" would have to be a bundle of
the CLISP vm plus the generated bytecodes. Again, folks with lots of
apps might appreciate an option that lets them install the VM once per
computer (or user per computer perhaps) rather than once per app.
Perhaps a starting point for this approach might be something like
what Sun has tried to do with the JRE (Java Runtime Environment).

Is that more or less how it would go or are there other better ways to
do it? And finally, does anyone who currently works on any of the free
Lisps care to say whether this kind of "enhancement" would in fact be
considered desirable in their implementation.

-Peter

-- 
Peter Seibel
·····@javamonkey.com
From: Pascal Costanza
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <auhgq2$14hi$1@f1node01.rhrz.uni-bonn.de>
Peter Seibel wrote:

> In a bytecode based implementation (CLISP, again, if I understand
> correctly) presumably a "standalone app" would have to be a bundle of
> the CLISP vm plus the generated bytecodes. Again, folks with lots of
> apps might appreciate an option that lets them install the VM once per
> computer (or user per computer perhaps) rather than once per app.
> Perhaps a starting point for this approach might be something like
> what Sun has tried to do with the JRE (Java Runtime Environment).

On Mac OS X, you can install CLISP and OpenMCL via fink. I don't think 
it's too much to ask for from users who want to use "free" software to 
require them to deal with fink. I guess, similar solutions are available 
for Linux. I don't know about other Unixes or Windows in that regard.


Pascal

-- 
Pascal Costanza               University of Bonn
···············@web.de        Institute of Computer Science III
http://www.pascalcostanza.de  R�merstr. 164, D-53117 Bonn (Germany)
From: Paolo Amoroso
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <13gMPqEoXv7V+iMiriV49N3rCgvO@4ax.com>
On Fri, 27 Dec 2002 06:29:51 GMT, Peter Seibel <·····@javamonkey.com>
wrote:

> was worth implementing), can anyone describe more or less what
> strategies are used in implementations that do generate standalone
> executables? Maybe someone (maybe me) will say, hmmm, that doesn't

One of the main techniques is called "tree shaking".


> FWIW, I imagine that in an implementation that compiles to native
> machine code (CMUCL and SBCL as I understand it) building an
> executable would presumably involve generating some wrapper code that

Some time ago there was some discussion in a CMU CL mailing list on
bundling together all the 3 application components (Lisp executable + core
+ launcher) mentioned in my other post in this thread. I can't remember
whether this has actually been done.


> do it? And finally, does anyone who currently works on any of the free
> Lisps care to say whether this kind of "enhancement" would in fact be
> considered desirable in their implementation.

I guess the maintainers of open-source Common Lisp implementations consider
desirable everything that someone is willing to contribute :)


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Peter Seibel
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <m3r8c3qr57.fsf@localhost.localdomain>
Paolo Amoroso <·······@mclink.it> writes:

> On Fri, 27 Dec 2002 06:29:51 GMT, Peter Seibel <·····@javamonkey.com>
> wrote:
> 
> > was worth implementing), can anyone describe more or less what
> > strategies are used in implementations that do generate standalone
> > executables? Maybe someone (maybe me) will say, hmmm, that doesn't
> 
> One of the main techniques is called "tree shaking".

I've heard that term before and infer that it means taking an image
and starting from some entry point (or points) shaking off all the
unused bits of code (anything not reachable from the transative
closure of called code starting from the entry points) so as to reduce
the size of the image and then dumping it. Is that more or less right?

> > do it? And finally, does anyone who currently works on any of the
> > free Lisps care to say whether this kind of "enhancement" would in
> > fact be considered desirable in their implementation.
> 
> I guess the maintainers of open-source Common Lisp implementations
> consider desirable everything that someone is willing to contribute
> :)

You think? Cool--I'll "contribute" my patch that "fixes" all those
annoying (to me) coding idioms in implementation X. Do you think the
maintainers of X will consider *that* patch desirable? ;-)

Seriously though, I asked because there are enough people here on
c.l.l. who seem to believe that "you don't really want that feature"
that it's possible there would be folks with strong philosophical
objections. Which is fine. I just wouldn't want anyone (maybe me) to
spend the time working on some code only to have it turned down
because it doesn't fit with the philosophy of a given project.

-Peter

-- 
Peter Seibel
·····@javamonkey.com
From: Rahul Jain
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <873coihyvk.fsf@localhost.localdomain>
Peter Seibel <·····@javamonkey.com> writes:

> You think? Cool--I'll "contribute" my patch that "fixes" all those
> annoying (to me) coding idioms in implementation X. Do you think the
> maintainers of X will consider *that* patch desirable? ;-)

For all values of X where X = SBCL, you have a good shot ;)

Actually, there is some talk of merging SBCL with CMUCL, since the
refactoring done by SBCL is resulting in some huge portability benefits,
and there are cases where people using SBCL want to use some of the
"extras" provided by CMUCL such as the UNIX package, etc.

--
Rahul Jain
From: Daniel Barlow
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87isxc6zo3.fsf@noetbook.telent.net>
Rahul Jain <·····@rice.edu> writes:

> Peter Seibel <·····@javamonkey.com> writes:
>
>> You think? Cool--I'll "contribute" my patch that "fixes" all those
>> annoying (to me) coding idioms in implementation X. Do you think the
>> maintainers of X will consider *that* patch desirable? ;-)
>
> For all values of X where X = SBCL, you have a good shot ;)

I think that Peter was talking in more general terms than the original
specific suggestion of a tree shaker, and I wouldn't want
comp.lang.lisp to think that SBCL will merge any and every patch that
comes along just because someone believes it will fix their
annoyances.  Because we won't  :-)

(Speaking from personal experience, I've been unable to convince
people that TRUENAME would be better off to _not_ follow symlinks, or
that LOAD-1-FOREIGN should be renamed)

But, specifically, if he (or anyone else reading) does want to
introduce a tree shaker, that would be very interesting.  I'd probably
suggest to anyone wanting to tackle it that they start by looking at
the new XREF functionality that Eric Marsden has introduced in CMUCL
and seeing how that translates.

(with-plug
 SBCL 0.7.11 will be released on or around New Year's Eve,
 all being well.  Changes are mostly bugfixes, but it also now builds on
 FreeBSD 5.0 - and apparently, is or will soon be in the freebsd
 "ports" system.

 Depending on how much time I have available, native threading
 (initially on Linux x86) will probably start landing in January or
 February, but that estimate's made wearing a developer hat, not a
 project manager hat.)


-dan

-- 

   http://www.cliki.net/ - Link farm for free CL-on-Unix resources 
From: Paolo Amoroso
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <qcsNPvMFAZiJa2DI2mHzjzmTUsMj@4ax.com>
On Fri, 27 Dec 2002 06:29:51 GMT, Peter Seibel <·····@javamonkey.com>
wrote:

> was worth implementing), can anyone describe more or less what
> strategies are used in implementations that do generate standalone
> executables? Maybe someone (maybe me) will say, hmmm, that doesn't

Another relevant source of information is the paper "Delivering the Goods
with Lisp", by Barber and Imlah, in the Sep 1991 issue of CACM.


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Tim Daly, Jr.
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <87smwhk2r6.fsf@bob.intern>
Paolo Amoroso <·······@mclink.it> writes:

> On Fri, 27 Dec 2002 06:29:51 GMT, Peter Seibel <·····@javamonkey.com>
> wrote:
> 
> > was worth implementing), can anyone describe more or less what
> > strategies are used in implementations that do generate standalone
> > executables? Maybe someone (maybe me) will say, hmmm, that doesn't
> 
> Another relevant source of information is the paper "Delivering the Goods
> with Lisp", by Barber and Imlah, in the Sep 1991 issue of CACM.
> 

Are you and I looking at the same "paper"?  They prattle for a mere
three pages about how slow their PS/2 was and how they could maybe
have used this product or maybe have used that product...  Not a word
about strategies for generating standalone executables.

Not that I can't sympathize about the PS/2.  Man, those things sucked.

-Tim
From: Paolo Amoroso
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <sE8PPjRBWRTtcCF0XsHaRm9DjLdN@4ax.com>
On 29 Dec 2002 02:55:25 +0100, ···@tenkan.org (Tim Daly, Jr.) wrote:

> Paolo Amoroso <·······@mclink.it> writes:
[...]
> > Another relevant source of information is the paper "Delivering the Goods
> > with Lisp", by Barber and Imlah, in the Sep 1991 issue of CACM.
> > 
> 
> Are you and I looking at the same "paper"?  They prattle for a mere
> three pages about how slow their PS/2 was and how they could maybe
> have used this product or maybe have used that product...  Not a word
> about strategies for generating standalone executables.

I think we are talking about the same paper. But while you probably have a
copy of it handy, I no longer have it, just a vague recollection. Hence its
alleged relevance and implied usefulness.


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Joe Marshall
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <fzsfmnt5.fsf@ccs.neu.edu>
Peter Seibel <·····@javamonkey.com> writes:

> Supposing for a moment that not everyone who *thinks* they need/want
> to deliver a standalone executable is totally on crack (and it must be
> of *some* use as all the commercial vendors seem to have thought it
> was worth implementing), can anyone describe more or less what
> strategies are used in implementations that do generate standalone
> executables? 

All lisp systems have *some* basic executable component.  The typical
bootstrapping strategy is to write the lisp runtime support as an
executable (for instance, as a C program) then using this basic
support, load the rest of the lisp system (either having
cross-compiled the lisp sources into fasl files, or by loading the
lisp compiler source code into a bootstrap interpreter and having it
compile itself first).  This is often called `cold loading' the lisp
system.

Once all the code is loaded into the system, the usual next step is to
produce a `dump' of the heap.  This usually involves something like
writing out all the allocated storage into a file.  When the basic
lisp executable is next run, you either provide it the location of the
image, or it searches for it in known locations.

A `standalone executable' (as what people seem to want) is a file that
the native operating system linker/loader can launch.  I.e., it is a
PECOFF file on Windows, an ELF file on Linux, etc. etc.  These file
formats are fairly flexible in that you can generally add arbitrary
`sections' to them.  The OS linker/loader will map each section into
memory and then the executable code can go about finding out where it
is.

So a trivial way of creating a `standalone executable' is to more or
less concatenate the `bootstrap executable' with a `dumped image'.  In
fact, there are lisps which do almost exactly that (you have to adjust
the resulting COFF or ELF file so that checksums and tables match up,
but otherwise they are just concatenating the files).

There is no material difference between providing a `standalone
executable' and providing a `set of files'.  The only difference is
that in one case the hierarchy is a `directory' composed of `files',
and the other case is a `COFF' or `ELF' file composed of `sections'.
If you deliver the package in a `zip' or `tar' archive, the hierarchy,
whatever it is, is hidden in a single compressed unit.

So when someone says `I *need* a standalone executable' what they are
saying is `it is essential to me that I don't see the software
components when I use the ls command.  Make me use some sort of ELF or
COFF image dumper instead.'  This is a ludicrous requirement.  If you
don't want to see the components, don't look at them.

Since these people seem to be reasonably comfortable with delivering
software as `dll's, the requirement seems even *less* sensical:  I
only want to see extensions like `exe' and `dll' in my software, I
don't want to see `fasl' and `dump' or `image'.
From: Frank A. Adrian
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <wB6Q9.1253$FV3.65575@news.uswest.net>
Joe Marshall wrote:

> So when someone says `I need a standalone executable' what they are
> saying is `it is essential to me that I don't see the software
> components when I use the ls command.  Make me use some sort of ELF or
> COFF image dumper instead.'  This is a ludicrous requirement.

In most cases, it IS a ludicrous requirement (and one not necessarily needed 
for most products).  But there are a few reasons why this requirement might 
be useful:

* A single file simplifies the installation and uninstallation of the 
program, as well as making it less easy for a naive user to screw up the 
system once installed (people will delete the oddest things).

* It's easier to download a single file than a bunch of them when 
desconstructing archival files is not possible.

* I really don't want to use Lisp and am trying to find a reason why it 
won't work for me.

I believe that the last reason is the main reason that causes the 
requirement to be added.

faa
From: Bulent Murtezaoglu
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <871y3ylayc.fsf@acm.org>
>>>>> "FAA" == Frank A Adrian <·······@ancar.org> writes:
[...]
    FAA> * A single file simplifies the installation and
    FAA> uninstallation of the program, as well as making it less easy
    FAA> for a naive user to screw up the system once installed
    FAA> (people will delete the oddest things).

    FAA> * It's easier to download a single file than a bunch of them
    FAA> when desconstructing archival files is not possible.

Yes to both.  I occasionally whip up mini gui's in Tcl/Tk to help clients
figure out what exactly it is they want.  I use prowrap to produce a 
standalone .exe which contains the libs and the script itself.  It is VERY
handy to stick a 3-4Mb exe somewhere shoot an e-mail saying "get this file
from this URL, save it onto your desktop and doubleclick on it."    


    FAA> * I really don't want to use Lisp and am trying to find a
    FAA> reason why it won't work for me.

    FAA> I believe that the last reason is the main reason that causes
    FAA> the requirement to be added.

Not for me, then again I don't post wishes for stand-alone executables 
either.  BTW, Lispworks and possibly Allegro do have this 'one file, stand 
alone executable' facility.

cheers,

BM
From: Paolo Amoroso
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <0HgMPu4iATDFdaVihEmI2b6K6kwP@4ax.com>
On Fri, 27 Dec 2002 09:19:04 +1100, "Coby Beck" <·····@mercury.bc.ca>
wrote:

> Delivering executables is the only thing there I don't know if you can get
> for free.  Usually it is not really what people need/want anyway.

Popular open-source implementations such as CLISP and CMU CL (and I also
guess SBCL) are able to deliver standalone applications. The process is not
automated, but it just involves shipping a few files to the customer:

1) the Lisp executable
2) the dumped image of the application
3) a small shell script that runs #1 and passes it #2 as an argument

The CLISP documentation explains how to deliver applications on all
supported operating systems.


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Paolo Amoroso
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <jM8NPkQxWA+ZEwexgiShT2TWpjUF@4ax.com>
On Fri, 27 Dec 2002 17:00:21 +0100, Paolo Amoroso <·······@mclink.it>
wrote:

> Popular open-source implementations such as CLISP and CMU CL (and I also
> guess SBCL) are able to deliver standalone applications. The process is not
> automated, but it just involves shipping a few files to the customer:

The procedure for CMU CL is explained in this comp.lang.lisp articles:

  ··············@orion.bln.pmsf.de

See also the messages in thread "Program Delivery" posted to the cmucl-help
mailing list, in particular:

  ··············@orion.bln.pmsf.de


Paolo
-- 
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
From: Thomas F. Burdick
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <xcv8yya5837.fsf@conquest.OCF.Berkeley.EDU>
"Lisp Newbie" <········@lisp.programmer> writes:

> Is there such thing as free Lisp compiler, that can work under Linux and
> Windows, develiver executables, have sockets implementation and support
> multithreading?

Just because it hasn't been pointed out already, it doesn't take a
genius to make something that feels like a native executable from a
native VM executable plus a core and one or more fasl's, at least on
unix.  Modifying CMUCL's vm to look for things in different places
isn't very hard.  Call the resulting binary "myapp" instead of "lisp",
and you're done.

-- 
           /|_     .-----------------------.                        
         ,'  .\  / | No to Imperialist war |                        
     ,--'    _,'   | Wage class war!       |                        
    /       /      `-----------------------'                        
   (   -.  |                               
   |     ) |                               
  (`-.  '--.)                              
   `. )----'                               
From: Lisp Newbie
Subject: Re: Looking for Lisp compiler
Date: 
Message-ID: <AkrP9.171940$6k.4125153@news1.west.cox.net>
I thought CMUCL doesn't work in Windows. Am I right?
CLISP doesn't qualify to my original requirements due to lack of
multithreading.
So what else is left among freebies?


"Thomas F. Burdick" <···@conquest.OCF.Berkeley.EDU> wrote in message
····················@conquest.OCF.Berkeley.EDU...
> "Lisp Newbie" <········@lisp.programmer> writes:
>
> > Is there such thing as free Lisp compiler, that can work under Linux and
> > Windows, develiver executables, have sockets implementation and support
> > multithreading?
>
> Just because it hasn't been pointed out already, it doesn't take a
> genius to make something that feels like a native executable from a
> native VM executable plus a core and one or more fasl's, at least on
> unix.  Modifying CMUCL's vm to look for things in different places
> isn't very hard.  Call the resulting binary "myapp" instead of "lisp",
> and you're done.
>
> --
>            /|_     .-----------------------.
>          ,'  .\  / | No to Imperialist war |
>      ,--'    _,'   | Wage class war!       |
>     /       /      `-----------------------'
>    (   -.  |
>    |     ) |
>   (`-.  '--.)
>    `. )----'