From: Zachary Turner
Subject: help!  absolute beginner
Date: 
Message-ID: <741sh7$lje$1@uuneo.neosoft.com>
i have never programmed lisp before in my life, so i have no idea what's
really going on here, but i plan on learning it over the next 6 months or
so.  what i need is a LISP compiler so that i can get started.  I'm looking
into Harlequin LispWorks 4.1 Professional.  I've downloaded the Personal
Edition (which is available free for download at their website) but i have
no idea how to get a simple program up and running.  Can someone tell me
exactly what i need to do and exactly what code to type just to write an
extremely simple program (the LISP equivalent of the typical C "Hello World"
perhaps?) that prints something on the screen, just so that i can see some
results and know that I'm actually on the right track?  I have no idea how
to use LispWorks, and no idea about the language itself, so if someone could
just hold my hand and treat me like a complete moron i'd really appreciate
it.  :-)  Thanks for any help.

Zachary Turner

From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <741siv$ljk$1@uuneo.neosoft.com>
by the way, i don't need the code explained to me, i'll learn all of that
soon enough, i just want to see something happen before I go shell out money
on this product.  Thanks

Zachary Turner

Zachary Turner wrote in message <············@uuneo.neosoft.com>...
>i have never programmed lisp before in my life, so i have no idea what's
>really going on here, but i plan on learning it over the next 6 months or
>so.  what i need is a LISP compiler so that i can get started.  I'm looking
>into Harlequin LispWorks 4.1 Professional.  I've downloaded the Personal
>Edition (which is available free for download at their website) but i have
>no idea how to get a simple program up and running.  Can someone tell me
>exactly what i need to do and exactly what code to type just to write an
>extremely simple program (the LISP equivalent of the typical C "Hello
World"
>perhaps?) that prints something on the screen, just so that i can see some
>results and know that I'm actually on the right track?  I have no idea how
>to use LispWorks, and no idea about the language itself, so if someone
could
>just hold my hand and treat me like a complete moron i'd really appreciate
>it.  :-)  Thanks for any help.
>
>Zachary Turner
>
>
From: Sunil Mishra
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <efyogpn1kno.fsf@flash.cc.gatech.edu>
Have a look at the Association of Lisp Users web site:

http://www.elwoodcorp.com/alu/

You will find a fair bit of information (tutorials as well) to get you up
to speed. Good luck!

Sunil

"Zachary Turner" <·······@elsitech.com> writes:

> by the way, i don't need the code explained to me, i'll learn all of that
> soon enough, i just want to see something happen before I go shell out money
> on this product.  Thanks
> 
> Zachary Turner
> 
> Zachary Turner wrote in message <············@uuneo.neosoft.com>...
> >i have never programmed lisp before in my life, so i have no idea what's
> >really going on here, but i plan on learning it over the next 6 months or
> >so.  what i need is a LISP compiler so that i can get started.  I'm looking
> >into Harlequin LispWorks 4.1 Professional.  I've downloaded the Personal
> >Edition (which is available free for download at their website) but i have
> >no idea how to get a simple program up and running.  Can someone tell me
> >exactly what i need to do and exactly what code to type just to write an
> >extremely simple program (the LISP equivalent of the typical C "Hello
> World"
> >perhaps?) that prints something on the screen, just so that i can see some
> >results and know that I'm actually on the right track?  I have no idea how
> >to use LispWorks, and no idea about the language itself, so if someone
> could
> >just hold my hand and treat me like a complete moron i'd really appreciate
> >it.  :-)  Thanks for any help.
> >
> >Zachary Turner
> >
> >
From: Toolshed37
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <19981201224608.08315.00000844@ng146.aol.com>
well, i took a brief look at it, and not much help.  i dont' really want a
tutorial right now, i just want someone to give me enough code that i can just
cut and paste into the window of LispWorks and then tell me what to do to
execute the program so that i can see some type of output on the screen.  i'm
not going to start trying to learn lisp until later, right now i just want to
see something on my screen.  if you could just post an _entire_ hello world
program and tell me what to do in LispWorks to get it to run, i'd really
appreciate it.
From: David B. Lamkins
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <4b492.16464$Sz4.9431935@news.teleport.com>
In article <·····························@ng146.aol.com> ,
··········@aol.com (Toolshed37) wrote:

>well, i took a brief look at it, and not much help.  i dont' really want a
>tutorial right now, i just want someone to give me enough code that i can just
>cut and paste into the window of LispWorks and then tell me what to do to
>execute the program so that i can see some type of output on the screen.  i'm
>not going to start trying to learn lisp until later, right now i just want to
>see something on my screen.  if you could just post an _entire_ hello world
>program and tell me what to do in LispWorks to get it to run, i'd really
>appreciate it.

Well, define what you mean by a hello world program...  Here are several
that meet your requirements.

Because strings are self-evaluating in Lisp, this is the the simplest
program:
"Hello, world!"

If you want to get closer to the canonical hello world program that does
explicit output, try these:
(write "Hello, world!")
(format t "Hello, world!")
(format nil "Hello, world!")

If you want to pop up a window that says hello world, that kind of output is
platform dependent (and I don't have a copy of LW at hand).  Typically, you
can create a window that behaves as a text output stream, and use it
something like this:
(let ((w (make-instance 'window)))
  (format w "Hello, world!"))
Again, the window example may not work in LW (if it does, it's a lucky guess
on the class name).  Why don't you take a look through the manual and see
whether they have a tutorial?

Finally, if you want to create a standalone application, you're really gonna
have to crack that manual.  Every Lisp is different, but it would probably
be fruitful if you searched the index for some hyphenated combination of
dump or save and image or application.

---
David B. Lamkins <http://www.teleport.com/~dlamkins/>

There are many ways to abbreviate something, but only one way not to.
From: Sunil Mishra
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <efy3e6zjei1.fsf@spanker.cc.gatech.edu>
"Hello World" is a rather futile exercise.

(princ "Hello World")
|Hello World|
"Hello World"

Any of these expressions at the command prompt will do what you want to,
and yet none of these is satisfying as a program.

If you want to write a function that simply does hello world, then

(defun foo ()
  (princ "Hello World"))

(foo)

This to me does not quite seem very satisfying as a program either.

I guess what I am trying to say is that Lisp is not very amenable to small
toy programs, simply because they are trivial to put together. Now, if you
want to deal with a large complex problem, Lisp tends to work great. There
are many, many problems that take loads of code in other languages, but end
relatively clean and nice in Lisp. An expressive syntax helps :-)

Going up yet another level, the "right way" to learn a language would be to
figure out what it is good for. Obviously, Lisp is not good for writing
Hello World, since you end up carrying around an environment at least 10MB
in size to get this to work. I don't know what your motives for learning
lisp are, but I would think hard about this question.

Now, if you want to ask what lisp is good for, I would suggest trekking
over to dejanews. This question has been asked enough times. If you have a
problem at hand, there are lots of *very* smart people on this newsgroup
that would be more than happy to help you out. (They've been kind to me,
certainly.)

Sunil

··········@aol.com (Toolshed37) writes:

> well, i took a brief look at it, and not much help.  i dont' really want a
> tutorial right now, i just want someone to give me enough code that i can just
> cut and paste into the window of LispWorks and then tell me what to do to
> execute the program so that i can see some type of output on the screen.  i'm
> not going to start trying to learn lisp until later, right now i just want to
> see something on my screen.  if you could just post an _entire_ hello world
> program and tell me what to do in LispWorks to get it to run, i'd really
> appreciate it.
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <743jh9$4up$1@uuneo.neosoft.com>
Sunil Mishra wrote in message ...
>Going up yet another level, the "right way" to learn a language would be to
>figure out what it is good for. Obviously, Lisp is not good for writing
>Hello World, since you end up carrying around an environment at least 10MB
>in size to get this to work. I don't know what your motives for learning
>lisp are, but I would think hard about this question.
>
>Now, if you want to ask what lisp is good for, I would suggest trekking
>over to dejanews. This question has been asked enough times. If you have a
>problem at hand, there are lots of *very* smart people on this newsgroup
>that would be more than happy to help you out. (They've been kind to me,
>certainly.)
Oh, I have very valid motives for learning LISP.  I'm a university student
and in
the Spring I'm doing an independent study course on Artificial Intelligence,
for which I've decided to learn LISP, since it is perfectly suited for doing
AI
programming.  I also realize that it is not explicitly for AI, but that it
does
suit the problem perfectly.  I was able to get something to run with the
help
of all the posts, thanks to everyone who posted.  I'll probably go shell out
some $$ for LispWorks now, I like the fact that you can write in LISP and
deliver it as a .DLL, which you can link to in C++ :-)

Zachary Turner
From: Joachim Achtzehnter
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <uc3e6ye6tq.fsf@ns.mercury.bc.ca>
"Zachary Turner" <·······@elsitech.com> writes:
>
> Oh, I have very valid motives for learning LISP.  I'm a university
> student...  I'll probably go shell out some $$ for LispWorks now...

If you just want to learn Lisp then why are you so keen on spending a
large sum of money? Last time I looked, both Harlequin and Franz had
full-featured versions of their Lisp environments free for personal
use. And there are a whole bunch of Lisp implementations that are free
software, in case you ever wanted to look at the inside :-)

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74466d$9i3$1@uuneo.neosoft.com>
I'm spoiled I guess?  I've been writing games in C/C++ using DirectX and
OpenGL
for some time now and it'd be nice if, after I get good with Lisp, I could
write all
my AI code in LISP, compile it to a DLL, and dynamically link to the DLL in
my
games, which will be written in C/C++.

Joachim Achtzehnter wrote in message ...
>"Zachary Turner" <·······@elsitech.com> writes:
>>
>> Oh, I have very valid motives for learning LISP.  I'm a university
>> student...  I'll probably go shell out some $$ for LispWorks now...
>
>If you just want to learn Lisp then why are you so keen on spending a
>large sum of money? Last time I looked, both Harlequin and Franz had
>full-featured versions of their Lisp environments free for personal
>use. And there are a whole bunch of Lisp implementations that are free
>software, in case you ever wanted to look at the inside :-)
>
>Joachim
>
>--
>·······@kraut.bc.ca      (http://www.kraut.bc.ca)
>·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Sunil Mishra
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <efy67bthk5r.fsf@hustle.cc.gatech.edu>
I suspect doing things the other way around - writing your games in lisp,
and using C/C++ routines for specific functions - will work out better. But
then that's just a guess. I'm only familiar with Harlequin Lispworks, and I
know they have an interface for automatically reading header files and
constructing an interface to C. So, as long as you have the header files,
constructing an OpenGL interface *ought* to be straightforward. (I haven't
done this, so I can't be certain.)

Sunil

"Zachary Turner" <·······@elsitech.com> writes:

> I'm spoiled I guess?  I've been writing games in C/C++ using DirectX and
> OpenGL
> for some time now and it'd be nice if, after I get good with Lisp, I could
> write all
> my AI code in LISP, compile it to a DLL, and dynamically link to the DLL in
> my
> games, which will be written in C/C++.
> 
> Joachim Achtzehnter wrote in message ...
> >"Zachary Turner" <·······@elsitech.com> writes:
> >>
> >> Oh, I have very valid motives for learning LISP.  I'm a university
> >> student...  I'll probably go shell out some $$ for LispWorks now...
> >
> >If you just want to learn Lisp then why are you so keen on spending a
> >large sum of money? Last time I looked, both Harlequin and Franz had
> >full-featured versions of their Lisp environments free for personal
> >use. And there are a whole bunch of Lisp implementations that are free
> >software, in case you ever wanted to look at the inside :-)
> >
> >Joachim
> >
> >--
> >·······@kraut.bc.ca      (http://www.kraut.bc.ca)
> >·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <7467qv$osp$1@uuneo.neosoft.com>
I'm not sure how practical that would be.  All graphics, sound, input, and
network programming I typically do in C++ so it wouldn't make much sense to
write the whole program in Lisp and link in all code relating to graphics,
sound, input, and network programming when I could write it in C++ and link
in only the AI stuff.  One thing that's just come to my attention though:
How do you prototype a Lisp function in C?  From what I've learned of Lisp
so far (not very much at all) it isn't typed at all.  The list (1 2 3 (foo
bar 37)) is quite different in memory size from the list (1).  So I guess
the two main things that you would need to know how to declare in C are an
atom and a list, taking into consideration that a list can contain other
lists.  Has anyone had any experience doing this?


Sunil Mishra wrote in message ...
>I suspect doing things the other way around - writing your games in lisp,
>and using C/C++ routines for specific functions - will work out better. But
>then that's just a guess. I'm only familiar with Harlequin Lispworks, and I
>know they have an interface for automatically reading header files and
>constructing an interface to C. So, as long as you have the header files,
>constructing an OpenGL interface *ought* to be straightforward. (I haven't
>done this, so I can't be certain.)
>
>Sunil
>
>"Zachary Turner" <·······@elsitech.com> writes:
>
>> I'm spoiled I guess?  I've been writing games in C/C++ using DirectX and
>> OpenGL
>> for some time now and it'd be nice if, after I get good with Lisp, I
could
>> write all
>> my AI code in LISP, compile it to a DLL, and dynamically link to the DLL
in
>> my
>> games, which will be written in C/C++.
>>
>> Joachim Achtzehnter wrote in message ...
>> >"Zachary Turner" <·······@elsitech.com> writes:
>> >>
>> >> Oh, I have very valid motives for learning LISP.  I'm a university
>> >> student...  I'll probably go shell out some $$ for LispWorks now...
>> >
>> >If you just want to learn Lisp then why are you so keen on spending a
>> >large sum of money? Last time I looked, both Harlequin and Franz had
>> >full-featured versions of their Lisp environments free for personal
>> >use. And there are a whole bunch of Lisp implementations that are free
>> >software, in case you ever wanted to look at the inside :-)
>> >
>> >Joachim
>> >
>> >--
>> >·······@kraut.bc.ca      (http://www.kraut.bc.ca)
>> >·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <746dv0$qij$1@uuneo.neosoft.com>
On second thought maybe this isn't such a bad idea..  Interesting at the
least...  I'll give it a shot in a few months after I get proficient with
Lisp.  On another note, does anyone know what the deal is with Franz, Inc.?
I've been trying to get in touch with these people for the last few weeks
and they don't return emails and when I call they say that the lady who
handles sales "is not currently in the office and wont' be in for the
remainder of the day."  That's happened twice now and the email thing has
happened twice also.  I evaluated Allegro CL 5.0 and I really like it, but
these people are basically sitting here with the "No we don't your money.
Go give it to Harelquin" attitude, which is really frustrating since I want
to purchase Allegro.  Has anyone else had better luck with them?

Zach


Sunil Mishra wrote in message ...
>I suspect doing things the other way around - writing your games in lisp,
>and using C/C++ routines for specific functions - will work out better. But
>then that's just a guess. I'm only familiar with Harlequin Lispworks, and I
>know they have an interface for automatically reading header files and
>constructing an interface to C. So, as long as you have the header files,
>constructing an OpenGL interface *ought* to be straightforward. (I haven't
>done this, so I can't be certain.)
>
>Sunil
>
>"Zachary Turner" <·······@elsitech.com> writes:
>
>> I'm spoiled I guess?  I've been writing games in C/C++ using DirectX and
>> OpenGL
>> for some time now and it'd be nice if, after I get good with Lisp, I
could
>> write all
>> my AI code in LISP, compile it to a DLL, and dynamically link to the DLL
in
>> my
>> games, which will be written in C/C++.
>>
>> Joachim Achtzehnter wrote in message ...
>> >"Zachary Turner" <·······@elsitech.com> writes:
>> >>
>> >> Oh, I have very valid motives for learning LISP.  I'm a university
>> >> student...  I'll probably go shell out some $$ for LispWorks now...
>> >
>> >If you just want to learn Lisp then why are you so keen on spending a
>> >large sum of money? Last time I looked, both Harlequin and Franz had
>> >full-featured versions of their Lisp environments free for personal
>> >use. And there are a whole bunch of Lisp implementations that are free
>> >software, in case you ever wanted to look at the inside :-)
>> >
>> >Joachim
>> >
>> >--
>> >·······@kraut.bc.ca      (http://www.kraut.bc.ca)
>> >·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Vassili Bykov
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <u1z92.2402$5n1.19100604@news.magma.ca>
For the kind of things you are describing, you might really like Corman Lisp
(http://www.corman.net/CormanLisp.html).

Zachary Turner wrote in message <············@uuneo.neosoft.com>...
>On second thought maybe this isn't such a bad idea..  Interesting at the
>least...  I'll give it a shot in a few months after I get proficient with
>Lisp.  On another note, does anyone know what the deal is with Franz, Inc.?
[...]
From: ···········@alcoa.com
Subject: Re: help! absolute beginner
Date: 
Message-ID: <7470f8$l5m$1@nnrp1.dejanews.com>
In article <·······················@news.magma.ca>,
  "Vassili Bykov" <·······@objectpeople.com> wrote:
> For the kind of things you are describing, you might really like Corman Lisp
> (http://www.corman.net/CormanLisp.html).

Corman Lisp is betaware. I recommend that if you have the money for real
commercial grade software that you stick with Franz or Harlequin. They have
been battle tested, implement the ANSI standards, and you will experience far
fewer fustrations with them. My experience with Corman Lisp involved frequent
crashes, missing features (like macrolet), and far slower performance. And
all I tried on Corman Lisp were a few small benchmarks. I hate to be negative
but you can't really expect a new Lisp environment to be competitive with
folks with a large staff and years behind their product? For a beginner on
Windows I recommend Harlequin Lispworks Personal Edition and when you get a
real project graduate to the Professional Edition (~$500) which costs only a
few hundred more than what Corman wants. If you need even more performance
(and have even more dollars) then consider Franz's Allegro CL(~$3000-4000).
(I have both and prefer ACL over Lispworks but I am biased by having used ACL
on Unix for years and the need for the absolute fastest floating point
speed.)

--
John Watton
Aluminum Company of America

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: Vassili Bykov
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74d0hq$iti$1@nnrp1.dejanews.com>
In article <············@nnrp1.dejanews.com>,
  ···········@alcoa.com wrote:

> Corman Lisp is betaware. I recommend that if you have the money for
> real commercial grade software that you stick with Franz or
> Harlequin. They have been battle tested, implement the ANSI standards,
> and you will experience far fewer fustrations with them. My experience
> with Corman Lisp involved frequent crashes, missing features (like
> macrolet), and far slower performance.

Very true.  But I do believe Corman Lisp has its place in the sun,
together with all the big commercial and non-commercial players,
because it scratches some itches other implementations do not.

> I hate to be negative but you can't
> really expect a new Lisp environment to be competitive with folks with
> a large staff and years behind their product?

Perhaps because there does not have to be a competition.  You
obviously speak from the viewpoint of a corporate developer who has
the money and expects return of investment.  This is great, but what
about the hobbyists?  The "Personal" and "Lite" editions of LWW and
ACLW are very different from Corman Lisp, and not always to their
advantage.  First thing, they are crippled, second, you don't have the
source, third, you are not allowed to redistribute them as parts of
your application.

> For a beginner on
> Windows I recommend Harlequin Lispworks Personal Edition and when you
> get a real project graduate to the Professional Edition (~$500) which
> costs only a few hundred more than what Corman wants.

What is better, open-source betaware or black box crippleware?  Before
answering, it is important to consider what for.

For a beginner, LWW Personal or ACLW Lite are indeed the best choice.

But, there is not necessarily a graduation with getting a real Lisp
project.  One may have Lisp-unrelated projects in real life they are
perfectly happy about, or unable to change for the time being.  So
what do you use if you want something that is not crippled, is still
free or very low cost, allows you to let others see and use your work
as an application, and allows you to see what is inside so that you
can add what is missing or fix what is broken?  None of the free
versions of the big commercial players fit the description.

As for "what Corman wants", you seem to have misread the license
conditions.  Roger does not want anything for the compiler and the
runtime system. Corman Lisp is free, and that is uncrippled version
with the right to non-commercially distribute binary applications
which may include the compiler.  None of the free versions of the
commercial stuff in Windows come close to this.  Add to that full
source code, and there is some weight to compare.

Now, on the technical side, Corman Lisp has a very interesting balance
between low-level and high-level stuff.  You can do everything a C
programmer can.  This was the reason I said the original poster may be
interested in Corman Lisp, since there was some talk about polygon
pushing and game programming.  The FFI in Corman Lisp is very flexible
and transparent.  The assembly-level stuff is fully open, to the point
that one can write inline assembly code in Lisp functions.  It is,
essentially, Lisp with all the capabilities of C when it comes to
communicating with the outside world.

So, my take on this (besides that I don't think Lisp community
should (or, even, can afford to) kick the underdog) is yes,
Corman Lisp is worse compared to the entrenched big commercial
players, but it just might happen to be the kind of worse that is
better here and there...

--Vassili

P.S.  BTW, Roger Corman will soon release an update with improved
thread stability (those crashes are due to Windows thread
implementation "features"), macrolet, and the full source--Lisp and C,
among some other things.

P.P.S.  No, I am not affiliated with Roger.

P.^3S. If you are interested in joining an online community of Corman
Lisp users, you are welcome to <http://www.dejanews.com/~cormanlisp>.

--
Vassili Bykov
The Object People
http://www.objectpeople.com

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: ···········@alcoa.com
Subject: Re: help! absolute beginner - Corman Lisp vs. ...
Date: 
Message-ID: <74eois$u59$1@nnrp1.dejanews.com>
In article <············@nnrp1.dejanews.com>,
  Vassili Bykov <·······@objectpeople.com> wrote:
>
> > As for "what Corman wants", you seem to have misread the license
> conditions.  Roger does not want anything for the compiler and the
> runtime system. Corman Lisp is free

He wants $300 for the development environment if memory serves me. The rest is
"free".

I want to correct something else I said earlier. Allegro CL Professional is
something like $4000 which gives you the ability to distribute runtime
internal to a corporation (external to corporation Franz wants to negotiate a
cut). There is an Allegro Personal which is in the $500-900 range but no
runtime ability which I had not mentioned.

--
John Watton
Aluminum Company of America

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: Rainer Joswig
Subject: Re: help! absolute beginner - Corman Lisp vs. ...
Date: 
Message-ID: <joswig-0612982156190001@194.163.195.67>
In article <············@nnrp1.dejanews.com>, ···········@alcoa.com wrote:

> In article <············@nnrp1.dejanews.com>,
>   Vassili Bykov <·······@objectpeople.com> wrote:
> >
> > > As for "what Corman wants", you seem to have misread the license
> > conditions.  Roger does not want anything for the compiler and the
> > runtime system. Corman Lisp is free
> 
> He wants $300 for the development environment if memory serves me. The rest is
> "free".
> 
> I want to correct something else I said earlier. Allegro CL Professional is
> something like $4000 which gives you the ability to distribute runtime
> internal to a corporation (external to corporation Franz wants to negotiate a
> cut). There is an Allegro Personal which is in the $500-900 range but no
> runtime ability which I had not mentioned.

So it seems that LispWorks for Windows is much cheaper?

-- 
http://www.lavielle.com/~joswig
From: Vassili Bykov
Subject: Re: help! absolute beginner - Corman Lisp vs. ...
Date: 
Message-ID: <74faug$d2m$1@nnrp1.dejanews.com>
In article <············@nnrp1.dejanews.com>, ···········@alcoa.com wrote:
> In article <············@nnrp1.dejanews.com>,
>   Vassili Bykov <·······@objectpeople.com> wrote:
> >
> > > As for "what Corman wants", you seem to have misread the license
> > conditions.  Roger does not want anything for the compiler and the
> > runtime system. Corman Lisp is free
>
> He wants $300 for the development environment if memory serves me. The rest is
> "free".

Right now it's $100.  This buys the IDE and the right to sell the applications
you develop.

I don't see the need to quote "free" talking about the rest, the rest has
enough value by itself :-).  It's not free in the sense of FSF, it is free in
the sense you don't have to pay.  (And in the more humourous senses suggested
by Rainer Joswig).  And quite frankly, the IDE is not that great, at least for
now, compared to simple Emacs, so taking it out does not turn the rest into
crippleware.

I never wanted to imply Corman Lisp would give Franz or Harlequin a run for
their money.  But, they are players of totally different leagues, and this
works both ways.  ACLW Lite and LWW Personal do not have some of the virtues
of Corman Lisp.  Those of the hobbyists experienced enough to not be afraid
of rough spots and looking for easy access to foreign and low-level stuff,
can have some great time with Corman Lisp.  It all boils down to where you
interests are.

--Vassili

--
Vassili Bykov
The Object People
http://www.objectpeople.com

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0612980615280001@194.163.195.67>
In article <············@nnrp1.dejanews.com>, Vassili Bykov
<·······@objectpeople.com> wrote:

> 
> As for "what Corman wants", you seem to have misread the license
> conditions.  Roger does not want anything for the compiler and the
> runtime system. Corman Lisp is free, and that is uncrippled version
> with the right to non-commercially distribute binary applications
> which may include the compiler.  None of the free versions of the
> commercial stuff in Windows come close to this.  Add to that full
> source code, and there is some weight to compare.

"Free" nowadays often seems to be:

- you are free to fix the bugs you never wanted to fix.
- you are free to implement the functionality you
  never wanted to program yourself
- you are free to wade through large amounts of complicated
  source code you never wanted to look at

Some people seem to really underestimate the massive amount of work
that is needed to write a really usable and balanced system. This
week I checked a small piece of code (two pages) in five Lisp systems.
a)  one *free* versions I checked was not able to run it
    (a class fixnum?) without modifications
b)  for the two free systems I gave up waiting for the result
    (P266 and SUN E250). My old MacIvory beats them. Ha!

So, people don't get your expectations to high and never underestimate
the work that is needed to develop a Lisp system. I said it some
time ago, I have ***high*** respect for guys like Roger Corman
who are trying to tame the beast.

-- 
http://www.lavielle.com/~joswig
From: Pierre Mai
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87soetm4mn.fsf@orion.dent.isdn.cs.tu-berlin.de>
······@lavielle.com (Rainer Joswig) writes:

> Some people seem to really underestimate the massive amount of work
> that is needed to write a really usable and balanced system. This
> week I checked a small piece of code (two pages) in five Lisp systems.
> a)  one *free* versions I checked was not able to run it
>     (a class fixnum?) without modifications

While the ANSI Standard does not mandate that the type specifier
FIXNUM have a corresponding class (unlike INTEGER), it does IMHO
allow implementations to define classes for other type specifiers
(Section 4.3.7):

   "Individual implementations may be extended to define other type
    specifiers to have a corresponding class."

CMU CL does define a class for FIXNUM.

> b)  for the two free systems I gave up waiting for the result
>     (P266 and SUN E250). My old MacIvory beats them. Ha!

If you could give me access to the source code, I´d be very interested
in looking into performance issues in CMU CL on Intel, so that the
quality of implementation increases for one of the free systems out
there ;)

> So, people don't get your expectations to high and never underestimate
> the work that is needed to develop a Lisp system. I said it some

Regs, Pierre.

-- 
Pierre Mai <····@acm.org>               http://home.pages.de/~trillian/
  "One smaller motivation which, in part, stems from altruism is Microsoft-
   bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0612982119200001@194.163.195.67>
In article <··············@orion.dent.isdn.cs.tu-berlin.de>, Pierre Mai
<····@acm.org> wrote:

> > b)  for the two free systems I gave up waiting for the result
> >     (P266 and SUN E250). My old MacIvory beats them. Ha!
> 
> If you could give me access to the source code, I=B4d be very interested
> in looking into performance issues in CMU CL on Intel, so that the
> quality of implementation increases for one of the free systems out
> there ;)

Nothing special. Just using generic functions instead of
ordinary functions. Calling methods seems to have a huge overhead
(also conses like hell) in CMU CL. The version without methods
ran as expected. I tried this test only
once on my MacIvory and actually methods were faster (!) - on other
implementations there was a 15% speed penalty using
methods.

-- 
http://www.lavielle.com/~joswig
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0612982121540001@194.163.195.67>
In article <··············@orion.dent.isdn.cs.tu-berlin.de>, Pierre Mai
<····@acm.org> wrote:

> While the ANSI Standard does not mandate that the type specifier
> FIXNUM have a corresponding class (unlike INTEGER), it does IMHO
> allow implementations to define classes for other type specifiers
> (Section 4.3.7):
> 
>    "Individual implementations may be extended to define other type
>     specifiers to have a corresponding class."
> 
> CMU CL does define a class for FIXNUM.

You are right.

-- 
http://www.lavielle.com/~joswig
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-0612981714020001@hugo.westfalen.de>
In article <·······················@194.163.195.67>, ······@lavielle.com
(Rainer Joswig) wrote:

Exactly. With commercial applications, this greatly differs:

>- you are free to fix the bugs you never wanted to fix.

You are not free to fix the bugs you never wanted to fix. In fact, often
you don't have the sources to fix it.

>- you are free to implement the functionality you
>  never wanted to program yourself

You might be possible to add the functionality you need but the company
never want's to implement. In fact, as soon as you want to fix things in
the lowest levels of the product, you are most often lost. It is better
with Lisp, since Lisp is much better customizeable than C++-Compilers, but
the problem still exists.

>- you are free to wade through large amounts of complicated
>  source code you never wanted to look at

Yes, that is definitely much better with commercial products where you
don't get source at all. You might get _some_ source, as with most Lisps,
but that is not a quite common case for commercial software.

No, I definitely prefer high-quality Open Source [TM] software. Makes
living easier, especial in my field of work. Of course, there is much crap
out there. But actually with free software, you don't have to pay for
crap, like is the case with commercial ones.

bye, Georg
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0712980917370001@194.163.195.67>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> In article <·······················@194.163.195.67>, ······@lavielle.com
> (Rainer Joswig) wrote:
> 
> Exactly. With commercial applications, this greatly differs:
> 
> >- you are free to fix the bugs you never wanted to fix.
> 
> You are not free to fix the bugs you never wanted to fix. In fact, often
> you don't have the sources to fix it.

Often source doesn't help you or access to source
isn't even the problem at all.

Will access to source magically give a Lisp system
a usable GC? Even getting a decent
UI seems not to be possible. Does "Open Source"
and Lisp mean: we will get nowhere?

This stuff is really complicated. It is **naive** to expect
that this will show up soon. What's out there is
light-years away from even the ancient
Xerox InterLisp-D systems.

> >- you are free to wade through large amounts of complicated
> >  source code you never wanted to look at
> 
> Yes, that is definitely much better with commercial products where you
> don't get source at all. You might get _some_ source,

So you lately have looked into your Genera file system, right?
Looked for the sources of Genera?
Maybe you even have installed the additional source folder
of MCL on your Mac. Did you need it? Much? Would you
have needed it?

> as with most Lisps,
> but that is not a quite common case for commercial software.

I better ask the experts.

> No, I definitely prefer high-quality Open Source [TM]

Now it is a trademark? Religious marketing bullshit.

> software.

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87hfv8jmoz.fsf@piracy.red-bean.com>
······@lavielle.com (Rainer Joswig) writes:

> Often source doesn't help you or access to source
> isn't even the problem at all.
> 
> Will access to source magically give a Lisp system
> a usable GC? 

Certainly not, and I don't think that anyone is making that claim.  It
has to be written wether the author puts it into public domain or not.
People are motivated by many different things, and not all of them may
be congruent to your motivations and goals, or mine.

I think the previous poster has different criteria for making his
decision about what system to use than you do, and both sides need to
be aware of that.  For some a liberal license for distribution is more
important than GC performance, and for others performance and features
are worth whatever price is asked.  Luckily, both sides can be
satisfied!  It's all good man, it's all lisp.

> > No, I definitely prefer high-quality Open Source [TM]
> 
> Now it is a trademark? Religious marketing bullshit.

Yes.  Bruce Perens and Eric Raymond have started an organization to do
Open Source license certification, and part of their plan is to
trademark the term that has been in public use since way before they
applied for the trademark, and "protect it" from misuse.  It's
actually ideologically free and makes extra effort to espouse the
proper party lines about pragmatism and realism that makes corporate
bottom-feeders salivate.  So it's Non-Religious marketing bullshit,
but marketing bullshit none the less.  Be sure not to confuse them
with the people who actually writing code.
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3pv9w16ez.fsf@todday.aiai.ed.ac.uk>
* Rainer Joswig wrote:

> So you lately have looked into your Genera file system, right?
> Looked for the sources of Genera?

Yes, I do that a lot.  I've even fixed bugs in it (not the file
system, the networking stuff).  I'm very glad they gave source out
(albeit it wasn't open source), I wish more people did, as it can
really save a bunch of time.

--tim
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0712981408520001@pbg3.lavielle.com>
In article <···············@todday.aiai.ed.ac.uk>, Tim Bradshaw
<···@aiai.ed.ac.uk> wrote:

> * Rainer Joswig wrote:
> 
> > So you lately have looked into your Genera file system, right?
> > Looked for the sources of Genera?
> 
> Yes, I do that a lot.  I've even fixed bugs in it (not the file
> system, the networking stuff).  I'm very glad they gave source out
> (albeit it wasn't open source),

Some of this stuff was by accident, AFAIK. ;-) Like the CLIM source. ;-)

> I wish more people did, as it can
> really save a bunch of time.

Open Genera should be shipped with even more source code, as I heard.

-- 
http://www.lavielle.com/~joswig
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122040629677186@naggum.no>
* ··@hugo.westfalen.de (Georg Bauer)
| Exactly. With commercial applications, this greatly differs:
| 
| >- you are free to fix the bugs you never wanted to fix.
| 
| You are not free to fix the bugs you never wanted to fix. In fact, often
| you don't have the sources to fix it.

  I have fixed some thirty-odd problems in Allegro CL with advise before I
  got source code, and once I got a load of source with the commercial
  support agreement, I fixed yet more problems.  it appears that my fixes
  are migrating into Franz Inc's own code base.  however, it often takes a
  whole day to write a good bug report, and only an hour or so to tweak
  some behavior in the source.  for some reason, navigating in the source
  code that came with Allegro CL had only one obstacle: the IF* macro.  the
  rest was a breeze.  :)

| >- you are free to implement the functionality you
| >  never wanted to program yourself
| 
| You might be possible to add the functionality you need but the company
| never want's to implement.  In fact, as soon as you want to fix things in
| the lowest levels of the product, you are most often lost.  It is better
| with Lisp, since Lisp is much better customizeable than C++-Compilers,
| but the problem still exists.

  well, _my_ Allegro CL 5.0 now does

(format nil "~,,' ,4:B" (get-universal-time))
=> "1011 1010 0001 0110 1000 0111 1110 1001"

  because I needed this behavior and didn't feel like writing my own FORMAT
  substitute or function to hack numbers and suck.  the changes were
  actually fairly simple.

| >- you are free to wade through large amounts of complicated
| >  source code you never wanted to look at
| 
| Yes, that is definitely much better with commercial products where you
| don't get source at all.  You might get _some_ source, as with most Lisps,
| but that is not a quite common case for commercial software.

  all Allegro CL licensees receive a lot of source code with their (signed)
  support agreement.  not the internals, of course, but the functions you
  would normally call or need to understand.

| No, I definitely prefer high-quality Open Source [TM] software.  Makes
| living easier, especial in my field of work.  Of course, there is much
| crap out there.  But actually with free software, you don't have to pay
| for crap, like is the case with commercial ones.

  I don't pay for crappy software to begin with.  why do other people?  if
  you are implying that commercial Common Lisp implementations are crap, I
  think you have an attitude problem so big you should be dismissed as a
  lunatic, like the other "Free Software" fanatic who posts occasionally to
  this newsgroup.

#:Erik
-- 
  The Microsoft Dating Program -- where do you want to crash tonight?
From: Vassili Bykov
Subject: Re: help! absolute beginner
Date: 
Message-ID: <FCVa2.2742$5n1.21717330@news.magma.ca>
Erik Naggum wrote in message <················@naggum.no>...

>  I don't pay for crappy software to begin with.  why do other people?  if
>  you are implying that commercial Common Lisp implementations are crap, I
>  think you have an attitude problem so big you should be dismissed as a
>  lunatic, like the other "Free Software" fanatic who posts occasionally to
>  this newsgroup.


Hmm, who would that be?
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <slrn76r151.a3.gb@jill.westfalen.de>
In article <·······················@194.163.195.67>, Rainer Joswig wrote:

Be warned, this get's long. Hey, _you_ started me on this, so you
deserve what you get ;-)

>Will access to source magically give a Lisp system
>a usable GC? Even getting a decent
>UI seems not to be possible.

Does being commercial help there? I say no. Just because there are
_loads_ of good Open Source software out there. Sure, there are parts
missing - everybody can feel free and add what's missing. The problem I
have with commercial products is that you are tied to what the company
delivers. And there the money plays and not the end user (except where
the end user is where the money is - often enough that is not the case).
If the user would be so important, there would be a CLIM implementation
on every commercial common lisp by now. And there won't be a fight
between SUN and M$ about what Java should be. And there wouldn't be
something like the Halloween Documents.

Ok, granted, the few Lisp companies left are doing quite good. But that
isn't symptomatic for commercial products. For example I would never try
to build an ISP on commercial systems and commercial software. Too much
problems - just look at parts like encryption software, interoperability
etc. No, M$ might try to _sell_ you that there software is
interoperable, but in reality it isn't. They change things like it is
fitting them, even if they break available standards with that.

Sure, sometimes Open Source Software breaks standards, too. But then I
am able to fix that. Not always myself, but it is easier to hire someone
with the right knowledge when the source is available and not a closed
secret. For example I would really like to have sources at least for
parts of Exchange Server, just to get this FPOS to run correctly.

From my point of view of administering Novell and NT servers now for
some years and using in parallel Linux based systems for as many
situations as possible (of course, the important part here is "possible"
- there are lot's of situations where I won't use Linux, just because
there still isn't software for that kind of situation), I would
definitily give Linux the thumbs-up and commercial support the
thumbs-down.

The main problem is, that commercial support is often not up to what
they try to sell. That's the case with every _big_ company (big in the
really big sense) I met. The smaller companies do much better, but it
still is a problem when you want to change things the way _you_ want it.
As soon as you hit the nerve of a programmer when you want to do things
that he doen not like, you _wish_ you had the source.

>So you lately have looked into your Genera file system, right?

You happened to notice that Symbolics is a bit of dead at the moment?
Just asking. No worries :-)

>Maybe you even have installed the additional source folder
>of MCL on your Mac. Did you need it? Much? Would you
>have needed it?

Sure. If not for patching, but for documentation. Because the available
documentation for this commercial product does much leave to be desired.
So I look where I know where the documentation is: in the source. As I
said, Lisp companies doing much better in this problem than others.
Where did you find the sources to the runtime library of Visual Basic?
And how about applying security patches to IIS or Netscape FastTrack
(the latter one might change some time, since Netscape itself joined the
Open Source idea).

>I better ask the experts.

Yes, you definitely should ask them. Hey, you practically _begged_ for 
this comment :-)

>Now it is a trademark? Religious marketing bullshit.

Nope. It is just a way to prohibit others to jump a bandwagon without
giving back to the community. There are several ways to solve that
problem. One is the GPL that has many problems (although most of them
are only psychologic problems and not real ones) in the commercial
world. Another one is the Open Source trademark: just a very simple and
easy to follow statement about what software is Open Source compliant
without the "idealistic" hassles of the GPL. So it is much easier to
talk about Open Source in a commercial surrounding than it is about GPL.

You _did_ notice that commercial companies begin to join the Open
Source idea? And some of them really without any force? Even for big
companies there are areas where Open Source pays the rent (like IBM and
Apache to give one example).

Ideologic blindfolding is happening on both sides - those supporting
Open Source and those refusing it. And it is bad on both sides. Open
Source is not the enemy of commercial products. It is just another way
of doing business with different goals.

BTW: Open Source is named in this way to avoid the confusion the "Free
Software" idea had. It is not important that the software is available
for free. It is important what level of freedom in using the available
source you have, to distinguish Open Source from simple giveaways.

To sum it up for me: for servers I prefer Linux. That's why I am still
interested in free Common Lisps, as they would make a perfect base for
CLHTTPD on Linux. On workstations, I still prefer my Mac running MacOS
just because of the IMO superior GUI. There are some promising new
developments for X, but up to now they are far behind what the Mac gives
me.

Sure, this is my personal resume, others might look different at the
situation. But your original posting that I commented in a somewhat
sloppy fashion just cried for some form of correction. Sure, usually I
am a Open Source advocate, but since I have something like a real life
in a real job, I have to look at things differently than RMS for
example. And funny enough, after now over 12 years of work, I think I
like the Open Source idea even more than before. When it wasn't for the
job, Open Source and GPL only gave me free access to good software. Now
I get high-quality tools to get my job done. The "for free" is not
important any more - but the "source included" is.

There are only few standing axioms in the computing business, and those
are not silly things like Moore's law but more things like:

1. All software sucks.
2. All hardware sucks.
3. Sales-Droids are your enemy.
4. There is only one way to get it done: do it yourself.
5. Management will choose the technologically inferior product.
6. Software is only as good as the next guy. And that is a moron.
7. If it is a stupid idea to ride a dead horse, management will do it.

Open Source doesn't solve those, but it helps to solve the problems
arising out of those :-)

And yes, there is a Open Hardware movement, too. And yes, there are
companies that join in. Not many, but it is a start. Make an end to
closed hardware where you don't get a driver for _your_ OS, just because
no one get's a description of the interfaces. And I really believe
that that has nothing to do with religion but everything with good common
sense.

bye, Georg
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122171306487142@naggum.no>
* ··@jill.westfalen.de (Georg Bauer)
| The problem I have with commercial products is that you are tied to what
| the company delivers.  And there the money plays and not the end user
| (except where the end user is where the money is - often enough that is
| not the case).

  I think this is your crucial mistake.

| If the user would be so important, there would be a CLIM implementation
| on every commercial common lisp by now.

  and this is an example of where the flaw in your logic becomes visible.
  it _is_ because users are important that there is a CLIM implementation
  on every commercial Common Lisp system by now, yet it only barely pays
  for itself.

| And there won't be a fight between SUN and M$ about what Java should be.
| And there wouldn't be something like the Halloween Documents.

  ahem, it's because they fight over who should get more _users_ that this
  happens.  all this is about market share and mind share.  it doesn't
  matter whether Bill Gates or James Gosling sit in their offices and think
  up evil or good things (respectively) unless they go out there and get
  hoardes of _users_ to agree with them with their checkbooks open.

  since you're a fellow Germanic European, I can perhaps say that our
  cultures are _very_ far from appreciating the customer-orientedness of
  American capitalism.  the typical "Staatskapitalismus" we find in Europe
  is evil because it actively limits the choices of the customers.  (and
  why did I use the German word for this phenomenon?  there is nothing like
  in American English.  that alone should be worth thinking about.)

| Ok, granted, the few Lisp companies left are doing quite good.  But that
| isn't symptomatic for commercial products.

  really?  perhaps I can compare this to the music industry.  most of the
  bands and musicians I like have a pretty limited market.  (many of them
  are German, by the way.)  I buy their CD's because I like their music and
  know that my music pusher is a good guy who doesn't cheat them or their
  labels out of money.  however, there's been a huge increase in music in
  MP3, and the people who don't want to pay the _musicians_ for their work
  are now targeting the increased sensitivity and hostility of the _large_
  music distributors and labels towards MP3 distribution of music.  not
  only do they stand to lose a significant amount of money, they can afford
  to fight, so I'm not worried about the large labels and distributors.
  I'm worried that if MP3 becomes the preferred medium of distribution, it
  will be very much harder for my favorite bands and musicians to buy the
  _fantastically_ expensive equipment they use to create their (electronic)
  music.  it is reasonable to argue that a transition to distribute music
  as MP3 will cause a bad case of _commercialism_ in what is available to
  the listening audience, which means: I don't get to listen to the music I
  like.  I might add that spent twenty years thinking I lacked appreciation
  for music until I found some labels that sell a couple thousand copies of
  each CD _at best_.  the CD medium made the music I like possible.  MP3
  might make it as unavailable as it once was.

  the commercial products you whine about are bad because the _customers_
  don't care.  it's because the f*cking lusers continue to buy shitware
  from Microsoft and put up with their fantastic policy of charging the
  customers for fixing their own mistakes, whereas the users of free
  software differ not in how much they want to pay or what they are
  actually buying, but because free software users _care_ more.

| For example I would never try to build an ISP on commercial systems and
| commercial software.

  well, I was an ISP back in 1987.  (that was when my Y2K drive started and
  I argued to the IETF WG on Host Requirements that RFC 822 be updated to
  use four-digit years -- RFC 1123 is the result.)  back then, if you
  wanted something that worked, you got it from companies that actually
  cared, and it cost a _lot_ of money.  (except the phone company -- the
  phone company cares about as much what happens to you as the IRS does.)
  getting free software is a _fantastic_ luxury.  I _love_ luxuries, of
  course, but I don't pretend they aren't luxuries, and I certainly don't
  take luxuries for granted.  (I do follow the advice of Lazarus Long and
  budget luxuries first, though.)

| No, M$ might try to _sell_ you that there software is interoperable, but
| in reality it isn't.

  I have a new .signature today.  paraphrased, Bill Gates understood early
  on that he should not call people who didn't want to understand computers
  idiots.  he decided to take their money.  so of _course_ you've been had
  if you buy any of the useless piece of shit he peddles to the ignorant
  masses.  likewise, do _you_ participate in the Staatslotterie?  or do
  you, like me, consider lotteries to be an extra tax on stupid people?

  if you believe Microsoft's propaganda and you wind up ripped off and
  naked, do you become an emperor?

| The main problem is, that commercial support is often not up to what
| they try to sell.  That's the case with every _big_ company (big in the
| really big sense) I met.  The smaller companies do much better, but it
| still is a problem when you want to change things the way _you_ want it.

  well, I'm like this one-person software company who does it because it's
  a nice way to avoid real work and still pay all the bills, and my clients
  don't get what they say they want, either.  if I don't agree with them,
  they get to pay somebody else to do it (and come back to me afterwards).
  most of the time, we come to terms on what they really want, and that's
  what they get.  if the client knows much better than me what he wants, he
  doesn't need my services.  I think this attitude scales very well.  as
  long as you're free to go elsewhere, I don't see the problem.

  now, you'll argue that you _aren't_ free to buy software from whichever
  software company you want to, but that's _also_ wrong.  it may cost more,
  but that means you either operate with a smaller margin, charge more, or
  leverage the costs across more volume -- or decide against doing it,
  which is no crime.  using Microsoft products is also much more expensive
  in real terms than using quality software, but it'll take a while before
  the suits who don't want to understand computers get sick and tired of
  having their money taken away.
 
| Ideologic blindfolding is happening on both sides - those supporting Open
| Source and those refusing it.  And it is bad on both sides.  Open Source
| is not the enemy of commercial products.  It is just another way of doing
| business with different goals.

  I think free software can be the enemy of commercial software the same
  way MP3 can be the enemy of commercial music.  we don't have to worry
  about the high-volume end of the market.  it's the nigh invisible, small
  players that need to be worried about.

| BTW: Open Source is named in this way to avoid the confusion the "Free
| Software" idea had.  It is not important that the software is available
| for free.  It is important what level of freedom in using the available
| source you have, to distinguish Open Source from simple giveaways.

  amusing.  "Free" in Free Software doesn't mean "available for free" the
  same way "Free Nelson Mandela" didn't mean you got a piece of him and a
  balloon at a parade.

#:Erik
-- 
  don't call people who don't understand statistics idiots.  take their money.
From: rusty craine
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74m1l5$1uu$1@excalibur.flash.net>
Erik Naggum wrote in message <················@naggum.no>...


>  don't call people who don't understand statistics idiots.  take their
money.

Our little community had a party to celebrate the season last week.  There
were 50 or so people there.  Computers were the topic of several
conversations during the evening.  As I listened to the topics, it became
clear to me that the "ignorant" masses didn't really care that they were
running software per say or bill gates by default.  Their collective concern
was only functionality of the total package.  ---"I can watch the teley,
play my DVD [sounds like a new brand of underware], play "space blaster",
write my term paper (or download a 'cheater' from the _NET_)".  Seemed the
quote of the night.

Movies, games, the internet, a spread sheet, and a little word processing
must be what sells comptures to the masses.  This must be the market that
Gates is after.  I would say he has done a remarkable job in that market,
Ayn Rand must've smiled.

At their worst Gates and Micorsoft  can't be all bad, the masses have voted
with the $$$.  To a family at the party, they all had at least one computer.
Many had several, upgarding and giving the old one to "Suzie" for her room.

Alas lisp was not on a single drive save mime,  VB was on several ["but i've
not learned to program yet only cost 90 bucks"],  java on none, IE on most,
and a few Netscape.

I think Bill Gates is here to stay, guess we should try to adjust. If you
think ya got it bad come over to the 390 for a while.  You will use CICS and
that is no shit.  Put out by one IBM, as they say, as they do....you and god
will comply.

me and god waiting for IBM to tell us what to do
rusty
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122211354918302@naggum.no>
* "rusty craine" <········@flash.net>
| At their worst Gates and Micorsoft can't be all bad, the masses have
| voted with the $$$.

  the _masses_ voted for some other things that turned out to remarkably
  evil, as well.  you can basically count on the masses to do whatever the
  best demagogue tells them to do.  Microsoft is about marketing and fraud,
  and they're very good at it.  it has never been about software.  had they
  spent their energy on the software, it would probably have been great,
  but they spend as little as humanly possible on the software to make as
  much money as inhumanly possible.

| I think Bill Gates is here to stay, guess we should try to adjust.

  lots of things are here to stay.  Hinduism, for instance.  biggest
  religion there is.  have you converted, yet?  if not, why not?

#:Erik
-- 
  don't call people who don't understand statistics idiots.  take their money.
From: rusty craine
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74mvmq$6sr$1@excalibur.flash.net>
Erik Naggum wrote in message <················@naggum.no>...
>  lots of things are here to stay.  Hinduism, for instance.  biggest
>  religion there is.  have you converted, yet?  if not, why not?

I hope someday we have an opportunity to discuss such topics.  It would make
for an interesting evening.  Once upon a time I think I found the _answers_
in such a discussion, alas the alcohol blood level was not conducive to
clear recollection.

Malt does more than Milton can to explain God's way to man.  (modus tollendo
pones)
Rusty
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122246776861675@naggum.no>
* "rusty craine" <········@flash.net>
| I hope someday we have an opportunity to discuss such topics.  It would
| make for an interesting evening.  Once upon a time I think I found the
| _answers_ in such a discussion, alas the alcohol blood level was not
| conducive to clear recollection.

  personally, I think most religions look a lot like such recollections.

#:Erik, who is just _begging_ for another "be nice to the humor-impaired"
	thread after the Stalin thread was buried.
-- 
  don't call people who don't understand statistics idiots.  take their money.
From: Gareth McCaughan
Subject: Re: help! absolute beginner
Date: 
Message-ID: <86pv9qn8rm.fsf@g.pet.cam.ac.uk>
Erik wrote:

[I mentioned that Hinduism is smaller than Christianity and Islam]
>   I'm rather unwilling to call people believers in the same religion when
>   they murder each other for their differences in belief -- it's a good way
>   to become a target they can agree on.  this reduces Christianity to a
>   huge number of different religions.

Well, maybe. I'm not sure there have been many Christians
killing each other for theological reasons for a while (the
obvious counterexample is Northern Ireland, but that's
really mostly a territorial thing; if the religious differences
weren't there they'd find something else to fight over).
I'm not sure what the situation is with Islam.

I conjecture (with no shred of evidence) that Roman Catholicism
is similar in size to Hinduism.

-- 
Gareth McCaughan       Dept. of Pure Mathematics & Mathematical Statistics,
·····@dpmms.cam.ac.uk  Cambridge University, England.
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3673bd17.950686995@news.newsguy.com>
On 11 Dec 1998 12:49:33 +0000, Gareth McCaughan
<·····@dpmms.cam.ac.uk> claimed or asked:
 
% I conjecture (with no shred of evidence) that Roman Catholicism
% is similar in size to Hinduism.

"Every sperm is sacred..." --- Monty Python's Meaning of Life

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Sashank Varma
Subject: Re: help! absolute beginner
Date: 
Message-ID: <varmas-1112981135580001@129.59.192.40>
In article <························@qtns00816.singnet.com.sg>,
·······@gol.com (Sudhir Shenoy) wrote:

[snip]
> 
> Sorry for the off-topic post but I couldn't resist (since I seem to be one of
> the only Hindus reading this ng).
> 
[snip]

you're not the only one, my brother!

(although i'm married to a protestant and attend church far more often
than i visit the local temple.)

sashank
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3672bc77.950527456@news.newsguy.com>
On 11 Dec 1998 01:47:24 +0000, Erik Naggum <····@naggum.no> claimed or
asked:


%   I'm rather unwilling to call people believers in the same religion when
%   they murder each other for their differences in belief -- it's a good way
%   to become a target they can agree on.  this reduces Christianity to a
%   huge number of different religions.  I think they want it that way, and I
%   am in no position to argue.  ditto for Islam.  now, Hinduism is the name
%   of a huge collective of religions, but at least they aren't at war over
%   it.  that's gotta count for something.

You know the joke about Hell being endothermic or exothermic?

If you don't, I can post it here even though it is way the heck off
topic.  Funny, but off topic.  I don't have it on hand right now, or I
would just go ahead and post it.  Actually, a DejaNews search should
find it (or multiple versions) on the humor groups.

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Espen Vestre
Subject: Re: help! absolute beginner
Date: 
Message-ID: <w6ogpdmkvb.fsf@gromit.nextel.no>
"rusty craine" <········@flash.net> writes:

> I think Bill Gates is here to stay, guess we should try to adjust. 

I don't know, Win2000 seems to be such a monster that they *must* be
programming themselves into a corner of an enormous room in a house
they don't longer know how they even got into...  MS has made remarkable
turns before (e.g. MSN -> Internet), but will they be able to get
out of the bloatware spiral without damage?

Speaking of your friends, I think they'll get tired of bugs at _some_
point.  People are remarkably patient with their PCs, but I think
that has to do with the facts that they don't know better and, most
important, that they're having such fun that they don't mind a few
problems.  But at some point, this will change, and they will demand
that the machines do real work for them.  Don't forget what happened 
to the american car industry, when consumers _understand_ that they
don't get the quality the deserve, they'll do something about it.
And the computer industry will learn that the hard way, just like
the american car industry!

Or so I hope...

-- 

  espen
From: Rob Warnock
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74th7o$2jplp@fido.engr.sgi.com>
Erik Naggum  <····@naggum.no> wrote:
+---------------
| * ····@rigden.engr.sgi.com (Rob Warnock)
| | Making the rounds lately:
| | 	Did you hear that Windows 2000 slipped again?  Yeah,
| | 	it's not gonna be out until the *second* half of 1901...
| 
|   I heard it was just going to be called Windows 1900.
+---------------

Yeah, but that was before it slipped another 18 months.


-Rob

-----
Rob Warnock, 8L-855		····@sgi.com
Applied Networking		http://reality.sgi.com/rpw3/
Silicon Graphics, Inc.		Phone: 650-933-1673
2011 N. Shoreline Blvd.		FAX: 650-964-0811
Mountain View, CA  94043	PP-ASEL-IA
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3671bb36.950206384@news.newsguy.com>
On 11 Dec 1998 04:10:43 GMT, ····@rigden.engr.sgi.com (Rob Warnock)
claimed or asked:

% [Yes, I know it's continuing an off-topic thread, but...]
% 
% Making the rounds lately:
% 
% 	Did you hear that Windows 2000 slipped again?  Yeah,
% 	it's not gonna be out until the *second* half of 1901...

ROFL!

Wouldn't it be funny if Bill just had his (what do you call cult
members?  Oh well.) programmers steal all the code to Linux and make
Windows 2000 simply be an X server with the Windows 98 L&F.  DCOM
would be a daemon process.  Hmmm.  That makes you think, eh?

With all the programmers MSFT has, that would be, what, a three week
job?

Actually, he should steal FreeBSD because of the license.

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Matthias Buelow
Subject: Re: help! absolute beginner
Date: 
Message-ID: <871zm3iq75.fsf@altair.mayn.de>
········@david-steuber.com (David Steuber "The Interloper") writes:

> Actually, he should steal FreeBSD because of the license.

Unless he sneaked up to cdrom.com with a truck at night and loaded in
some hundred boxes with cdroms without paying for them I can't see much
stealing involved.

-- 
  ``Round about the terminal go; / In the poisoned upgrade throw.
  Code, which by a student done / In minutes numbering sixty-one.
  Run-time error, protection fault / Crash ye first, crash ye shalt.''
	(The Oracle, MacBeth95)
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3lnkhpo55.fsf@haystack.aiai.ed.ac.uk>
* Georg Bauer wrote:
> Does being commercial help there? I say no. Just because there are
> _loads_ of good Open Source software out there. Sure, there are parts
> missing - everybody can feel free and add what's missing. The problem I
> have with commercial products is that you are tied to what the company
> delivers. And there the money plays and not the end user (except where
> the end user is where the money is - often enough that is not the case).
> If the user would be so important, there would be a CLIM implementation
> on every commercial common lisp by now. And there won't be a fight
> between SUN and M$ about what Java should be. And there wouldn't be
> something like the Halloween Documents.

Well, the flip side of this is: has there ever been an OSS Lisp which
has a seriously good GC (comparable with most recent commercial
Lisps)?  I'm not aware of one, although I don't know what CMUCL does
on X86 these days.

And, actually, which commercial Lisps *don't* have CLIM?

--tim
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0912981407490001@pbg3.lavielle.com>
In article <················@jill.westfalen.de>, ··@jill.westfalen.de
(Georg Bauer) wrote:

> >Will access to source magically give a Lisp system
> >a usable GC? Even getting a decent
> >UI seems not to be possible.
> 
> Does being commercial help there? I say no.

Isn't it funny that GCs of commercial systems are lightyears
ahead? That I have yet to see more than a handful
"free" software products with a *usable* user interface?
To create a great product it takes more than
just naming it "open", "free", "open source", ...
What we end up with are geek products. Ugly,
idiosyncratic, complicated, fragile, maintenance
intensive, unsupported, ...

> Just because there are
> _loads_ of good Open Source software out there.

Where? Most of the Open Source stuff is written by geeks
for geeks. Most of the stuff is ***extremely*** user hostile.
A lot of the stuff is of extraordinary poor design.

I'm not saying that there are no good "Open Source"
tools - but saying that this is the path
to the software nirvana is ridiculous.

> Sure, there are parts
> missing - everybody can feel free and add what's missing.

I don't have neither the time, the knowledge nor the will
to need to deal with this stuff.

> The problem I
> have with commercial products is that you are tied to what the company
> delivers.

Come on.

> And there the money plays and not the end user (except where
> the end user is where the money is - often enough that is not the case).
> If the user would be so important, there would be a CLIM implementation
> on every commercial common lisp by now.

We have a CLIM implementation on every commercial Lisp by now.
I can use CLIM on Mac/Unix/Lispm/Windows.

But not that there is some possible user somewhere makes things
magically happen - the demand will drive it. Then
you need to be able to deliver.

> Ok, granted, the few Lisp companies left are doing quite good.

Where are you living? I have the complete opposite view.
How can it be that we look at the same thing - yet I get
a radically different impression. The Lisp market
is silent. Example: How many Lisp projects have you
heard/seen/... about that
were started in Germany in the last two years?
One? Two? How big is Germany?

> For example I would never try
> to build an ISP on commercial systems and commercial software.

We do. We are using SUNs and Solaris, Ascends, Ciscos, ...
Most commercial ISPs I know are built around this with
a lot of commercial and a lot of "free" software.

>Too much problems

With crap like SENDMAIL, etc.? Yes. Thanks for the source.
I would have preferred not to look at the source.
Mayordomo?

> >So you lately have looked into your Genera file system, right?
> 
> You happened to notice that Symbolics is a bit of dead at the moment?
> Just asking. No worries :-)

No, they are still alive. 
I'm still using my Lisp machine.

> To sum it up for me: for servers I prefer Linux. That's why I am still
> interested in free Common Lisps, as they would make a perfect base for
> CLHTTPD on Linux.

But there are not and they are years away from that.

> Sure, this is my personal resume, others might look different at the
> situation. But your original posting that I commented in a somewhat
> sloppy fashion just cried for some form of correction. Sure, usually I
> am a Open Source advocate, but since I have something like a real life

So your COBOL programs are Open Source? The company
who pays you will release them as Open Source?

> 1. All software sucks.
> 2. All hardware sucks.
> 3. Sales-Droids are your enemy.
> 4. There is only one way to get it done: do it yourself.
> 5. Management will choose the technologically inferior product.
> 6. Software is only as good as the next guy. And that is a moron.
> 7. If it is a stupid idea to ride a dead horse, management will do it.

Come on, this list is plain stupid.

What I'd like to see are more people with a vision, who really
want to move things forward, making using computers more
fun and more productive. I'm also pretty sure that
the Lisp community can provide a lot of ideas to make
that happen.

> Open Source doesn't solve those, but it helps to solve the problems
> arising out of those :-)

Actually the current stuff that is being distributed
as Open Source or whatever are not really part
of the solution. Often they are part of the problem.

Example: Linux sucks. Big time. Stuff like that frustrates me.

> And yes, there is a Open Hardware movement, too.

Another religious marketing bullshit.

-- 
http://www.lavielle.com/~joswig
From: Larry Hunter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <rbr9u92wfb.fsf@work.nlm.nih.gov>
I don't really want to get involved in the war over commercial vs. "open
source" (like most of these X vs. Y battles, my opinion is "both, please")
but I couldn't let this response by Rainer Joswig to Georg Bauer's comment
go by:


  Bauer: "To sum it up for me: for servers I prefer Linux. That's why I am
  still interested in free Common Lisps, as they would make a perfect base
  for CLHTTPD on Linux."

  Joswig: "But there are not and they are years away from that."

Rainer, you of all people should know better.  The CL-HTTP web page
(http://www.ai.mit.edu/projects/iiip/doc/cl-http/home-page.html) says:

  Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
  the 18b release of CMUCL, and is considered beta-test software when
  running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

As with many open source projects, it may take some technical effort to get
the desired level of performance, but it is clearly possible to make it
happen today.

Larry

-- 
Lawrence Hunter, PhD.
National Library of Medicine               phone: +1 (301) 496-9303
Bldg. 38A, 9th fl, MS-54                   fax:   +1 (301) 496-0673
Bethesda. MD 20894 USA                     email: ······@nlm.nih.gov
From: Mike McDonald
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74mhuu$9dg$1@spitting-spider.aracnet.com>
In article <··············@work.nlm.nih.gov>,
	Larry Hunter <······@nlm.nih.gov> writes:

>   Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
>   the 18b release of CMUCL, and is considered beta-test software when
>   running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

  But for all practical purposes, there is NO 18B release of CMUCL.

  Mike McDonald
  ·······@mikemac.com
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3emq8rz8v.fsf@todday.aiai.ed.ac.uk>
* Mike McDonald wrote:

>   But for all practical purposes, there is NO 18B release of CMUCL.

?.  My CMUCL is 18b, and I'm on a sparc/Solaris not intel/Linux box,
and I didn't have to work that hard to find it (I think I just had a
look on cons.org, noticed a new one and installed it). It works.

--tim
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0912981650170001@pbg3.lavielle.com>
In article <··············@work.nlm.nih.gov>, Larry Hunter
<······@nlm.nih.gov> wrote:

>   Bauer: "To sum it up for me: for servers I prefer Linux. That's why I am
>   still interested in free Common Lisps, as they would make a perfect base
>   for CLHTTPD on Linux."

Read: "perfect base"?

>   Joswig: "But there are not and they are years away from that."
> 
> Rainer, you of all people should know better.  The CL-HTTP web page
> (http://www.ai.mit.edu/projects/iiip/doc/cl-http/home-page.html) says:
> 
>   Douglas Thomas Crosher has ported CL-HTTP to CMU Common Lisp which runs on
>   the 18b release of CMUCL, and is considered beta-test software when
>   running multi-threaded on Intel X86 hardware under FreeBSD and Linux"

Read: "beta-test software".

> As with many open source projects, it may take some technical effort to get
> the desired level of performance, but it is clearly possible to make it
> happen today.

Linux and CMUCL are not a perfect base for CL-HTTP right now.

More work is needed. It is nice that it is kind of running at all.

-- 
http://www.lavielle.com/~joswig
From: Paul Rudin
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m31zm9l4qf.fsf@shodan.demon.co.uk>
······@lavielle.com (Rainer Joswig) writes:

> > Does being commercial help there? I say no.
> 
> Isn't it funny that GCs of commercial systems are lightyears
> ahead? That I have yet to see more than a handful
> "free" software products with a *usable* user interface?
> To create a great product it takes more than
> just naming it "open", "free", "open source", ...
> What we end up with are geek products. Ugly,
> idiosyncratic, complicated, fragile, maintenance
> intensive, unsupported, ...
> 


Wandering a little off topic here, but a few couterexamples that
spring to mind:

A goodly percentage of software infrastructre of the internet is free,
open source software.

Linux is a more realiable OS than Windows.

Mozzilla is at least as good as Internet Explorer

The best document preperation system for scientific/mathematical
material is (La)TeX by a long way..

A little more on topic, (X)emacs/ilisp is the best lisp editor around
IMO.
From: Steve Gonedes
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m267blgkuq.fsf@KludgeUnix.com>
Paul Rudin <·····@shodan.demon.co.uk> writes:
 
< Wandering a little off topic here, but a few couterexamples that
< spring to mind:
< 
< A goodly percentage of software infrastructre of the internet is free,
< open source software.
<
< Linux is a more realiable OS than Windows.

Please, don't misinterpret what I'm saying as I like linux very much -
it works great for me but...

What is reliability? Will linux support/do x, y, and z tomorrow?
If you're answer is yes then _that is my point_! 

There is no security offered by linux in this and other areas because
_it is a response_ to a `lack of' and was not, and simply cannot
(currently?) be a method to drive and create (partially due to the
attitude which linux thrives on - that is the meeting of a challenge).

Not to say that linux is not doing new things etc. I like it very much
and it works great for me. For example no hardware vendor will create
new hardware due to the perceived flexability that linux offers them
(this is much the same for most every `potential solution' in the
`computer' industry right now, no reason to stop trying or anything
though, IMHO). Even better - no new internet protocols are written
because of the benefits that linux offers - if anything new protocols
are old ideas just clarified to better understand why people keep
failing to solve the same problem (again and again and ...).

Rest assured, Microsoft will not be creating new internet protocols
due to linux. The halloween documents were a joke to demonstrate just
how much control MS has over the media. (Remeber also, microsoft
cannot create software - they never have and never will. They'll soon
have no new software to steal and will being selling people _access_
to their own information, just watch! It's already happening.)

< Mozzilla is at least as good as Internet Explorer

They are both pieces of shit in that neither of them _SOLVE ANY
PROBLEMS_; their total benfit is being derived from their ability to
_create new problems_! For instance, the newer versions of internet
exporer keep getting slower - which demonstrates that it _is not
addressing any problems!_ The right thing to do is _create_ a method
to represent information is a new, efficient (in terms of clarity and
physical space) mannor which will eventually allow for new ways to
express ideas on the internet. Instead we have reoccuring `religious
experiences' guided by reoccuring mistakes... 

< The best document preperation system for scientific/mathematical
< material is (La)TeX by a long way..

The worse thing about TeX is how it promotes stagnation (read:
regression) and treats any idea of progression as undesirable and
therefor non-existant. TeX is such a horrendous piece of shit that
people _are scared_ of changing it! There have been no, and there
_NEVER WILL BE_, new ideas or developments in the TeX area of
typesetting. The simple fact that there are numerous versions of latex
proves that the only solution is to dump the cruft and move on
- same goes with X-windows, extract the xfree drivers and dump the
cruft! You see, latex failed to address the problem. The problem, TeX,
was readily recognized and admitted at one point, but people are
beginning to forget this.

< A little more on topic, (X)emacs/ilisp is the best lisp editor
< around IMO.

This is not because it is free software - it is because it solves some
of your problems (whatever they may be)...

[I'm not against free software - in fact I like it and respect it
_very much_, I just get very annoyed when people forget what the
benefits really are and when they criticize other people's beliefs
without valid justification. (and here I am, is this ironic?) It makes
them sound like baptists screaming how evil the moromons are (if you
follow religion at all) :) ]

Also, do you notice how your acceptance of `best' may prevent you from
accepting the new, hence potentially better? You will probably look at
every new editor in terms of features it lacks in terms of xemacs. All
new features that a new editor would introduce to you will most likely
result in the assimilation of new _into the old_.

The danger, you see, is that you may have come to accept some of the
original limitations as features/desirable functionality, and any
attempt to free you from them will be percieved as an attack on your
self due to your familiarity => stability, `_we_ like it', etc etc.

We're all going to die, it's sucks - I know.

[Not trying to sound nasty - just raising some potentially valid points?]
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-0912981700110001@pbg3.lavielle.com>
In article <··············@shodan.demon.co.uk>, Paul Rudin
<·····@shodan.demon.co.uk> wrote:

> A goodly percentage of software infrastructre of the internet is free,
> open source software.

Depending on the definition of "free" the percentage varies.

> Linux is a more realiable OS than Windows.

In general or the individual C program? Does BIND crash
less often or more often than a Windows NT DNS?

> Mozzilla is at least as good as Internet Explorer

I haven't seen a released open source Mozilla yet. Where is it?
The thing in development was highly buggy last time I checked.

> The best document preperation system for scientific/mathematical
> material is (La)TeX by a long way..

Yes.

> A little more on topic, (X)emacs/ilisp is the best lisp editor around
> IMO.

I'm not using it as a Lisp editor. Well, the X-Windows UI
is like a stone tied to the foot. But there are easier to use
alternatives.

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <871zm9i623.fsf@piracy.red-bean.com>
······@lavielle.com (Rainer Joswig) writes:

> > Linux is a more realiable OS than Windows.
> 
> In general or the individual C program? Does BIND crash
> less often or more often than a Windows NT DNS?

Uhm, both the kernel is more reliable, and for the large part, most of
the programs in the various distributions are more reliable.  BIND is
definetly better than MS DNS, I have horror stories about it if you
want.

> > Mozzilla is at least as good as Internet Explorer
> 
> I haven't seen a released open source Mozilla yet. Where is it?
> The thing in development was highly buggy last time I checked.

That's because it's in "development".  Enter, "Gecko", the result of
the open source effort which is being hailed as an extremely fast and
small layout engine by most major media outlets (outlets, or
sphincters, I can't decide what they are).  Mind you that Netscape has
been paying alot of developers of their own to work on the open code
base, so I think this may be a bad example for the argument the poster
was intended to make.

Apache is the obvious example.  I've worked with web servers ranigng
from IIS, to Netscape (all generations) to IBM GoServer and all it's
weird Lotus-bound contraptions for torture.  Apache is basically
the best hands down, and if you want a particular "type" of server,
"single-process multi-threaded" or "multi-process" or whatever, there
are close to a dozen free servers which fit those models, and kick
much ass.

I guess I would argue that the management of the project, the talent
working on it, and it's ability to capture the interest of users and
developers is a primary determinant wether it is open-source or
commercial.  The Lisp vendors are extremely competent, but the same
cannot be said for commercial developers in a large number of other
markets.  Thankfully, the Lisp developers working with free software
are extremely competent as well.
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1012980134050001@194.163.195.67>
In article <··············@piracy.red-bean.com>, Craig Brozefsky
<·····@onshore.com> wrote:

> Uhm, both the kernel is more reliable, and for the large part, most of
> the programs in the various distributions are more reliable.

Maybe gradually - but not in principle. Is there
any revolutionary coding practice in effect that
suddenly makes Linux software more reliable?
I mean are the developers switching to Eiffel/Ada/SML and
are really taking reliability into account (runtime assertions,
test suites, checking adherance to specifications,
exception handling infrastructure, ...)?

Or are they just hacking as ever?

>  BIND is
> definetly better than MS DNS, I have horror stories about it if you
> want.

I guess there are enough horror stories about BIND, too.
Given the massive amount of work that has gone
into these things, they are still pathetic.
 
> commercial.  The Lisp vendors are extremely competent, but the same
> cannot be said for commercial developers in a large number of other
> markets.  Thankfully, the Lisp developers working with free software
> are extremely competent as well.

Yep.

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87vhjkhir5.fsf@piracy.red-bean.com>
······@lavielle.com (Rainer Joswig) writes:

> > Uhm, both the kernel is more reliable, and for the large part, most of
> > the programs in the various distributions are more reliable.
> 
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?

Yah, massive peer review, large-scale beta testing, and use of an
educated user base to produce high quality bug reports.  It's a
stretch to call these revolutionary tho, but that doesn't change the
fact that they have rather consistently resulted in high quality
software.  Does there NEED to be a revolutionary coding practice to
make the reliability worth noting?  I mean the issue here after all is
that there is some reliable free software out there, not that free
software is developed with revolutionary coding practices or in
whatever safe programming language is en vogue this year.

The poor sods might be C luzerz and all, but give them credit where
credit is due.  At least they're writing what they want rather than
whining about being marginalized and not recognized for their obvious
intellectual superiority, or blasting people for not supporting the
dwindling ranks of the commercial vendors for their favorite
languages.  Well fuck, they do that stuff too, but in addition to
writing realiable softwre 8^)

> I guess there are enough horror stories about BIND, too.
> Given the massive amount of work that has gone
> into these things, they are still pathetic.

Yah, I got those too, I got horror stories about any peice of software
I have ever used for more than 15 minutes, that doesn't make them all
pieces of shit.  I've seen it randomly die, I've seen exploits to root
it, I've seen it start throwing out random bits of garbage in
responses.  I've also see it run name service for universities and
large corporations nearly flawless for years.
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1012980411380001@194.163.195.67>
In article <··············@piracy.red-bean.com>, Craig Brozefsky
<·····@onshore.com> wrote:

> ······@lavielle.com (Rainer Joswig) writes:
> 
> > > Uhm, both the kernel is more reliable, and for the large part, most of
> > > the programs in the various distributions are more reliable.
> > 
> > Maybe gradually - but not in principle. Is there
> > any revolutionary coding practice in effect that
> > suddenly makes Linux software more reliable?
> 
> Yah, massive peer review, large-scale beta testing,

Sometimes I have the feeling I'm living in a beta world.
Everything now is beta. When do I get final stuff?
What once was pre-alpha now is beta.

> and use of an
> educated user base to produce high quality bug reports.  It's a
> stretch to call these revolutionary tho,

Yep, Microsoft, Apple, SUN, etc. are doing the same.

> The poor sods might be C luzerz and all, but give them credit where
> credit is due.  At least they're writing what they want rather than
> whining about being marginalized and not recognized for their obvious
> intellectual superiority, or blasting people for not supporting the
> dwindling ranks of the commercial vendors for their favorite
> languages.

Just look at CL-HTTP. Only one free Lisp is up to the task
running it. On one platform. With early support for
threads. But it has been used on five commercial
Lisp on different platforms. Hey, I'm not saying
that people should only use commercial software -
just that it seems simply not obvious to some people that things
are not black and white. "Open Source" is not equivalent
to "good" and "proprietary" is not equivalent to "bad".
"User" is not equivalent to "looser". It is simply
neither necessary nor sufficient for "good software"
to be "Open Source".

That people are still hacking with CLisp, CMU CL or now
with Corman Lisp is a good sign. Just don't be constrained
in your mind. There is enough room for improvement
and cool ideas. Look at Corman Lisp. Seems like
it has great potential to reach a lot of users. But this
is the result of the hard work of its developer
with a lot of experience.

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87r9u8hbn2.fsf@piracy.red-bean.com>
······@lavielle.com (Rainer Joswig) writes:

> > Yah, massive peer review, large-scale beta testing,
> 
> Sometimes I have the feeling I'm living in a beta world.
> Everything now is beta. When do I get final stuff?
> What once was pre-alpha now is beta.

Well, Debian hamm is not beta, but slink is just frozen and in beta
before release.  Linux 2.0.X is not beta, but 2.1.X is.  Apache 1.3.X
is no longer beta, but there were several beta versions of Apache.
They do early beta releases, and often have a very large number of
people running their beta software, but they do eventually (in most
cases at least) have full non-beta releases.  

> > and use of an
> > educated user base to produce high quality bug reports.  It's a
> > stretch to call these revolutionary tho,
> 
> Yep, Microsoft, Apple, SUN, etc. are doing the same.

Well, by some stretch of word, yes.  They just now are getting hang of
releasing betas to the general public for testing.  It used to be that
they charged for them.  But now we're just mincing words.

> Just look at CL-HTTP. Only one free Lisp is up to the task
> running it. On one platform. With early support for
> threads. 

And I've been running it, in CMUCL, on Linux, with multi-processing.
I've also run it in ACL on Linux.  For the most part it was
multithreading that was the major hurdle for free implementations to
get over.  That is not in the CL standard tho, so I think it's
reasonable, but definetly not desirable, that some implementations
don't support it.  There are alot of other CL packages that run in
free lisps, Loom, Ontolingua(I could have been fantasizing), and
other KR apps, the part of speech tagger from Parc, garnet, and
zebu...
From: Rob Warnock
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74torc$2g3li@fido.engr.sgi.com>
Rainer Joswig <······@lavielle.com> wrote:
+---------------
| Sometimes I have the feeling I'm living in a beta world.
| Everything now is beta. When do I get final stuff?
| What once was pre-alpha now is beta.
+---------------

Ah, how quickly they forget!  :-{

	"The Rise of Worse Is Better",
	Section 2.1 of "Lisp: Good News / Bad News / How to Win Big"
	by Richard P. Gabriel, 1991.
	http://www.naggum.no/worse-is-better.html

[Note: Despite its pathname, this URL contains all of "Good News/Bad News"...]


-Rob

-----
Rob Warnock, 8L-855		····@sgi.com
Applied Networking		http://reality.sgi.com/rpw3/
Silicon Graphics, Inc.		Phone: 650-933-1673
2011 N. Shoreline Blvd.		FAX: 650-964-0811
Mountain View, CA  94043	PP-ASEL-IA
From: Matthias Buelow
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ye0n24w74m6.fsf@atlantis.informatik.uni-wuerzburg.de>
In article <··············@piracy.red-bean.com>
Craig Brozefsky <·····@onshore.com> writes:

>> Maybe gradually - but not in principle. Is there
>> any revolutionary coding practice in effect that
>> suddenly makes Linux software more reliable?
>
>Yah, massive peer review, large-scale beta testing, and use of an
>educated user base to produce high quality bug reports.  It's a

Peer review?  Isn't it with linux that a rather dumb mass cheers
and praises every fart that comes from a couple of self-glorifying
"gurus" (even when the same old boring wheel gets reinvented poorly
over and over again)?  Educated user base producing high quality
bug reports?  That's really a minority, imho.

-- 
  Matthias K. Buelow      * Boycott Micro$oft, see http://www.vcnet.com/bms/ *
From: Joachim Achtzehnter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ucyaog7nhn.fsf@ns.mercury.bc.ca>
······@lavielle.com (Rainer Joswig) writes:

> Craig Brozefsky <·····@onshore.com> wrote:
> 
> > Uhm, both the kernel is more reliable, and for the large part, most
> > of the programs in the various distributions are more reliable.
> 
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?

No, of course not. But neither is the opposite true: The fact that the
source code is kept secret doesn't imply that it is well-designed or
that good coding practices were used. What I am trying to get at
here is that it is futile to compare the quality of "propriatary
software" with that of "free software" in general terms, just doesn't
make sense. We have to realise that the question of whether software is
free or not, is orthogonal to the quality issue.

Note also, that "free software" can be commercial software. There
seems to be this common misconception that somehow free software is
always produced by amateurs, or that proprietary software is always
produced by professionals. Much of the heated arguments and flaming
about this topic would be avoided if people didn't always mixup the
issues of freedom and quality. Free software can be of high or low
quality, and the same is true for proprietary software.

What we can reasonably argue about is whether it is easier to build
and maintain high quality software using the proprietary model or the
free software model. Everything else being equal (most importantly, if
the people doing the development are of comparible caliber) which
model might be more successful for a given project? This is the only
kind of question that makes sense in my mind. So in terms of the Linux
kernel, would it have been easier or more difficult for Linus, Alan
Cox, and a handful of others to build (and maintain!) something like
Linux using a proprietary approach?

And from a user's perspective, you have to weigh the alternatives: put
all eggs in one basket and accept that you depend totally on one
vendor to fix things or extend things? Or have the freedom to hire any
professional software developer to work on freely available source if
the "official" maintainers don't do their job?

And finally, from a vendor's perspective: Would a software company
be more or less successful in generating revenue with either model?
Clearly, only the proprietary model lends itself to the Microsoft
way of making money. But there can only be one (or at most a few)
Microsofts. How about companies like Franz or Harlequin? Is it so
clearcut whether they would be any worse off if they followed the free
software model? 

Again, trying to compare "quality" of proprietary software with free
software out of context makes no sense. Let us not forget that most
existing "serious" software was produced as proprietary software,
while a large portion of free software was done as a hobby. Let us not
compare apples with oranges.

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1012980419290001@194.163.195.67>
In article <··············@ns.mercury.bc.ca>, Joachim Achtzehnter
<·······@kraut.bc.ca> wrote:

> What we can reasonably argue about is whether it is easier to build
> and maintain high quality software using the proprietary model or the
> free software model. Everything else being equal (most importantly, if
> the people doing the development are of comparible caliber) which
> model might be more successful for a given project? This is the only
> kind of question that makes sense in my mind. So in terms of the Linux
> kernel, would it have been easier or more difficult for Linus, Alan
> Cox, and a handful of others to build (and maintain!) something like
> Linux using a proprietary approach?

Aren't/weren't they mostly "copying" what is/was already out there
grounded on massive experience and preexisting tools?
This seems to be easier than to develop something completely
new.

> And from a user's perspective, you have to weigh the alternatives: put
> all eggs in one basket and accept that you depend totally on one
> vendor to fix things or extend things?

Or a bunch of vendors competing with each other?

> Microsofts. How about companies like Franz or Harlequin? Is it so
> clearcut whether they would be any worse off if they followed the free
> software model? 

No. It would be interesting to explore the idea. Netscape
seems to see it as a way to survive as a company
(note that they are only releasing the browser software,
not the servers). How do you think would an "Open Source"
approach would look like for a Lisp vendor?

Rainer Joswig

-- 
http://www.lavielle.com/~joswig
From: Gareth McCaughan
Subject: Re: help! absolute beginner
Date: 
Message-ID: <863e6mopnj.fsf@g.pet.cam.ac.uk>
Erik wrote:

>                                                           with the rabidly
>   anti-commercial attitudes the "Free Software" movement has recently
>   adopted,

I'm puzzled. I can see rabidly anti-commercial attitudes in the
"free software" movement, but not *recently adopted* ones. The
only really big recent changes I can think of are (1) Netscape
going into "free software", and (2) the whole "open source"
business. How is either of those "rabidly anti-commercial"?

-- 
Gareth McCaughan       Dept. of Pure Mathematics & Mathematical Statistics,
·····@dpmms.cam.ac.uk  Cambridge University, England.
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122429737934433@naggum.no>
* Gareth McCaughan <·····@dpmms.cam.ac.uk>
| I'm puzzled.  I can see rabidly anti-commercial attitudes in the "free
| software" movement, but not *recently adopted* ones.  The only really big
| recent changes I can think of are (1) Netscape going into "free
| software", and (2) the whole "open source" business.  How is either of
| those "rabidly anti-commercial"?

  ok, so maybe I have not been paying 100% attention for the past year or
  so, but in my opinion, the Free Software Foundation has changed its view
  on commercial software.  it used to be that GNU/Linux was good because it
  was free software.  _something_ has changed to make it bad for commercial
  operators to use GNU/Linux and to ship commercial (non-free) software for
  GNU/Linux.  this came as a surprise to me, as did the pronouncement that
  non-free software was "evil".  I don't want to be evil, and one way not
  to be is simply to leave the people who might call me that.

  "open source" (whatever it actually _is_) is a competitor to the Free
  Source Foundation, for good or bad.  I'm beginning to think that maybe
  people lost track of their real purpose, like the IRS has lost track of
  financing the government, and government in general has lost track of
  providing for people's safety.  in becoming institutions, something died.

  I'm frankly worried that Open Source is a symptom of something that
  doesn't really work rather than a cure for a problem.

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Joachim Achtzehnter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <uc7lvue54q.fsf@ns.mercury.bc.ca>
Erik Naggum <····@naggum.no> writes:
>
>   ...in my opinion, the Free Software Foundation has changed its
>   view on commercial software.  it used to be that GNU/Linux was
>   good because it was free software.  _something_ has changed to
>   make it bad for commercial operators to use GNU/Linux and to ship
>   commercial (non-free) software for GNU/Linux.

Don't think that the FSF's view has changed on this issue at all, at
least not in the sense you discuss. The central argument is still
that software should be free according to the interpretation of free
software given in the GPL. This was always the FSF's view.

Regarding "commercial operators" we should watch our language more
carefully because the FSF has absolutely nothing against commercial
software. Commercial software is not the opposite of free software,
these two aspects of software development are orthogonal issues. What
the FSF doesn't like is proprietary software, software where the
copyright owner denies you the right to use the source code in one way
or another.

This argument against proprietary software is not new, don't know why
you think "_something_ has changed"? Note, in the view of the FSF,
proprietary products that run on Linux are not any worse than the same
proprietary products running on NT.

Perhaps, one aspect has changed: It is now possible to build a
practical general-purpose system without using any proprietary
software. This wasn't feasable until recently.

>   this came as a surprise to me, as did the pronouncement that
>   non-free software was "evil".  I don't want to be evil, and one way not
>   to be is simply to leave the people who might call me that.

Don't think the FSF claims that users of proprietary software are
"evil". They believe proprietary software is a bad thing and would
prefer if people used free software when there is a choice. The FSF is
about freedom, it certainly doesn't want to deny you the right to use
whatever software you like.

>   "open source" (whatever it actually _is_) is a competitor to the Free
>   Source Foundation, for good or bad.  I'm beginning to think that maybe
>   people lost track of their real purpose...
> 
>   I'm frankly worried that Open Source is a symptom of something that
>   doesn't really work rather than a cure for a problem.

Agree with you about "open source". IMHO, the open source initiative (OSI)
is an attempt by certain quarters to advance personal ambitions. It
totally misses the point about free software and has caused a lot of
confusion. By the way, have you seen the OSI home page at URL
http://www.osi.org ? :-)

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122671247752912@naggum.no>
* Joachim Achtzehnter <·······@kraut.bc.ca>
| This argument against proprietary software is not new, don't know why you
| think "_something_ has changed"?

  well, it certainly appears that you aren't paying much attention to
  what's going on, only reiterating old rhetoric.  all this is wasted on
  me, Joachim.  I've been working from _inside_ the FSF for many years,
  supporting GNU Emacs development with many thousands of hours.  now, what
  would make somebody (you) believe that somebody else (I) would do that
  without understanding and appreciation of the stated goals?  my guess is
  that _you_ don't actually understand the implications of "free software".

| Perhaps, one aspect has changed: It is now possible to build a practical
| general-purpose system without using any proprietary software.  This
| wasn't feasable until recently.

  precisely, and hence a dangerous arrogance results that non-free software
  is no longer "necessary", and people no longer have any "excuse" to use
  it.  I take exception to this arrogance, and so does many others, it
  seems, so it may well be the downfall of the entire movement.

| Don't think the FSF claims that users of proprietary software are "evil".

  RMS has referred to programmers who write non-free software as "evil" at
  a conference.  shortly thereafter, I removed my entry from the SERVICES
  file.  writing free software is a luxury, as all freedom is luxury, which
  is not to say it should be denied, but it doesn't _come_ for free.
  writing non-free software for a lot of money could very well be a means
  to an end, just as most people have paying jobs for the sole reason of
  supporting their real interests.  RMS has denied the entire world of
  programmers this option.

| Agree with you about "open source".  IMHO, the open source initiative
| (OSI) is an attempt by certain quarters to advance personal ambitions.
| It totally misses the point about free software and has caused a lot of
| confusion.

  I don't care about your opinions when they take this shape.  I also see
  no evidence that you have got _the_ point with free software right.

  my concern, hope, and goal with working to further what I have come to
  call "unrestricted source availability" was (1) to educate programmers by
  showing them good coding practice and effectively provide a "canon" of
  code, and (2) to weed out the bad code that inevitably creeps into large
  projects where lots and lots of people participate.

  well, have my hopes come to fruition?  (1) is not a goal shared by
  people, anymore.  all they want is for some fast computer to do stupid
  software tricks.  that means fewer very good people to learn from, and
  fewer new good people grow up in this environment.  I was into free
  software because I have a passion for competence and I saw it as a
  vehicle for collective betterment and a platform for the truly great
  potential to reach higher.  and, let's be honest, (2) never happens.
  people don't fix other people's code except when it stops "working",
  whatever the hell _that_ means these days.  e.g., GNU Emacs itself became
  infested with the worst crap I have ever seen when the MULE shit drowned
  out the good.  instead of being very good at what it does and improving
  in interesting ways, Emacs is heading for popularity, such as
  WYSIWYGitude and that kind of crap, which I find to be manifestly
  uninspiring, and which cannot be done really well without seriously
  rethinking what we want to _get_ before we start _seeing_ it -- WYSIWYG
  reinforces an "I do not want what I cannot see" attitude, which is
  _incredibly_ stifling.  the whole GUI business is forcing this on all of
  us: if you can't select it from a menu, you shouldn't _want_ it.

  in a nutshell, the free software movement has attracted people who want
  something for free and who don't give a damn about the purported freedom
  that comes with it.  I don't see any need to work to give people freedom
  they don't actively want.  the programmers in the free software movement
  have become _less_ skilled over time because the really good people get
  soaked up by commercial operations as soon as a smart manager realizes
  what a fantastic recruitment gold-mine they've discovered, and the
  movement itself is no longer capable of generating goals with sufficient
  "emotional gravity" to keep people hanging around.

  when _my_ goals are better served by working with "closed source", I'd
  say the Free Software Foundation has a serious problem.

  incidentally, I don't think we will discuss software in terms of openness
  or closedness in a decade or so.  software is gradually becoming so big
  and complex that "open source" may well become a necessity to survival,
  whereas "free software" was a protest movement.  like women's liberation
  is becoming a necessity to maintain our economies, it loses some of its
  force as a protest movement.  the biggest problem with protest is that it
  sometimes succeeds, and then you get to discover if you had any _real_
  goals with what you were doing.  when the free software movement turns to
  arrogance towards their much reduced "enemy" of the past, I get a very
  bad feeling that the people behind it weren't _constructive_ in their
  desire for changes.  furthermore, I believe that really competent people
  want to build things that last, they don't want to change things for no
  good reason -- that's what we have politicians for.  the lithmus test,
  then, should be whether the best people hang around when the protest is
  waning or whether they loyalties start to change.

  as you can see, I once felt strongly for free software.  I wouldn't just
  change my mind without something motivating it.  I know it isn't getting
  kids and accumulating debts and all that crap that turns many a protester
  into a salary slave with a difficult past, so I think it's something in
  the free software foundation that is changing.  no amount of regurgitated
  rhetoric is going to chance that impression.

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Joachim Achtzehnter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <uclnk9w5y6.fsf@ns.mercury.bc.ca>
On Dec 15, 1998 Erik Naggum <····@naggum.no> wrote::
> 
>   well, it certainly appears that you aren't paying much attention to
>   what's going on, only reiterating old rhetoric.  all this is wasted
>   on me, Joachim.

Noted without comment.

>   I've been working from _inside_ the FSF for many years,

Not having worked from the inside of the FSF I will take your word as
far as inside information is concerned.

>   now, what would make somebody (you) believe that somebody else (I)
>   would do that without understanding and appreciation of the stated
>   goals?

Remember, on Dec 12, 1998 Erik Naggum wrote:
>
>   the Free Software Foundation has changed its view on commercial
>   software.

Since I don't have the inside scoop about the FSF, my interpretation
of their views about "commercial software", or anything else for that
matter, is based on the FSF's official documents and statements, not
on anything particular individuals, e.g. rms, may have said in private
or in speeches I don't know about. My point was not whether _you_ do
or don't appreciate the stated goals, my point was that in terms of
the FSF's official goals I was surprised that you would say they have
changed.

And further on Dec 15, 1998 Erik Naggum wrote:
>
>   my guess is that _you_ don't actually understand the implications
>   of "free software".

Noted without comment.

>   precisely, and hence a dangerous arrogance results that non-free
>   software is no longer "necessary", and people no longer have any
>   "excuse" to use it.  I take exception to this arrogance, and so
>   does many others, it seems, so it may well be the downfall of the
>   entire movement.

Arrogance is always a bad thing, and I agree with you that the
representatives of the free software movement, and _everybody_ in
general, is well-advised to watch that they don't come across as being
arrogant :-)

>   RMS has referred to programmers who write non-free software as
>   "evil" at a conference.

Didn't know that, but I will take your word for it. On this particular
matter I strongly disagree with him then. I'll have to, if I don't want
to refer to _myself_ as an evil person :-(

>   shortly thereafter, I removed my entry from the SERVICES
>   file.

That was your choice, and I don't think anybody will blame you. As far
as I'm concerned though, an individual making a statement in a speech
with which I disagree does not necessarily cause me to abandon support
for an organisation that this individual happens to represent. I don't
think there is any individual with whom I would agree on every point
even when discussion is limited to a narrow topic like the software
industry. Life is a compromise...  Of course, different people will
attach different significance to such a statement...

>   writing free software is a luxury, as all freedom is luxury, which
>   is not to say it should be denied, but it doesn't _come_ for free.

Agree with this statement, at least at this point in time when free
software is the exception to the rule.

>   writing non-free software for a lot of money could very well be a
>   means to an end, just as most people have paying jobs for the sole
>   reason of supporting their real interests.

Agree 100%. I do believe that non-free software is a bad thing
compared to free software, but in comparison with other evils in this
world I consider it rather unimportant. Life is not perfect, and if
one has to compromise, better do it in a less important area.

>   RMS has denied the entire world of programmers this option.

Well, IMHO this is far too strong a statement.

> | Agree with you about "open source".  IMHO, the open source initiative
> | (OSI) is an attempt by certain quarters to advance personal ambitions.
> | It totally misses the point about free software and has caused a lot of
> | confusion.
> 
>   I don't care about your opinions when they take this shape.

What exactly about my opinions do you have in mind here? Is it that
you think the OSI is a good thing? Do you think the OSI's leadership,
in comparison with the FSF's representatives, is less motivated by
personal ambition and only has the "general good" in mind? Do you
think the introduction of a new term for free software has reduced
confusion?

>   I also see no evidence that you have got _the_ point with free
>   software right.

Noted without comment.

>   my concern, hope, and goal with working to further what I have
>   come to call "unrestricted source availability" was (1) to educate
>   programmers by showing them good coding practice and effectively
>   provide a "canon" of code,

This is a noble goal, and I would think that the free software model
is rather helpful for achieving it. Of course, it is a separate goal,
and it won't happen simply by making software free.

>   and (2) to weed out the bad code that inevitably creeps into large
>   projects where lots and lots of people participate.

Again, a good objective, and one where one would think the free
software model can help.

>   well, have my hopes come to fruition?  (1) is not a goal shared by
>   people, anymore.  all they want is for some fast computer to do
>   stupid software tricks.  that means fewer very good people to
>   learn from, and fewer new good people grow up in this environment.

Perhaps your expectations were too high? Progress never happens over
night.

>   I was into free software because I have a passion for competence
>   and I saw it as a vehicle for collective betterment and a platform
>   for the truly great potential to reach higher.

Agree with these lofty goals 100%, but have to say that in reality,
like always, we'll have to live with compromises. Not every piece of
code will be as good as it could be. And it doesn't really have to be
either, if it does the job. Writing software is not a goal in itself,
most people write software to address a practical requirement.

>   and, let's be honest, (2) never happens.

This is simply not true. Bad code has been removed from the Linux
kernel on many occasions, and the same is true for other projects.

>   people don't fix other people's code except when it stops
>   "working",

Correct. But I don't see any big problem with this. If it really works
then why fix it? The real problem with bad code is that it often
doesn't work, especially in the face of changing requirements. And
after trying band-aid fixes a few times it is certainly advisable to
do it right.

>   WYSIWYG reinforces an "I do not want what I cannot see" attitude,
>   which is _incredibly_ stifling.  the whole GUI business is forcing
>   this on all of us: if you can't select it from a menu, you
>   shouldn't _want_ it.

Disagree with your view of WYSIWYG, but I'd rather not get into a
discussion about that topic, it has little to do with our discussion
here.

>   in a nutshell, the free software movement has attracted people who
>   want something for free and who don't give a damn about the
>   purported freedom that comes with it.

The way I see it these people seem to gather around the open source
initiative...

>   the programmers in the free software movement have become _less_
>   skilled over time because the really good people get soaked up by
>   commercial operations...

Could it be, in part, also simply because there are more of them now?
While in the past, only really good people had the insight to see the
advantages of free software?

>   when _my_ goals are better served by working with "closed source",
>   I'd say the Free Software Foundation has a serious problem.

I wouldn't focus too much on _you_ :-), but certainly if your view was
shared by many former free software supporters then the FSF would have
a problem.

>   as you can see, I once felt strongly for free software.  I
>   wouldn't just change my mind without something motivating it.

Certainly. Of course I don't know you personally, but is it so
outlandish to consider the possibility that _you_ may have changed a
little too?

>   I know it isn't getting kids and accumulating debts and all that
>   crap that turns many a protester into a salary slave with a
>   difficult past,

Well, I must say that having a family did change my approach to these
things very dramatically.

>   so I think it's something in the free software foundation that is
>   changing.

No doubt, the FSF is changing too, like everything else.

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122847815898530@naggum.no>
* Joachim Achtzehnter <·······@kraut.bc.ca>
| Agree 100%.  I do believe that non-free software is a bad thing compared
| to free software, but in comparison with other evils in this world I
| consider it rather unimportant.  Life is not perfect, and if one has to
| compromise, better do it in a less important area.

  well, I don't believe that "non-free software" is a bad thing.  period.

  I believe that blocking people's access to the whatever they need to do a
  good job and to understand what's going on is bad, but, and this is my
  gripe: I get _better_ results in this exact regard from signing NDAs and
  license agreements than I get from stamping GPL on my code or using
  GPL'ed code, and in some cases, purportedly free code contaminates a
  commercial product to the point where one just cannot use it.  that's why
  I have come to talk about "unrestricted source availability", since what
  is free to you may be unfree to me, and what is unfree to you may be free
  to me.  e.g., because I cannot agree on behalf of my paying client that
  their code be subject to the GPL, I am _not_ free to use GPL'ed code in
  the development of our software.  however, because I do agree to a lot of
  other terms, I _am_ free to use source code that few other people are
  free to use because they don't agree to those terms (or haven't had the
  opportunity to).

  _real_ freedom is not subject to what you agree to, but in the ability to
  agree or disagree with whatever you want and proceed from there.  ergo,
  in at least one sense, "free software" is antithetical to real freedom.
  there is no such thing as arguing that people "should agree", and that's
  the change in attitude that I see from the Free Software movement: the
  freedom to choose commercial, proprietary, closed, or whatever software
  is no longer respected as much as it used to be.

| What exactly about my opinions do you have in mind here?

  "advancing personal ambitions".  you need hard evidence that that is what
  people are really after before you go public with such comments.  people
  have ambitions of all sorts and shapes, but you're denying them the
  opportunity to be constructive by saying what you did, because one whose
  goals are "advancing personal ambitions" _will_ be illoyal to his peers
  and causes if he thinks he can profit on and get away with it.

  always look for the constructive element in what people do.  you might be
  surprised how often it is not the "personal amibition" it appears to be.
  competent people very seldom place their person above their merits -- I
  have found it to be a disturbingly recurring trait of the incompetent to
  do so, perhaps because their merits don't quite cut it.  again, I am not
  willing to judge people that harshly without significant evidence.

| Is it that you think the OSI is a good thing?

  I have no opinion on OSI.

| Perhaps your expectations were too high?  Progress never happens over
| night.

  I tend to invest at least half a decade in something before I start to
  look for returns on investment.  I don't know many other people who are
  equally patient.  however, it doesn't appear that you need much evidence
  to imply that people believe in "overnight progress" nor that they might
  "advance personal ambitions", and your vagueness and non-committal form
  is quite annoying coupled with the vaguely derogatory style.  it could be
  that "progress never happens over night" is just a meaningless phrase to
  you that seemed to fit, but why repeat content-less phrases, and out of
  context, even?

| Agree with these lofty goals 100%, but have to say that in reality, like
| always, we'll have to live with compromises.  Not every piece of code
| will be as good as it could be.  And it doesn't really have to be either,
| if it does the job.  Writing software is not a goal in itself, most
| people write software to address a practical requirement.

  what _do_ you actually say here?  there are several paragraphs like this
  in both this and your previous messages.  I can't spot the contents or
  the issue you want to raise -- it's all very comfortably non-committal.

  I don't think life is about compromises, it's about goals and values and
  dealing with physical reality and other people's goals and values, and in
  this compromise is _sometimes_ a necessary evil, but reaching goals and
  upholding values is what makes it all worth it.  for me, competence is
  such a high value that I'm unwilling to compromise it against anything.
  we don't _have_ to live with compromises -- it's a choice as good as any,
  and you are free to walk away.  I do that sometimes, and I get this weird
  look from people who think they have to take all the shit that's given to
  them because "life's all about compromises".

| This is simply not true. Bad code has been removed from the Linux
| kernel on many occasions, and the same is true for other projects.

  only when it failed to "work".  I haven't seen people go over free code
  in the "review" sense.  but that's what it all needs.  (BTW, it appears
  that the FreeBSD people are actively engaged in code review internally.)

| But I don't see any big problem with this.  If it really works then why
| fix it?  The real problem with bad code is that it often doesn't work,
| especially in the face of changing requirements.  And after trying
| band-aid fixes a few times it is certainly advisable to do it right.

  it's bad because "if it ain't broke, don't fix it" has a bad habit of
  turning into "if we can't fix it, it ain't broke"�, or, in more software
  terms, "it isn't a bug, it's a feature".

| Certainly. Of course I don't know you personally, but is it so outlandish
| to consider the possibility that _you_ may have changed a little too?

  as I said, I have ruled that out.  what changes have occurred have to do
  with reaching the goals that I had with my free software work elsewhere,
  and it appears that I'm not at all alone in this regard, probably because
  the protest movement against closed source and no access has succeeded in
  giving people access, but at much more _reasonable_ terms than the GPL.

| No doubt, the FSF is changing too, like everything else.

  I don't know what this statement means.  I don't know what the entirely
  equivalent statement that everything stays the same means, either.

#:Erik
-------
1 attributed to Lt.Col. Walt Weir, USArmy
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Pierre Mai
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87lnk713e6.fsf@orion.dent.isdn.cs.tu-berlin.de>
Erik Naggum <····@naggum.no> writes:

>   only when it failed to "work".  I haven't seen people go over free code
>   in the "review" sense.  but that's what it all needs.  (BTW, it appears
>   that the FreeBSD people are actively engaged in code review internally.)

In this area, the OpenBSD group has also been notable, in that one or
two years ago they performed a full security audit on their source
base.  Other OSes (open- and closed-source) have mostly taken on the
attitude, that fixing security holes one at a time as they appear on
the diverse security alert fora is sufficient... ;-(

>   it's bad because "if it ain't broke, don't fix it" has a bad habit of
>   turning into "if we can't fix it, it ain't broke"¹, or, in more software
>   terms, "it isn't a bug, it's a feature".

It's also problematic, in that broken code will slowly start infecting
many other parts of the system, since others will have to code around
the broken bits.

Regs, Pierre.

-- 
Pierre Mai <····@acm.org>               http://home.pages.de/~trillian/
  "One smaller motivation which, in part, stems from altruism is Microsoft-
   bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]
From: Joachim Achtzehnter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <uc7lvq8jz1.fsf@ns.mercury.bc.ca>
Erik Naggum <····@naggum.no> writes:
> 
>   well, I don't believe that "non-free software" is a bad thing.
>   period.
>
>   I believe that blocking people's access to the whatever they need
>   to do a good job and to understand what's going on is bad,

This is an interesting point of view. The point of the "free software"
versus "non-free software" debate is, of course, exactly about making
sure that people get this access to what they need. Sorry about
repeating "old rhetoric" again, but your contradictory statements
provoke such repetition.

>   I get _better_ results in this exact regard from signing NDAs and
>   license agreements than I get from stamping GPL on my code or using
>   GPL'ed code,

You should have highlighted the word _I_ in this statement. You will
have to admit that not everybody who needs access will get access in
this way. With non-free software access is not a right, access is a
privilege given to whoever the people controlling the software want to
give that privilege. Effectively, the people who control the software
get to decide who needs access. It is in this sense that non-free
software is bad.

>   and in some cases, purportedly free code contaminates a commercial
>   product to the point where one just cannot use it.

On this point I have had my own doubts about the GPL for a long
time. Clearly, the GPL restricts freedom in the sense that GPL'd
software cannot be used in certain circumstances. The way I look at
this is that I accept this restriction because it is a means by which
an author who makes the source available to everybody is trying to
ensure that her work is not exploited by others who don't provide such
access. On the other hand, I have no problem with software licenses
that give totally unrestricted access to source code.

>   that's why I have come to talk about "unrestricted source
>   availability", since what is free to you may be unfree to me, and
>   what is unfree to you may be free to me.  e.g., because I cannot
>   agree on behalf of my paying client that their code be subject to
>   the GPL, I am _not_ free to use GPL'ed code in the development of
>   our software.  

Not sure whether I get your point here. If you are advocating that
there should be absolutely no restriction on access to source code,
then I agree. But then, you said above that non-free software is not
a bad thing? I'm confused.

>   however, because I do agree to a lot of other terms, I _am_ free
>   to use source code that few other people are free to use because
>   they don't agree to those terms (or haven't had the opportunity
>   to).

You agree then, that source code access via an NDA is not the kind of
unrestricted source availability we need?

>   _real_ freedom is not subject to what you agree to, but in the
>   ability to agree or disagree with whatever you want and proceed
>   from there.  ergo, in at least one sense, "free software" is
>   antithetical to real freedom.

Yes, one aspect of the GPL is problematic in this regard, namely the
insistence that "any work based on" the software must fall under the
same license. An acceptable objective here is that somebody granting
access to source code wants to prevent others from exploiting her
work. The problem is, that with software components and systems
becoming more and more interrelated and integrated, it is becoming
increasingly difficult to apply the mechanisms in the GPL/LGPL that
try to differentiate between mere _use_ of a software package (in the
way libraries are linked with programs) and a derivative work.

>   there is no such thing as arguing that people "should agree", and
>   that's the change in attitude that I see from the Free Software
>   movement: the freedom to choose commercial, proprietary, closed,
>   or whatever software is no longer respected as much as it used to
>   be.

Agree in the sense that nobody should be forced to agree with, or use
the license of, the Free Software Foundation. People should continue
to have the right to use and create proprietary, closed software.
Whether or not it is true that respect for these rights is
less than what it used to be is open to question.

Note, that the right to create proprietary, closed software implies
that freedom of access is explicitly denied, which I accept. If
somebody doesn't agree with this restriction they can choose not to
use proprietary software, but this is the only remedy. Equally though,
it must be acceptable for people to put certain restrictions on the
use of _their_ free software. And again, the only remedy for people who
disagree with such restrictions is to not use the software.

>   "advancing personal ambitions".  you need hard evidence that that
>   is what people are really after before you go public with such
>   comments.

I'll accept your criticism on this point, I did indeed make these
statements without providing specific evidence. Instead of providing
such evidence now and starting another long debate I prefer to retract
my statements, especially because I think this isn't the most
important problem with the OSI (and I don't want to start a discussion
about the OSI, period!). These issues are being discussed almost weekly
on Slashdot, if anybody is interested...

>   always look for the constructive element in what people do.

Thanks, for reminding me. _Everybody_ should heed this advise :-)

>   I am not willing to judge people that harshly without significant
>   evidence.

Since you are giving out free advise, let me contribute some of my
own: You may want to re-read some of your posts, one can detect an
ever so slight tendency to "harshly judge people without significant
evidence" even in _your_ posts. :-)

> | Perhaps your expectations were too high?  Progress never happens over
> | night.
> 
>   it doesn't appear that you need much evidence to imply that people
>   believe in "overnight progress" ... and your vagueness and
>   non-committal form is quite annoying

The simple reason for the vagueness was that I didn't want to claim
that I knew the definitive reason for your believe that there was a
failure. Was only suggesting that perhaps it hasn't failed with the
finality you seemed to imply.

> | Agree with these lofty goals 100%, but have to say that in reality, like
> | always, we'll have to live with compromises.  Not every piece of code
> | will be as good as it could be.  And it doesn't really have to be either,
> | if it does the job.  Writing software is not a goal in itself, most
> | people write software to address a practical requirement.
> 
>   what _do_ you actually say here?

I am saying that the purpose of software is to do a job, to work
according to requirements. Whether a particular piece of source code
is in bad style, is irrelevant as long as it complies with
requirements and does the job. I disagree with the contention that it
is worth going through working software with the only purpose to
replace badly written (but correct) code with well-written code.

Now, before I am mis-interpreted here. I am not advocating bad code or
incompetence. On the contrary, I have shaken my head in disbelief many
times when reading bad code. When problems with such bad code do occur,
then by all means take the opportunity to do it right.

>   I don't think life is about compromises, it's about goals and
>   values and dealing with physical reality and other people's goals
>   and values, and in this compromise is _sometimes_ a necessary
>   evil,

That was more or less what I was trying to say. I might want to add
one more reason for compromise with regard to the competence thing:
Not every software project has the luxury to only employ people from
the exclusive group of highly competent software developers like
yourself. This requires certain compromises in terms of the kind of
source code that is "acceptable", or "good enough".

>   for me, competence is such a high value that I'm unwilling to
>   compromise it against anything.

Can't resist making another one of these "vaguely derogatory" remarks:
Might this explain the tone in some of your posts? Being surrounded by
incompetence must be hard to take! :-)

>   we don't _have_ to live with compromises -- it's a choice as good
>   as any, and you are free to walk away.

We don't have to compromise on everything. But I still claim that the
statement "we have to live with compromises" is very true if you want
to achieve anything that requires collaboration with other people.

> | This is simply not true. Bad code has been removed from the Linux
> | kernel on many occasions, and the same is true for other projects.
> 
>   only when it failed to "work".  I haven't seen people go over free
>   code in the "review" sense.

Depends on the purpose of the review. If the only objective is to
replace "bad code" with "good code" for the sake of purity then I
firmly believe that most substantial software systems cannot afford
such luxury, it usually does more harm than good. Of course, often bad
code does, in fact, _not_ work, and then, yes! it should be fixed
properly and band-aid solutions are short-sighted.

>   it's bad because "if it ain't broke, don't fix it" has a bad habit
>   of turning into "if we can't fix it, it ain't broke"�, or, in
>   more software terms, "it isn't a bug, it's a feature".

Then let us avoid the habit. The reluctance to fix broken software
doesn't invalidate the argument that things that work don't need
fixing.

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Dwight Hughes
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3679C5E5.87C6B82D@ipa.net>
Joachim Achtzehnter wrote:
> 
> Erik Naggum <····@naggum.no> writes:

> >   it's bad because "if it ain't broke, don't fix it" has a bad habit
> >   of turning into "if we can't fix it, it ain't broke"�, or, in
> >   more software terms, "it isn't a bug, it's a feature".
> 
> Then let us avoid the habit. The reluctance to fix broken software
> doesn't invalidate the argument that things that work don't need
> fixing.

Perhaps, but what _does_ invalidate it is that virtually the entire
history of technological advancement is based on fixing things that
weren't broken.

-- Dwight
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122936797876902@naggum.no>
* Erik Naggum
| well, I don't believe that "non-free software" is a bad thing.  period.
| 
| I believe that blocking people's access to the whatever they need to do a
| good job and to understand what's going on is bad,

* Joachim Achtzehnter <·······@kraut.bc.ca>
| This is an interesting point of view.  The point of the "free software"
| versus "non-free software" debate is, of course, exactly about making
| sure that people get this access to what they need.  Sorry about
| repeating "old rhetoric" again, but your contradictory statements provoke
| such repetition.

  contradictory statements?  the point is there are _a multitude_ of means
  to get "access to whatever [you] need to do a good job" besides an
  organized free software protest movement, and that the criticism I levy
  against the movement is that it _doesn't_ achieve that goal because they
  protest too much and reasonably business-oriented people cannot use free
  software in their own products, thereby creating pockets of resistance,
  instead of a real movement.  it is even contentious whether a programmer
  can look at free software while working on software that is intended to
  be sold as such, as opposed to in-house tools.  free software works for
  and between programmers, but even we cross the border to userland once in
  a while, and different rules and conditions apply there.

  it appears that the only way you imagine getting access to source is if
  it's "free" in the FSF sense.  this is just not so in the real world.

| You should have highlighted the word _I_ in this statement.  You will
| have to admit that not everybody who needs access will get access in this
| way.

  I will have to admit no such thing.  have you ran out of arguments so
  badly you need to tell _me_ what to do, now?

  since I started working with computers in 1978, I have obtained a lot of
  "special favors" from people all over the place, only to realize that
  they weren't special at all.  I have had something to offer, and people
  have responded in kind.  remember my criticism against the people in the
  free software movement, that they just _want_ and don't care about what
  they get?  well, consider this possibility: that they don't have anything
  to offer anybody, so free software is their only hope.  how about that?

  I have previously argued that free software denies us the opportunity to
  use money as the currency of exchange of last resort.  that is, any two
  people who disagree can resort to exchanging money, because most values
  boil down to money sooner or later.  for free software, that is no longer
  an option.  nobody can buy their way out of the GPL.  however, the legal
  profession is _all_about_ getting into and out of various agreements in
  exchange for money.  what does this say about how the GPL is supposed to
  work in our society?  think about it.

| With non-free software access is not a right, access is a privilege given
| to whoever the people controlling the software want to give that
| privilege.

  you obviously haven't read the GNU Public License carefully or discussed
  it with a intellectual rights lawyer.  I have both, other businesses even
  consult me on what it means and what they can and cannot do.  (it is in
  part because of this that I argue against the idea that free software is
  intrinsically "good" and that non-free software is intrinsically "bad".
  first, values aren't intrinsic to begin with, but second, values relate
  to a context of other values, yet some programmers don't seem to get it.)

  in a nutshell: GPL is an agreement between the author and the user such
  that the user must refrain from a _lot_ of rights otherwise obtained by
  default once you buy or use something.  this is not unlike the
  shrink-wrap licensing that goes on in the rest of the software business
  and which is in for a rough time as the "contracts" get so involved that
  people cannot reasonably be expected to understand them.  in the same
  way,   "free software" is very aptly named.  the software may be free,
  but the _people_ using it really aren't.  it's a big difference.

| Effectively, the people who control the software get to decide who needs
| access.  It is in this sense that non-free software is bad.

  listen to yourself!  bombs over bagdad, this is so freaking annoying.
  free software is _more_ controlled in exactly this sense than source that
  does not lie in the open -- it has to, _because_ it is in the open.  do
  think about this, _please_!

| Clearly, the GPL restricts freedom in the sense that GPL'd software
| cannot be used in certain circumstances.

  precisely my point, and it's still "free"?  I don't think so.  there are
  some legal terms that are much stronger than "free".  one is "unfettered"
  and the other is "unencumbered".  (I use "unrestricted" because most
  people don't understand the two other words.)
  
| The way I look at this is that I accept this restriction because it is a
| means by which an author who makes the source available to everybody is
| trying to ensure that her work is not exploited by others who don't
| provide such access.

  and I accept the restrictions imposed on me by other licenses.  what's
  your beef with non-free software if you really understand that you do
  have to accept a number of restrictions?

| Not sure whether I get your point here.  If you are advocating that there
| should be absolutely no restriction on access to source code, then I
| agree.  But then, you said above that non-free software is not a bad
| thing?  I'm confused.

  commercial operators can agree to things amongst themselves on a case by
  case basis.  in contrast to this, the GNU _Public_ License affects people
  who are not party to the agreement, and they cannot negotiate further.
  GPL is like consumer law in Europe, where businesses are forbidden from
  entering special deals or rejecting customers on a case by case basis if
  they have announced something to be open for the general public.  the GPL
  is very much a take it or leave it proposition.  however, programmers and
  people creating software aren't consumers, they're businesses, and every
  country on earth has recognized the need for businesses to protect their
  interests in deals between themselves differently from consumers, whom
  some countries think the government should protect.

| You agree then, that source code access via an NDA is not the kind of
| unrestricted source availability we need?

  stop telling me what I agree to!  damn it, aren't you intellectually
  honest enough to handle the fact that I _reject_ the foundations for your
  beliefs without having me "admit" or "agree" to _your_ views!?  you don't
  undestand what I need, and yet claim to speak for both of us.  just quit
  that, OK?

  an NDA is no more restrictive than the GPL is, in each their senses.  if
  my goal is to do a good job, I do whatever it takes, and if it means
  having a client cough up a lot of money or me signing an NDA, so be it.

  do you understand _why_ there is non-free software?  I don't think so.

| Agree in the sense that nobody should be forced to agree with, or use
| the license of, the Free Software Foundation.

  then why can't they negotiate their own licenses?

| Note, that the right to create proprietary, closed software implies
| that freedom of access is explicitly denied, which I accept.

  you've got it all upside down.  access is _granted_, even in the GPL.  no
  wonder you're confused.  there exist no right to "access" what somebody
  else has created.  that person must obviously publish it first, and then
  our societies recognize that creators of things published need to make a
  living, too, and therefore protect them from all the people who just
  _want_ what they publish.  the GPL is a horribly involved legal document
  because it attempts to _deny_ what our societies have set up as the
  default and as an extension of the _right_ of any author or creator to
  maintain control over what he has written or created.

  the GPL is like a note on a door of a house saying "you are free to come
  inside and take whatever you want, but you promise that once you do so,
  you will open the door to your own house the same way and impose on any
  who enters your house the same requirement.  I will force you to comply
  with this requirement if necessary, and you agree to force anyone to
  comply with this requirement, or be in non-compliance."  this flies in
  the face of the way our societies are organized.  (that it might well
  reduce what would otherwise have been theft to zero because nobody dares
  steal anything is another matter entirely. :)

| Since you are giving out free advise, let me contribute some of my
| own: You may want to re-read some of your posts, one can detect an
| ever so slight tendency to "harshly judge people without significant
| evidence" even in _your_ posts. :-)

  amusing twist.  but notice what I said: "I am not willing to judge people
  that harshly without significant evidence."  now stress "that".  got you,
  didn't I?  :)

| Can't resist making another one of these "vaguely derogatory" remarks:
| Might this explain the tone in some of your posts? Being surrounded by
| incompetence must be hard to take! :-)

  we are USENET.  you will be derogated.  resistance is futile.

  in _fact_, I am positively _surrounded_ by competent people.  because I
  expect people to be competent and reward it with profound recognition, I
  very seldom have to deal with incompetents in real life (I think they
  just leave).  USENET is an exception, probably because it is very hard to
  emanate recognition in this medium, but so amazingly easy to communicate
  the lack thereof.

| We don't have to compromise on everything.  But I still claim that the
| statement "we have to live with compromises" is very true if you want
| to achieve anything that requires collaboration with other people.

  tell that to Saddam and Clinton.  funny how those who don't compromise
  get to run big corporations and countries and such, when all we're told
  all through our "education" is to accept compromise, innit?  could it be
  that compromise is the _best_ way to keep the public pacified and in line
  with what the non-compromisers want?  my guess is it's a lot easier to
  run a country (or a religion, as the case may be) of compromisers.

| Depends on the purpose of the review.  If the only objective is to
| replace "bad code" with "good code" for the sake of purity then I firmly
| believe that most substantial software systems cannot afford such luxury,
| it usually does more harm than good.

  huh?  replacing bad code with better code does more harm than good?

  I do code review _all_ the time.  I have other people go over my code to
  comment on it and understand it.  this summer, I had my client pay a very
  promising young programmer to look over everything I did with me and help
  me make it understandable and maintainable and correct.  it's the best
  work I have probably ever done.

  have you ever heard any author claim to have written all his stuff alone
  and never had anybody read it before it was published all at once?  in my
  years in the publishing world, I came to appreciate that the author is
  but a small, albeit crucial, part of the process from idea to finished
  publication.  this is, unfortunately, lost on the WWW generation where
  everybody thinks they can be a publisher, just like everybody thinks they
  can program a computer just because it doesn't ridicule their efforts.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Joachim Achtzehnter
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ucbtl1ff4u.fsf@ns.mercury.bc.ca>
Erik Naggum <····@naggum.no> writes:
>
> In response to my statements about NDAs:
> | You should have highlighted the word _I_ in this statement.  You
> | will have to admit that not everybody who needs access will get
> | access in this way.
> 
>   I will have to admit no such thing.  have you ran out of arguments so
>   badly you need to tell _me_ what to do, now?

I stick to my guns here. Somebody seriously claiming that _everybody_
will get access to source code via an NDA is joking, period.

> | Effectively, the people who control the software get to decide who
> | needs access.  It is in this sense that non-free software is bad.
> 
>   listen to yourself!  bombs over bagdad,...

You posted this to comp.lang.lisp, man! And you accuse me of running
out of arguments?

> | The way I look at this is that I accept this restriction because it is a
> | means by which an author who makes the source available to everybody is
> | trying to ensure that her work is not exploited by others who don't
> | provide such access.
> 
>   and I accept the restrictions imposed on me by other licenses.  what's
>   your beef with non-free software if you really understand that you do
>   have to accept a number of restrictions?

As far as I am concerned, people can create and use non-free software
as long as they like. At the same time I have every right to say that
I consider non-free software to be a bad thing. This is not so
difficult to understand, isn't it? People make their own decisions,
but we are trying to influence each other while doing so.

> | You agree then, that source code access via an NDA is not the kind of
> | unrestricted source availability we need?
> 
>   stop telling me what I agree to!  damn it, ...

Did you miss the question mark? I didn't understand what you were
trying to say, hence I _asked_ for clarification.

>   ... and yet claim to speak for both of us.  just quit that, OK?

What did you read? Speaking for _you_? That would never occur to me.

> | Agree in the sense that nobody should be forced to agree with, or use
> | the license of, the Free Software Foundation.
> 
>   then why can't they negotiate their own licenses?

I didn't say they can't. But I don't have to like it, or do I?

> | Since you are giving out free advise, let me contribute some of my
> | own: You may want to re-read some of your posts, one can detect an
> | ever so slight tendency to "harshly judge people without significant
> | evidence" even in _your_ posts. :-)
> 
>   amusing twist.  but notice what I said: "I am not willing to judge people
>   that harshly without significant evidence."  now stress "that".  got you,
>   didn't I?  :)

You didn't get me, you totally lost me. I have to admit that I don't
understand your joke here.

> | We don't have to compromise on everything.  But I still claim that the
> | statement "we have to live with compromises" is very true if you want
> | to achieve anything that requires collaboration with other people.
> 
>   tell that to Saddam and Clinton.

Plonk!

> | Depends on the purpose of the review.  If the only objective is to
> | replace "bad code" with "good code" for the sake of purity then I
> | firmly believe that most substantial software systems cannot
> | afford such luxury, it usually does more harm than good.
> 
>   huh?  replacing bad code with better code does more harm than
>   good?

In practise, when ordinary people write software (not _you_ perhaps)
they make mistakes. By replacing code that works perfectly well,
although it may not live up to certain standards of good programming,
with supposedly better code, you risk introducing bugs. Now, with
unlimited resources and time you can fix those. Real projects don't
have either, hence my suggestion that they can't _afford_ to fix code
that works.

>   I do code review _all_ the time.

While code is still actively developed a review is very useful. I was
talking about working software, and in particular, about code that is
no longer the focus of development.

Joachim

-- 
·······@kraut.bc.ca      (http://www.kraut.bc.ca)
·······@mercury.bc.ca    (http://www.mercury.bc.ca)
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123055729210311@naggum.no>
* Joachim Achtzehnter <·······@kraut.bc.ca>
| I stick to my guns here.  Somebody seriously claiming that _everybody_
| will get access to source code via an NDA is joking, period.

  since this is clearly a projection of your mindset onto my arguments, and
  proof of the utter lack of ability to understand contrary viewpoints, I
  consider this discussion to be permanently over; I thought you would have
  a reasonable foundation for your beliefs and opinions, but it turns out
  you are incapable of understanding any alternatives, so there's no point
  in trying to figure out how you reached your conclusions.

  first, the facts: I am not the person who is joking, since I have never
  made any claims, serious or otherwise, about _everbody_ getting anything
  by any means whatsoever.  I don't talk about everybody.   people who talk
  about "everybody" are most often trying to avoid being specific about the
  people they are actually talking about in a negative sense: whoever is
  "refusing" to be part of their nice little "everybody" notion, and if
  they go on to talk about everybody, maybe those people will somehow cease
  to exist; or they are trying to "futilize" their dreams and defend their
  procrastionation and sense of hopelessness in that  no matter how you
  define "everybody", you will _never_ get enough people to count as
  "everybody" to want the same thing, no matter what it is (except sex).

  second, I would, although I haven't, seriously claim that anybody who
  wants to do a good job would get access to source code via licenses or
  other agreements, including NDAs.  this happens to be how the world
  actually works, today, but to Joachim Achtzehnter this is joking.

  third, free software is manifestly _not_ for everybody.  there are some
  rather interesting legal encumbrances on those who work with it.  for
  instance, those who want to reimplement something into free software are
  strongly encouraged not to look at the other software, but have someone
  else tell them what it does.  someone who wants to use the results from
  free software would likewise be restricted from actually looking at the
  code, if only to avoid the repercussions of the GPL tainting the work.
  that is, if you really expect to make money off of writing software that
  somebody will want to keep to themselves, for whatever reason, you'd
  better not be a free software champion.

  as I have actually said, which you ignore when you can't find any more
  material in what I do say to attack so you have to invent something that
  actually contradicts what I said, my goal with free software was to help
  educate people, help discover the real talents and bring them into very
  productive positions.  clearly, my desire to see good source code in the
  open is not changed by my growing disdain for the legal machinery that
  surrounds free software.  perhaps "Open Source" is better, I haven't had
  time to look.  perhaps what Franz Inc does in releasing to their paying
  customers a significant fraction of their source code with their support
  agreements is sufficient.  perhaps books with source code where the
  author gets paid by the buying public is an even better solution -- I
  already have several such books, and the authors have made a lot of
  money, which I think is a lot better combination than giving out stuff
  for free to people who don't appreciate what they are actually getting.


  let me touch on some of the other puzzling parts of your message:

| And you accuse me of running out of arguments?

  you pick the strangest things to react to.  I can only assume that by
  failing to get the *RIDICULE* about two pathetic losers each caught with
  their pants down and trying to be tough, you don't think they are, in
  fact, pathetic losers, but somehow should be respected for what they do!
  in my country, we have an embarrassment for finance minister who actually
  costs Norway a billion Norwegian crowns in lost international confidence
  every time he opens his mouth and says something immensely stupid.  we
  have a prime minister who has been defeated on all possible counts, yet
  who clings to powr like a leech and says "we can live with that", and
  he's a _priest_.  Norway is the second country on the planet to be run by
  a man of the cloth, yet ours is a spineless wimp who chooses idiots for
  ministers.  our spineless wimp has also voiced his whole-hearted support
  for the bombing of Iraq, so you might appreciate that I cannot but think
  it is all a giant joke.  perhaps my disrespect for the stupid leaders of
  other countries is like the black humor in Eastern Europe where people
  _had_ to laugh at their miserable conditions to survive.  however, I
  think every reasonable man has an _obligation_ to laugh at incompetent
  fools who get into (political) power.  that way, everything they do is
  the source of another joke, and that's the _best_ guarantee that the
  public will know what they're doing and have time to protest before it
  has hurt them.  think about it.

| At the same time I have every right to say that I consider non-free
| software to be a bad thing.

  "just because you're constitutionally entitled to a personal opinion
  doesn't mean you're constitutionally entitled to a professional opinion."
						-- Michael Padlipsky

| This is not so difficult to understand, isn't it?

  well, speaking of which, precisely how much do you think you _understand_
  of what I have been trying to communicate to you?  I don't think you want
  me to _understand_ you, either, or you would have done a hell of a lot
  better job of providing me with something that would, in fact, have
  communicated your reasoning and experiences, instead of your emotional
  responses to not getting what you want.

| People make their own decisions, but we are trying to influence each
| other while doing so.

  sure, and by refusing to understand your opposition, even given your
  whining just a paragraph ago, you communicate that your position is one
  that others should _not_ adopt, because it is based on a willfully
  incomplete and romanticized view of the world you live in, which, BTW, is
  _very_ frequently coupled with a perceived need to "compromise" when the
  world isn't as nice as you think it is and you refuse to understand it
  and adjust to it, instead continuing to prefer your romanticized view + a
  compromise or two.  like our priest for prime minister: "you can live
  with that".

  compromise is what you do when you don't know what you want.
  experience is what you get when you don't get what you want.

| Did you miss the question mark?

  oh, how I _love_ this line.  thank you, thank you, thank you.  now I can
  insult you any time I want, as long as I use a question mark.  that's
  like, a General Insulting License, and we can call it "free insults".
  I'm glad you didn't use a smiley, though, because that's the General
  Insulting License, version 2, and it's not as easy to comply with.

| >   tell that to Saddam and Clinton.
| 
| Plonk!

  darn.  had I known you were a political _activist_ nutcase, too, I would
  have made _much_ more out of this pathetic warmongering incident than I
  have.  "hey, they're going to have their religious holidays soon, so
  let's at _least_ be nicer than George Bush was in 1991 and bomb them
  _before_ the Ramadan.  that'll tell'em _we're_ basically good guys!"


  and now for some really ridiculous expressions of _deep_ confusion:

| > | Depends on the purpose of the review.  If the only objective is to
| > | replace "bad code" with "good code" for the sake of purity then I
| > | firmly believe that most substantial software systems cannot
| > | afford such luxury, it usually does more harm than good.
| > 
| >   huh?  replacing bad code with better code does more harm than
| >   good?
| 
| In practise, when ordinary people write software (not _you_ perhaps)
| they make mistakes.  By replacing code that works perfectly well,
| although it may not live up to certain standards of good programming,
| with supposedly better code, you risk introducing bugs.  Now, with
| unlimited resources and time you can fix those.  Real projects don't
| have either, hence my suggestion that they can't _afford_ to fix code
| that works. 

  let's see, "better code" doesn't really mean "better code", it means
  "worse code" because most people are incompetent, and if I don't agree
  with this immeasurably stupid line, it's because I think I'm perfect,
  which you, because you are a true believer in incomptence, knows is
  really an _insult_ in Incompetenceville, which, by the way, is soon going
  to be bombed out of existence because you don't want to cooperate with
  code inspectors.

  you know, I can sort of understand how your mind works (or not) and how
  you come to your outlandish conclusions based on this valuable insight
  into the way you think (or not) about the world.  by this newspeak twist,
  you have effectively removed "better" from the language, and now we only
  have "bad code that works" and "better code that doesn't work".  isn't
  that's just the cutest hallmark of a "better mind" at work.  (that is,
  one that doesn't work, but which has replaced a bad one that did.  hm.
  that could be what psychiatry is about.)

  now, why am I _not_ surprised that you are a latter-day free software
  champion?  if I'm not mistaken, your belief system _centers_ around
  incompetent people, and that's why you made that derogatory remark about
  me having a hard time being surrounded by incompetents, too: that's how
  _your_ world would be if you had begun to look for competence, because
  you wouldn't know competence if it came up and bit you in the nose, but
  you _do_ know incompetence and how prevalent it is.

| While code is still actively developed a review is very useful.  I was
| talking about working software, and in particular, about code that is no
| longer the focus of development.

  review is even more important during the maintenance phase.  that's why I
  became very happy when my colleagues want to understand Lisp and I can
  explain things to them and see if something is so complex that it should
  be rewritten to be easier to understand.  maintenance is, after all,
  about getting back into the mindset of the code and make changes without
  fucking it up.  code review helps that process immensely.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Brad Knotwell
Subject: Re: help! absolute beginner
Date: 
Message-ID: <slrn77o1ae.he.brad@speedy.knotwell.dom>
>  third, free software is manifestly _not_ for everybody.  there are some 
>  rather interesting legal encumbrances on those who work with it. [1]

>  for instance, those who want to reimplement something into free software 
>  are strongly encouraged not to look at the other software, but have someone
>  else tell them what it does.  [2]

>  someone who wants to use the results from free software would likewise be 
>  restricted from actually looking at the code, if only to avoid the 
>  repercussions of the GPL tainting the work. [3]

>  that is, if you really expect to make money off of writing software that
>  somebody will want to keep to themselves, for whatever reason, you'd
>  better not be a free software champion. [4]

1)  I agree with this statement.  You need to understand the ramifications
    of the various licenses. . .GPL, LGPL, BSD. . .etc.

    Other than a coupla minor administrative requirements--"don't say you
    wrote it" and "give me credit as a third-party contributor", I don't
    really see the problem with the BSD or perl licensing schemes.  Similarly,
    given a bit more minor restrictions, I don't really see unworkable 
    problems with the LGPL.   

    In general, I believe use of BSDized, LGPL'd or artistically licensed
    code can be a good technical _and_ business decision.  Using GPL'd
    code would for a pretty tough business case to make except in a few 
    instances (see below).

2)  This seems (to me at least) to be an inherent requirement for free 
    software.  The requirement is driven by the need to release the source 
    code without limitation.  However, it's not a requirement unique only to 
    free software.  As an example, it's also a requirement for any commercial 
    developer wanting the freedom to do whatever they want with their 
    application.  In the case of a commercial company, it may be that they 
    don't wish to pay a run-time license fees for an implementation of a 
    "hot" methodology (the new Windows client protocol ICA? comes to mind). 

3)  Perhaps you could elaborate on this further, I don't really understand
    what you mean.

4)  Perhaps my reading of the GPL is incorrect, but the GPL doesn't require
    you make your source code available to everyone.  It only requires that
    you "You must cause any work that you distribute or publish,. . .to
    be licensed as a whole at no charge to all third parties under the
    terms of this License."  

    In my reading (IANAL), the last 9 words are key for the custom developer.  
    I read this as saying that you only need to give source and liberal 
    licensing terms to the third parties to whom you distributed the code. 
    Thus, it appears you could write custom software modifying GPL'd software 
    and give it to the client.  It would then be up to _them_ and _you_ how 
    it's distributed.  If it needs to be kept secret, it will only be used 
    within the company.  Assumption:  the company would enjoin employees from 
    releasing code to the outside world.
    
    The caveat to the above:  I only see it working with custom software 
    companies and companies who generate a large percentage of revenue from 
    support agreements, i.e. IBM.  Within the custom arena, the residual 
    revenue model then becomes one of charging a support cost based the 
    number of installations as well as charging for further follow-on work.  
    IOW, the model isn't that much different for custom companies than it 
    is now.  

Personally, I think the GPL requirements are usually too strict for companies 
designing products  However, I believe the LGPL restrictions are not a problem 
for many projects.  Furthermore, I believe that most commercial developers 
could easily work effectively and profitably work within the the BSD and 
artistic licenses.

>
>#:Erik
>-- 
>  Attention, Republican members of the House Judiciary Committee!  We have
>  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
>  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
>  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!

--Brad (ducking for cover)
From: Sudhir Shenoy
Subject: Re: help! absolute beginner
Date: 
Message-ID: <sshenoy-2012981029480001@qtns04419.singnet.com.sg>
In article <··················@speedy.knotwell.dom>,
········@ix.netcom.com wrote:

> Personally, I think the GPL requirements are usually too strict for companies 
> designing products  However, I believe the LGPL restrictions are not a
problem 
> for many projects.  Furthermore, I believe that most commercial developers 
> could easily work effectively and profitably work within the the BSD and 
> artistic licenses.
> 

This argument has been flogged to death on gnu.misc.discuss. However, please
note that

1) If all software was released under the GPL, there would be no problem.

2) Since all software is not released under the GPL, there is a problem
w.r.t. the conflict between the alternative models of release of source.

3) The GPL is designed to prevent authors' from releasing source and
later revoking the rights to use it. This has happened in several cases
e.g. oo-browser in Emacs. It *cannot* happen if the source is GPLed.

4) The argument that you can arrive at a separate license with the
commercial software vendor and thereby get the source to modify it for
your own purposes is specious since you could (theoretically) do exactly
the same with the holder of a GPL copyright.

5) If one is worried about GPL 'tainted' (aka GPL 'virus') software, one
is at liberty (free) not to use it. This is identical to the situation
with commercial software (not available) or commercial software source
obtained with an NDA (non-disclosure ...). Although this sounds arrogant
it is precisely the *same* argument which is employed wherever there is
a concept of "intellectual property" - that of ownership. It also bears
mentioning that the GPL has been stated to be necessary only because we
live in an imperfect world where access to the source code is not auto-
matic even though you have purchased a binary. Note that the GPL does
not prevent you from using and modifying code. It only prevents you from
re-distributing that *same* code under NDAs.

It may also be worth stating the obvious - just like every programmer
has written a "hello, world" program, every commercial software vendor
has re-implemented hundreds of common functions simply in order to
avoid breaking licenses. If we think about the millions of hours of
wasted effort, there is very little to recommend it except for the
dubious educational value of reimplementing the wheel.

On a related note, this is probably why we have a constant stream of
newbie questions on comp.lang.lisp. There is no publicly available
code in Lisp which does useful common tasks. When I was learning Lisp,
I looked in vain for a regexp matcher. I finally wrote one and released
it (with no restrictions other than you couldn't sell it). There are
some serious bugs and it bears rewriting (I was a Lisp newbie when I
wrote it). However, *noone* actually has (although one person is trying
to fix the serious bugs). Had the code been GPLed I would be sure that
this was only due to the few Lisp programmers actually interested in
regular expressions. Since it isn't GPLed, some other uncharitable
conclusions can also be arrived at.

Cheers
Sudhir

·······@gol.com
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123161087997428@naggum.no>
* ·······@NOSPAMgol.com (Sudhir Shenoy)
| On a related note, this is probably why we have a constant stream of
| newbie questions on comp.lang.lisp.  There is no publicly available
| code in Lisp which does useful common tasks.

  if there were, the newbies would only ask different questions.

| When I was learning Lisp, I looked in vain for a regexp matcher.

  this is one reason why I'm using and favoring commercial Lisp systems.
  it was there for the picking in the previous release, and had a pretty
  nice interface.  it's included by default, now.  it's because you
  insisted on free software that you had to reinvent the wheel.  this
  happens a lot, and I find it immensely ironic.  also note that the whole
  free software plan _was_ to rewrite the wheel so it could be free.  when
  they can't keep up with the commercial world, and commercial users can
  get access to better source, it's time to reassess the goals and whether
  it is still worth pursuing them, or to declare _victory_ in that source
  is now more easily accessible, thanks in large part to user expectations
  that they should be and in large part to the complexity of software so it
  doesn't make sense to ask people to pay for finding bugs -- it makes more
  sense to reward them for fixing them for you.

  the world has changed since the free software movement started.  I think
  the free software movement should realize that people aren't as much
  their enemies as they used to be, and find ways to be less combative
  where there is actually grounds for cooperation.

  instead, we get "non-free software is bad" from some quarters who have
  learned _nothing_ about the purpose of the protest and how to measure the
  success thereof.  *sigh*

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Christopher R. Barry
Subject: Re: help! absolute beginner
Date: 
Message-ID: <871zlu4i4c.fsf@2xtreme.net>
Erik Naggum <····@naggum.no> writes:

> * ·······@NOSPAMgol.com (Sudhir Shenoy)
> | On a related note, this is probably why we have a constant stream of
> | newbie questions on comp.lang.lisp.  There is no publicly available
> | code in Lisp which does useful common tasks.
> 
>   if there were, the newbies would only ask different questions.
> 
> | When I was learning Lisp, I looked in vain for a regexp matcher.

The CMU AI repository has some public domain regexp code and a lot of
little code bits to do some common tasks. But it's pretty lame
compared to what the, say, perl community has got going for it. There
are a lot of people that put up useful code on their web pages to do
some things but you've got to invest some serious altavista time to
find them. A central repository or categorized links page to where
it's all at with brief discriptions would be really helpful.

Christopher
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123189967380187@naggum.no>
* ······@2xtreme.net (Christopher R. Barry)
| The CMU AI repository has some public domain regexp code and a lot of
| little code bits to do some common tasks.  But it's pretty lame compared
| to what the, say, perl community has got going for it.  There are a lot
| of people that put up useful code on their web pages to do some things
| but you've got to invest some serious altavista time to find them.  A
| central repository or categorized links page to where it's all at with
| brief discriptions would be really helpful.

  but why compare to Perl?  what is it that made Perl a "success" to make
  it worth comparing to?  why not look at MS-DOS?  there are literally
  millions of small utilities out there that _had_ to be created because
  the core system was so fantastically useless.  similarly, I think the
  Perl community _had_ to build a whole lot of "useful" stuff -- they had
  already decided to reinvent the wheel with parts from all other Unix
  wheels, so of course they'll need to do everything over again in their
  new syntax.  Perl is a "success" in this regard because people who want
  to use it have no choice.

  Common Lisp has a _lot_ of tools built in, and it isn't a reinvention of
  a wheel, so there's less need for this stuff to begin with.  I think
  that's a _good_ thing, actually.  Perl is also used for all sorts of
  stuff that Common Lisp is not, such as systems administration of wacky
  systems, parsing stupid log file formats, glueing together uncooperative
  software, etc.  none of this is the Common Lisp way.

  the question I ask people who ask me whether Common Lisp can do what Perl
  does, is: "do you use Perl for these taks because you want to or because
  you have to?"  very few admit to wanting to, and actually think Perl is
  good because it lets them do things they already think is useless and
  ugly fairly quickly.  so the followup question: "would you want to use a
  language in which you didn't have to do those tasks?" and the result is
  sometimes enlightenment on part of the Perl hacker.  not that this
  converts people or anything, but it tends to communicate my view that
  Perl is a cure that extends and prolongs the problem rather than fixing
  it.  "oh, but sometimes you can't fix it!", some Perl hackers exclaim,
  and that's probably right, so Perl has a place in the universe, after
  all, but that still doesn't mean one shouldn't try to fix them.

  e.g., instead of trying to live with the multitude of incompatible time
  formats in the Unix world (especially in logs), let's do something in
  Common Lisp that really deals with the situation.  I think ISO 8601 is a
  brilliant standard, despite its many abbreviated forms, so I have written
  a fairly extensive package that uses it as the basis of the input and
  output of time.  as long as the order year-month-day-hour-minute-second
  is maintained and all fields are numeric, this package can deal with it.
  (if you want something else, feel free to hack like Perl in the language
  of your choice.)  this is a package I have created for a client on my own
  time, so I can release it when I feel comfortable in its ability to deal
  with everything I need.  would you set up that central repository or
  categorized links page to where it's all at with brief discriptions in
  the meantime?

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Christopher R. Barry
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87ww3l3nu4.fsf@2xtreme.net>
Erik Naggum <····@naggum.no> writes:

>  would you set up that central repository or categorized links page to
>  where it's all at with brief discriptions in the meantime?

Well, if I was to do that it would be at somehost.2xtreme.net/~cbarry/
or however my ISP does it, but yes, I suppose at some point I could
start work on such a thing, if there would be genuine interest. I
think my ISP only gives me 50MB of space or so, so there is not much
possibility of local source mirroring, which I think would be best.

Christopher
From: Steve Gonedes
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m2ww3lbd1e.fsf@KludgeUnix.com>
······@2xtreme.net (Christopher R. Barry) writes:
 
< Well, if I was to do that it would be at somehost.2xtreme.net/~cbarry/
< or however my ISP does it, but yes, I suppose at some point I could
< start work on such a thing, if there would be genuine interest. I
< think my ISP only gives me 50MB of space or so, so there is not much
< possibility of local source mirroring, which I think would be best.

That sounds really neat. (I would have replyied via e-mail but it
doesn't seem to be working for me.)

Also, doen anyone have the url for that jpeg source? I seemed to have
lost it and can't find it will deja news. Thanks...
From: Sudhir Shenoy
Subject: Re: help! absolute beginner
Date: 
Message-ID: <sshenoy-2112981054000001@qtns00960.singnet.com.sg>
In article <················@naggum.no>, Erik Naggum <····@naggum.no> wrote:

> * ·······@NOSPAMgol.com (Sudhir Shenoy)
> | When I was learning Lisp, I looked in vain for a regexp matcher.
> 
>   this is one reason why I'm using and favoring commercial Lisp systems.
>   it was there for the picking in the previous release, and had a pretty
>   nice interface.  it's included by default, now.  it's because you
>   insisted on free software that you had to reinvent the wheel. 

The fallacy in your argument is that I have to choose one *particular*
commercial
system (I presume you are referring to ACL). I use Macintosh Common Lisp (the
only commercial system on the Macintosh) and sometimes CMUCL (on Solaris).
There is no option ...

Choosing one commercial Lisp over another just to get my hands on a regexp (or
any other) package surely isn't what you are recommending.

Cheers
Sudhir

·······@gol.com
From: Mike McDonald
Subject: Re: help! absolute beginner
Date: 
Message-ID: <75ktt4$lrc$1@spitting-spider.aracnet.com>
In article <························@qtns00960.singnet.com.sg>,
	·······@gol.com (Sudhir Shenoy) writes:
> In article <················@naggum.no>, Erik Naggum <····@naggum.no> wrote:
> 
>> * ·······@NOSPAMgol.com (Sudhir Shenoy)
>> | When I was learning Lisp, I looked in vain for a regexp matcher.
>> 
>>   this is one reason why I'm using and favoring commercial Lisp systems.
>>   it was there for the picking in the previous release, and had a pretty
>>   nice interface.  it's included by default, now.  it's because you
>>   insisted on free software that you had to reinvent the wheel. 
> 
> The fallacy in your argument is that I have to choose one *particular*
> commercial
> system (I presume you are referring to ACL). I use Macintosh Common Lisp (the
> only commercial system on the Macintosh) and sometimes CMUCL (on Solaris).
> There is no option ...
> 
> Choosing one commercial Lisp over another just to get my hands on a regexp (or
> any other) package surely isn't what you are recommending.

  Didn't we just have this discussion with this example?

  Mike McDonald
  ·······@mikemac.com
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123226396167055@naggum.no>
* ·······@gol.com (Sudhir Shenoy)
| The fallacy in your argument is that I have to choose one *particular*
| commercial system (I presume you are referring to ACL). I use Macintosh
| Common Lisp (the only commercial system on the Macintosh) and sometimes
| CMUCL (on Solaris).

  you really need to look up "fallacy" in a dictionary.  you appear to
  disagree (it is unclear precisely with what, though), but there is
  nothing inherently contradictory in the argument even if you don't feel
  like agreeing to it.

  the fallacy in your argument is that you have decided to assign blame
  before you know what harm was actually been done.  (yes, this _is_ an
  inherent flaw in your argument -- it's called "begging the question".)

| There is no option ...

  and why did you rule out "complain to vendor about missing functionality"?

  things just don't get better if you don't let the people who make them
  know what you would like them to be like.

| Choosing one commercial Lisp over another just to get my hands on a
| regexp (or any other) package surely isn't what you are recommending.

  where do you draw the line?  when _would_ you choose another commercial
  Common Lisp implementation?  I'd say you have a serious attitude problem
  in that you accept without protest that some product is the way it is.
  it's that attitude that let Bill Gates defraud the entire globe.  with
  just a little less complacent customers and just a little more awareness
  of actual customer needs, his scam would never have been possible.  not
  that I blame you personally for their criminal and immoral activities,
  but it's worth noting that by not getting your money's worth, _you_ are
  at fault for not upholding your end of the deal.

  sell me shitty software once, shame on you.
  sell me shitty software twice, shame on me.

#:Erik
-- 
  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123200405037716@naggum.no>
* ····@speedy.knotwell.dom (Brad Knotwell)
| 2)  This seems (to me at least) to be an inherent requirement for free
|     software.  The requirement is driven by the need to release the
|     source code without limitation.  However, it's not a requirement
|     unique only to free software.  As an example, it's also a requirement
|     for any commercial developer wanting the freedom to do whatever they
|     want with their application.  In the case of a commercial company, it
|     may be that they don't wish to pay a run-time license fees for an
|     implementation of a "hot" methodology (the new Windows client
|     protocol ICA? comes to mind).
| 
| 3)  Perhaps you could elaborate on this further, I don't really
|     understand what you mean.

  I think I would have to answer pretty much what you answered in 2) above.
  free software isn't, and you have to treat it like any other licensed
  software to remain at liberty to do whatever you want yourself.

| 4)  Perhaps my reading of the GPL is incorrect, but the GPL doesn't
|     require you make your source code available to everyone.

  this is correct.  I don't have to make it available to anyone.

|     In my reading (IANAL), the last 9 words are key for the custom
|     developer.  I read this as saying that you only need to give source
|     and liberal licensing terms to the third parties to whom you
|     distributed the code.

  this is correct.

|     Thus, it appears you could write custom software modifying GPL'd
|     software and give it to the client.  It would then be up to _them_
|     and _you_ how  it's distributed.

  this is incorrect.  it is no longer up to the author of GPL'ed code how
  it is distributed once it leaves the author's domain.  that is the key.

|     If it needs to be kept secret, it will only be used within the
|     company.  Assumption: the company would enjoin employees from
|     releasing code to the outside world.

  it is not quite resolved whether this is possible under the GPL.

| Furthermore, I believe that most commercial developers could easily work
| effectively and profitably work within the the BSD and artistic licenses.

  I would agree with that.

  I have asked my lawyer to architect a license agreement that I can use on
  source code that I plan to release without giving people the opportunity
  to distribute changes to it or make money off of it without my explicit
  cooperation.  I also want to know who is redistributing it so I can make
  sure that changes propagate properly.  (I'm sure he'll be happy about the
  extra income, but it may turn out to cost too much to give something away
  under reasonable terms.)

#:Erik
-- 
  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!
From: Hrvoje Niksic
Subject: Re: help! absolute beginner
Date: 
Message-ID: <kigyao6dtur.fsf@jagor.srce.hr>
Erik Naggum <····@naggum.no> writes:

>   contradictory statements?  the point is there are _a multitude_ of
>   means to get "access to whatever [you] need to do a good job"
>   besides an organized free software protest movement, and that the
>   criticism I levy against the movement is that it _doesn't_ achieve
>   that goal because they protest too much and reasonably
>   business-oriented people cannot use free software in their own
>   products, thereby creating pockets of resistance, instead of a
>   real movement.

Hmm.  If I follow your line of thought correctly, then you should
probably like the recent developments in "open source" community.  The
point of open source, according to their spokesmen, is to avoid what
you call "protesting too much".

>   however, the legal profession is _all_about_ getting into and out
>   of various agreements in exchange for money.  what does this say
>   about how the GPL is supposed to work in our society?

The GPL is supposed to change the society -- its creator hasn't kept
that a secret.  Is that why you call the FSF and their followers a
"protest movement"?

>   I have previously argued that free software denies us the opportunity to
>   use money as the currency of exchange of last resort.  that is, any two
>   people who disagree can resort to exchanging money, because most values
>   boil down to money sooner or later.  for free software, that is no longer
>   an option.  nobody can buy their way out of the GPL.
[...]
> | Agree in the sense that nobody should be forced to agree with, or use
> | the license of, the Free Software Foundation.
> 
>   then why can't they negotiate their own licenses?

It is my understanding that the owner of the code can always relicense
their code under a different license, thus making it eligible for the
recipient's purposes.  This should make it possible to negotiate your
own license, or buying your way out of the GPL.  If you are not the
owner of the code in question, then you cannot do these things -- but
doesn't the same apply to non-free software as well?
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123014884919222@naggum.no>
* Hrvoje Niksic <·······@srce.hr>
| The GPL is supposed to change the society -- its creator hasn't kept that
| a secret.  Is that why you call the FSF and their followers a "protest
| movement"?

  hey, I'm all for changing society, but there's fighting _for_ something
  and there's fighting _against_ something.  it's a hell of a lot easier to
  get people to agree on what they don't want and fight against it, and
  maybe you succeed in defeating it, but then what do you do?  that's a
  protest movement.  in contrast, if you fight _for_ something, you never
  really have that question.

  (the worst you can do to anyone is to fulfill all their dreams.)

| It is my understanding that the owner of the code can always relicense
| their code under a different license, thus making it eligible for the
| recipient's purposes.

  I'm concerned with what this would mean.  if there is a public license
  and a private license for the same code, the two can never really be
  merged, so there would have to be a very firm split at the time when the
  two different licenses were created, perhaps to be glued together by the
  owner of the source.  considering the fact that one of these are in the
  open, that must mean the other is precluded from any form of publication,
  and is also barred from adopting changes to the same source made under
  the GPL.  this is very different from what the case would be if the
  GPL'ed version did not exist.

  in the usual understanding of "relicense", such dependence is not an
  issue, since neither licenses grant anything to the entire public, but
  lets the owner retain control over it, and would presumably preclude
  conflicts of interests and conflicting licenses for the same people.

  it is, therefore, my understanding that once GPL'ed, it has to be
  retracted and only later developments may be relicensed.  I'm not sure
  how this would be done in practice and ensured to be fully available as a
  legal option.  I'd like to hear about people who have done this before I
  would speculate on how it would actually be done.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: rusty craine
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3675D070.A9786474@flash.net>
> * Gareth McCaughan <·····@dpmms.cam.ac.uk>
> | I'm puzzled.  I can see rabidly anti-commercial attitudes in the "free
> | software" movement, but not *recently adopted* ones.
>
>
> Erik wrote:
>    like the IRS has lost track of
>   financing the government, and government in general has lost track of
>   providing for people's safety.  in becoming institutions, something died.
>

I have tried to guess at the system the respondents work with and on in this
news group.  Some times it seems the view regrading software is seems very
narrow,  pointed to internet servers, small networks, and single work
stations.  In large hospital systems (the only systems I really know), the
systems are a large composite of 390's, dec's (main frame and alpha's),  rs
6000's, nt's and a few old netware net works, all hooked up in this one giant
house of cards.  In October I go to the "board" and ask for around 15 million
dallors to keep the house of cards running another year.  I  have little
understanding where free ware of anykind fits in this world.  For example we
have 3 $60,000 alpha's (hardware + software)  that do nothing but act as
gateways to the mainframe for our pc networks.  We have two $80,000 nt's that
our nothing but interface engines between different systems (lab, xray,
pharmacy, lifetime clinical record [db2 390], etc).  Kinda HL7 servers.

The venders behing these products are a great deal more important to us in the
long run than the sms (shit masquerading as software) they sale.  When the
house of cards blows down who ya gona call "gnu"?  I think in the enterprise
world it is about vendors not software....that is in my opinion.   To put it
another way if the lab system is down the doctor's don't get results, can't
treat patients in a timely manner.  The lab system...what does that mean.  Well
the lab system per say (dec VM), the interface engine (nt) from the lab system
to 390 db2  (where the results are stored), CICS thread that LU6.2 uses to
process a db2 request,  main frame gateways (aplpha)  to the pc nt net work
then to  nt ms  sql server( temp storage data the doctor has requested to view)
to nt network and VB program called "physican's view"  that actually displays
the patient data.  I don't see any place for freeware in a very comples system
like this.

Freeware may have it's place but maybe not in the real wrold, that is if the
above is anywhere close to the real world. (real world = where the money is
spent)

rusty
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122685146348197@naggum.no>
* rusty craine <········@flash.net>
| Freeware may have it's place but maybe not in the real wrold, that is if
| the above is anywhere close to the real world. (real world = where the
| money is spent)

  my current client is a financial news agency.  they have some annoying
  real-time and uptime requirements.  we run this system on two Intel 133
  MHz Pentium� machines that used to be disk servers and which were ready
  for generational garbage collection (i.e., old trash) because none of the
  Microsoft crap could run reasonably on them.  they run Linux 2.0.35 and
  Allegro CL 5.0.  no need to upgrade or buy expensive hardware (more of
  them are lying around), and all the money could be invested in software.
  so far, they've been doing great.

  I'd certainly rate this a real world application.  without Linux, we
  would have had to request a 50% larger initial budget for new hardware,
  which wouldn't have been a useful suggestion to the board at the time.

#:Erik
-------
� 52 bogomips, compared to 400 bogomips for my development machine.  sigh.
-- 
  man who cooks while hacking eats food that has died twice.
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <36781dcf.6493477@news.newsguy.com>
On 15 Dec 1998 04:32:26 +0000, Erik Naggum <····@naggum.no> claimed or
asked:

%   I'd certainly rate this a real world application.  without Linux, we
%   would have had to request a 50% larger initial budget for new hardware,
%   which wouldn't have been a useful suggestion to the board at the time.

I'm confused.  Didn't you just flame free software a few posts back?
Or did you just flame the idea that all software must be free?

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122774512682606@naggum.no>
* ········@david-steuber.com (David Steuber "The Interloper")
| I'm confused.  Didn't you just flame free software a few posts back?
| Or did you just flame the idea that all software must be free?

  I did neither.  pay attention.

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Christopher R. Barry
Subject: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <87d85kih7j.fsf_-_@2xtreme.net>
This is kinda off topic, but after all your talk Erik about your
involvement with the FSF and Emacs, and all the things you had to say
about what's wrong with Emacs, I'm kinda curious about what your "from
the _inside_ of the FSF" view of XEmacs is. In the XEmacs FAQ, they
mention:

 "There are currently irreconcilable differences in the views about
 technical, programming, design and organizational matters between RMS
 and the XEmacs development team which provide little hope for a merge
 to take place in the short-term future."

I can see in your X-Newsreader header that you're using Stallmacs
instead of XEmacs, and I'm wondering if this is just because you
prefer the FSF style mousing behavior or some similar detail or if
it's because you believe that Stallmacs offers some type of technical
superiority to XEmacs or is perhaps superior by virtue of not being
irreconcilably fubared (perhaps you have this view?).

I used to use Stallmacs whenever I wanted a text-mode or terminal
buffer until I realized that XEmacs actually has the (IMHO) superior
text mode to.

At any rate, I love your idea of a CL-Emacs which would enable one to
compile everything to native code for a speed advantage and also not
have to maintain an inferior (with respect to CL) and seperate
interpreter for an inferior dialect implemented in an inferior
language.

Just earlier this week tried out Hemlock, and it's a nice
little editor, but it only seems to have like 2 modes right now and
obviously no gnus or any other comparable packages, and ilisp seems
to be a better listener interface at the moment anyways (IMHO).

Christopher
From: Erik Naggum
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <3122837945282209@naggum.no>
* ······@2xtreme.net (Christopher R. Barry)
| This is kinda off topic, but after all your talk Erik about your
| involvement with the FSF and Emacs, and all the things you had to say
| about what's wrong with Emacs, I'm kinda curious about what your "from
| the _inside_ of the FSF" view of XEmacs is.

  well, nobody cares.  XEmacs is a bunch of hateful lunatics who appear to
  take pleasure in attacking Emacs users and maintainers for no good reason
  yet aren't anywhere near as good at what they're doing as the Emacs
  people are (except in the MULE division, where XEmacs got it right and
  the MULE team in Emacs act as if they think that because Japan suffered
  exclusion for a lot of years, now Europe can suffer, instead).

| I can see in your X-Newsreader header that you're using Stallmacs instead
| of XEmacs,

  you can?  really?  "Stallmacs", even?  look, you should know by now that
  I hate idiots.  do me a favor and haul your ass over to Bagdad, OK?  but
  since you bring it up, look a little closer:

X-Newsreader: Gnus v5.5/Emacs 19.2003

  doesn't look much like "Stallmacs" to me.  I get the impression that you
  aren't very careful in collecting evidence before you jump to conclusions.
  I need some evidence that XEmacs proponents aren't _all_ fucking morons,
  and you're not it, so let me just can this off topic discussion right now.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Christopher R. Barry
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <871zlzvcws.fsf@2xtreme.net>
Damn Erik, mellow out. I guess you think using "Stallmacs" to refer
the the FSF maintained version of Emacs is somehow deragatory, and I
didn't know that you would take it in that sense, because I didn't
mean to attack it or anything. I still have a copy of it on my machine
and I've set up XEmacs to use a lot of the (IMHO) superior FSF
defaults and behavior. At least from my experience though, XEmacs is
kinda nicer in some ways, and that is all I really said. I didn't
specifically downtalk Emacs, because I don't think there is anything
to downtalk. You had plenty to downtalk about it though about how
asian languages support was implemented and you also downtalked the
XEmacs people saying that the FSF Emacs people are way better at what
the're doing, but didn't really give any (IMO) satisfying or unheated
reasons why. I just wanted to know your opinion on a few things, and I
got it, but I would have thought that since this is comp.lang.lisp it
would have been civil and unheated.

"I need some evidence that XEmacs proponents aren't _all_ [...], and
you're not it"

I'm not really an XEmacs proponent. I have both ACL 5.0 and CMUCL on
my machine here, and (until very recently) I used ACL 5.0 99% of the
time. I'm now using CMUCL slightly more, but have yet to really form
any opinions about really favoring one over the other. And if I did,
and said that I say, like CMUCL more, but not specifically downtalked
ACL 5, and also wanted your opinion as someone that has worked with
some of the ACL 5 sources, would that have also made your blood boil?

Christopher
[posted with Emacs 20.2.2]
From: Erik Naggum
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <3122861705479499@naggum.no>
* ······@2xtreme.net (Christopher R. Barry)
| I guess you think using "Stallmacs" to refer the the FSF maintained
| version of Emacs is somehow deragatory, and I didn't know that you would
| take it in that sense, because I didn't mean to attack it or anything.

  then why not call it by its name?  notice that the problems started when
  the XEmacs crowd called their Emacs version just that.  users everywhere
  have serious problems talking about Emacs as opposed to XEmacs because of
  _their_ name.  so those who think XEmacs is entitled to usurp the name
  space choose some other means to refer to Emacs, such as "Stallmacs",
  which I cannot fathom how you can even believe is _not_ derogatory --
  think about how you came up with it, why you cannot call Emacs by its
  name, and what the real problem is.  OK?

  and if you know about the history of XEmacs and Emacs at all, you would
  _know_ that the stupid name collision is one of the contentious points.
  and of course I assume that people aren't completely ignorant when they
  ask about contentious issues.

| I'm not really an XEmacs proponent.

  well, _excuse_ me for thinking that one looks, walks, and quacks like a
  stupid XEmacs proponent is one.  my patience with that kind was exhausted
  years ago.  there's a reason I don't talk about XEmacs.  deal with it.

| I have both ACL 5.0 and CMUCL on my machine here, and (until very
| recently) I used ACL 5.0 99% of the time. I'm now using CMUCL slightly
| more, but have yet to really form any opinions about really favoring one
| over the other. And if I did, and said that I say, like CMUCL more, but
| not specifically downtalked ACL 5, and also wanted your opinion as
| someone that has worked with some of the ACL 5 sources, would that have
| also made your blood boil?

  if you found cause to use a derogatory name for either of them, sure.
  incidentally, I'm not aware of any comparable hostility between ACL and
  CMUCL that you seem to take for granted or ignore between Emacs and
  XEmacs -- you just _can't_ compare two things without understanding
  whether they have similar natures at all.

  BTW, XEmacs has another good thing going for it that I forgot besides
  getting MULE right (that is, the ability to compile without it): it is
  not scheduled for conversion to GUILE, which I consider to be another
  serious mistake in the FSF camp.  all the language version chaos that
  reigns in Emacs Lisp and especially between XEmacs Lisp and Emacs Lisp is
  not getting any better with a language that has serious growth problems
  built into it.  my first reason for wanting a Common Lisp Emacs was that
  it would have a stable language underneath it.  trying to keep a lot of
  Emacs customization working is a lot more work than it should be, and
  I'll wager a bet that this is _because_ it is free software and there
  aren't very many reasons to do the boring work of maintaining a stable
  and a development release and actual specifications people can trust.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: Barry Margolin
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <0qae2.55$pr1.5636@burlma1-snr1.gtei.net>
In article <················@naggum.no>, Erik Naggum  <····@naggum.no> wrote:
>* ······@2xtreme.net (Christopher R. Barry)
>| I guess you think using "Stallmacs" to refer the the FSF maintained
>| version of Emacs is somehow deragatory, and I didn't know that you would
>| take it in that sense, because I didn't mean to attack it or anything.
>
>  then why not call it by its name?  notice that the problems started when
>  the XEmacs crowd called their Emacs version just that.  users everywhere
>  have serious problems talking about Emacs as opposed to XEmacs because of
>  _their_ name.  so those who think XEmacs is entitled to usurp the name
>  space choose some other means to refer to Emacs, such as "Stallmacs",
>  which I cannot fathom how you can even believe is _not_ derogatory --
>  think about how you came up with it, why you cannot call Emacs by its
>  name, and what the real problem is.  OK?

Are you saying that the name "Emacs" refers to FSF Emacs, or that people
should say "FSF Emacs" or "GNU Emacs" rather than using a nickname like
"Stallmacs"?  If the former, I don't think that's right.  "Emacs" refers to
an entire family of editors, including: PDP-10 EMACS, Multics Emacs,
Gosling/Unipress Emacs, Micro Emacs, Zimmerman's Emacs, GNU Emacs, Xemacs,
and even Epsilon.

And how does the derivation of "Stallmacs" imply that it's derogatory?  It
seems to be an abbreviation of Stallman's Emacs, which is an accurate
description of GNU Emacs.  I admit that I'd never heard the term before
this thread, but I instantly understood it and didn't think it was intended
as an insult.  Do you think there's something wrong with being associated
with RMS?  Or that the people who came up with that term do?  Maybe it's
like the word "nigger" -- if a white person uses it, it's an insult, but
African Americans can use it among themselves with no such implications.

-- 
Barry Margolin, ······@bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Don't bother cc'ing followups to me.
From: Tim Bradshaw
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <ey3pv9iso5m.fsf@todday.aiai.ed.ac.uk>
* Barry Margolin wrote:

> Are you saying that the name "Emacs" refers to FSF Emacs, or that people
> should say "FSF Emacs" or "GNU Emacs" rather than using a nickname like
> "Stallmacs"?  If the former, I don't think that's right.  "Emacs" refers to
> an entire family of editors, including: PDP-10 EMACS, Multics Emacs,
> Gosling/Unipress Emacs, Micro Emacs, Zimmerman's Emacs, GNU Emacs, Xemacs,
> and even Epsilon.

If I remember correctly, RMS's point of view was that Lucid Emacs (as
it then was) was a GNU Emacs as well as the version looked after by
the FSF.  So he didn't like us (Lucid emacs people) calling the FSF
variant `FSF Emacs'.  It may have been that he was happy with `Lucid
GNU Emacs' and `FSF GNU Emacs', but I'm not sure.  He was pretty
difficult about the whole thing, IMO.  In any case the name change to
XEmacs was pretty unfortunate, and was, I think, because Sun &c who
were putting money into it then didn't want it called `Lucid'.  I
voted for Lambda Emacs (so the binary would still be lemacs...), but I
don't think anyone listened.  The ultimate irony is that Lucid then
went bankrupt, so it could have been called Lucid GNU Emacs after all.

(Perhaps this should be on an emacs newsgroup).

--tim
From: Hannu Koivisto
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <t2w90g6r8p0.fsf@lehtori.cc.tut.fi>
Barry Margolin <······@bbnplanet.com> writes:

| Are you saying that the name "Emacs" refers to FSF Emacs, or that people
| should say "FSF Emacs" or "GNU Emacs" rather than using a nickname like

Well, RMS once asked me (in private email) not to call Emacs
"FSF Emacs". I have understood from other discussions in various
Emacs newsgroups that the same wish applies to "GNU Emacs" too
(although he didn't say anything about that). Thus, I use Emacs
for Emacs and XEmacs for XEmacs. Simple.

| And how does the derivation of "Stallmacs" imply that it's derogatory?  It
| seems to be an abbreviation of Stallman's Emacs, which is an accurate
| description of GNU Emacs.  I admit that I'd never heard the term before
| this thread, but I instantly understood it and didn't think it was intended
| as an insult.  Do you think there's something wrong with being associated

Perhaps it wasn't intended that way, but, personally, I wouldn't
want to be on this planet when RMS hears Emacs being called
"Stallmacs" :) 

//Hannu
From: Hrvoje Niksic
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <kigaf0mticr.fsf@jagor.srce.hr>
Hannu Koivisto <·····@iki.fi.ns> writes:

> Barry Margolin <······@bbnplanet.com> writes:
> 
> | Are you saying that the name "Emacs" refers to FSF Emacs, or that people
> | should say "FSF Emacs" or "GNU Emacs" rather than using a nickname like
> 
> Well, RMS once asked me (in private email) not to call Emacs
> "FSF Emacs". I have understood from other discussions in various
> Emacs newsgroups that the same wish applies to "GNU Emacs" too
> (although he didn't say anything about that).

FWIW, Stallman uses the term ``GNU Emacs'' all the time.
From: Christopher R. Barry
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <87yao6ciw0.fsf@2xtreme.net>
Hrvoje Niksic <·······@srce.hr> writes:

> Hannu Koivisto <·····@iki.fi.ns> writes:
> 
> > Barry Margolin <······@bbnplanet.com> writes:
> > 
> > | Are you saying that the name "Emacs" refers to FSF Emacs, or that people
> > | should say "FSF Emacs" or "GNU Emacs" rather than using a nickname like
> > 
> > Well, RMS once asked me (in private email) not to call Emacs
> > "FSF Emacs". I have understood from other discussions in various
> > Emacs newsgroups that the same wish applies to "GNU Emacs" too
> > (although he didn't say anything about that).
> 
> FWIW, Stallman uses the term ``GNU Emacs'' all the time.

Also FWIW, when I start up Emacs on my machine it displays:

  Welcome to GNU Emacs, one component of a Linux-based GNU system.
  [newbie commands snipped]

  GNU Emacs 20.2.2 (i386-debian-linux-gnu, X toolkit)
   of The Jul 16 1998 on raven
  Copyright (C) 1997 Free Software Foundation, Inc.

  GNU Emacs comes with ABSOLUTELY NO WARRANTY; [rest snipped]

Christopher
From: William Paul Vrotney
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <vrotneyF43sor.5JM@netcom.com>
In article <··············@2xtreme.net> ······@2xtreme.net (Christopher
R. Barry) writes:
> 
> "I need some evidence that XEmacs proponents aren't _all_ [...], and
> you're not it"
> 

Well, recently I have been developing an Elisp package and more than once
have had to rig some of the perfectly reasonable code just so that it
would work for our XEmacs users.  That is to say that XEmacs clearly does
not work according to specification.





-- 

William P. Vrotney - ·······@netcom.com
From: Barry Margolin
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <_rae2.56$pr1.5636@burlma1-snr1.gtei.net>
In article <·················@netcom.com>,
William Paul Vrotney <·······@netcom.com> wrote:
>Well, recently I have been developing an Elisp package and more than once
>have had to rig some of the perfectly reasonable code just so that it
>would work for our XEmacs users.  That is to say that XEmacs clearly does
>not work according to specification.

There's an Emacs specification that's intended to apply across different
implementations?  Since when?

-- 
Barry Margolin, ······@bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Don't bother cc'ing followups to me.
From: Mike McDonald
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <75bi3u$m99$1@spitting-spider.aracnet.com>
In article <·················@burlma1-snr1.gtei.net>,
	Barry Margolin <······@bbnplanet.com> writes:
> In article <·················@netcom.com>,
> William Paul Vrotney <·······@netcom.com> wrote:
>>Well, recently I have been developing an Elisp package and more than once
>>have had to rig some of the perfectly reasonable code just so that it
>>would work for our XEmacs users.  That is to say that XEmacs clearly does
>>not work according to specification.
> 
> There's an Emacs specification that's intended to apply across different
> implementations?  Since when?

  I thought the spec was Zmacs. :-)

  Mike McDonald
  ·······@mikemac.com
From: William Paul Vrotney
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <vrotneyF45DwH.n2p@netcom.com>
In article <·················@burlma1-snr1.gtei.net> Barry Margolin
<······@bbnplanet.com> writes:

> 
> In article <·················@netcom.com>,
> William Paul Vrotney <·······@netcom.com> wrote:
> >Well, recently I have been developing an Elisp package and more than once
> >have had to rig some of the perfectly reasonable code just so that it
> >would work for our XEmacs users.  That is to say that XEmacs clearly does
> >not work according to specification.
> 
> There's an Emacs specification that's intended to apply across different
> implementations?  Since when?
> 

I consider that a quibble.  The Elisp info nodes represent a specification
as to what the basic Elisp functions should do.  I'm sure and I've read that
the Xemacs folks strive to be compatible with Elisp in this sense.  If you
then want to quibble that Xemacs has functionality that GNU Elisp does not
have and so on then I'll bow out, since I find such argument futile.  Xemacs
implementors have to decide whether they want Xemacs to be compatible with
GNU Emacs and to what degree and then live with that decision.  When Xemacs
has advertised that it is compatible with GNU Emacs the intended
specification that applies is implicit.  And this was since the beginning of
Lucid Emacs.

-- 

William P. Vrotney - ·······@netcom.com
From: Hrvoje Niksic
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <kigd85hepwh.fsf@jagor.srce.hr>
·······@netcom.com (William Paul Vrotney) writes:

> In article <·················@burlma1-snr1.gtei.net> Barry Margolin
> <······@bbnplanet.com> writes:
> 
> > 
> > In article <·················@netcom.com>,
> > William Paul Vrotney <·······@netcom.com> wrote:
> > >Well, recently I have been developing an Elisp package and more than once
> > >have had to rig some of the perfectly reasonable code just so that it
> > >would work for our XEmacs users.  That is to say that XEmacs clearly does
> > >not work according to specification.
> > 
> > There's an Emacs specification that's intended to apply across different
> > implementations?  Since when?
> > 
> 
> I consider that a quibble.

I don't think Barry was quibbling; your mentioning a "specification"
was indeed a bit ambiguous, although I got your meaning.

> The Elisp info nodes represent a specification as to what the basic
> Elisp functions should do.  I'm sure and I've read that the Xemacs
> folks strive to be compatible with Elisp in this sense.

Yes.  The incompatibilities are in the area of more esoteric features, 
or fairly recent ones.
From: William Paul Vrotney
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <vrotneyF46s9n.B6z@netcom.com>
In article <···············@jagor.srce.hr> Hrvoje Niksic <·······@srce.hr>
writes:

> 
> ·······@netcom.com (William Paul Vrotney) writes:
> 
> > In article <·················@burlma1-snr1.gtei.net> Barry Margolin
> > <······@bbnplanet.com> writes:
> > 
> > > 
> > > In article <·················@netcom.com>,
> > > William Paul Vrotney <·······@netcom.com> wrote:
> > > >Well, recently I have been developing an Elisp package and more than once
> > > >have had to rig some of the perfectly reasonable code just so that it
> > > >would work for our XEmacs users.  That is to say that XEmacs clearly does
> > > >not work according to specification.
> > > 
> > > There's an Emacs specification that's intended to apply across different
> > > implementations?  Since when?
> > > 
> > 
> > I consider that a quibble.
> 
> I don't think Barry was quibbling; your mentioning a "specification"
> was indeed a bit ambiguous, although I got your meaning.
> 

I'm happy that you agree and got my meaning, but I just wanted to say a few
words about the ambiguity of the term "specification".

From my training and reading in the science of software life cycles every
program has an implicit requirements, specification and implementation,
regardless if they are written or not.  The only question is whether they
are in fact written or how well they are written.  In cases of incomplete
specifications we have to make do with what we've got.  So if what I've
learned is correct then using the term "specification" to refer to that
property of any program regardless of how well it is written can never be
ambiguous from this theoretical perspective.  The notion of the existence,
intention, or ambiguity of a given written specification is a different
matter.

Sheesh, now I'm quibbling! :-)


-- 

William P. Vrotney - ·······@netcom.com
From: Chris Kirkwood-Watts
Subject: Re: Off topic: FSF, Emacs and XEmacs (was Re: help! absolute beginner)
Date: 
Message-ID: <k76g1adpcrw.fsf@cna0790340.rsc.raytheon.com>
Christopher R. Barry <······@2xtreme.net> writes:
> I can see in your X-Newsreader header that you're using Stallmacs instead
> of XEmacs,

Erik Naggum <····@naggum.no> responds:
>   you can?  really?  "Stallmacs", even?  look, you should know by now that
>   I hate idiots.  do me a favor and haul your ass over to Bagdad, OK?  but
>   since you bring it up, look a little closer:
> 
> X-Newsreader: Gnus v5.5/Emacs 19.2003
> 
>   doesn't look much like "Stallmacs" to me.  I get the impression that you
>   aren't very careful in collecting evidence before you jump to conclusions.
>   I need some evidence that XEmacs proponents aren't _all_ fucking morons,
>   and you're not it, so let me just can this off topic discussion right now.
> 
> #:Erik

Since it is highly unlikely that anyone will actually allow be allowed to
"can" an off topic discussion, and lest we continue to talk past one
another, I'll quote from http://sourcery.naggum.no/emacs/ for those who are
too lazy to figure it out for themselves:

    If Emacs 20.3 doesn't work for users of true 8-bit character sets, I'll
    publish my internal, MULE-free version, dubbed Emacs 19.2003.

Chris.

-- 
Chris Kirkwood-Watts
········@ti.com
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <36796f94.61956298@news.newsguy.com>
On 16 Dec 1998 05:21:52 +0000, Erik Naggum <····@naggum.no> claimed or
asked:

% * ········@david-steuber.com (David Steuber "The Interloper")
% | I'm confused.  Didn't you just flame free software a few posts back?
% | Or did you just flame the idea that all software must be free?
% 
%   I did neither.  pay attention.

I thought I was paying attention.  Perhaps you weren't being clear?

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122937011752312@naggum.no>
* ········@david-steuber.com (David Steuber "The Interloper")
| On 16 Dec 1998 05:21:52 +0000, Erik Naggum <····@naggum.no> claimed or
| asked:
| 
| % * ········@david-steuber.com (David Steuber "The Interloper")
| % | I'm confused.  Didn't you just flame free software a few posts back?
| % | Or did you just flame the idea that all software must be free?
| % 
| %   I did neither.  pay attention.
| 
| I thought I was paying attention.  Perhaps you weren't being clear?

  excuse me?  you're presenting some horribly contorted confusions as a
  means of telling _me_ I'm not being clear?  really, now.

#:Erik
-- 
  Attention, Republican members of the House Judiciary Committee!  We have
  intercepted a coded transmission from Bill Clinton to Saddam Hussein that
  puts your life in jeopardy.  Clinton is prepared to cease fire if all of
  you are killed by Iraqi terrorists, whom he won't prosecute.  Be warned!
From: David Steuber "The Interloper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <36805ec9.2134308@news.newsguy.com>
On 18 Dec 1998 02:30:11 +0000, Erik Naggum <····@naggum.no> claimed or
asked:

% * ········@david-steuber.com (David Steuber "The Interloper")
% | On 16 Dec 1998 05:21:52 +0000, Erik Naggum <····@naggum.no> claimed or
% | asked:
% | 
% | % * ········@david-steuber.com (David Steuber "The Interloper")
% | % | I'm confused.  Didn't you just flame free software a few posts back?
% | % | Or did you just flame the idea that all software must be free?
% | % 
% | %   I did neither.  pay attention.
% | 
% | I thought I was paying attention.  Perhaps you weren't being clear?
% 
%   excuse me?  you're presenting some horribly contorted confusions as a
%   means of telling _me_ I'm not being clear?  really, now.

Oh how silly of me!  If you talk with your mouth full, I should have
no difficulty understanding you.

Hmmm.  This is rather a late response.  I wonder if I should even be
posting it, as meaningless as it is?  I'll leave that to the
philosophers.

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3123396410441615@naggum.no>
* ········@david-steuber.com (David Steuber "The Interloper")
| Oh how silly of me!  If you talk with your mouth full, I should have
| no difficulty understanding you.

  it's interesting to notice that you aren't even able to recognize that
  the two choices you gave me are fantastically hostile and don't even
  recognize when you're being slapped for it.  merry christmas.

#:Erik
-- 
  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!  Nie wieder KrF!
From: Marco Antoniotti
Subject: Re: help! absolute beginner
Date: 
Message-ID: <lw67bdlr46.fsf@copernico.parades.rm.cnr.it>
rusty craine <········@flash.net> writes:

> The venders behing these products are a great deal more important to us in the
> long run than the sms (shit masquerading as software) they sale.  When the
> house of cards blows down who ya gona call "gnu"?

AFAIK, you can call Cygnus.

    ...
> Freeware may have it's place but maybe not in the real wrold, that is if the
> above is anywhere close to the real world. (real world = where the money is
> spent)
> 

Not that I  do not sympathize with your concerns.  I believe that you
are right when you point out all these problems you have dealing with
a very large heterogenous network, where a lot of the system is
being maintained by largish scale software/hardware vendors.

However, I also believe that your conclusion does not hold. A Linux
system *is* very reliable and you can set up a business relationship
with a software maintainance house, pretty much in the same way you
set up your business relationship with the DB2 software house.  Since
you do not sue your vendors (of course there is always the evenience
of a major mishap), and instead you ask them to fix the problems, the
underlying software used or provided makes a difference only in you
bottom line. Most likely, since free software is most of the time near
gratis, your bottom line will also improve.  Not bound to happen, but
it may.

Cheers

-- 
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - (0)6 - 68 10 03 17, fax. +39 - (0)6 - 68 80 79 26
http://www.parades.rm.cnr.it
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3empxjgrt.fsf@todday.aiai.ed.ac.uk>
* rusty craine wrote:

> The venders behing these products are a great deal more important to
> us in the long run than the sms (shit masquerading as software) they
> sale.  When the house of cards blows down who ya gona call "gnu"?  I
> think in the enterprise world it is about vendors not
> software....that is in my opinion.  

I think that's pretty much right.  What you need, if you are serious,
is someone you can pay money, and then when you call them and say
`this is broken' either they fix it, or, ultimately, they start paying
*you* money to cover your losses.  And that means all sorts of things
-- for instance they have to be stable enough that you're reasonably
sure that they won't just go bankrupt when you try this.  But it's the
*vendor* you sign the contract with not the package.  It's buying
insurance really.

But that's not *necessarily* incompatible with free software.  If,
say, Sun were to start offering the kind of support for Linux on sparc
that they do for Solaris on sparc now then why would Solaris be better
than Linux?  I think it would then come down to a choice of features.
Similarly some other company could start offering support for any
random free package: I bet you can buy serious support for Apache now
(don't IBM support it or something?), and you can get support for
sendmail or bind (from almost any Unix vendor!).

And actually, for big companies, free software (for suitable
definition fo `free') might be preferable.  For instance it might at
some point become better to hire your own support people for Linux and
stop paying money to the vendor.  With free software you have the
option of insuring yourself this way too.  That doesn't really work
for small companies.

I think there are some interesting cases in the middle.  For instance,
if I'm a small company (a startup, say), and I want to sell my stuff
to very large, serious people, then they're going to say `you aren't
big enough: if something bad happens you'll just go bankrupt and walk
away and we'll be left with broken software we can't maintain' (at
least, that's what *I'd* say if I was a big serious person).  So maybe
you want to make source available, or make some kind of promise to
make source freely available if you do go down the tubes, and promise
to write in very standard languages like CL, I don't know

What is *really* mysterious, I think, is commercial people buying
commercial software where you have no realistic hope of support -- for
instance Windows 95/98.  Free software is trivially better than that.

--tim
From: rusty craine
Subject: Re: help! absolute beginner
Date: 
Message-ID: <367AE081.F8EA5265@flash.net>
Tim Bradshaw wrote:

> [snip]
> definition fo `free') might be preferable.  For instance it might at
> some point become better to hire your own support people for Linux and
> stop paying money to the vendor.  With free software you have the
> option of insuring yourself this way too.  That doesn't really work
> for small companies......
>

You stated the stituation much better than I did.  I would make one point
about hiring support people, as programmers or systems.  The company hires
a bright young go-getter,  spends several thousand dollars on farther
education to get the new hire up to speed.  Lets say for this example the
new programmer is hired with a starting salary of $40K/yr.  After the
company has invested a couple of years in training, mistakes, education,
our new hire can now advance his salary quicker by take a new job outside
your company than he can by staying with our company.  This is because
most companies have some type of evaluation process that makes sure raise
stay in an expected range.

On the other hand if you hire a hot shot fully fledged applicant at an
appropriate salary that indeed proves to be all you expected, he will (in
our shop) be quickly moved to management.

Thus is the life in a shop, where the first order of business is to keep
vender soft ware running and the second order of business is to produce
code to patch everything together.  Ya have management writting code and
making very valuable empolyees for someone elese

What I think would be a great deal is if some of the newsgroup
participant. could consults, write code, patch systems, etc remotely.  We
certainly have all the necessary hardware and software in place to do it.
I'm not sure how our legal department would look upon such an arrangment
and it might be a jolt for accounts payable to pay a consultant from who
knows where that never set foot in our shop. I'm sure we have some lisp
consultants in this group.  Have you seen any arrangement like this.  If
so how did they work out for the mutual benefit?   Geeez just think of
it...get the best lispes around and not have to pay travel and on the
other side get consultant money and never leave your study at home.
hmmmmm sounds reasonable.

Rusty.
From: Rob Warnock
Subject: Re: help! absolute beginner
Date: 
Message-ID: <75f63n$4hra6@fido.engr.sgi.com>
Tim Bradshaw  <···@aiai.ed.ac.uk> wrote:
+---------------
| I think there are some interesting cases in the middle.  For instance,
| if I'm a small company (a startup, say), and I want to sell my stuff
| to very large, serious people, then they're going to say `you aren't
| big enough: if something bad happens you'll just go bankrupt and walk
| away and we'll be left with broken software we can't maintain' (at
| least, that's what *I'd* say if I was a big serious person).  So maybe
| you want to make source available, or make some kind of promise to
| make source freely available if you do go down the tubes...
+---------------

This happens all the time, not only for software but also for hardware!
The magic word here is "escrow": The small vendor deposits archival
copies of their "source" (software, the build tools needed for it, hardware
schematics, Verilog source, PAL designs, everything) with a trusted-by-
both-sides 3rd-party escrow agent, who will release the source to the
customer if the vendor doesn't live up to certain conditions (e.g., doesn't
stay in business, doesn't keep manufacturing/selling/supporting the product,
whatever).

I worked for Digital Communications Assoc. in Atlanta many, many years ago
and when we were just starting out, several of our large customers required
us to do this for our whole product line. [No, we never defaulted, but if
we had, they would have been *somewhat* protected... maybe...]


-Rob

-----
Rob Warnock, 8L-855		····@sgi.com
Applied Networking		http://reality.sgi.com/rpw3/
Silicon Graphics, Inc.		Phone: 650-933-1673
2011 N. Shoreline Blvd.		FAX: 650-964-0811
Mountain View, CA  94043	PP-ASEL-IA
From: Steve Gonedes
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m2soeowj2a.fsf@KludgeUnix.com>
Joachim Achtzehnter <·······@kraut.bc.ca> writes:

< And from a user's perspective, you have to weigh the alternatives: put
< all eggs in one basket and accept that you depend totally on one
< vendor to fix things or extend things? Or have the freedom to hire any
< professional software developer to work on freely available source if

  ^^^^^^^^^^^^ ^^^^^^^^ ^^^^^^^^^

No such thing.
From: Paul Rudin
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m3zp8xjl3h.fsf@shodan.demon.co.uk>
······@lavielle.com (Rainer Joswig) writes:

> In article <··············@shodan.demon.co.uk>, Paul Rudin
> <·····@shodan.demon.co.uk> wrote:
> Depending on the definition of "free" the percentage varies.
> 
> > Linux is a more realiable OS than Windows.
> 
> In general or the individual C program? Does BIND crash
> less often or more often than a Windows NT DNS?

I mean the operating system kernel itself. Programs may fall over on
either platform, but on linux you can kill the offending program
and restart it. On windows you often need to turn off the machine and
start again. My experience is that in 7 years of running a number of
different linux distributions on several machines, the OS itself
has crashed once. OTOH rebooting a windows box because the OS has
fallen over happens every couple of days.

> 
> > Mozzilla is at least as good as Internet Explorer
> 
> I haven't seen a released open source Mozilla yet. Where is it?
> The thing in development was highly buggy last time I checked.

I agree that the latest open source releases are buggy. But the binary
only "netscape" releases are free in any case..
 

> > A little more on topic, (X)emacs/ilisp is the best lisp editor around
> > IMO.
> 
> I'm not using it as a Lisp editor. Well, the X-Windows UI
> is like a stone tied to the foot. But there are easier to use
> alternatives.

Which X windows UI? I've got about 7 or 8 installed on the machine I'm
sitting at just now; and there are plenty of others around. If you
like windows then something like KDE is not that dissimilar in a lot
of respects; and FVWM-95 is a deliberate attempt to emulate some of
the features of the W95 desktop.


"easier to use" is perhaps the key phrase here. The nice thing about
using emacs as an program editor is that you can get more or less the
functionality you want if you're prepared to put the effort in. This
may not be "easy", and the initial learning curve to get up to speed
with using emacs may also not be "easy". But in the long run you get
tools and techniques that you can use across different programming
languages (and other kinds of data).


(And incidentally, I'm reading this newsgroup and preparing this post
inside gnus, an emacs newsreader, and since I have X running on three
machines I have several emacs windows open on three separate monitors,
giving me various presentations of the thread I'm participating in;
the article I'm replying to and the article I'm typing. Each X server
has its own console, and yet the emacs windows are all under the same
emacs process so the underlying data it the same. There isn't a
commercial product available that I'm aware of that would allow
anything close to similar functionality.)


Don't get me wrong: I don't have anything against commercial software
(as of a few weeks ago I'm a full time common lisp programmer, for a
company that sells commercial software). There is also some great
commercial software out there (where would we be without Doom and its
imitators :-) ). But I do think your characterisation of free software
is wide of the mark.
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1012980105570001@194.163.195.67>
In article <··············@shodan.demon.co.uk>, Paul Rudin
<·····@shodan.demon.co.uk> wrote:

> has crashed once. OTOH rebooting a windows box because the OS has
> fallen over happens every couple of days.

whoever compares an OS in terms of stability with Windows
can't be really helped. I mean anybody with some self
respect doesn't do that. "Hey my Hyundai is faster
than your Trabant." Still it doesn't give you a Mercedes
(or whatever these cars will be called now. ;-) ).

> Which X windows UI?

Any of the stuff I have seen so far is not overwhelming.

> If you like windows

No. Not really. Maybe "hate" would more correctly describe
my relationship with Windows.

> then something like KDE is not that dissimilar in a lot
> of respects; and FVWM-95 is a deliberate attempt to emulate some of
> the features of the W95 desktop.

Which already was an emulation of the Mac UI, which already
took elements from Xerox UIs. Doesn't give you the same
sophistication of a Mac UI, though.

> (And incidentally, I'm reading this newsgroup and preparing this post
> inside gnus, an emacs newsreader, and since I have X running on three
> machines I have several emacs windows open on three separate monitors,
> giving me various presentations of the thread I'm participating in;
> the article I'm replying to and the article I'm typing. Each X server
> has its own console, and yet the emacs windows are all under the same
> emacs process so the underlying data it the same. There isn't a
> commercial product available that I'm aware of that would allow
> anything close to similar functionality.)

My Symbolics Lisp machine is accessible the same way. Yet, it provides
a complete environment.

-- 
http://www.lavielle.com/~joswig
From: Louis Glassy
Subject: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <74nr63$kbi$1@netra.msu.montana.edu>
Rainer Joswig <······@lavielle.com> wrote:
; In article <··············@shodan.demon.co.uk>, Paul Rudin
; <·····@shodan.demon.co.uk> wrote:

[Linux is more stable than MS Windows]

[Rainer replies]
; whoever compares an OS in terms of stability with Windows
; can't be really helped. I mean anybody with some self
; respect doesn't do that. [i.e. MS Windows is hopeless]

True. :-)

;> [Paul] Which X windows UI?

; [Rainer] Any of the stuff I have seen so far is not overwhelming.

[...]

;> [Paul says: I read news with Gnus/Emacs/X, and commercialware
;>  doesn't give similar functionality]

[Rainer replies- ]
; My Symbolics Lisp machine is accessible the same way. Yet, it provides
; a complete environment.

I've heard in various newsgroup conversations that LispM's were/are
the cat's meow [=Oregon dialect "the best"] of Lisp environments.  
Unfortunately, I only have access to very conventional 
Intel-based PC hardware.  Are there any software approximations 
to a LispM that run under a free Unixy OS?  

What is it in particular that makes LispM's special?

(Note- I'm not doubting that they're neat, I just have never used one,
so don't know why people who like them, do like them.)

Without having actually seen a LispM, I'm guessing they have 
superior integration of the various pieces (apps, hardware, OS),
relative to other Lisp-oriented software environments.

Be that as it may (and now the head of the LispOS dragon rears its
ugly head!) - suppose you wanted to develop a nice-to-use
Lisp-friendly environment, subject to the following (possibly
crazy) constraints-

* runs on Intel hardware (say, 486/66 or later, 8+ MB RAM)
* runs as a layer on top of a free Unix (*BSD, Linux)

Which apps would you like to see?  Alternatively, should 
functionality you would like to use be bundled in terms of 
what we now think of as "apps"?

If you do think dividing functionality into "apps" is proper,
then here's (roughly) my list... [not in order of importance]

* a way to read/send/manage email.
* a way to transfer data via ftp (upload & download).
* a way to log in to a remote machine via something like telnet.
* a way to read web stuff (e.g., a web browser).
* a way to manage files, or a better way of managing
  persistent data, if you know of one.
* a way to read/write/store/manipulate text
  (think "text editor").

and of course, I would like these all to be seamlessly 
integrated.  Of course.

<<Some people have told me before, Emacs/Guile/X/Scsh/foo already 
  do all this.  Why reinvent the wheel?  Well, it's a question
  of integration.  Of the commercial OS's+app bundles, the Mac
  stands out in my mind as an example of a good set of ideas,
  although [as Rainer points out] there are cracks in the plaster, 
  and some of the cracks run deep.>>
  
I've phrased the list of apps above
as "a way to" and not as "ftp client" etc
because maybe there is a better way to do things than what I'm
used to ("what I'm used to" = "Unix, in whatever flavor").

So my question for you is:  what apps would you like to see
in a Lisp-friendly environment built on top of a Unix?  and 
which would you like to see first?  (ie. could you put your
short list in order of importance?)

I'm not a kernel hacker, and while I appreciate 
(in some spiritual sense :-) 
the idea of a completely Lisp-based OS/VM, I don't 
see that implementing LispOS is within my current technical
ability or free-time constraints.  But if I had a decent 
roadmap, I'd like to try my hand at making some small but
useful pieces of the puzzle-- both as a learning exercise
(any excuse to program in Lisp is a good one) and also in 
the hopes of making something generally useful...
Kelly Murray had (has) some ideas on this.  Do others?

-Lou

-- 
Louis Glassy (······@cs.montana.edu)
Department of Computer Science
Montana State University
Bozeman, Montana 59717 USA
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1212981851040001@194.163.195.67>
In article <············@netra.msu.montana.edu>, Louis Glassy
<······@acheta.nervana.montana.edu> wrote:

> I've heard in various newsgroup conversations that LispM's were/are
> the cat's meow [=Oregon dialect "the best"] of Lisp environments.

In some respect they are still very advanced.
  
> Unfortunately, I only have access to very conventional 
> Intel-based PC hardware.  Are there any software approximations 
> to a LispM that run under a free Unixy OS?

Not that I heard of. Hmm, the old Xerox Interlisp environment
has been ported to PCs. I don't think it is still available.
 
> What is it in particular that makes LispM's special?

A complete environment written in object-oriented Lisp
with a multitude of tightly coupled tools.

> * a way to read/send/manage email.
> * a way to transfer data via ftp (upload & download).
> * a way to log in to a remote machine via something like telnet.
> * a way to read web stuff (e.g., a web browser).
> * a way to manage files, or a better way of managing
>   persistent data, if you know of one.
> * a way to read/write/store/manipulate text
>   (think "text editor").

mail server, DNS, namespace system, gui builder,
command interpreter, finder, ...

> So my question for you is:  what apps would you like to see
> in a Lisp-friendly environment built on top of a Unix?  and 
> which would you like to see first?  (ie. could you put your
> short list in order of importance?)

0) command interpreter
1) editor
2) gui builder (menus, icons,
3) diverse inspector, browser, debugger interfaces
4) finder
5) web browser
6) office environment (paint, draw, text, calc, ...)
7) multimedia environment (port sk8 then you're done)

-- 
http://www.lavielle.com/~joswig
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1212982032380001@194.163.195.67>
In article <·······················@194.163.195.67>, ······@lavielle.com
(Rainer Joswig) wrote:

> In article <············@netra.msu.montana.edu>, Louis Glassy
> <······@acheta.nervana.montana.edu> wrote:
> 
> > I've heard in various newsgroup conversations that LispM's were/are
> > the cat's meow [=Oregon dialect "the best"] of Lisp environments.
> 
> In some respect they are still very advanced.

To communicate the basic concepts of the Symbolics Genera
operating system I have converted the an extract
from the Genera Documentation to HTML. This is really
old stuff. Still interesting, though.

http://www.lavielle.com/~joswig/genera1/genera.html

Other Lisp stuff at: 
 http://www.lavielle.com/~joswig/lisp.html

-- 
http://www.lavielle.com/~joswig
From: Paolo Amoroso
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <3673b4bf.56481@news.mclink.it>
On 10 Dec 1998 06:53:55 GMT, Louis Glassy
<······@acheta.nervana.montana.edu> wrote:

> I've heard in various newsgroup conversations that LispM's were/are
> the cat's meow [=Oregon dialect "the best"] of Lisp environments.  
[...]
> What is it in particular that makes LispM's special?

Since I didn't have the luck of using such a machine, I was wondering the
same thing. I've started collecting some information. You may start at the
Online Symbolics Museum:

	http://home.brightware.com/~rwk/symbolics/

Symbolics Technology, Inc. is the company that still develops and sells the
Genera environment:

	http://stony-brook.scrc.symbolics.com/www/index.html

Symbolics Genera comes with a powerful document creation and management
system. It includes three tools: Concordia for authors, Document Examiner
for readers and one--I don't remember its name--for document designers.
Some info on the system is available at:

	http://www.win.tue.nl/2L670/static/concordia.html

You may also check the following papers:

- "Document Examiner: Delivery Interface for Hypertext Documents", J.H. 
  Walker; 1st ACM Conference on Hypertext, Nov. 1987, pagg. 307-323
- "Supporting document development with Concordia", J.H. Walker; IEEE 
  Computer 21:1, Jan. 1988, pagg. 48-59

Concordia, by the way, was used to write the excellent book
"Object-Oriented Programming in Common Lisp - A Programmer's Guide to CLOS"
by Sonya E. Keene (Addison-Wesley, 1989; ISBN 0-201-17589-4). The preface
includes a short description of Concordia in section "About this Book".

It would be _great_ if some Symbolics users posted to a Web/FTP site a few
(of course, feel free to post many ;-) screen shots showing the neat things
that can be done with their Genera environment. The Symbolics Technology
Web site offers only a couple of unimpressive login screen shots :)


Paolo
-- 
Paolo Amoroso <·······@mclink.it>
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1312981610100001@194.163.195.67>
In article <··············@news.mclink.it>, ·······@mclink.it (Paolo
Amoroso) wrote:

> Since I didn't have the luck of using such a machine, I was wondering the
> same thing. I've started collecting some information.

Overview stuff:
http://www.lavielle.com/~joswig/symbolic-computing.html
http://www.lavielle.com/~joswig/genera1/genera.html

Screen shots:
http://www.lavielle.com/~joswig/docex.html
http://www.lavielle.com/~joswig/listener/listener.html

More on the UI aspects:
http://kogs-www.informatik.uni-hamburg.de/~moeller/uims-clim/clim-intro.html


O.k., here a some more screen shots. I did these from an NXP1000 (a Lisp
machine without
graphics card) via X from my Mac at home.

The Debugger:
http://www.lavielle.com/~joswig/genera2/debugger.gif

An example for using the file system maintenance tool:
http://www.lavielle.com/~joswig/genera2/file-system-maintenance.gif

The Inspector:
http://www.lavielle.com/~joswig/genera2/inspector.gif

Creating layouts for applications frames:
http://www.lavielle.com/~joswig/genera2/layout.gif

Screen space divided between an editor (Zmacs) and a listener:
http://www.lavielle.com/~joswig/genera2/listener-zmacs.gif

Login Listener:
http://www.lavielle.com/~joswig/genera2/login.gif

The Namespace Editor as a interface to a database of users/hosts/networks/...
http://www.lavielle.com/~joswig/genera2/namespace-editor.gif

Browse the system's notifications:
http://www.lavielle.com/~joswig/genera2/notifications.gif

Peek gives overviews about system activities (processes, windows, routing
tables, ...)
http://www.lavielle.com/~joswig/genera2/peek.gif

The context sensitive UI (Dynamic Windows) needs special inspectors:
http://www.lavielle.com/~joswig/genera2/presentation-inspector.gif

Restoring CL-HTTP over the Internet from an FTP server directly into the
file system:
http://www.lavielle.com/~joswig/genera2/restore-distribution.gif

Applications can be invoked by pressing to keys:
http://www.lavielle.com/~joswig/genera2/select-key.gif

Simple tool to divide screen space:
http://www.lavielle.com/~joswig/genera2/split-screen.gif

The mail tool (Zmail):
http://www.lavielle.com/~joswig/genera2/zmail.gif


Have fun.

Greetings,

Rainer Joswig

-- 
http://www.lavielle.com/~joswig
From: Hannu Koivisto
Subject: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <t2www3vet83.fsf_-_@lehtori.cc.tut.fi>
······@lavielle.com (Rainer Joswig) writes:

| Overview stuff:
| http://www.lavielle.com/~joswig/symbolic-computing.html
| http://www.lavielle.com/~joswig/genera1/genera.html
| 
| Screen shots:
| http://www.lavielle.com/~joswig/docex.html
| http://www.lavielle.com/~joswig/listener/listener.html

*drool*

Especially the document examiner and the professionally written
documentation (stuff behind ...genera.html link) did impress me.
Looks like years ahead of things like Visual C++ IDE and its
documentation browser (no, I'm not using MS software but I've
seen the horrors on my friends' machines).

Does someone here in Finland happen to have a retired Symbolics
machine and want to make me the happiest kid of this Christmas? :)

Oh well, better stop dreaming and get back to hacking Elisp,
//Hannu
From: Steve Gonedes
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <m2zp8rv7jz.fsf@KludgeUnix.com>
Hannu Koivisto <·····@iki.fi> writes:

< Looks like years ahead of things like Visual C++ IDE and its
< documentation browser (no, I'm not using MS software but I've
< seen the horrors on my friends' machines).

Have you had a chance to see the new Visual C++? The reason I ask is
because I was just reading about it and some of it's features. It said
that you can re-define functions without having to re-compile and
re-link the _entire program_, but this will only work for a select few
functions and after 50 times you must recompile the complete program.
Was just wondering about it, sounded interesting. From the reading it
sounded like gdb. Some weird limitiations I thought; such as flat out
refusing to redefine some functions on the fly, the code cannot be
optimized and the debugger must be running etc...

It would sound like C/++ IDEs are gaining some of the same features
that lisp compilers have. I was wondering if all it's really doing is
caching some of the compilation to a file (there is a new file type
`.db' or something that the compiler will dump), or if it's like a
CVB++ interpreter? The idea of having a limit on the number of
modifcations seemed rather weird too, was just wondering as I haven't
ever had the opportunity to see MSVBC++ in action.
From: Pierre Mai
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <87soeizupc.fsf@orion.dent.isdn.cs.tu-berlin.de>
Steve Gonedes <········@worldnet.att.net> writes:

> Have you had a chance to see the new Visual C++? The reason I ask is
> because I was just reading about it and some of it's features. It said
> that you can re-define functions without having to re-compile and
> re-link the _entire program_, but this will only work for a select few
> functions and after 50 times you must recompile the complete program.

Well, IBM's new Visual Age for C++ is supposed to be able to do
incremental recompilation of single functions any number of times, by
keeping really detailed dependency information.  Only changes to
header files will necessitate larger recompilation it seems.  Also VA
for C++ is supposed to be the first compiler to support the new C++
standard fully (Want to bet? ;)...

> It would sound like C/++ IDEs are gaining some of the same features
> that lisp compilers have. I was wondering if all it's really doing is

Well, if I look at current mainstream trends, most seem to be trying
to implement Lisp features on the much more brittle base of current
languages...  See Java's introspection and inner classes, some of
CORBA's features, the current trend to use XML for limited object
persistency (along the lines of Lisp's read/write), CINT (a C++
interpreter Library for C++), etc.

Regs, Pierre.

[Don't ask me for details on C++ compilers, I just heard about Visual
Age in comp.compilers]

-- 
Pierre Mai <····@acm.org>               http://home.pages.de/~trillian/
  "One smaller motivation which, in part, stems from altruism is Microsoft-
   bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]
From: Paolo Amoroso
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <36789712.2487662@news.mclink.it>
On 14 Dec 1998 15:03:59 +0100, Pierre Mai <····@acm.org> wrote:

> Well, IBM's new Visual Age for C++ is supposed to be able to do
> incremental recompilation of single functions any number of times, by
> keeping really detailed dependency information.  Only changes to

The research work that led to that feature is illustrated in "CodeStore and
Incremental C++ - Why wait for slow builds?", by Lee R. Nackman (Dr. Dobb's
Journal, Dec. 1997; page 92).

The headline in the table of contents says: "IBM CodeStore technology
promises fast builds by taking the source code, the previous build's target
files, and any other state the system chooses to save to produce the
desired target files. For ``incremental'' builds such as these, build time
is proportional to the impact of the source code changed since the last
build".


Paolo
-- 
Paolo Amoroso <·······@mclink.it>
From: Hannu Koivisto
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <t2wpv9mef9e.fsf@lehtori.cc.tut.fi>
Steve Gonedes <········@worldnet.att.net> writes:

| Have you had a chance to see the new Visual C++? The reason I ask is
| because I was just reading about it and some of it's features. It said
| that you can re-define functions without having to re-compile and
| re-link the _entire program_, but this will only work for a select few
..
| optimized and the debugger must be running etc...

I have also heard about the feature you mention and the
limitation that the debugger must be running. I haven't seen
this in practise, though. I'll have to ask my colleagues to
demonstrate it when I'll see them next time. What I already have
seen and heard is that its code generator makes sometimes
absolutely insane code sequences and its C++ standard support
lacks features when compared to, for example, egcs. One
colleague of mine managed to get internal compiler errors with
certain templatized code. No response of any kind to his bug
reports. 

| It would sound like C/++ IDEs are gaining some of the same features
| that lisp compilers have. I was wondering if all it's really doing is

Yes, so it seems, like Pierre already mentioned.

I guess that's enough of C++ horrors for this time in c.l.l,
//Hannu
From: Paolo Amoroso
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <3675134b.2952838@news.mclink.it>
On 13 Dec 1998 21:29:16 +0200, Hannu Koivisto <·····@iki.fi> wrote:

> Does someone here in Finland happen to have a retired Symbolics
> machine and want to make me the happiest kid of this Christmas? :)

If you like the development environment and you have access to a Mac or an
Alpha workstation, you can probably purchase just Genera.


Paolo
-- 
Paolo Amoroso <·······@mclink.it>
From: Rainer Joswig
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <joswig-1412981504330001@pbg3.lavielle.com>
In article <················@news.mclink.it>, ·······@mclink.it (Paolo
Amoroso) wrote:

> On 13 Dec 1998 21:29:16 +0200, Hannu Koivisto <·····@iki.fi> wrote:
> 
> > Does someone here in Finland happen to have a retired Symbolics
> > machine and want to make me the happiest kid of this Christmas? :)
> 
> If you like the development environment and you have access to a Mac

For a Mac version you would need a 68k Mac with lots of slots
and a MacIvory (preferably model 2 or 3) Nubus card.

-- 
http://www.lavielle.com/~joswig
From: Paolo Amoroso
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <36754fd4.1858707@news.mclink.it>
On Mon, 14 Dec 1998 15:04:33 +0100, ······@lavielle.com (Rainer Joswig)
wrote:

> For a Mac version you would need a 68k Mac with lots of slots
> and a MacIvory (preferably model 2 or 3) Nubus card.

Will the card be available for PowerPC systems?


Paolo
-- 
Paolo Amoroso <·······@mclink.it>
From: Rainer Joswig
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <joswig-1512982346180001@194.163.195.67>
In article <················@news.mclink.it>, ·······@mclink.it (Paolo
Amoroso) wrote:

> On Mon, 14 Dec 1998 15:04:33 +0100, ······@lavielle.com (Rainer Joswig)
> wrote:
> 
> > For a Mac version you would need a 68k Mac with lots of slots
> > and a MacIvory (preferably model 2 or 3) Nubus card.
> 
> Will the card be available for PowerPC systems?

No. :-(

-- 
http://www.lavielle.com/~joswig
From: Mike McDonald
Subject: Re: Now I know what I want for a Christmas present :) (was: Re: Apps in Lisp)
Date: 
Message-ID: <753l2q$8jq$1@spitting-spider.aracnet.com>
In article <················@news.mclink.it>,
	·······@mclink.it (Paolo Amoroso) writes:
> On 13 Dec 1998 21:29:16 +0200, Hannu Koivisto <·····@iki.fi> wrote:
> 
>> Does someone here in Finland happen to have a retired Symbolics
>> machine and want to make me the happiest kid of this Christmas? :)
> 
> If you like the development environment and you have access to a Mac or an
> Alpha workstation, you can probably purchase just Genera.
> 
> 
> Paolo

  I wonder if you actually can. 18 months ago when I purchased my XL1201, they
were about out of 8.3 CDs. And since 8.5 was going to be released any day now,
Symbolics encouraged us to wait for that. I'm still waiting!

  Mike McDonald
  ·······@mikemac.com
From: Paolo Amoroso
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <36751de3.5665675@news.mclink.it>
The shots and the other material are very interesting. I'm studying them
carefully. I feel like an archaeologist discovering the wonders of a rich,
advanced and refined lost civilization. Thank you very much.


On Sun, 13 Dec 1998 16:10:10 +0100, ······@lavielle.com (Rainer Joswig)
wrote:

> http://www.lavielle.com/~joswig/listener/listener.html

From what I understand, the listener accepts a special-purpose command
language different from Lisp. Is this correct?


> Creating layouts for applications frames:
> http://www.lavielle.com/~joswig/genera2/layout.gif

Is this a sort of UI builder?


> Browse the system's notifications:
> http://www.lavielle.com/~joswig/genera2/notifications.gif

If I get things right, this is something similar to syslog under Unix.


Paolo
-- 
Paolo Amoroso <·······@mclink.it>
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1412981947370001@pbg3.lavielle.com>
In article <················@news.mclink.it>, ·······@mclink.it (Paolo
Amoroso) wrote:

> The shots and the other material are very interesting. I'm studying them
> carefully. I feel like an archaeologist discovering the wonders of a rich,
> advanced and refined lost civilization. Thank you very much.
> 
> 
> On Sun, 13 Dec 1998 16:10:10 +0100, ······@lavielle.com (Rainer Joswig)
> wrote:
> 
> > http://www.lavielle.com/~joswig/listener/listener.html
> 
> From what I understand, the listener accepts a special-purpose command
> language different from Lisp. Is this correct?

Same concepts available in CLIM.
Every application can incorporate a command interpreter.
The command interpreter may accept Lisp forms and/or
commands in the for "Show File foo:bar;test.lisp.3".
The document examiner for example has a command interpreter, too.
A command interpreter has a command table. A command table
can inherit commands from other command tables.
The parsing of the commands, completion, help, suggestions,
prompting, defaulting, building command dialogs, ...
is all built into the system and has not to be rewritten
in every application.

> > Creating layouts for applications frames:
> > http://www.lavielle.com/~joswig/genera2/layout.gif
> 
> Is this a sort of UI builder?

Kind of. You can specify the layout of your application
frame, some of its parameters, its pane and the panes
parameters. 

> > Browse the system's notifications:
> > http://www.lavielle.com/~joswig/genera2/notifications.gif
> 
> If I get things right, this is something similar to syslog under Unix.

Yep.

"Converse" is "talk". The mail server has a UI. The file server
has a UI. The tape dumper/reader has a UI. ...

-- 
http://www.lavielle.com/~joswig
From: Kelly Murray
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <3674733C.6A4FD9FC@IntelliMarket.Com>
Louis Glassy wrote:
>...
> So my question for you is:  what apps would you like to see
> in a Lisp-friendly environment built on top of a Unix?  and
> which would you like to see first?  (ie. could you put your
> short list in order of importance?)
> 
> I'm not a kernel hacker, and while I appreciate
> (in some spiritual sense :-)
> the idea of a completely Lisp-based OS/VM, I don't
> see that implementing LispOS is within my current technical
> ability or free-time constraints.  But if I had a decent
> roadmap, I'd like to try my hand at making some small but
> useful pieces of the puzzle-- both as a learning exercise
> (any excuse to program in Lisp is a good one) and also in
> the hopes of making something generally useful...
> Kelly Murray had (has) some ideas on this.  Do others?
> 

Guess the others are only mikemac and Rainer.. ?

>mikemac:
>Well, obviously, the first should be Erik's CL Emacs. :-) 
>You need an editor..
>..what you'd use for a windowing system? 

A editor written in lisp will be neede eventually, but that 
is hardly the place to start.  Been doing just by fine without it.

There is plenty to do that doesn't require a windowing system.

>Rainer:
> 0) command interpreter
> 1) editor
> 2) gui builder (menus, icons,
> 3) diverse inspector, browser, debugger interfaces
> 4) finder
> 5) web browser
> 6) office environment (paint, draw, text, calc, ...)
> 7) multimedia environment (port sk8 then you're done)

Good suggestions, but perhaps jobs too big to start with.

I put forth a lisp implementation of GZIP is just such a thing,
and a very useful utility that needs to written in lisp sooner
rather than later.  Same goes for a TAR utility.
And I'm still offering $500 for a JPEG utility, since Chris Vogt
excellent implementation can't be used commercially, at least not
for only $500..  Of course this utility is absolutely needed
to implement a web browser.

For one wants to "feel the bits" a TCP implemementation in Lisp
would be cool.

But really Louis (and anyone else), find something that interests
YOU, something you want to USE, and just do it.

-Kelly Murray   ···@intellimarket.com
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1412980418090001@194.163.195.67>
In article <·················@IntelliMarket.Com>, Kelly Murray
<···@IntelliMarket.Com> wrote:

> > Kelly Murray had (has) some ideas on this.  Do others?
> > 
> 
> Guess the others are only mikemac and Rainer.. ?

I hope not!

> A editor written in lisp will be neede eventually, but that 
> is hardly the place to start.  Been doing just by fine without it.

Currently we have the freely available Hemlock for X.

> There is plenty to do that doesn't require a windowing system.

You are right.

> I put forth a lisp implementation of GZIP is just such a thing,
> and a very useful utility that needs to written in lisp sooner
> rather than later.  Same goes for a TAR utility.
> And I'm still offering $500 for a JPEG utility, since Chris Vogt
> excellent implementation can't be used commercially, at least not
> for only $500..  Of course this utility is absolutely needed
> to implement a web browser.

Some base utils:
- Compression: gzip
- Encryption: ???
- Archives: tar
- Graphic: jpeg, png, gif, tiff, pict
- Translation and de/encoding: Unicode, base64, MIME, uuencode, ...
- Sound: WAV, MP3
- Video: MPEG

Internet Protocols/Utilities:
- IP
- TCP, UDP, ICMP
- DNS, LDAP
- Telnet, RSH, REXEC, FTP, HTTP, NNTP, SMTP, ESMPT, POP3, SMNP, IRC, ...

Graphics:
- bitmaps
- base primitives: color, points, lines, rects, circles, ...
- bitblt
- Transformations: rotation, scaling, ...
- fonts: bitmap fonts, scalable fonts, ...

...

-- 
http://www.lavielle.com/~joswig
From: Raymond Toy
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <4n7lvuvkvz.fsf@rtp.ericsson.se>
>>>>> "Rainer" == Rainer Joswig <······@lavielle.com> writes:

    Rainer> In article <·················@IntelliMarket.Com>, Kelly Murray
    Rainer> <···@IntelliMarket.Com> wrote:

    >> I put forth a lisp implementation of GZIP is just such a thing,
    >> and a very useful utility that needs to written in lisp sooner
    >> rather than later.  Same goes for a TAR utility.
    >> And I'm still offering $500 for a JPEG utility, since Chris Vogt
    >> excellent implementation can't be used commercially, at least not
    >> for only $500..  Of course this utility is absolutely needed
    >> to implement a web browser.

    Rainer> Some base utils:
    Rainer> - Compression: gzip
    Rainer> - Encryption: ???
    Rainer> - Archives: tar
    Rainer> - Graphic: jpeg, png, gif, tiff, pict
    Rainer> - Translation and de/encoding: Unicode, base64, MIME, uuencode, ...
    Rainer> - Sound: WAV, MP3
    Rainer> - Video: MPEG

Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp?  I
can see why graphics and translation would be useful in lisp, but
what's wrong with just running the external utilities for gzip, tar,
etc.?  Or are you expecting to manipulate these within lisp?

    Rainer> Internet Protocols/Utilities:
    Rainer> - IP
    Rainer> - TCP, UDP, ICMP
    Rainer> - DNS, LDAP
    Rainer> - Telnet, RSH, REXEC, FTP, HTTP, NNTP, SMTP, ESMPT, POP3, SMNP, IRC, ...

Forgive me for being slow, but what does it mean to put all of these
in lisp?

Ray
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1512982354370001@194.163.195.67>
In article <··············@rtp.ericsson.se>, Raymond Toy
<···@rtp.ericsson.se> wrote:

> Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp?

Yes, this is questionable. The ultimate goal would
be to have it on a Lisp machine in a sane way. But
we will not get it soon - just because there is
a lot of work needed for this software. Still I haven't seen
a really satisfying strategy to incorporate this stuff into
a Lisp environment.

>     Rainer> Internet Protocols/Utilities:
>     Rainer> - IP
>     Rainer> - TCP, UDP, ICMP
>     Rainer> - DNS, LDAP
>     Rainer> - Telnet, RSH, REXEC, FTP, HTTP, NNTP, SMTP, ESMPT, POP3,
SMNP, IRC, ...
> 
> Forgive me for being slow, but what does it mean to put all of these
> in lisp?
> 
> Ray

The Symbolics has IP, TCP, UDP, ICMP, DNS, Telnet, REXEC, FTP, HTTP,
NNTP client, SMTP, POP3 client, NFS, X, ...

I would want to have a real extensible object-oriented network computer
programmed, maintained, scripted in a high-level language.
*Easy* to understand and debug.

-- 
http://www.lavielle.com/~joswig
From: Raymond Toy
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <4n1zm0utdp.fsf@rtp.ericsson.se>
>>>>> "Rainer" == Rainer Joswig <······@lavielle.com> writes:

    Rainer> In article <··············@rtp.ericsson.se>, Raymond Toy
    Rainer> <···@rtp.ericsson.se> wrote:

    >> Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp?

    Rainer> Yes, this is questionable. The ultimate goal would
    Rainer> be to have it on a Lisp machine in a sane way. But
    Rainer> we will not get it soon - just because there is
    Rainer> a lot of work needed for this software. Still I haven't seen
    Rainer> a really satisfying strategy to incorporate this stuff into
    Rainer> a Lisp environment.

So, for gzip, you have a lisp function that takes a stream, say,
produces a gzipped version of the stream to another?  (Replace stream
with your favorite data type).

For writing a tar, you want a to hand it a list of files and it creates a tar
file?  For reading, you hand it a tar file and it lists or extracts
the contents?

I think a tar file handler would be fairly easy.  Emacs has a tar mode
for reading a tar file (format) so that would be a start.

Ray
From: Rainer Joswig
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <joswig-1612981549560001@pbg3.lavielle.com>
In article <··············@rtp.ericsson.se>, Raymond Toy
<···@rtp.ericsson.se> wrote:

> >>>>> "Rainer" == Rainer Joswig <······@lavielle.com> writes:
> 
>     Rainer> In article <··············@rtp.ericsson.se>, Raymond Toy
>     Rainer> <···@rtp.ericsson.se> wrote:
> 
>     >> Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp?
> 
>     Rainer> Yes, this is questionable. The ultimate goal would
>     Rainer> be to have it on a Lisp machine in a sane way. But
>     Rainer> we will not get it soon - just because there is
>     Rainer> a lot of work needed for this software. Still I haven't seen
>     Rainer> a really satisfying strategy to incorporate this stuff into
>     Rainer> a Lisp environment.
> 
> So, for gzip, you have a lisp function that takes a stream, say,
> produces a gzipped version of the stream to another?  (Replace stream
> with your favorite data type).
> 
> For writing a tar, you want a to hand it a list of files and it creates a tar
> file?  For reading, you hand it a tar file and it lists or extracts
> the contents?

Correct. If you have functionality you can use as black box, you
can easily define a wrapper (or a calling convention) and use
it in some way.

If you need to make it using internal data structures (say, a
stream buffer) you have on your Lisp side - then you are losing.

My question was how to generate the appropriate wrapper
automagically. Say, you have Quicktime as a library
and you want all the provided interfaces (data structure,
algorithms, ..) packaged via CLOS. The software may
or may not be written in an OO-style. But you may want
an easy to use interface and excellent performance.

-- 
http://www.lavielle.com/~joswig
From: Raymond Toy
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <4nww3sszxc.fsf@rtp.ericsson.se>
>>>>> "Rainer" == Rainer Joswig <······@lavielle.com> writes:

    Rainer> Correct. If you have functionality you can use as black box, you
    Rainer> can easily define a wrapper (or a calling convention) and use
    Rainer> it in some way.

Ok, after 30 minutes of hacking (plus some more to find out the format
of a tar file) I have a very, very simple tar file reader.  It
basically returns the tar header for each entry and the contents of
each entry.  The header is a Lisp structure with the sanitized header
information (filename, mod time, uid, gid, checksum, etc.).  The
contents is just a giant array of (unsigned-byte 8).  What would you
want to do with that?

A related question: Is the character type the same for all lisps
(8-bit unsigned byte, essentially)?  For portability I assumed that
this is not true, so I open the file with element-type (unsigned-byte
8) and massage that appropriately to get what I want.

    Rainer> If you need to make it using internal data structures (say, a
    Rainer> stream buffer) you have on your Lisp side - then you are losing.

I'm sorry, I didn't follow this last sentence.  What are you trying to 
say?

Ray
From: Erik Naggum
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <3122753560354984@naggum.no>
* Raymond Toy <···@rtp.ericsson.se>
| Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp?

  the key is "mind share".  when you have deal with a number of things in
  some clearly inferior language, such as C, you lose in a big way.  for
  instance, I'm currently struggling with MD5 checksums.  there's a C
  function I can call with a block of data, but the interface is all wrong
  and very hard to use from a Lisp point of view.  MD5 was obviously�
  designed in assembly language on an Intel processor, and I thik it would
  make sense to publish that code instead of the god-awful C code they did
  publish.  however, the MD5 algorithm itself, which works only on 512-bit
  blocks of data, doesn't need a support apparatus in C.  but because it
  was written in C, the support apparatus was also written in C.  _that's_
  what's wrong.  I can certainly live with a low-level algorithm in C or
  assembly or whatever, but I can't use it today because the interface was
  designed to be usable from C, not designed to be usable from any
  language.

| I can see why graphics and translation would be useful in lisp, but
| what's wrong with just running the external utilities for gzip, tar,
| etc.?

  that they are external (in the Unix process sense).

| Forgive me for being slow, but what does it mean to put all of these in
| lisp?

  leveraging the implementation experience so no new protocol needs to be
  written in C just because it uses stuff that has already been implemented
  in C.

  I'm all for using black boxes through foreign function interfaces, but
  when I actually need to do something that the people who thought in the
  inferior language couldn't think of, I have to do it my own way.  e.g.,
  there's a great domain name resolver out there, but, not surprisingly, it
  is written in C and has this stupid interface which you can't just give a
  list of requests to process in parallel, which really is quite easy given
  the protocol design, and return once it's done.  it's _too_ low-level,
  and it was written in the days of non-threading code that only needed to
  look up individual domain names at a time.  another simple example is the
  `ident' protocol, which is trivial, but which is also implemented in a
  stupid locking fashion.  I ended up writing an implementation for it in
  Allegro CL where I send the stream (socket) to a separate Lisp process as
  a run-reason, and it does its work and fills in a slot in the subclass of
  stream that keeps this information in a way that suspends the process if
  it requests the slot-value of this slot until it is known, which means
  the caller can predict whether it will block or not, and doesn't have to
  time out waiting for a machine that won't answer before it can service
  the request.  all very stream-lined, and not something you would have
  thought of in C in the first place.

  so, in brief, the differences in philosophy between Lisp and C make it
  quite important to have a Lisp handle on things.  sometimes, C is enough
  and the function is simple enough, but not very often, in my experience.

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Espen Vestre
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <w6ogp4wm2u.fsf@gromit.nextel.no>
Erik Naggum <····@naggum.no> writes:

>   the key is "mind share".  when you have deal with a number of things in
>   some clearly inferior language, such as C, you lose in a big way.  for
>   instance, I'm currently struggling with MD5 checksums.  

There's MD5 lisp code in the sources of CL-HTTP, check it out (I haven't
looked at it myself, but I know it's there).

-- 

  (espen)
From: Erik Naggum
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <3122814423567602@naggum.no>
* Espen Vestre <··@nextel.no>
| There's MD5 lisp code in the sources of CL-HTTP, check it out (I haven't
| looked at it myself, but I know it's there).

  I know it's there, too, but because of CL-HTTP's licensing terms (or,
  more precisely, the utter lack thereof), I can't use it, and would have
  some problems if I were to study it.

  incidentally, my MD5 implementation has been working in production mode
  since about 06:30 this morning -- I spent almost 8 hours debugging a
  stupid little typo on the initial context.  its only problem is that it's
  _really_ slow (800 �s CPU per block) and conses like mad because 32-bit
  numbers aren't fixnums.  I'll have to look into that.

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Kelly Murray
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <3676ADF7.6D4E04D9@IntelliMarket.Com>
Raymond Toy wrote:
> Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp? 

The simple answer is because everything needs to be in lisp
on a real lisp machine, which is the ultimate goal.
-k
From: William Paul Vrotney
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <vrotneyF41CGK.K1C@netcom.com>
In article <·················@IntelliMarket.Com> Kelly Murray
<···@IntelliMarket.Com> writes:

> 
> Raymond Toy wrote:
> > Can someone remind my why gzip, tar, MPEG, MP3 need to be in lisp? 
> 
> The simple answer is because everything needs to be in lisp
> on a real lisp machine, which is the ultimate goal.

Including the machine itself.  :-)



-- 

William P. Vrotney - ·······@netcom.com
From: Christopher C Stacy
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <x8l1zm1aw29.fsf@world.std.com>
In general, you might want those utilities to be modularized in a way
that they can be easily called by any programs on the system.

Having them as external callable program (of course you need some kind
of interprocess stream facility like pipes in the system) is one way
of doing that.

The Lisp Machine had an object-oriented (and easily extensible)
streams facility.  That notion is more important than whether the
terminal functions (such as eg., gunzip and tar) are written in
Lisp or not.  Possibly you would get moer flexible functionality
from them by rewriting them to be libraries (which you might as well
do in Lisp), but the important thing is to be able to hook them up.

It's nice to be able to do:
  (with-open-stream (tar (make-instance 'tar:stream :compression 9))
    (format s "Here is some compressed data...")
    (stream-copy-until-eof some-other-data-stream tar))

It's also easy to make a new kind of stream that incorporates the
TAR:STREAM functionality, but adds additional behaviour or whatever.
Maybe a better example would be a TAR stream that did not have the
compression feature built in, but you wrap around a GZIP stream.
Or you just mix the streams together into a new class.

By the way, the Lisp Machine had a tar and compression stream library;
this has all already been implemented.  I don't remember exactly what
the names of the things were or how they were modulrarized, but hopefully
you get the idea.
From: Simon Leinen
Subject: Re: Apps in Lisp (was: Re: help! absolute beginner)
Date: 
Message-ID: <aak8zp1ndy.fsf@limmat.switch.ch>
>>>>> "rj" == Rainer Joswig <······@lavielle.com> writes:
> Internet Protocols/Utilities:
> - IP
> - TCP, UDP, ICMP
> - DNS, LDAP
> - Telnet, RSH, REXEC, FTP, HTTP, NNTP, SMTP, ESMPT, POP3, SMNP, IRC, ...

A freely distributable SNMP implementation in Common Lisp is available
under http://www.switch.ch/misc/leinen/snmp/lisp/ .  We use it in
production under Allegro 5.0, but at some point in time it had run
under Genera, CMU CL, and earlier Allegro versions.  Since it includes
BER encoding/decoding (at least the subset that is required for SNMP),
it may be useful for implementing other protocols that are based on
ASN.1/BER.

(There's even an SNMP agent for Genera, so that you can integrate your
Lisp Machines into a network management system :-)

Have fun,
-- 
Simon Leinen				       ·····@babar.switch.ch
SWITCH				   http://www.switch.ch/misc/leinen/

	    Who is General Failure & why's he reading my disk?
From: ··········@scientia.com
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74o3mc$te1$1@nnrp1.dejanews.com>
In article <·······················@194.163.195.67>,
  ······@lavielle.com (Rainer Joswig) wrote:
> In article <··············@shodan.demon.co.uk>, Paul Rudin
> <·····@shodan.demon.co.uk> wrote:
>
> > has crashed once. OTOH rebooting a windows box because the OS has
> > fallen over happens every couple of days.
>
> whoever compares an OS in terms of stability with Windows
> can't be really helped. I mean anybody with some self
> respect doesn't do that. "Hey my Hyundai is faster
> than your Trabant." Still it doesn't give you a Mercedes
> (or whatever these cars will be called now. ;-) ).

well, OK, but the point is that Windows is commercial software and Linux is
free and in many repsects superior.

>
> > Which X windows UI?
>
> Any of the stuff I have seen so far is not overwhelming.

Don't be shy, name names :-)

> My Symbolics Lisp machine is accessible the same way. Yet, it provides
> a complete environment.
>

Sounds good, what does such a device cost?

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1112981439070001@pbg3.lavielle.com>
In article <············@nnrp1.dejanews.com>, ··········@scientia.com wrote:

> > > Which X windows UI?
> >
> > Any of the stuff I have seen so far is not overwhelming.
> 
> Don't be shy, name names :-)

Try to use a Mac for a while. You'll might see the difference.
 
> > My Symbolics Lisp machine is accessible the same way. Yet, it provides
> > a complete environment.

$5000 as an emulator for a DEC Alpha running Digital Unix.
Real hardware is no longer being produced, but
you might be able to buy a used system cheap.

-- 
http://www.lavielle.com/~joswig
From: Paul Rudin
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m3d85qitzp.fsf@shodan.demon.co.uk>
······@lavielle.com (Rainer Joswig) writes:

> In article <············@nnrp1.dejanews.com>, ··········@scientia.com wrote:

> 
> Try to use a Mac for a while. You'll might see the difference.

Oh, I have. I do see a difference and would chose unix and X every
time (although incidentally linux is now available on maC hardware I
believe).



>  
> > > My Symbolics Lisp machine is accessible the same way. Yet, it provides
> > > a complete environment.
> 
> $5000 as an emulator for a DEC Alpha running Digital Unix.
> Real hardware is no longer being produced, but
> you might be able to buy a used system cheap.


OTOH at home I have a three machine network, using hardware with a
value of maybe $2000 and free software (well, one of the machines is
dual boot W95 for playing games and for the kids..).


Sounds like your setup requires an alpha, digital unix, the symbolics
emulator, say $10,000 ? and you've still only got one display, and
very much less available software I'd bet. I'm sure it's a great lisp
environment; at that price it certainly should be.


I trust you see the irony in describing open source software as
unsupported, whilst at the same time advocating the use of a machine
that is now presumably unsupported? 


Even if one did buy the emulator I wouldn't have a lot of confidence
that it would continue to be supported indefinitely...
From: Simon Raahauge DeSantis
Subject: OT: UNIX on Mac
Date: 
Message-ID: <slrn772vd1.s00.xiamin@ghostpriest.rakis.net>
In article <··············@shodan.demon.co.uk>, Paul Rudin wrote:
>······@lavielle.com (Rainer Joswig) writes:
>
>> In article <············@nnrp1.dejanews.com>, ··········@scientia.com wrote:
>
>> 
>> Try to use a Mac for a while. You'll might see the difference.
>
>Oh, I have. I do see a difference and would chose unix and X every
>time (although incidentally linux is now available on maC hardware I
>believe).

There's also NetBSD, which works wonderfully on Mac68k (there's also a PPC
port, but I've never used it).

-- 
-Simon Raahauge DeSantis
From: Mike McDonald
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <74s290$ev$1@spitting-spider.aracnet.com>
In article <·····················@ghostpriest.rakis.net>,
	······@scdesantis.ne.mediaone.net (Simon Raahauge DeSantis) writes:
> In article <··············@shodan.demon.co.uk>, Paul Rudin wrote:
>>······@lavielle.com (Rainer Joswig) writes:
>>
>>> In article <············@nnrp1.dejanews.com>, ··········@scientia.com wrote:
>>
>>> 
>>> Try to use a Mac for a while. You'll might see the difference.
>>
>>Oh, I have. I do see a difference and would chose unix and X every
>>time (although incidentally linux is now available on maC hardware I
>>believe).
> 
> There's also NetBSD, which works wonderfully on Mac68k (there's also a PPC
> port, but I've never used it).
> 

  Does MacCL run under these versions of Unix?

  Mike McDonald
  ·······@mikemac.com
From: Rainer Joswig
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <joswig-1212980047150001@194.163.195.67>
In article <···········@spitting-spider.aracnet.com>, ·······@mikemac.com wrote:

> > There's also NetBSD, which works wonderfully on Mac68k (there's also a PPC
> > port, but I've never used it).
> > 
> 
>   Does MacCL run under these versions of Unix?

Macintosh Common Lisp only runs on MacOS.

-- 
http://www.lavielle.com/~joswig
From: Mike McDonald
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <74sjcv$505$2@spitting-spider.aracnet.com>
In article <·······················@194.163.195.67>,
	······@lavielle.com (Rainer Joswig) writes:
> In article <···········@spitting-spider.aracnet.com>, ·······@mikemac.com wrote:
> 
>> > There's also NetBSD, which works wonderfully on Mac68k (there's also a PPC
>> > port, but I've never used it).
>> > 
>> 
>>   Does MacCL run under these versions of Unix?
> 
> Macintosh Common Lisp only runs on MacOS.
> 

  Are you sure? Digitool may claim it only runs under MacOS but has anyone
verified that it doesn't under Linux or NetBSD? The reason I ask is that Linux
on other systems claims to run the native apps ala DEC Unix and Irix. I was
wondering whether MacLinux makes the same claim.

  Mike McDonald
  ·······@mikemac.com
From: David B. Lamkins
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <FLkc2.1405$9n4.791325@news.teleport.com>
In article <············@spitting-spider.aracnet.com> , ·······@mikemac.com
(Mike McDonald) wrote:

>In article <·······················@194.163.195.67>,
> ······@lavielle.com (Rainer Joswig) writes:
>> In article <···········@spitting-spider.aracnet.com>, ·······@mikemac.com wrote:
>> 
>>> > There's also NetBSD, which works wonderfully on Mac68k (there's also a PPC
>>> > port, but I've never used it).
>>> > 
>>> 
>>>   Does MacCL run under these versions of Unix?
>> 
>> Macintosh Common Lisp only runs on MacOS.
>> 
>
>  Are you sure? Digitool may claim it only runs under MacOS but has anyone
>verified that it doesn't under Linux or NetBSD? The reason I ask is that Linux
>on other systems claims to run the native apps ala DEC Unix and Irix. I was
>wondering whether MacLinux makes the same claim.

MCL makes calls to the Mac toolbox, and (maybe, in some versions)
manipulates the MMU directly.  AFAIK, there is no Mac OS emulator for either
NetBSD or mkLinux.

--
David B. Lamkins <http://www.teleport.com/~dlamkins/>

Recently undead Isabelle to the archangel Gabriel in "The Prophecy II": 
"So, you're keeping me alive because you don't know DOS?"
From: Christopher R. Barry
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <36723825.130D181@2xtreme.net>
David B. Lamkins wrote:
[...]
> AFAIK, there is no Mac OS emulator for either
> NetBSD or mkLinux.
> 

http://www.sheepshaver.com/
From: Georg Bauer
Subject: Re: OT: UNIX on Mac
Date: 
Message-ID: <gb-1212981322070001@hugo.westfalen.de>
In article <·····················@news.teleport.com>, "David B. Lamkins"
<········@teleport.com> wrote:

>manipulates the MMU directly.  AFAIK, there is no Mac OS emulator for either
>NetBSD or mkLinux.

One is in the works - SheepShaver. It runs under BeOS currently, but they
are working on a LinuxPPC version. Although I wasn't able to run MCL under
the BeOS version last time I tried (SheepShaver is a bit flaky).

bye, Georg

-- 
http://www.westfalen.de/hugo/
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1212980046340001@194.163.195.67>
In article <··············@shodan.demon.co.uk>, Paul Rudin
<·····@shodan.demon.co.uk> wrote:

> > Try to use a Mac for a while. You'll might see the difference.
> 
> Oh, I have. I do see a difference and would chose unix and X every
> time (although incidentally linux is now available on maC hardware I
> believe).

Try to look at "Inside Macintosh". You'll see that there are some
really great things inside Mac OS. Stuff that is clearly
unmatched on any platform (like the resource concept,
the excellent UI guide line, extensibility, internationalization,
great care to get UI details right, support for disabled
people, printing architecture, font architecture, Finder, 
the open scripting architecure, AppleEvents, ...). The UI also
looks much better compared to the depressing Windows/CDE/KDE/Motif UIs.
It is also very easy to use and to configure. There are
****many**** lessons to be learned from the Mac OS
(yes, I know that it also has problems in other areas).

> Sounds like your setup requires an alpha, digital unix, the symbolics
> emulator, say $10,000 ? and you've still only got one display, and
> very much less available software I'd bet. I'm sure it's a great lisp
> environment; at that price it certainly should be.

If you don't need it - don't buy it.

> I trust you see the irony in describing open source software as
> unsupported,

I'm not saying that it is generally unsupported. I'm just
saying that often the effect is that it is unsupported.
Nobody will implement something that comes close
to a Symbolics Lisp environment for Linux in the
next decade. **I** don't hold my breath. For ***some*** hard
problems such a thing might be the necessary tool
to solve them. I think have some of these problems.
You might not.

> whilst at the same time advocating the use of a machine
> that is now presumably unsupported? 

The president of United States runs his publication web
site on such a system. 

? (http:show-url-headers "http://www.pub.whitehouse.gov/")
Date: Fri, 11 Dec 1998 23:47:30 GMT
Server: CL-HTTP/67.99 (Symbolics Common Lisp)
Cache-Control: no-cache
Content-type: text/html; charset=ISO-8859-1
Content-Location: http://www.pub.whitehouse.gov/

> Even if one did buy the emulator I wouldn't have a lot of confidence
> that it would continue to be supported indefinitely...

You get source with it. A lot of source.

Greetings,

Rainer Joswig

-- 
http://www.lavielle.com/~joswig
From: Paul Rudin
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m3btl9is1i.fsf@shodan.demon.co.uk>
······@lavielle.com (Rainer Joswig) writes:

> In article <··············@shodan.demon.co.uk>, Paul Rudin
> <·····@shodan.demon.co.uk> wrote:

[snip Mac UI features]

All the features you cite are present under X, as far as I can see
your main complaint with X is that you don't like any of the windows
managers that you've seen

[symbolic emulator stuff]

> 
> If you don't need it - don't buy it.

It's not a question of need, Its a question of whether it's a cost
effective solution way of delivering all the computing services I
need/want.

And whether I like it or not this involves exchanging
data/code/executables with the x86/Windows dominated world. I do need
a machine that runs windows from time to time. Most of the time
however I run linux at home because for most of the things I do this
is a superior platform.

(At work I have to use windows, but I also take in my linux laptop and
plug it into the network so that I can use that for some tasks.)




> 
> > I trust you see the irony in describing open source software as
> > unsupported,
> 
> I'm not saying that it is generally unsupported. I'm just
> saying that often the effect is that it is unsupported.
> Nobody will implement something that comes close
> to a Symbolics Lisp environment for Linux in the
> next decade. **I** don't hold my breath. For ***some*** hard
> problems such a thing might be the necessary tool
> to solve them. I think have some of these problems.
> You might not.

Well, I don't really know anything about the Symbolics environment.
However I don't believe that *any* one piece of software is *necessary*
to solve a particular problem, although it might be a be useful tool.
One question is whether it's a cost effective tool.


> 
> > whilst at the same time advocating the use of a machine
> > that is now presumably unsupported? 
> 
> The president of United States runs his publication web
> site on such a system. 

And Slick Willy is presumably in the enviable position of being able
to spend practically unlimited funds on
hardware/software/programmer/consulting support for his setup.

> 
> > Even if one did buy the emulator I wouldn't have a lot of confidence
> > that it would continue to be supported indefinitely...
> 
> You get source with it. A lot of source.

Well, I'm not about to buy an alpha and the emulator, but if I see an
old machine going real cheap second hand, I might get one to have a
look at.
From: David Cooper
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3672D265.C8A89E44@genworks.com>
Paul Rudin wrote:
> 
> ······@lavielle.com (Rainer Joswig) writes:
>
> > The president of United States runs his publication web
> > site on such a system.
> 
> And Slick Willy is presumably in the enviable position of being able
> to spend practically unlimited funds on
> hardware/software/programmer/consulting support for his setup.
> 

Um, I forget what all this hoo-haa was about in the first place,
but I definitely do not believe that machine and programmer
resources are requiring unlimited funds for the White House
publication server. From what I gather, they have a very small
team (1 - 2 programmers) tending to one or two Symbolics machines,
serving the whole thing, with constant enhancements and extenstions.

If you would take the time to actually visit this
server, you will notice that it is actually one of the more
sophisticated web servers out there. To achieve those results
with more mundane technology would require some decent hardware
and significant programmer resources, no matter how you slice
it.
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1212982120330001@194.163.195.67>
> Paul Rudin wrote:
> > 
> > ······@lavielle.com (Rainer Joswig) writes:
> >
> > > The president of United States runs his publication web
> > > site on such a system.
> > 
> > And Slick Willy is presumably in the enviable position of being able
> > to spend practically unlimited funds on
> > hardware/software/programmer/consulting support for his setup.

It is not "his setup". It is a service for the people.
Additionally one of the building blocks - the dynamic
web server - is available for download and reuse.
Sounds like a good thing to me.

-- 
http://www.lavielle.com/~joswig
From: Marius Vollmer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87k8zyoaex.fsf@zagadka.ping.de>
······@lavielle.com (Rainer Joswig) writes:

> I'm not saying that there are no good "Open Source" tools - but
> saying that this is the path to the software nirvana is ridiculous.

It's not open source *versus* proprietary software.  It's open source
*and* proprietary software and has been since the beginning.  Open
source is getting bigger recently, just like the rest of the software
`industry', me thinks.

I see it more as two waves: open source chasing proprietary software.
Those that aren't moving fast enough with their proprietary stuff get
swamped eventually.  That's a good thing.

I'm very happy riding the open source wave.  I'm not religous about
not using `unfree' software and I can see where using closed stuff is
a win for me.  But unless something is open source, I don't consider
it really to be part of the `success story of computing'.  For
example, PostScript clearly is here and belongs to the all-time
achievements of the information age, because it is freely available.
Every can study the spec and produce conforming products, even by
building upon a complete, open implementation.  On the other hand, the
MS Word document format is no achievement at all, it's only Microsofts
finger up your ass.  Boy, you should have seen me raving when I was
told to deliver a paper in Word 7 format.
From: Alexey Goldin
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m1iufkycsy.fsf@flight.uchicago.edu>
······@lavielle.com (Rainer Joswig) writes:

> In article <··············@piracy.red-bean.com>, Craig Brozefsky
> <·····@onshore.com> wrote:
> 
> > Uhm, both the kernel is more reliable, and for the large part, most of
> > the programs in the various distributions are more reliable.
> 
> Maybe gradually - but not in principle. Is there
> any revolutionary coding practice in effect that
> suddenly makes Linux software more reliable?
> I mean are the developers switching to Eiffel/Ada/SML and
> are really taking reliability into account (runtime assertions,
> test suites, checking adherance to specifications,
> exception handling infrastructure, ...)?
> 
> Or are they just hacking as ever?
> 

The only real change is that there is way more testers, all bugs are
readily exposed and easily reported because of source code
availability. And it helps a lot. I am sure if they did more of the
stuff you suggest software would be even more reliable but even first
step helps a lot. Linux is at least two orders of magnitude more
reliable then your favorite MacOS (from personal experience). Some
developers are switching to (yikes!) Perl and Python (which is
actually nice) which eliminate a lot of problems of C (like memory
leaks, bounds checking). I am sure Lisp would make these developers
even more productive because even my favorite Python is way below Lisp
on every account but I afraid they will be put off by attitude of most
competent people on this newsgroup, high cost of commercial products
(How much do you have to pay for CLIM?) and unavailability of meny
libraries they are used to in other languages.

You may want to thing long and hard before you blaiming Linux
developers for not using Lisp. They are usually very open minded and
like new stuff. Redhat standart distribution includes more languages
then any other commercial OS by order of magnitude and many of them
are necessary for functioning of an OS. If they are not using Lisp
this is because Lisp has a problem. And I do not think this problem is
syntax. 

P.S. compare progress of free Smalltalk (Squeak ) in last year with
progress of any commercial Lisp and may be you will understand why
open source is important.

P.P.S. I greatly admire your professional skills (which are evident
from your posts) so please do not consider it as personal flame.


Putting on asbestos underwear anyway...
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1012980444560001@194.163.195.67>
In article <··············@flight.uchicago.edu>, Alexey Goldin
<······@flight.uchicago.edu> wrote:

> Linux is at least two orders of magnitude more
> reliable then your favorite MacOS (from personal experience).

Depends. But it is clear that the current Mac OS has
inherent design weaknesses (libs not thread safe,
no real preemptive multitasking OS, no memory protection,
static memory allocation for programs, ...) . Apple hasn't delivered
anything that really addresses these issues yet. MacOS X
might be a step forward. When it arrives. I'm
not holding my breath.

> Some
> developers are switching to (yikes!) Perl and Python (which is
> actually nice) which eliminate a lot of problems of C (like memory
> leaks, bounds checking).

Python also seems to be a great step forward from the
usual shell scripts.

> I am sure Lisp would make these developers
> even more productive

I'm not that sure about that.

> because even my favorite Python is way below Lisp
> on every account but I afraid they will be put off by attitude of most
> competent people on this newsgroup,

I don't think the attitude of some Python people is that much
different.

> high cost of commercial products

Some.

> (How much do you have to pay for CLIM?)

For example LWW with CLIM seems to be reasonably priced.
Commercial Unix Lisps are too expensive, IMHO, though.

> and unavailability of meny
> libraries they are used to in other languages.

And vice versa. I'm using libraries in Lisp that are
hardly available in other languages or are much
harder to use in these languages.

> You may want to thing long and hard before you blaiming Linux
> developers for not using Lisp. They are usually very open minded and
> like new stuff. Redhat standart distribution includes more languages
> then any other commercial OS by order of magnitude and many of them
> are necessary for functioning of an OS.

Maybe concentrating on fewer tools might help to remove
bloat and clutter. ;-)

> If they are not using Lisp
> this is because Lisp has a problem. And I do not think this problem is
> syntax.

I see some problems, too. It might be helpful to name some of these
problems and to see if they can be addressed. This
surely depends on the system we would look at. But
I think there are also general guidelines which could
improve the overall user experience. 

> P.S. compare progress of free Smalltalk (Squeak ) in last year with
> progress of any commercial Lisp and may be you will understand why
> open source is important.

I don't think this is valid. Some commercial Lisp made good
progress last year and they are starting from a different position.
Btw., remove the support from Disney for Squeak and let's
see what happens. Somebody pays the show.

> ... so please do not consider it as personal flame.
> 
> 
> Putting on asbestos underwear anyway...

No reason to do that. A "Framethrower" is much more useful
than a "Flame thrower". ;-)

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87ogpchaen.fsf@piracy.red-bean.com>
······@lavielle.com (Rainer Joswig) writes:

> > Some
> > developers are switching to (yikes!) Perl and Python (which is
> > actually nice) which eliminate a lot of problems of C (like memory
> > leaks, bounds checking).
> 
> Python also seems to be a great step forward from the
> usual shell scripts.

I have a lame solution for CL scripting that I have cooked up in my
secret labratories.  I designed a little protocol for a process to
connect to a long running lisp process and do various things, like
setup an environment for itself, load a file in a batch/script mode,
or run a repl.  I have a small C program which acts as the front end,
and in effect gives you a startup time for scripting in under oneq
second, persistent data storage in the long running lisp, and supports
the Unix #! convention for scripts.

The protocol is not lisp specific and needs some work to add better
authentication (right now it's just plain-text token based) and I have
to complete the client-server protocol for configuring environments
for the execution of scripts.  I use it for shell scripting work when
I just have to have my CL fix, and once I get the environment passing
done, I will be able to do CGI scripts with it.

The "scripting" model really has advantages.  You flush your state
every time (most of the time I hate this, but certain tasks are made
easier by it), it integrates with the rest of the host system fairly
well, and it's procedural style supports one-off programming, quickies
if you will.

It's hardly releasable code, as I've been doing a bit of scheme
hacking lately, but if people reinterested I will post my little
description of the protocol, let everyone pick it to shreds, and then
update my client/server code to handle a goodportion of it and release
the code.

The long running process bit might scare some people off, but I really
like it.  I can load all my favorite libraries once, and I can keep
data hanging around in it for quick access if I'm dealing with script
that execute frequently, or I have scripts that needto share data.  My
present system is CMUCL (multi-proc version only) but should work with
any multi-processing lisp.  I'm debating wether I should have some
sort of harness or other mechanism to ensure the long running process
is always available.  I want to have it start on boot, and stay around
forever.

> > (How much do you have to pay for CLIM?)
> 
> For example LWW with CLIM seems to be reasonably priced.
> Commercial Unix Lisps are too expensive, IMHO, though.

No kidding.  I did some pricing awhile back and found the disparity
between windows versions and unix versions dispicable.  This has been
gone over before in this newsgroup tho.

> > and unavailability of meny
> > libraries they are used to in other languages.
> 
> And vice versa. I'm using libraries in Lisp that are
> hardly available in other languages or are much
> harder to use in these languages.

Perhaps it might be better to say, the perceived monolithic nature of
Lisp is the real detriment there.  At the same time every
implementation I have used has had a FFI that rocks.  Python and Perl
have so many library interfaces because they are being used as glue
languages, simplifying access to functionality embedded in other
languages, and libraries.  Very little if any of the Lisp programming
I do is like this. Often I am working entirely within the realm of
Lisp's capabilities, and would rather spend a day or so re-writing some
functionality, then dickering about with some library interface and
importing alot of cruft.

I'm not sure there is really away to solve this, given the present
user profile of the lisp community, at least as it appears on this
newsgroup.  The lack of a standardized FFI makes it a bit difficult to
be implementation agnostic.  Perhaps a few flagship library interfaces
for a particular implementation would be enough to spark some interest
in Lisp for this class of programming.
From: David Cooper
Subject: Our Attitude Problem (was: Re: help! absolute beginner)
Date: 
Message-ID: <366F56EE.7B714040@genworks.com>
Alexey Goldin wrote:
> 
> I afraid they will be put off by attitude of most
> competent people on this newsgroup,...
> 

I don't want this to become any kind of a flame exchange,
but I'm genuinely interested in exactly what our attitude 
problem is and how we might go about adjusting it.

I sense the same kind of perceived attitude problem out in
the field where I work (ICAD/KBE people vs. the various CAD
empires vs. traditional C/C++ and Oracle IT folks). They
seem unwilling/unable to work with us. Part of the problem is
that we are absolutely coding circles around them as far as
getting any kind of practical applications up and running.

(By the way, ICAD is a KBE system based on Allegro CL,
not to be confused with AutoCAD).

I have been in this business for five years now, and the same
drama plays out again and again. The CAD people  raise
objections or ``roadblocks,'' we overcome the ``roadblocks,''
and they just find more roadblocks. It becomes apparent that the 
roadblocks are not really roadblocks at all, but just excuses.

We roll applications into production despite them. But at greater
cost and lower volume than if we had more cooperation. This keeps
the cost of the technology high, and makes it difficult to get
out of this rut.

This is not the way it is supposed to work.

Note that we are still surviving, because upper-level management
continues to be blown away by the actual results we produce. 
But working with the middle-level managers and engineers is
a constant struggle. 

Obviously, some of this has to do with protecting of empires -
a manager of a ``Design-Advisor'' department at a certain
of the Big 2 auto companies here in Detroit has about 45 ``C'' 
programmers in his sweatshop, producing reams of undocumented
unmaintainable ``AI'' applications which run as ``Design Advisors''
in conjunction with their CAD system. He has Ph.D's writing 
C programs which do sophisticated feature-recognition against
the CAD database.

Meanwhile, a half-dozen ICAD programmers turn out Generative
KBE applications which make his stuff look like crude school
projects. The ICAD apps generate geometry from Lisp objects 
based on rules, obviating the need for this kind of 
feature-recognition because you can just reference-chain to 
any attribute of any object in the whole model tree. You know 
everything about the geometry because your code generated it.

It's called a labor-intensive vs. capital-intensive approach
to doing this stuff. I thought the U.S. is supposed to be a 
capital-intensive economy. I guess no one told this guy. 

So anyway, how does all this constitute an attitude problem?

Too close to the situation, I'm sure I am one of the primary 
instigators of this attitude problem, but it is very difficult
for me to see it in myself because I miss the forest for the trees.

How to fix the attitude problem, and yet stay true to principles?

Personally, I would not be able to live the lie of not calling a 
Spade a Spade.
From: Lars Lundback
Subject: Re: ICAD
Date: 
Message-ID: <3674B0DE.335BF1F@eralslk.ericsson.se>
David,

and the others, you must have been puzzled by my response. I have
received a most kind notification on the word "apprehensive", suggesting
that I probably meant "enthusiastic".

Of course I meant that!

Apart from that, I must blame Netscape. I usually don't save my postings
for future editing before I send them. This time I did, and that
"Netscape draft" somehow got sent, probably through my fumbling. I
didn't know, until I visited the newsgroup. Some of my remarks were
mystifying indeed. Sigh.

I do want to stress that I still find ICAD a _very_ good system for the
areas you mention. That was our findings when ICAD appeared. Also,
things look rosier today, because the traditional barriers between the
designer as the "creative artist", and engineering as being "booring and
repetitive", is disappearing.

Regards, Lars
From: Reini Urban
Subject: Re: ICAD Newsgroup?
Date: 
Message-ID: <3675604c.851254@judy>
David Cooper <········@genworks.com> wrote:
>By the way, how many ICAD people are lurking out there?
>
>I really would like to get a comp.lang.lisp.icad (or, using the new
>name, comp.lang.lisp.kbo) group going. Pipe dream? Perhaps, but
>stranger things have happened on this wondrous spinning sphere.

comp.cad.icad seems to be more appropriate in my eyes.
although all of you do more lisp programming instead of doing "normal"
cad stuff and the icad defpart system has it's own little language, more
problems, questions and discussion would go into the cad field.
kbo would fit there as well.

but for this new group you would need a proper traffic argument which
doesn't hold for now. people at news.announce.newgroups really prefer
traffic, best done with a RFD.

--                                         
Reini
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1112981431570001@pbg3.lavielle.com>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> Eclipse? Golden Common Lisp? Lucid Common Lisp (although I am not sure
> about that one)?

LCL, too.

> The "market-leaders" deliver CLIM - _now_. Look about 2 years back and you
> will find CLIM only for Unix. I still remember the rant about a missing
> CLIM on the MCL-wish-page at Digitool. It was written by Rainer, IIRC.

CLIM 2. Many years ago we were already using CLIM 1 on Macs.

-- 
http://www.lavielle.com/~joswig
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1112981436230001@pbg3.lavielle.com>
In article <··············@flight.uchicago.edu>, Alexey Goldin
<······@flight.uchicago.edu> wrote:

> ······@lavielle.com (Rainer Joswig) writes:
> 
> > In article <··············@ns.mercury.bc.ca>, Joachim Achtzehnter
> > <·······@kraut.bc.ca> wrote:
> > > 
> > > And from a user's perspective, you have to weigh the alternatives: put
> > > all eggs in one basket and accept that you depend totally on one
> > > vendor to fix things or extend things?
> > 
> > Or a bunch of vendors competing with each other?
> 
> Where many is two?

Harlequin, Franz, Symbolics, Digitool, Elwood Corp, Gold Hill, ISR, ...
and probably some more who still have something CL-like to sell
or are having own implementations (Gensym?, ...).

-- 
http://www.lavielle.com/~joswig
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1112981433310001@pbg3.lavielle.com>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> In article <·······················@pbg3.lavielle.com>,
> ······@lavielle.com (Rainer Joswig) wrote:
> 
> >Another religious marketing bullshit.
> 
> Seems it is useless to discuss such a topic with you. Too bad.

I just don't have time to wait for Open Hardware. I have
to use stuff now. Really. Not as a toy or a hobby.

-- 
http://www.lavielle.com/~joswig
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-1112981905230001@hugo.westfalen.de>
In article <·······················@pbg3.lavielle.com>,
······@lavielle.com (Rainer Joswig) wrote:

>I just don't have time to wait for Open Hardware. I have
>to use stuff now. Really. Not as a toy or a hobby.

So? I do use Open Source software here and now - our whole Internet
project is built on Open Source software. And if I would have the need for
a 8-port Serial board, I would use Open Hardware. Not as a toy, but as a
dialout-server. I am not talking about hobby.

As I said, it is wrong to bash all OSS just because OSS doesn't deliver in
some particular area (and that happens to be the area oneself is
interested in). There is a lot of business going on that builds on OSS.
Linux is _not_ only a hobbyist system. And it is not just "from geeks for
geeks". To say that implies ignoring all business success it already had
up to now. 

And there are areas where I definitely would count on OSS instead on
proprietary software - one example being Firewall systems. Why should I
trust a system whose source is not available to me or trusted third
parties? Using the company that sells the firewall as that "trusted" third
party is something we in Germany call "den Bock zum G�rtner machen".
Another example is encryption software. I only trust encryption software
that is available for public review (and several incidents around NTs
"security" models only strengthen this believe).

bye, Georg

-- 
http://www.westfalen.de/hugo/
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1212980227550001@194.163.195.67>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:


> And there are areas where I definitely would count on OSS instead on
> proprietary software - one example being Firewall systems.

I have no problems using software written by experts, maintained by experts
and proven in thousands of absolutely mission critical installations.

> that is available for public review (and several incidents around NTs
> "security" models only strengthen this believe).

But you believe in Unix-based security?

Source for various of this stuff has been available for years.
Unix-based software itself has been stress tested for years on the
Internet. Still security is a joke.

let's see the latest ciac bulletines from 8th december 1998:

- Unix (peripherals running some Unix brand)
- Win NT (HTML viruses)
- Unix (vacation on HP/UX)
- Unix (autofsd on IRIX)
- Unix (Tooltalk)
- Unix (rdist)

go and look in the ciac archives for more fun.

-- 
http://www.lavielle.com/~joswig
From: Erik Naggum
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3122431253724438@naggum.no>
* ··@hugo.westfalen.de (Georg Bauer)
| As I said, it is wrong to bash all OSS just because OSS doesn't deliver
| in some particular area (and that happens to be the area oneself is
| interested in).

  excuse me, but aren't _you_ knocking commercial products as such because
  you don't think some particular commercial product doesn't deliver in
  some particular area?  why the hell should I take a doofus seriously who
  can't even get his principled arguments straight?  go get lost, Georg
  Bauer.

| And there are areas where I definitely would count on OSS instead on
| proprietary software - one example being Firewall systems.  Why should I
| trust a system whose source is not available to me or trusted third
| parties?  Using the company that sells the firewall as that "trusted"
| third party is something we in Germany call "den Bock zum G�rtner
| machen".  Another example is encryption software.  I only trust
| encryption software that is available for public review (and several
| incidents around NTs "security" models only strengthen this believe).

  anyone who _trusts_ Microsoft is seriously in need of psychiatric care.
  I must wonder why you bring them up all the time.  they are _not_ the
  only player in town.  in fact, they are but _one_ player, and you don't
  need to deal with them at all.  I never have, and I have been in this
  business since 1984.  (I have helped people overcome their limitations
  when using Microsoft products, but I don't consider helping people out of
  burning buildings to be playing with fire.)

  if you are really raging against Microsoft, name them.  it isn't a crime
  to be specific.  it _is_ a crime to blame honest and good people for the
  crimes of the Microsoft corporation, which you do by implication.  I
  don't think you have dealt with any other company, but that's just like
  raging and raving about low quality food if you only eat McDonald's stuff
  and then go on to raise your own cattle instead of discovering some of
  the many alternatives.  only really stupid people would do such a thing,
  and, frankly, we don't _need_ stupid people to support free software.

  "just because you think you need steel-belted radial tires and the store
  only has polyglas-belted ones at present, is still no excuse for you
  going off in a corner and reinventing the travois." --Michael A Padlipsky
  (in "The Elements of Networking Style and other essays and animadversions
  on the Art of Intercomputer Networking".  ISBN 0-13-268111-0.)

#:Erik
-- 
  man who cooks while hacking eats food that has died twice.
From: Christopher R. Barry
Subject: Re: help! absolute beginner
Date: 
Message-ID: <367250A5.B9FC9CF6@2xtreme.net>
Kelly Murray wrote:
> 
> Georg Bauer wrote:
> >
> > In article <···············@haystack.aiai.ed.ac.uk>, Tim Bradshaw
> > <···@aiai.ed.ac.uk> wrote:
> >
> > >And, actually, which commercial Lisps *don't* have CLIM?
> >
> 
> CLIM was a dismal failure, and in many ways
> represents what is wrong with lisp, not the least of which is
> CLIM proponents can't acknowledge it was a failure and move on,
> and moreover let the short-term-customer-driven vendors move on too.
> -k
[...]

And what would moving on be? I haven't had a chance to use CLIM, nor
likely ever will (I'm getting started with Garnet though right now), but
from what I've heard of CLIM it is a Good Thing - some miraculous tool
that let's you build sophisticated interfaces in days that would take a
month with C++ or Java and the usual tools. I've never seen it
downtalked before.

But, to hear it downtalked is comforting in a way in that maybe I'm not
missing out on much. :)

Christopher
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-1212981330330001@hugo.westfalen.de>
Hi!

>I have no problems using software written by experts, maintained by experts
>and proven in thousands of absolutely mission critical installations.

Sorry, but thousands of installations prove nothing but a good marketing.

>But you believe in Unix-based security?

Nope. That's why I use Linux. I believe in security of a system where
literally thousands of experts do source reviews. (If you don't like
Linux, go for OpenBSD - they did a marvelous job with their source review
of the base system). Actually the high quality of Linux and OpenBSD in
this area is only possible because of the OSS model. That's why I brought
this up in the first place. No commercial Unix with closed sources (or any
other OS with closed sources) can achieve something similar - this is
immanent to the closed source.

Look at BugTraq for fun: usually Linux-bugs are fixed the next day, or
Unix-bugs don't apply to Linux, since they were fixed there before. Don't
believe that other systems have less problems than Unix - they are only
found later, that's all. Look at Nomad Research Center for some examples
on NT-Security or Novell-Security. Read NT-Bugtraq for examples how M$
screws up on Security and Stability.

Yes, Linux and OpenBSD do have security holes from time to time - but the
difference is, they are discovered _and_ closed, often within days (for
quite low values for "days"). Look at the same problems with NT or MacOS
or Novell and you will see how some of those problems linger around for
months after discovery. That is inacceptable, if you for example build a
Firewall system. Need a good example for MacOS? Look at Personal Web
Sharing. Was in Bugtraq some days ago: if your URL you send the server is
longer than about 4K, the server crashs. Where is the fix from Apple?

>go and look in the ciac archives for more fun.

Go and look in the bugtraq archives for even more fun. I explicitly said
"OSS" and not commercial Unix. There is a reason for that :-)

bye, Georg

-- 
http://www.westfalen.de/hugo/
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1212981601080001@194.163.195.67>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> Hi!
> 
> >I have no problems using software written by experts, maintained by experts
> >and proven in thousands of absolutely mission critical installations.
> 
> Sorry, but thousands of installations prove nothing but a good marketing.

Excellent products and good support maybe.

> >But you believe in Unix-based security?
> 
> Nope. That's why I use Linux.

Good luck. To expensive for me.

> I believe in security of a system where
> literally thousands of experts do source reviews. (If you don't like
> Linux, go for OpenBSD - they did a marvelous job with their source review
> of the base system).

I would more trust in a minimal system that would be build on carefully
designed/enforced specifications and with the right abstractions. It might
even help **not** to open the source to arbitrary people.

> Look at BugTraq for fun: usually Linux-bugs are fixed the next day, or
> Unix-bugs don't apply to Linux, since they were fixed there before.

hey, this is great.  So I can send people arbitrary bug reports
and everything gets fixed the next day. Cool.

> Yes, Linux and OpenBSD do have security holes from time to time

Seems to me immanent in an Unix-like OS. Everything else would surprise
me.

> Firewall system. Need a good example for MacOS? Look at Personal Web
> Sharing. Was in Bugtraq some days ago: if your URL you send the server is
> longer than about 4K, the server crashs. Where is the fix from Apple?

I don't know - ask Apple. 

> Go and look in the bugtraq archives for even more fun. I explicitly said
> "OSS" and not commercial Unix. There is a reason for that :-)

You seem to forget that even commercial Unix sources were and
are available to a wide range of people. That users
have the source of the software they are using is nothing
new. Some companies are giving you wide or complete access to the source
either as part of the software (Symbolics) or for an extra price.
New is that people are trying to "fight" against some large
companies and that this is getting religious. New is that a lot
of people are interested in getting source. Old is, that the
sudden existance of the source of a rule-based engine (SENDMAIL)
on one's own disk drive doesn't make myself into an
AI-expert or (worse) into an Email-expert.

-- 
http://www.lavielle.com/~joswig
From: Christopher Browne
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74sfun$qb8$11@blue.hex.net>
On 11 Dec 1998 02:03:14 +0000, Erik Naggum <····@naggum.no> wrote:
>* ··@hugo.westfalen.de (Georg Bauer)
>| But if commercial support would be really that good for the customer, why
>| wasn't there a CLIM 2 years ago?
>
>  and if free software and open source was so great, why didn't you start
>  in 1880?

And if liberal democracy is so great, why didn't it start a couple of
thousand years back?  If computers were so great, they should have been
developed a couple hundred years earlier... 

-- 
..you could spend *all day* customizing the title bar.  Believe me.  I
speak from experience." -- Matt Welsh
········@hex.net- <http://www.hex.net/~cbbrowne/lsf.html>
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-1312981401210001@hugo.westfalen.de>
Hi!

In article <·······················@194.163.195.67>, ······@lavielle.com
(Rainer Joswig) wrote:

>Excellent products and good support maybe.

Life would be fun if it worked like that.

>I would more trust in a minimal system that would be build on carefully
>designed/enforced specifications and with the right abstractions. It might
>even help **not** to open the source to arbitrary people.

Right approach with the carefull design/specification but wrong approach
with the closed source. Actually you _want_ public reviews of your source
by as many people as possible. Sure, people are free to add source. But
there is no force in anyone going with the newest bells-and-whistles
version. That's why there exist two kernel tracks for Linux: the stable
version only get's bug fixes and security fixes (and so isn't that much a
moving target now) and the developer version get's all the bells and
whistles. Some time the developer version will become the stable version
and a new developer version will spawn. Works much better than most
commercial products where they throw a beta version as "stable" version on
the market.

>hey, this is great.  So I can send people arbitrary bug reports
>and everything gets fixed the next day. Cool.

That's exactly how the Linux user kernel (stable kernel) is maintained.
Alan Cox actively monitors important lists like Bugtraq and fixes bugs
mentioned there. Although there is not much to be fixed in the network
code now, as one can see in the newest 2.0.36 version announcement on
bugtraq - only a handfull quite obscure problems were fixed. It is btw.
same with for example the Debian distribution: the stable release only
get's important bugfixes and especially security fixes. And security fixes
have very high priority there - they are usually fixed very fast, or they
document a work-around, if the fix can't be made ad-hoc.

>Seems to me immanent in an Unix-like OS. Everything else would surprise
>me.

It is immanent to _any_ OS that is a bit more complex than DOS. It is the
same with all other systems, be it MacOS, NT or even Genera. Only that
problems with some of the more obscure OSs are not discovered that much,
because only few people know that OS. Look at the regurlar bugs in Cisco
IOS. There is no such thing as a bugfree OS. It's ridiculous to dream of
one and throw away the working alternatives due to some "to be secure" OS
that's not availabe. As youself said: "I want do use it _now_". We are not
talking about hobby here.

>I don't know - ask Apple. 

There shouldn't be a need to ask a company for security fixes. They have
to put them out themselves and _announce_ them actively. Apple has the
worst security politics on the OS market in comparance with other big
companies: often Bugs are fixed only in the next release of MacOS, and
that is _far_ to late for security problems. Due to the lazy security
politics of Apple I would never put any trust in a Mac-server. M$ did
learn a bit in the last times, they try to  put out fixes quite fast, but
actually they fail the easy integration of those fixes due to the high
complexity of their OS (Unix is much better modularized in that aspect -
it might be ugly, but at least bugs are quite local to some compontent and
so can be much easier fixed with less sideeffects than is the case with
NT). The unix vendors usually do quite good.

>You seem to forget that even commercial Unix sources were and
>are available to a wide range of people.

Not as much as Linux or *BSD. A company can only employ so much developers
as is payed by the selling. OSS doesn't care for selling and so has much
more people available. And it is kind of a sport to find bugs and security
holes in OSS projects. You underestimate the motivation of those
contributers. Read up on some of the articles at opensource.org, it is
quite interesting. Normal "customers" don't help there, either: they are
lacking the hacking desire of the usual OSS people.

>New is that people are trying to "fight" against some large
>companies and that this is getting religious.

Sure, some people get religious. But actually I didn't see any OSS fanatic
in _this_ discussion. Only practical arguments why OSS works in some parts
better than close source. Oh, and I did see much religious arguments
against OSS and pro commercial products, especially in the Lisp area.

>sudden existance of the source of a rule-based engine (SENDMAIL)
>on one's own disk drive doesn't make myself into an
>AI-expert or (worse) into an Email-expert.

Oh, sendmail is much better than you might think. It actually might be the
most secured mail agent out there, just because it is used that much. I
don't like it and use Exim instead, but I have to accept that the work
that went into sendmail security-wise is much more elaborate than what was
invested into Exim. There is an option with qmail, if you can accept the
author, though. Should be quite good security-wise, as the authors main
field of activity is exactly in that area. All MTAs mentionend are OSS.
And ways better than any of the available commercial alternatives. And
that's not religous belief but hard and sad fact (have a look at Exchange
and shudder).

bye, Georg

-- 
http://www.westfalen.de/hugo/
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1312981634410001@194.163.195.67>
In article <···················@hugo.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> Hi!
> 
> In article <·······················@194.163.195.67>, ······@lavielle.com
> (Rainer Joswig) wrote:
> 
> >Excellent products and good support maybe.
> 
> Life would be fun if it worked like that.

It is sometimes.

> >I would more trust in a minimal system that would be build on carefully
> >designed/enforced specifications and with the right abstractions. It might
> >even help **not** to open the source to arbitrary people.
> 
> Right approach with the carefull design/specification but wrong approach
> with the closed source. Actually you _want_ public reviews of your source
> by as many people as possible.

No, I don't think this is necessary.

> >hey, this is great.  So I can send people arbitrary bug reports
> >and everything gets fixed the next day. Cool.
> 
> That's exactly how the Linux user kernel (stable kernel) is maintained.

May work for a some areas. Well, since Linux is old stuff brewed
up again, there is less design work. More like
copying existing designs and debug this stuff. When
they start trying to innovate, we'll see how this model
works.

> >Seems to me immanent in an Unix-like OS. Everything else would surprise
> >me.
> 
> It is immanent to _any_ OS that is a bit more complex than DOS. It is the
> same with all other systems, be it MacOS, NT or even Genera.

IBM for example has developed software that fullfils much higher 
security needs. They did it without open source.

> >I don't know - ask Apple. 
> 
> There shouldn't be a need to ask a company for security fixes. They have
> to put them out themselves and _announce_ them actively. Apple has the
> worst security politics on the OS market in comparance with other big
> companies: often Bugs are fixed only in the next release of MacOS,

sometimes you get bug fixes sooner.

> complexity of their OS (Unix is much better modularized in that aspect -
> it might be ugly, but at least bugs are quite local to some compontent and
> so can be much easier fixed with less sideeffects than is the case with
> NT). The unix vendors usually do quite good.

But the design is so broken that bugs a coming in continuously.

> >You seem to forget that even commercial Unix sources were and
> >are available to a wide range of people.
> 
> Not as much as Linux or *BSD.

Hmm, I thought diverse Unix versions were long time available
for Universities in source. Even Solaris is available, AFAIK.

> >New is that people are trying to "fight" against some large
> >companies and that this is getting religious.
> 
> Sure, some people get religious. But actually I didn't see any OSS fanatic
> in _this_ discussion. Only practical arguments why OSS works in some parts
> better than close source.

Hmm, I found reading the recent discussion on gnu.misc.discuss
more refreshing, relaxed and less evangelist than your contribution.
I also found the following posting to the Future Apple Systems Technologies
mailing list quite entertaining:

 Subject: Re: Why Mac OS X Should be Open
 From:...
 Date: Wed, 09 Dec 1998 23:53:16 -0600
 X-Message-Number: 3
 
 At 03:21 PM 11/29/98 -0800, ... wrote:
 >
 >In the past, I have criticized open-source technology. I was wrong. The
 >industry is clearly moving to open source, and this helps innovation, as can
 >be witnessed by Linux.
 >
 The industry is trying to combat Microsoft. Open source is the current way
 to do that and it's working reasonably well against NT in the server
 market. Companies going open source would sooner not release their money
 making code to the competition I think.
 
 There's nothing innovative about Linux. It developed nothing that's
 impacting the industry, except for the concept of open source on large
 software projects. It's a clone of Minux/Unix. It uses a clone of X11 for
 its window system/GUI. It will use Corba as its ORB. It uses clones or
 clean room copies of almost every innovative piece of software it has. It
 has not made any contributions in object programming, GUI, networking,
 multimedia or system administration that I know of.
 
 >Mac OS X should be open-source.
 >
 I don't care much either way. But if Apple is to do it, they need to it for
 the right reasons. Doing it on the belief that it'll make the operating
 system better is not a good one. The quality of code is dependent on the
 quality of its coders and management. Open source sacrifices development
 speed for a fool proof management system. Somewhere out there, there will
 be a good programmer contributing good code. But a quality programming team
 under good management (and Steve Jobs is a magnet for good coders, in both
 ways) should be able to out-code and out-innovate an open source project in
 speed, polish and quality any day. An open source team can be organized the
 same way, but I don't think altruism goes that far.


> Oh, and I did see much religious arguments
> against OSS and pro commercial products, especially in the Lisp area.

Some of these arguments seem to come from people who actually
earn money with Lisp tools. Actually I think these
arguments might have some weight behind them just because
of this fact. Talking from a position of somebody who
really doesn't use a Lisp system (?), doesn't need the continuing support
of this tool or the further development of this tool, seems
to be easy. If the strategy fails one can just say "sorry
I was wrong". I would like to be much more careful and
not fall into the "open source" trap. I have seen
a lot of things coming and going. Always it was the
thing to boost productivity (or whatever) - remember component
frameworks?. I'll wait and see when the dust settles (say
in three years from now) what is still out there and how it performs.
Well, I expect my Common Lisp code from today working even
in three years.



Rainer Joswig

-- 
http://www.lavielle.com/~joswig
From: Christopher Browne
Subject: Re: help! absolute beginner
Date: 
Message-ID: <754lg5$gi1$6@blue.hex.net>
On Sun, 13 Dec 1998 16:34:32 +0100, Rainer Joswig <······@lavielle.com>
wrote: 
>> That's exactly how the Linux user kernel (stable kernel) is maintained.
>
>May work for a some areas. Well, since Linux is old stuff brewed
>up again, there is less design work. More like
>copying existing designs and debug this stuff. When
>they start trying to innovate, we'll see how this model
>works.

Linux arguably represents an alternative "embrace and extend" approach
to that used by Microsoft.  There is indeed relatively little that has
been *newly* invented as part of Linux; happily there is ample value to
"doing things again, right."  And unlike Microsoft, Linux doesn't have
the same capricious tendancy to try to lock everyone else out...

This is even somewhat useful to the Lisp community, if it results in the
availability of a reasonably stable platform on which to deploy Lisp
systems.  It might be nicer in some respects to have something
implemented more deeply using Lisp, however that has the disadvantage
that it means having to worry about the low level stuff that changes
almost too fast to track.  

For instance, even XFree86 has a hard enough time tracking new video
boards; the only folks that overall seem able to keep up with new video
hardware are MSFT, Xi Graphics, and MetroLink. 

-- 
Who needs fault-tolerant computers when there's obviously an ample
market of fault-tolerant users?
········@ntlug.org- <http://www.ntlug.org/~cbbrowne/lsf.html>
From: Paul Dietz
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3675E846.ECE01E87@interaccess.com>
Christopher Browne wrote:

> This is even somewhat useful to the Lisp community, if it results in the
> availability of a reasonably stable platform on which to deploy Lisp
> systems.  It might be nicer in some respects to have something
> implemented more deeply using Lisp, however that has the disadvantage
> that it means having to worry about the low level stuff that changes
> almost too fast to track.

More directly, it might be useful to hack the Linux kernel to
more efficiently support Lisp, say by providing additional
ways of interacting with the virtual memory system.

It might also be useful to have a Lisp that could run on
a microkernel like L4 or Fiasco, with Linux running beside
it (as another microkernel task.) 

	Paul
From: Christopher Browne
Subject: Re: help! absolute beginner
Date: 
Message-ID: <75a77h$4lv$2@blue.hex.net>
On Tue, 15 Dec 1998 04:40:38 +0000, Paul Dietz <·····@interaccess.com> wrote:
>Christopher Browne wrote:
>
>> This is even somewhat useful to the Lisp community, if it results in the
>> availability of a reasonably stable platform on which to deploy Lisp
>> systems.  It might be nicer in some respects to have something
>> implemented more deeply using Lisp, however that has the disadvantage
>> that it means having to worry about the low level stuff that changes
>> almost too fast to track.
>
>More directly, it might be useful to hack the Linux kernel to
>more efficiently support Lisp, say by providing additional
>ways of interacting with the virtual memory system.

... And if such hacks proved worthwhile, it would be entirely
appropriate to submit such code for general inclusion in "production"
versions of the Linux kernel. 

>It might also be useful to have a Lisp that could run on
>a microkernel like L4 or Fiasco, with Linux running beside
>it (as another microkernel task.) 

That is more useful for microkernels that are well-supported; 'tis not
clear where L4 and Fiasco stand at this point.  L4 having been made
unavailable due to licensing issues not entirely unlike those associated
with FluxOS, and Fiasco being "somewhat early in its history" and of
unknown stability. 

I would suggest the thought that it may be preferable to have a Lisp
that runs on a highly supported kernel like Linux than to require that
people get a research microkernel running.  A reasonably competent UNIX
person can have a PC, a floppy, and a CD, and reasonably expect to have
Linux up and running within a half hour. 

I've struggled unsuccessfully to try to get a purportedly "reasonably
well-packaged" Hurd to boot.  Initial browsing of L4 docs suggest that
it is "fairly challenging" to install. 

The point being that if the overall purpose is to Hack Lisp Code, time
spent messing around with kernel installation is counterproductive, as
is time spent trying to keep a research kernel somewhat in sync with
developments in the Linux production kernels vis-a-vis new PC
hardware... 

-- 
"And 1.1.81 is officially BugFree(tm), so if you receive any bug 
reports on it, you know they are just evil lies." (By Linus Torvalds)
········@hex.net- <http://www.ntlug.org/~cbbrowne/oses.html>
From: Paul Dietz
Subject: Re: help! absolute beginner
Date: 
Message-ID: <3679C8F7.9669F4E4@interaccess.com>
Christopher Browne wrote:

> The point being that if the overall purpose is to Hack Lisp Code, time
> spent messing around with kernel installation is counterproductive, as
> is time spent trying to keep a research kernel somewhat in sync with
> developments in the Linux production kernels vis-a-vis new PC
> hardware...

Oh, certainly, certainly, but it would be interesting if it were
in fact possible to significantly speed up Lisp by working
with the kernel.  Integrating the changes back into mainline
Linux would be the best outcome.

	Paul
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3iuf9k7i7.fsf@todday.aiai.ed.ac.uk>
* Paul Dietz wrote:

> Oh, certainly, certainly, but it would be interesting if it were
> in fact possible to significantly speed up Lisp by working
> with the kernel.  Integrating the changes back into mainline
> Linux would be the best outcome.

But we know that's not true.  You might perhaps get a factor of 2 if
you really had some special VM-level support, but you probably don't
want to do all that.  Assuming you have a decent implementation you're
already pretty close to the limit of how fast you can go.  If you want
speed spend time on the compiler and the GC, not on writing kernel
support.  Or just wait a few months and buy a new machine, which is
probably cheaper than your time!

--tim
From: Paul Dietz
Subject: Re: help! absolute beginner
Date: 
Message-ID: <367A5D3E.8B1CB346@interaccess.com>
Tim Bradshaw wrote:
> 
> * Paul Dietz wrote:
> 
> > Oh, certainly, certainly, but it would be interesting if it were
> > in fact possible to significantly speed up Lisp by working
> > with the kernel.  Integrating the changes back into mainline
> > Linux would be the best outcome.
> 
> But we know that's not true.  You might perhaps get a factor of 2 if
> you really had some special VM-level support, but you probably don't
> want to do all that.  Assuming you have a decent implementation you're
> already pretty close to the limit of how fast you can go.  If you want
> speed spend time on the compiler and the GC, not on writing kernel
> support.  Or just wait a few months and buy a new machine, which is
> probably cheaper than your time!

I'd consider a factor of 2 very significant.

Doing this just for one's self would be silly, but I
don't think that is what was being proposed.

	Paul
From: Christopher R. Barry
Subject: Re: help! absolute beginner
Date: 
Message-ID: <87soedcran.fsf@2xtreme.net>
Paul Dietz <·····@interaccess.com> writes:

> Tim Bradshaw wrote:
> > 
> > * Paul Dietz wrote:
> > 
> > > Oh, certainly, certainly, but it would be interesting if it were
> > > in fact possible to significantly speed up Lisp by working
> > > with the kernel.  Integrating the changes back into mainline
> > > Linux would be the best outcome.
> > 
> > But we know that's not true.  You might perhaps get a factor of 2 if
> > you really had some special VM-level support, but you probably don't
> > want to do all that.  Assuming you have a decent implementation you're
> > already pretty close to the limit of how fast you can go.  If you want
> > speed spend time on the compiler and the GC, not on writing kernel
> > support.  Or just wait a few months and buy a new machine, which is
> > probably cheaper than your time!
> 
> I'd consider a factor of 2 very significant.
> 
> Doing this just for one's self would be silly, but I
> don't think that is what was being proposed.
> 
> 	Paul

He mentioned "VM-level support". There's no way that using a
specialized kernel is going to increase the speed of something like
ACL, CMUCL or MCL by a factor of 2. I believe he was talking about
something comparable to what the Java OS people are doing.

Christopher
From: Tim Bradshaw
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ey3g1adjqwg.fsf@todday.aiai.ed.ac.uk>
* Christopher R Barry wrote:

> He mentioned "VM-level support". There's no way that using a
> specialized kernel is going to increase the speed of something like
> ACL, CMUCL or MCL by a factor of 2. I believe he was talking about
> something comparable to what the Java OS people are doing.

That's a good point -- you might want to do some special
instruction-set thing, but I think that history (lispms) and current
experience (risc) really shows that this is not worth it.

But what I *meant* was special Virtual Memory (not Virtual Machine)
support.  And I think I was wrong: you can't get a factor of 2.  If a
typical lisp system spends 10% of it's time in GC-related stuff you
can only *possibly* get a 10% speedup, even if you make the GC cost
just go away.  (I just ran a hugely consy program I have under CMUCL:
1800 secs of real time, 14 seconds of GC runtime, 4Gb of consing, so
for at least some programs which cons a lot, GC is under 1%
*already*.)

--tim
From: Steve Gonedes
Subject: Re: help! absolute beginner
Date: 
Message-ID: <m2ogp0ev9d.fsf@KludgeUnix.com>
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
 
< That's a good point -- you might want to do some special
< instruction-set thing, but I think that history (lispms) and current
< experience (risc) really shows that this is not worth it.
< 
< But what I *meant* was special Virtual Memory (not Virtual Machine)
< support.  And I think I was wrong: you can't get a factor of 2.  If a
< typical lisp system spends 10% of it's time in GC-related stuff you
< can only *possibly* get a 10% speedup, even if you make the GC cost
< just go away.  (I just ran a hugely consy program I have under CMUCL:
< 1800 secs of real time, 14 seconds of GC runtime, 4Gb of consing, so
< for at least some programs which cons a lot, GC is under 1%
< *already*.)

Would the mlock or one of the sched_setscheduler system calls work for
this? You could lock the lisp process into ram by pid or address.
Dunno if this would prevent paging or not though.
From: Paul Dietz
Subject: Re: help! absolute beginner
Date: 
Message-ID: <367B11A5.4942BA72@interaccess.com>
Tim Bradshaw wrote:

> But what I *meant* was special Virtual Memory (not Virtual Machine)
> support.  And I think I was wrong: you can't get a factor of 2.  If a
> typical lisp system spends 10% of it's time in GC-related stuff you
> can only *possibly* get a 10% speedup, even if you make the GC cost
> just go away.  (I just ran a hugely consy program I have under CMUCL:
> 1800 secs of real time, 14 seconds of GC runtime, 4Gb of consing, so
> for at least some programs which cons a lot, GC is under 1%
> *already*.)

I also think the factor of two you brought up may be
an overestimate of the potential for improvement, but
it's too pessimistic to just look at GC time.  There are
other costs introduced by storage layout, most particularly
costs related to the cache and TLB.  These costs show
up in non-GC running time.

	Paul
From: Kelly Murray
Subject: Re: Kernel for Lisp
Date: 
Message-ID: <367AED33.55E35B6@IntelliMarket.Com>
>Christopher:
> He mentioned "VM-level support". There's no way that using a
> specialized kernel is going to increase the speed of something like
> ACL, CMUCL or MCL by a factor of 2. I believe he was talking about
> something comparable to what the Java OS people are doing.
> 

I believe one could potentially achieve significant performance
improvement, perhaps 10-100-1000 times, by having the VM system know
about Lisp. SunOS adding a system call for Lucid CL that told it to
change
it's paging algorithm used, because when doing a stop-n-copy GC the
reference pattern is completely backwards from normal LRU operation.
That is, a page that was just scanned will not be referenced again,
but the pages that haven't been touched recently are the next ones to be
scanned.  This can have a huge thrashing effect using LRU when
the physical memory is even a little smaller than the virtual size.
Today we don't see it much with 64-128-256mb physical memories
and generational GC, but these effects rear their ugly heads
once again when I have my OODB that wants to fit 2gb of
data into 256mb.  In this case, I really need my own VM and GC
working in sync.  I've run into major GC-thrashing in some cases.

Of course, these kinds of differences are highly dependent on
many factors, which essentially makes kernel VM tweaking a case of
"premature optimization" as discussed here recently,
and certainly putting the cart before the horse since there
isn't a bunch of big lisp applications that are thrashing 
that need fixing...  "it's the applications, stupid" "it's the
applications, stupid" "it's the applications, stupid" 

-Kelly Murray
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-1312982206230001@jill.westfalen.de>
In article <·······················@194.163.195.67>, ······@lavielle.com
(Rainer Joswig) wrote:

I stop this discussion exactly here (actually it is long off track for
this newsgroup already). No use in discussing with closed-minded people.

-- 
http://www.westfalen.de/hugo/
From: Rainer Joswig
Subject: Re: help! absolute beginner
Date: 
Message-ID: <joswig-1412980211510001@194.163.195.67>
In article <···················@jill.westfalen.de>, ··@hugo.westfalen.de
(Georg Bauer) wrote:

> In article <·······················@194.163.195.67>, ······@lavielle.com
> (Rainer Joswig) wrote:
> 
> I stop this discussion exactly here (actually it is long off track for
> this newsgroup already). No use in discussing with closed-minded people.

relax ;-)

-- 
http://www.lavielle.com/~joswig
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <877lvva4t1.fsf@piracy.red-bean.com>
··@hugo.westfalen.de (Georg Bauer) writes:

> In article <·······················@194.163.195.67>, ······@lavielle.com
> (Rainer Joswig) wrote:
> 
> I stop this discussion exactly here (actually it is long off track for
> this newsgroup already). No use in discussing with closed-minded people.

Just a pointer here Georg, you stop discussion by not posting to
them, not by claiming to stop them and then making derogatory
insinuations.
From: Georg Bauer
Subject: Re: help! absolute beginner
Date: 
Message-ID: <gb-1512982201210001@hugo.westfalen.de>
In article <··············@ns.mercury.bc.ca>, Joachim Achtzehnter
<·······@kraut.bc.ca> wrote:

>the open source initiative (OSI)

Uhm - are you sure they are the same? Open Source is something with a
homepage at http://www.opensource.org. OSI on the other hand is something
completely different :-)

bye, Georg

-- 
http://www.westfalen.de/hugo/
From: ············@mediaone.net
Subject: Re: help! absolute beginner
Date: 
Message-ID: <366c2462.54035378@news.ne.mediaone.net>
Another opinion.

.. On Lisp, Free and Commercial ..

Lisp is a fairly complicated language and environment.  It has always been my
opinion that the best lisp is a supported lisp, and that the quality of one's
experience with lisp is proportional to the quality of support.

Some free lisps are supported, but not to the level of a commercial lisp.

Even for my own personal endeavors, I purchase a supported lisp.  I don't like
spending the dollars, but for the value I receive it certainly is justified when
compared to things like Visual C++, etc.  

And I definitely prefer to write my code, and not spend my time doctoring
someone else's code in some free lisp implementation.  I've written GC's, but
that doesn't mean I want to write another one, for no pay, when I have more
pressing code to write, and throw it into some free lisp.

On the other hand, sometimes getting a reproducible bug report to a lisp vendor
is vastly more work without the sources, and I'm not happy when I spend days
essentially acting as a QA resource for a lisp vendor.  But these things happen.
At  least with a lisp vendor I get (a) the fix, and (b) some degree of
expectation that the fix will be maintained in future versions.  That's more
than I can say for bugs I find in the Sun JDK.  There it's a case of "maybe I'll
get a fix, some day".

I'm really glad to see new lisps like Corman's lisp come about, simply because I
believe in lisp and want to see it stick around.  (I like the .sig I see in some
poster's messages in this group, "those who don't know lisp are destined to
repeat it", or something like that).  I'm also glad there are free lisps, and I
don't mean them any injustice.  Maybe they're very good.  I haven't used them,
except for Kyoto Common Lisp many years ago, which is more or less when I
adopted my motto: "The only good lisp is a supported lisp".  KCL was adequate,
but nothing to write home about.  That doesn't mean I wouldn't use it, I just
prefer a more supported lisp and am willing to spend the money.  But I'm
dedicated to my lisp applications, and they're not just academic toys for me, so
I'm also dedicated to using a well supported lisp.

.. On Franz ..

I've been a Franz customer for years, their support is very good.  Sure,
sometimes there are things they don't fix because it's too hard, or they just
can't cost justify it.  But I've never been left with a mission critical
application failure by them.  If it's critical they fix it, at least in their
mainstream products (Lisp). 

Some of their lesser-known products (such as Allegrostore) receive less support
because there is less demand for the product, and less revenue to justify fixes.
It's a chicken-and-egg problem, since they'd sell more if they supported it
better, in my opinion.

That Franz wants to make money is fine with me.  They don't give product away
and that's okay,  I'd hate to see them go out of business for being
revenue-stupid.  I've found that they're open to negotiation if you're
developing non-commercial applications and want a higher powered version of
their product.   They'd rather sell you some lisp for less than not sell it at
all, as long as you're not trying to take advantage of them (their good support
has real costs, after all). 

 I admit that even as a satisfied customer, I have found one or two acts of
pricing by them to be rather predatory, so I'm not always sanguine about the
money, but by and large they're sensible, negotiable, and provide value for your
dollar.  

I expect Harlequin is too.  I only hope they're not about to go six-feet-under
as a lisp vendor, it'd be terrible for the lisp community both by removing a
source of lisp (threatening the lisp supply), and allowing supply and demand to
raise my costs (threatening *my* lisp supply).  Any analogy you draw between
lisp and certain recreational drugs is strictly your own :-)

... On commercial application deployment ...

I strongly dislike the current royalty requirement for commercially deployed
lisp applications.  I believe that lisp should be like other languages, allowing
me to deploy my application without owing the lisp vendor a cut of the deployed
application.  ACLPC 3.0.2 used to allow this, but ACL5 does not.  If Franz is
too demanding in the royalty clause for my current project, it could kill use of
their lisp as the tool to make it happen.  It could also kill my project.  But
again, they're open to negoation, and I view them as a strategic partner and
enabler, not as someone trying to kill my project.

A free lisp doesn't necessarily have this restriction, but the GNU copyleft can
equally kill use of the free lisp for deployment of COMMERCIAL applications.

Your mileage varies with various free lisps or lisp vendors in this regard,
Franz isn't alone and so I don't mean to single them out.  If I used Symbolics
platforms to deliver my app, then I'd either need to ship Ivory machines, or
perhaps license OpenGenera for all the platforms.  I'm not sure about the
latter, but it's a good bet they're not going to let you deploy a Genera app
without a Genera license.

Either way, I feel that this particular thing hurts the viability of lisp as a
commercial application deployment vehicle.  Tough call though, since it also
helps current lisp vendors remain in business.  Ah well...

Apologies for the longwinded reply.
D. Tenny
············@mediaone.net - no spam please
From: Craig Brozefsky
Subject: Re: help! absolute beginner
Date: 
Message-ID: <877lw3k9vu.fsf@piracy.red-bean.com>
············@mediaone.net writes:

> A free lisp doesn't necessarily have this restriction, but the GNU
> copyleft can equally kill use of the free lisp for deployment of
> COMMERCIAL applications.

CMUCL is public domain, you could do whatever you want with it.

The GPL would only kill your application if you did not want to, or
couldn't, distribute the application under the GPL, and you were using
a GPLed lisp.  I can think of many applications for which this would
be a project killer, but I can also think of many where it would not
make a difference.  
From: Tim Bradshaw
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <ey3btllyyhq.fsf@todday.aiai.ed.ac.uk>
* Zachary Turner wrote:
> On second thought maybe this isn't such a bad idea..  Interesting at the
> least...  I'll give it a shot in a few months after I get proficient with
> Lisp.  On another note, does anyone know what the deal is with Franz, Inc.?
> I've been trying to get in touch with these people for the last few weeks
> and they don't return emails and when I call they say that the lady who
> handles sales "is not currently in the office and wont' be in for the
> remainder of the day."

They've probably been very busy organising a conference (very well, I
thought).

--tim
From: John Atwood
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <7471vp$ghk$1@news.NERO.NET>
if you're into games, ya might also be interested in this scheme product
that has an integrated solid modeler:
	http://www.schemers.com/3ds20.html


John Atwood
-------------------------------------------
Zachary Turner <·······@elsitech.com> wrote:
>>> I'm spoiled I guess?  I've been writing games in C/C++ using DirectX and
>>> OpenGL
>>> for some time now and it'd be nice if, after I get good with Lisp, I
>could
>>> write all
>>> my AI code in LISP, compile it to a DLL, and dynamically link to the DLL
>in
>>> my
>>> games, which will be written in C/C++.
From: Erik Naggum
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <3121816954816086@naggum.no>
* "Zachary Turner" <·······@elsitech.com>
| On another note, does anyone know what the deal is with Franz, Inc.?

  well, yes, I do.  it appears that you walked into a small restaurant with
  Burger King expectations.  in so doing, you might have triggered some
  strong negative responses based solely in a cultural conflict that you
  might not even be aware of.  phone numbers and e-mail addresses look very
  much alike, in contrast to storefronts, offices, and counters.

  you said you know nothing about (Common) Lisp -- one of the things you
  will experience is that the community is very different from the Windows
  communities, whence it appears you come.  for instance, the Common Lisp
  market is not marketing-driven, it is not a pyramid game that requires
  ever new people nor a bug-and-upgrade scam, and it is not leveraging its
  operational costs across a huge volume of sales.  rather, it is a pretty
  mature market of long-term partnerships with a steady growth.  the quick
  sale is not unlike a one-night-stand in this setting and you _may_ just
  have appeared much less than serious than you believed you were.

| I evaluated Allegro CL 5.0 and I really like it, but these people are
| basically sitting here with the "No we don't your money.  Go give it to
| Harelquin" attitude, which is really frustrating since I want to purchase
| Allegro.  Has anyone else had better luck with them?

  yes, I have.  the fact that I have happy Common Lisp clients today is
  probably due mostly to the excellent and welcoming attitude at Franz Inc
  when I first approached them.  they have continued to be very helpful in
  making my projects succeed, both for me and for my clients.  I think what
  you write is grossly unfair, so I have to reiterate my impression that
  you have stumbled on a cultural conflict; not all restaurants serve fast
  food, some cater to a very different audience and their tastes and needs.

#:Erik
-- 
  The Microsoft Dating Program -- where do you want to crash tonight?
From: Vassili Bykov
Subject: Re: help! absolute beginner
Date: 
Message-ID: <74bn5m$i4d$1@nnrp1.dejanews.com>
In article <················@naggum.no>,
  Erik Naggum <····@naggum.no> wrote:
> * "Zachary Turner" <·······@elsitech.com>
> | I evaluated Allegro CL 5.0 and I really like it, but these people are
> | basically sitting here with the "No we don't your money.  Go give it to
> | Harelquin" attitude, which is really frustrating since I want to purchase
> | Allegro.  Has anyone else had better luck with them?
>
>   yes, I have.  the fact that I have happy Common Lisp clients today is
>   probably due mostly to the excellent and welcoming attitude at Franz Inc
>   when I first approached them.  they have continued to be very helpful in
>   making my projects succeed, both for me and for my clients.

I can add that I have had a totally different (from Zachary's) experience in
communicating with Franz, in spite of never even having been their paying
customer.  Virtually every e-mail inquiry sent to the sales (out of
curiosity, to keep an eye on the situation in case I do get a chance to
become a paying customer), as well as ordering of ACL 4.3 for Linux, was
followed up by a phone call from them.	I have no idea what was the secret of
"looking serious", but I have always had an impression they indeed have a
very enthusiastic and welcoming attitude.

--Vassili

--
Vassili Bykov
The Object People
http://www.objectpeople.com

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
From: Jon S Anthony
Subject: Re: help! absolute beginner
Date: 
Message-ID: <ufn24w6h2d.fsf@synquiry.com>
Vassili Bykov <·······@objectpeople.com> writes:

> 
> In article <················@naggum.no>,
>   Erik Naggum <····@naggum.no> wrote:
> > * "Zachary Turner" <·······@elsitech.com>
> > | I evaluated Allegro CL 5.0 and I really like it, but these people are
> > | basically sitting here with the "No we don't your money.  Go give it to
> > | Harelquin" attitude, which is really frustrating since I want to purchase
> > | Allegro.  Has anyone else had better luck with them?
> >
> >   yes, I have.  the fact that I have happy Common Lisp clients today is
> >   probably due mostly to the excellent and welcoming attitude at Franz Inc
> >   when I first approached them.  they have continued to be very helpful in
> >   making my projects succeed, both for me and for my clients.
> 
> I can add that I have had a totally different (from Zachary's) experience in
> communicating with Franz, in spite of never even having been their paying
> customer.  Virtually every e-mail inquiry sent to the sales (out of

I'm a paying customer and (in general) I think Franz has been one of
the top 2 or 3 software companies I've ever had to deal with in terms
of

  a) timeliness of responses

  b) knowledge and information content of responses

  c) proactive behavior to requests (for example, during the various
     beta releases of ORBlink I would often fire off several bug
     reports and/or requests for changes and often these were
     addressed within a day or two with downloadable versions
     reflecting the results immediately available.)

I've had some goofs surrounding certain licensing issues and
deliveries of purchased items, but really nothing major.  And they
always seem to have an open ear.

Certainly compared to even the best of the ordinary vendors (take your
pick, but I mean folk like Sun, Symantec, Aonix, etc.) Franz is truly
a qualitative jump beyond.  No comparison really.

/Jon


-- 
Jon Anthony
Synquiry Technologies, Ltd. Belmont, MA 02478, 617.484.3383
"Nightmares - Ha!  The way my life's been going lately,
 Who'd notice?"  -- Londo Mollari
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74goi3$kfm$1@uuneo.neosoft.com>
>  you said you know nothing about (Common) Lisp -- one of the things you
>  will experience is that the community is very different from the Windows
>  communities, whence it appears you come.  for instance, the Common Lisp
>  market is not marketing-driven, it is not a pyramid game that requires
>  ever new people nor a bug-and-upgrade scam, and it is not leveraging its
>  operational costs across a huge volume of sales.  rather, it is a pretty
>  mature market of long-term partnerships with a steady growth.  the quick
>  sale is not unlike a one-night-stand in this setting and you _may_ just
>  have appeared much less than serious than you believed you were.
>
>| I evaluated Allegro CL 5.0 and I really like it, but these people are
>| basically sitting here with the "No we don't your money.  Go give it to
>| Harelquin" attitude, which is really frustrating since I want to purchase
>| Allegro.  Has anyone else had better luck with them?
>
>  yes, I have.  the fact that I have happy Common Lisp clients today is
>  probably due mostly to the excellent and welcoming attitude at Franz Inc
>  when I first approached them.  they have continued to be very helpful in
>  making my projects succeed, both for me and for my clients.  I think what
>  you write is grossly unfair, so I have to reiterate my impression that
>  you have stumbled on a cultural conflict; not all restaurants serve fast
>  food, some cater to a very different audience and their tastes and needs.

How is it unfair?  Was there ever a time when you knew *nothing* about
Common Lisp?  I'm inclined to believe that there was.  Just call it a hunch.
Judging by your other posts to this NG it seems like you want to come across
as the allmighty Lisp programmer who knows all, but you were just a baby
wearing a diaper sucking out of a bottle at one point too.  So does that
mean that you deserved to be treated differently than someone who had been
programming in Lisp longer than you?  You'll probably say yes for argument's
sake but the _correct_ answer is no.  The bottom line is that you have no
idea how serious I am about all my motives behind purchasing a commercial
Lisp implementation.  I could be purchasing it so that I can do Lisp on my
spare time, or I could be trying to be the next allmighty Erik Naggum.  So
please, don't presume to make claims about the validity of my complaints,
because they were very valid complaints.  However, none of this matters
anymore because the issue has since been resolved.  I have been contacted by
two different people from Franz, Inc. both of whom have been extremely
helpful.  And the bottom line still remains that if they weren't selling
licenses they wouldn't be in business.  That's generally how it works.
Whether I want to buy 1 license or 100 licenses doesn't change the situation
any.  I'm still trying to give them money.

Zach
From: Erik Naggum
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <3122041717984112@naggum.no>
* "Zachary Turner" <·······@elsitech.com>
| Judging by your other posts to this NG it seems like you want to come
| across as the allmighty Lisp programmer who knows all, but you were just
| a baby wearing a diaper sucking out of a bottle at one point too.

  newsgroups are not the proper place to discuss your personal problems,
  and especially not projecting them onto others in pathetic desperation.

| So does that mean that you deserved to be treated differently than
| someone who had been programming in Lisp longer than you?

  I do my homework and work hard to be competent in what I do.  that
  generally _does_ have a positive effect on other competent people,
  actually no matter what they are competent in.  I assume you wouldn't
  know what I'm talking about.

| You'll probably say yes for argument's sake but the _correct_ answer is
| no.

  OK, I won't confuse your mind with facts.  it would probably hurt a lot.

| The bottom line is that you have no idea how serious I am about all my
| motives behind purchasing a commercial Lisp implementation.

  I never claimed to have.  I'm sorry that you have this personal problem
  that you have to take out on me, but if you could please go back and read
  what I wrote and stop imputing all sorts of insanities to me just because
  _you_ would think that way, maybe you'll figure out something important.

| So please, don't presume to make claims about the validity of my
| complaints, because they were very valid complaints.

  sure.  in the context of your personally experience, but not universally.
  since you seem to be the kind of guy who makes universal claims out of
  your personal experience, I assume you won't grasp the difference nor the
  fact that your fucking stupid "allmighty" crap should tell everybody that
  they should stay _far_ away from you.

| I have been contacted by two different people from Franz, Inc. both of
| whom have been extremely helpful.  And the bottom line still remains that
| if they weren't selling licenses they wouldn't be in business.  That's
| generally how it works.  Whether I want to buy 1 license or 100 licenses
| doesn't change the situation any.  I'm still trying to give them money.

  I wouldn't sell anything to you.  nor do I think anybody else should.

#:Erik
-- 
  The Microsoft Dating Program -- where do you want to crash tonight?
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74h5vb$osl$1@uuneo.neosoft.com>
You don't have many friends do you?
From: rusty craine
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74h7lf$p8k$1@excalibur.flash.net>
Zachary Turner wrote in message <············@uuneo.neosoft.com>...
>You don't have many friends do you?
>
>
I like the guy.  He takes no prisoners and ask for no quarter.  There are
several  guys in the NG (hmmm no women?) that I try to learn all the lisp I
can from.  He's one of them.  If ya piss him off just duck your head, wait
till it's over and keep on learning [I'm still a few posting from know it
all :)].

If ya decided to get in a verbable joust with him wear your flack jacket.
He seems to come prepared.

Rusty
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74h8s6$ph4$1@uuneo.neosoft.com>
I dont' want to get in a verbal joust with anyone, and perhaps my rant was
partially unjustified, but I take offense to being made feel like i don't
belong in the Lisp community simply because I am a beginner.  If that's not
what he was saying, then I apologize for going off.  Otherwise, there's no
reason for me to sit here and have someone tell me that i'm not worthy of
programming in Lisp.  In any case, that's how it came across.

Zach

rusty craine wrote in message <············@excalibur.flash.net>...
>
>Zachary Turner wrote in message <············@uuneo.neosoft.com>...
>>You don't have many friends do you?
>>
>>
>I like the guy.  He takes no prisoners and ask for no quarter.  There are
>several  guys in the NG (hmmm no women?) that I try to learn all the lisp I
>can from.  He's one of them.  If ya piss him off just duck your head, wait
>till it's over and keep on learning [I'm still a few posting from know it
>all :)].
>
>If ya decided to get in a verbable joust with him wear your flack jacket.
>He seems to come prepared.
>
>Rusty
>
>
>
>
From: Vassili Bykov
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <hMVa2.2743$5n1.21722932@news.magma.ca>
Zachary Turner wrote in message <············@uuneo.neosoft.com>...
>I dont' want to get in a verbal joust with anyone, and perhaps my rant was
>partially unjustified, but I take offense to being made feel like i don't
>belong in the Lisp community simply because I am a beginner.  If that's not
>what he was saying, then I apologize for going off.  Otherwise, there's no
>reason for me to sit here and have someone tell me that i'm not worthy of
>programming in Lisp.  In any case, that's how it came across.


Zach, I think all you have to do is to give another call to Franz and/or
send an e-mail to the sales.  I really have had a very positive experience
with their attitude.  They obviously don't have a staff of ten to handle
sales, and you might have just hit the rough patch the first time you tried.

--Vassili
From: Zachary Turner
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <74hb46$pv3$1@uuneo.neosoft.com>
>Zach, I think all you have to do is to give another call to Franz and/or
>send an e-mail to the sales.  I really have had a very positive experience
>with their attitude.  They obviously don't have a staff of ten to handle
>sales, and you might have just hit the rough patch the first time you
tried.


Yeah, like I said a few posts ago, they got back to me and have been very
helpful and accommodating.  I am quite happy with them.

Zach
From: Erik Naggum
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <3122087050762733@naggum.no>
* "Zachary Turner" <·······@elsitech.com>
| I dont' want to get in a verbal joust with anyone, and perhaps my rant
| was partially unjustified, but I take offense to being made feel like i
| don't belong in the Lisp community simply because I am a beginner.

  your being a beginner had nothing to do with it.  your being arrogant out
  of your ignorance does.  ignorant people can learn.  arrogant ignorant
  people can't.  I welcome anyone who wants to learn.  I'd rather those who
  can't just go away.

| If that's not what he was saying, then I apologize for going off.

  accepted.

| Otherwise, there's no reason for me to sit here and have someone tell me
| that i'm not worthy of programming in Lisp.

  you will find no grounds for his paranoid delusion if you actually read
  what people write instead of reacting as if they wrote what you think it
  looks like they did.

| In any case, that's how it came across.

  ... to you.  there's no universality in this.  don't think there is.

#:Erik
-- 
  The Microsoft Dating Program -- where do you want to crash tonight?
From: David Steuber "The Interloper
Subject: Re: help!  absolute beginner
Date: 
Message-ID: <366e9241.611980541@news.newsguy.com>
On Mon, 7 Dec 1998 12:15:36 -0000, "Zachary Turner"
<·······@elsitech.com> claimed or asked:

% You don't have many friends do you?

Enhance your calm, Zachary and Erik.

While your little flame war is quite entertaining, it is something of
a waste of bandwidth.

Back on topic, I am finding that Graham's "ANSI Common Lisp" is a good
book.  It is a bit steep.  No running up that hill.  But the book is
thin like K&R C and contains exercises relevant to the material that
was introduced.  There is always this news group for additional
information if Graham proves too difficult.

BTW, that little remark about Erik seems uncalled for.  Erik has never
held back when posting to this group, and I for one am glad of it.

--
David Steuber (ver 1.31.3a)
http://www.david-steuber.com
To reply by e-mail, replace trashcan with david.

May the source be with you...