Hi!
I have to write a few servlets and I'm planning to use CL. I
was wondering if there's some library that encapsulates state
retention, something like Java's HttpSession, available? O:-)
TIA
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
hey,
if you're using CL-HTTP, I'm sure that you can use the servlet model
somehow. What kinds of things, specifically, do you want these
servlets to do? i.e. what "state" do you want to save, etc?
dave
On 10 May 2000 09:00:15 -0400, David Bakhash <·····@alum.mit.edu>
wrote:
>hey,
>
>if you're using CL-HTTP, I'm sure that you can use the servlet model
I'm afraid that won't be an option: I'll have to use Apache.
>somehow. What kinds of things, specifically, do you want these
>servlets to do? i.e. what "state" do you want to save, etc?
It must handle a sort of PIM. Basically I need to know who's
the user to show his personalized content. Since not everybody
accepts cookies, I would have encode this info in hidden parameters in
forms or in the links. I was looking for some library that would
encapsulate this thing and do the dirty work for me...
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
Fernando <·······@must.die> writes:
> On 10 May 2000 09:00:15 -0400, David Bakhash <·····@alum.mit.edu>
> wrote:
>
> >hey,
> >
> >if you're using CL-HTTP, I'm sure that you can use the servlet model
>
> I'm afraid that won't be an option: I'll have to use Apache.
Then please take a look at IMHO, which uses the Apache JServ module to
provide a lisp servlet interface. IMHO also includes session
management and component based interface construction supporting the
use of templates. It's inspired by WebObjects, which if you're not
familiar with it is simply the best commercial web app environment
around. We intend to one-up it in the near future. IMHO is licensed
under an MIT/X license, so it's Free Software.
http://alpha.onshore.com/lisp-software
> It must handle a sort of PIM. Basically I need to know who's
> the user to show his personalized content. Since not everybody
> accepts cookies, I would have encode this info in hidden parameters in
> forms or in the links. I was looking for some library that would
> encapsulate this thing and do the dirty work for me...
IMHO session management does not rely upon cookies, instead it uses a
session identifier encoded in URLs.
Note: I'm one of the developers for IMHO and UncommonSQL (also at the
URL I gave you) and I work for onShore.
--
Craig Brozefsky <·····@red-bean.com>
Free Scheme/Lisp Software http://www.red-bean.com/~craig
"Hiding like thieves in the night from life, illusions of
oasis making you look twice. -- Mos Def and Talib Kweli
On 11 May 2000 08:54:56 -0700, Craig Brozefsky <·····@red-bean.com>
wrote:
>Fernando <·······@must.die> writes:
>
>> On 10 May 2000 09:00:15 -0400, David Bakhash <·····@alum.mit.edu>
>> wrote:
>>
>> >hey,
>> >
>> >if you're using CL-HTTP, I'm sure that you can use the servlet model
>>
>> I'm afraid that won't be an option: I'll have to use Apache.
>
>Then please take a look at IMHO, which uses the Apache JServ module to
>provide a lisp servlet interface. IMHO also includes session
>management and component based interface construction supporting the
>use of templates. It's inspired by WebObjects, which if you're not
>familiar with it is simply the best commercial web app environment
>around. We intend to one-up it in the near future. IMHO is licensed
>under an MIT/X license, so it's Free Software.
>
>http://alpha.onshore.com/lisp-software
Seams _very_ interesting. :-) BTW, the (tutorial)
http://alpha.onshore.com/lisp-software/imho/doc/imho-tutorial.html
gives a 404 error... :-?
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
Fernando <·······@must.die> writes:
> Seams _very_ interesting. :-) BTW, the (tutorial)
> http://alpha.onshore.com/lisp-software/imho/doc/imho-tutorial.html
> gives a 404 error... :-?
Oh poo!
It should be available now, just built it.
I would note that the best course of action to getting IMHO and
UncommonSQL running is to snag the latest revisions out of anonymous
CVS. Instructions for doing so are on the home page. The tar and
debian distributions do not work reliably as we severely
underestimated the complexity of their construction an presently do
not have time to work the rest out.
--
Craig Brozefsky <·····@red-bean.com>
Free Scheme/Lisp Software http://www.red-bean.com/~craig
"Hiding like thieves in the night from life, illusions of
oasis making you look twice. -- Mos Def and Talib Kweli
On 12 May 2000 08:49:45 -0700, Craig Brozefsky <·····@red-bean.com>
wrote:
>Fernando <·······@must.die> writes:
>
>> Seams _very_ interesting. :-) BTW, the (tutorial)
>> http://alpha.onshore.com/lisp-software/imho/doc/imho-tutorial.html
>> gives a 404 error... :-?
>
>Oh poo!
>
>It should be available now, just built it.
OK Thanks. :-)
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
David Bakhash <·····@alum.mit.edu> writes:
> if you're using CL-HTTP, I'm sure that you can use the servlet model
> somehow. What kinds of things, specifically, do you want these
> servlets to do? i.e. what "state" do you want to save, etc?
Does CL-HTTP support the mod_jserv or FastCGI APIs? Those might be a
good compromise. A web search returned nothing.
Tom.
tom wrote:
> David Bakhash <·····@alum.mit.edu> writes:
> > if you're using CL-HTTP, I'm sure that you can use the servlet model
> > somehow. What kinds of things, specifically, do you want these
> > servlets to do? i.e. what "state" do you want to save, etc?
>
> Does CL-HTTP support the mod_jserv or FastCGI APIs? Those might be a
> good compromise. A web search returned nothing.
>
I believe IMHO uses JServe. These interfaces are not supported yet under
CL-HTTP (to the best of my knowledge). I would do it myself but I
discovered that it is just as easy to talk HTTP over a socket as it is to
talk fastcgi or other protocol. If you or anyone else can show me a
working httpd.conf file for Apache that shows how to configure mod_fastcgi
to talk to a standalone self-managed application process, then I would be
interested in hacking on an interface.
Andrew K. Wolven
P.S.: Saving state information of clients in a Lisp web server is a
no-brainer. (when to GC is the hard part) I believe you already proposed
the solution: Simply stuff a client-id into emitted anchors and forms.
Don't call it a session, however, or you might be violating some
dickhead's patent. (On that logic, you can't use lisp, because people
patent CL facilities all the time [I am thinking about starting a
bogus-patent web ring for fun, please send your bogus patents to me])
>
> Tom.
"Andrew K. Wolven" wrote:
> tom wrote:
>
> > David Bakhash <·····@alum.mit.edu> writes:
> > > if you're using CL-HTTP, I'm sure that you can use the servlet model
> > > somehow. What kinds of things, specifically, do you want these
> > > servlets to do? i.e. what "state" do you want to save, etc?
> >
> > Does CL-HTTP support the mod_jserv or FastCGI APIs? Those might be a
> > good compromise. A web search returned nothing.
> >
>
> I believe IMHO uses JServe. These interfaces are not supported yet under
> CL-HTTP (to the best of my knowledge). I would do it myself but I
> discovered that it is just as easy to talk HTTP over a socket as it is to
> talk fastcgi or other protocol. If you or anyone else can show me a
> working httpd.conf file for Apache that shows how to configure mod_fastcgi
> to talk to a standalone self-managed application process, then I would be
> interested in hacking on an interface.
>
> Andrew K. Wolven
>
> P.S.: Saving state information of clients in a Lisp web server is a
> no-brainer. (when to GC is the hard part) I believe you already proposed
> the solution: Simply stuff a client-id into emitted anchors and forms.
> Don't call it a session, however, or you might be violating some
> dickhead's patent. (On that logic, you can't use lisp, because people
> patent CL facilities all the time [I am thinking about starting a
> bogus-patent web ring for fun, please send your bogus patents to me])
>
I am going to Patent the concept of a bogus-patent web ring
>
> >
> > Tom.
On Mon, 15 May 2000 11:57:35 -0500, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
>P.S.: Saving state information of clients in a Lisp web server is a
>no-brainer. (when to GC is the hard part) I believe you already proposed
Please, enlighten us. O:-) I also have to save some client
related stuff into a database (results of a search, so I can show the
first 10 matches and latter show the rest without repeating the
query), and I'm not sure when to "GC" the temporary tables holding
that info. I was planning to use a chron process to cleanup every
entry older than X minutes... :-?
>the solution: Simply stuff a client-id into emitted anchors and forms.
>Don't call it a session, however, or you might be violating some
>dickhead's patent. (On that logic, you can't use lisp, because people
Is this what Jeff "Dickhead" Bezos patented (saving client-ids
into forms and anchors)??!!! =:-O I guess the real dickheads are at
the patents office then...
>patent CL facilities all the time [I am thinking about starting a
>bogus-patent web ring for fun, please send your bogus patents to me])
Fine, I'll patent the use of ">" for quoting text.
BTW, is there a more appropriate NG for discussing web server
programming? Most of this isn't really Lisp specific, and I'm afraid
I might be boring other readers... O:-)
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
Fernando> I also have to save some client
Fernando> related stuff into a database (results of a search, so I can show the
Fernando> first 10 matches and latter show the rest without repeating the
Fernando> query),
Is this really necessary? If your data is (made) local and indexes are good,
chances are that re-running the query is faster ... This is what SRS
(sophisticated format-parser cum indexing system used to interconnect
hundreds of biological 'databases' (read: ascii files) does; see
http://www.lionbio.co.uk/), and it was never any problem. Cheers,
Philip
--
/dev/brain: character special (53/0)
-----------------------------------------------------------------------------
Philip Lijnzaad, ········@ebi.ac.uk | European Bioinformatics Institute,rm A2-24
+44 (0)1223 49 4639 | Wellcome Trust Genome Campus, Hinxton
+44 (0)1223 49 4468 (fax) | Cambridgeshire CB10 1SD, GREAT BRITAIN
PGP fingerprint: E1 03 BF 80 94 61 B6 FC 50 3D 1F 64 40 75 FB 53
Philip Lijnzaad <········@ebi.ac.uk> writes:
> Fernando> first 10 matches and latter show the rest without repeating the
> Fernando> query),
>
> Is this really necessary? If your data is (made) local and indexes are good,
> chances are that re-running the query is faster ...
One problem is that with a dynamic database, rerunning the query can
give you a different result, which will look rather odd.
Stig Hemmer,
Jack of a Few Trades.
Stig> Philip Lijnzaad <········@ebi.ac.uk> writes:
Fernando> first 10 matches and latter show the rest without repeating the
Fernando> query),
>>
>> Is this really necessary? If your data is (made) local and indexes are good,
>> chances are that re-running the query is faster ...
Stig> One problem is that with a dynamic database, rerunning the query can
Stig> give you a different result, which will look rather odd.
yes, granted. And anyway, even when re-running the query, there is still
client state to be kept track of, namely which window on the (stored or
re-queried) data needs to be served as the 'next page'. But query
results are usually more voluminous than the queries (query strings)
themselves, so at least the amount of state-to-be-stored can be minimized,
and perhaps even access is faster.
Philip
--
/dev/brain: character special (53/0)
-----------------------------------------------------------------------------
Philip Lijnzaad, ········@ebi.ac.uk | European Bioinformatics Institute,rm A2-24
+44 (0)1223 49 4639 | Wellcome Trust Genome Campus, Hinxton
+44 (0)1223 49 4468 (fax) | Cambridgeshire CB10 1SD, GREAT BRITAIN
PGP fingerprint: E1 03 BF 80 94 61 B6 FC 50 3D 1F 64 40 75 FB 53
Fernando wrote:
> On Mon, 15 May 2000 11:57:35 -0500, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> >P.S.: Saving state information of clients in a Lisp web server is a
> >no-brainer. (when to GC is the hard part) I believe you already proposed
>
> Please, enlighten us. O:-) I also have to save some client
> related stuff into a database (results of a search, so I can show the
> first 10 matches and latter show the rest without repeating the
> query), and I'm not sure when to "GC" the temporary tables holding
> that info. I was planning to use a chron process to cleanup every
> entry older than X minutes... :-?
>
> >the solution: Simply stuff a client-id into emitted anchors and forms.
> >Don't call it a session, however, or you might be violating some
> >dickhead's patent. (On that logic, you can't use lisp, because people
>
> Is this what Jeff "Dickhead" Bezos patented (saving client-ids
> into forms and anchors)??!!! =:-O I guess the real dickheads are at
> the patents office then...
>
> >patent CL facilities all the time [I am thinking about starting a
> >bogus-patent web ring for fun, please send your bogus patents to me])
>
> Fine, I'll patent the use of ">" for quoting text.
>
> BTW, is there a more appropriate NG for discussing web server
> programming? Most of this isn't really Lisp specific, and I'm afraid
> I might be boring other readers... O:-)
>
Apparently there is not a better place. There used to be the mailing list
www-cl but that was a cl-http mailing list and has since fizzled for political
reasons. Perhaps we should start a mailing list.
let me go to starbucks and get my coffee, and then I will write you on the list
and try to give you some good information on the best way to get started doing
web with lisp.
AKW
back in few
>
> //-----------------------------------------------
> // Fernando Rodriguez Romero
> //
> // frr at mindless dot com
> //------------------------------------------------
"Andrew K. Wolven" <·······@redfernlane.org> writes:
> Apparently there is not a better place. There used to be the
> mailing list www-cl but that was a cl-http mailing list and has
> since fizzled for political reasons. Perhaps we should start a
> mailing list.
Taking this as my cue, I setup a list called "lispweb" which is for
discussion of web programming in lisps of various flavors. If you
would like to a subscription please send mail to lispweb-request, or
me.
I will make a seperate announcement for this in the next hour or so.
--
Craig Brozefsky <·····@red-bean.com>
Free Scheme/Lisp Software http://www.red-bean.com/~craig
"Hiding like thieves in the night from life, illusions of
oasis making you look twice. -- Mos Def and Talib Kweli
On Tue, 16 May 2000, Andrew K. Wolven wrote:
> Apparently there is not a better place. There used to be the mailing list
> www-cl but that was a cl-http mailing list and has since fizzled for political
> reasons. Perhaps we should start a mailing list.
So what's up with that? Even the cl-http mailing list has been extremely
quiet in the last month. Are people moving away from cl-http? If so,
why?
Tim
Tim Moore wrote:
> On Tue, 16 May 2000, Andrew K. Wolven wrote:
>
> > Apparently there is not a better place. There used to be the mailing list
> > www-cl but that was a cl-http mailing list and has since fizzled for political
> > reasons. Perhaps we should start a mailing list.
>
> So what's up with that? Even the cl-http mailing list has been extremely
> quiet in the last month. Are people moving away from cl-http? If so,
> why?
>
> Tim
It is simple really, John Mallery, who wrote 95-100% of the core components of
CL-HTTP was intentionally vague about the commercial license. He dragged his feet
in the decision making process on the commercial licensing issue, and failed to
respond to email on this subject.
Basically people got fed up and found other solutions.
Technically, CL-HTTP is a solid server, especially on MCL and Genera. I wouldn't
recommend it for Franz however. It has bugs dealing with the threads on that
platform, and I doubt anyone will actually take the time to fix them. (The people
at Franz apparently felt it is just as easy to write a server, rather than working
the bugs out of CL-HTTP.)
The distribution is also starting to get unwieldy due to size and lack of
defsystem. (there is a defsystem but it is not actually used for the distribution)
AKW
I hate to join what may be a "bashing" thread, and so far as I know
CL-HTTP is an interesting, useful, and innovative system, but when
I needed to provide an HTTP interface to a Lisp application, I decided
against CL-HTTP, for a couple of reasons.
First, the license terms. I forget exactly what the issues were, but
it looked like I'd have to deal with lawyers before I could seriously
consider using it, or even parts of it, such as the XML parser.
Second, even though I'm an experienced Lisp programmer, the code was
dauntingly large and complex - and not well commented. It looked like
it would be far easier for me to write an adequate server myself, from
scratch, than to understand CL-HTTP well enough so that I could use it
with confidence. I couldn't take the chance that I'd find myself up
against a deadline while I was trying to fix some obscure bug.
So I wrote a little server myself in GCL (with some small amounts of
C code for select and a few other things), and I've been pretty happy
with it.
There are probably things I'd like to do that would be easier or
more elegant in CL-HTTP, but the effort required to get to the point
where I could do them was just too great.
-- jeff
In article <·················@redfernlane.org>, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
> Basically people got fed up and found other solutions.
Seems that *some* people are developing CL-HTTP based
applications. Others are copying some of its concepts.
> (The people
> at Franz apparently felt it is just as easy to write a server, rather
> than working
> the bugs out of CL-HTTP.)
I know about who this says something...
> The distribution is also starting to get unwieldy due to size and lack of
> defsystem.
I just compiled/loaded the latest CL-HTTP
devo on my Powerbook G3 by loading one file, started
the server with one menu item invocation and got the top server web
page in Netscape with another menu item invocation. Total time:
less than one minute. Doesn't sound too complicated to me.
> (there is a defsystem but it is not actually used for the
> distribution)
CL-HTTP doesn't use a defsystem???
Try "Show System Definition CL-HTTP" on Genera. How do you
think patches are being maintained? I'm also just doing "Restore
Distribution" over the net to keep my local sources current.
Load CL-HTTP into MCL and look at the CL-HTTP menu under "Edit CL-HTTP
System", "Compile CL-HTTP System",
"Load CL-HTTP System" and "Recompile CL-HTTP system".
In LispWorks browse your systems using the usual facilities
after loading CL-HTTP. What do you see?
To compare anything to they way it works on Genera is cheating.
No there is no freaking defsystem to load CL-HTTP on Franz.
CL-HTTP promptly crashes ACL within 24 hours. CL-HTTP loads nicely on Franz
using one command, It also loads a lot of stuff you don't need. (Therefore you
must write your own build-cl-http command)
Mallery isn't doing anything for the benefit of the Lisp Community itself, and
is only alienating supporters if he thinks I am going to make any customer
cough up more than two hundred bucks for CL-HTTP when Apache is free. If he
wants to have bad blood with Franz, then I am sorry, because I am going to side
with Franz.
Rainer Joswig wrote:
> In article <·················@redfernlane.org>, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> > Basically people got fed up and found other solutions.
>
> Seems that *some* people are developing CL-HTTP based
> applications. Others are copying some of its concepts.
>
> > (The people
> > at Franz apparently felt it is just as easy to write a server, rather
> > than working
> > the bugs out of CL-HTTP.)
>
> I know about who this says something...
>
> > The distribution is also starting to get unwieldy due to size and lack of
> > defsystem.
>
> I just compiled/loaded the latest CL-HTTP
> devo on my Powerbook G3 by loading one file, started
> the server with one menu item invocation and got the top server web
> page in Netscape with another menu item invocation. Total time:
> less than one minute. Doesn't sound too complicated to me.
>
> > (there is a defsystem but it is not actually used for the
> > distribution)
>
> CL-HTTP doesn't use a defsystem???
>
> Try "Show System Definition CL-HTTP" on Genera. How do you
> think patches are being maintained? I'm also just doing "Restore
> Distribution" over the net to keep my local sources current.
>
> Load CL-HTTP into MCL and look at the CL-HTTP menu under "Edit CL-HTTP
> System", "Compile CL-HTTP System",
> "Load CL-HTTP System" and "Recompile CL-HTTP system".
>
> In LispWorks browse your systems using the usual facilities
> after loading CL-HTTP. What do you see?
It should also be noted that large chunks of code live only in the platform
dependent directories and is about as organized as a plate of spaghetti.
"Andrew K. Wolven" wrote:
> To compare anything to they way it works on Genera is cheating.
> No there is no freaking defsystem to load CL-HTTP on Franz.
> CL-HTTP promptly crashes ACL within 24 hours. CL-HTTP loads nicely on Franz
> using one command, It also loads a lot of stuff you don't need. (Therefore you
> must write your own build-cl-http command)
> Mallery isn't doing anything for the benefit of the Lisp Community itself, and
> is only alienating supporters if he thinks I am going to make any customer
> cough up more than two hundred bucks for CL-HTTP when Apache is free. If he
> wants to have bad blood with Franz, then I am sorry, because I am going to side
> with Franz.
>
> Rainer Joswig wrote:
>
> > In article <·················@redfernlane.org>, "Andrew K. Wolven"
> > <·······@redfernlane.org> wrote:
> >
> > > Basically people got fed up and found other solutions.
> >
> > Seems that *some* people are developing CL-HTTP based
> > applications. Others are copying some of its concepts.
> >
> > > (The people
> > > at Franz apparently felt it is just as easy to write a server, rather
> > > than working
> > > the bugs out of CL-HTTP.)
> >
> > I know about who this says something...
> >
> > > The distribution is also starting to get unwieldy due to size and lack of
> > > defsystem.
> >
> > I just compiled/loaded the latest CL-HTTP
> > devo on my Powerbook G3 by loading one file, started
> > the server with one menu item invocation and got the top server web
> > page in Netscape with another menu item invocation. Total time:
> > less than one minute. Doesn't sound too complicated to me.
> >
> > > (there is a defsystem but it is not actually used for the
> > > distribution)
> >
> > CL-HTTP doesn't use a defsystem???
> >
> > Try "Show System Definition CL-HTTP" on Genera. How do you
> > think patches are being maintained? I'm also just doing "Restore
> > Distribution" over the net to keep my local sources current.
> >
> > Load CL-HTTP into MCL and look at the CL-HTTP menu under "Edit CL-HTTP
> > System", "Compile CL-HTTP System",
> > "Load CL-HTTP System" and "Recompile CL-HTTP system".
> >
> > In LispWorks browse your systems using the usual facilities
> > after loading CL-HTTP. What do you see?
In article <·················@redfernlane.org>, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
> It should also be noted that large chunks of code live only in the
> platform
> dependent directories and is about as organized as a plate of spaghetti.
Platform maintainers can reduce the amount of code duplication
- just see how little changes are necessary for the LispWorks port.
Read "http:lw;server;sysdcl.lisp" for the system
declarations for the LispWorks port.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.netWWW: http://www.ision.net/
"Andrew K. Wolven" <·······@redfernlane.org> writes:
> CL-HTTP promptly crashes ACL within 24 hours.
there is no question that CL-HTTP in ACL isn't as polished as it is on
MCL, but it doesn't crash within 24 hours, either.
ACL running CL-HTTP is serving www.alu.org, and it's been running
continuously since january 8th. (but nothing very complicated is
going on).
-matt
"R. Matthew Emerson" wrote:
> "Andrew K. Wolven" <·······@redfernlane.org> writes:
>
> > CL-HTTP promptly crashes ACL within 24 hours.
>
> there is no question that CL-HTTP in ACL isn't as polished as it is on
> MCL, but it doesn't crash within 24 hours, either.
>
> ACL running CL-HTTP is serving www.alu.org, and it's been running
> continuously since january 8th. (but nothing very complicated is
> going on).
>
> -matt
Please send me your latest patches.
Thanks,
Andrew K. Wolven
"Andrew K. Wolven" <·······@redfernlane.org> writes:
> > ACL running CL-HTTP is serving www.alu.org, and it's been running
> > continuously since january 8th. (but nothing very complicated is
> > going on).
> >
> > -matt
>
> Please send me your latest patches.
vanilla cl-http 70-23 includes the fixes.
also, before you start cl-http:
(setf http:*resolve-ip-addresses* nil)
(setf acl-socket:*print-hostnames-in-stream* nil)
this prevents CL-HTTP from doing a reverse lookup when it tries to put
the hostname of the connecting system into the process name string.
previous versions didn't pay attention to http:*resolve-ip-addresses*
and had a buggy wait function which would cause processes to get
arrested and stack up.
-matt
Make sure you look at the log and see how many hits your getting.
I am pretty sure that if you try to do any heavy developing it will puke
at least every couple of days.
Could be an ACL501 thing I guess. (or worse: a windows thing) [it was
puking every couple of days on sun too, I believe]
"R. Matthew Emerson" wrote:
> "Andrew K. Wolven" <·······@redfernlane.org> writes:
>
> > > ACL running CL-HTTP is serving www.alu.org, and it's been running
> > > continuously since january 8th. (but nothing very complicated is
> > > going on).
> > >
> > > -matt
> >
> > Please send me your latest patches.
>
> vanilla cl-http 70-23 includes the fixes.
>
> also, before you start cl-http:
> (setf http:*resolve-ip-addresses* nil)
> (setf acl-socket:*print-hostnames-in-stream* nil)
>
> this prevents CL-HTTP from doing a reverse lookup when it tries to put
> the hostname of the connecting system into the process name string.
>
> previous versions didn't pay attention to http:*resolve-ip-addresses*
> and had a buggy wait function which would cause processes to get
> arrested and stack up.
>
> -matt
"Andrew K. Wolven" <·······@redfernlane.org> writes:
> Make sure you look at the log and see how many hits your getting.
> I am pretty sure that if you try to do any heavy developing it will puke
> at least every couple of days.
> Could be an ACL501 thing I guess. (or worse: a windows thing) [it was
> puking every couple of days on sun too, I believe]
There are some known race conditions in the mulitprocess code in ACL
501. I would imagine that they may be the culprit.
I also should not just expect CL-HTTP to behave wonderfully for debugging
on a Nichimen Mirai-devo image.
I was missing my Lisp machine having everything up in one memory space.
Thus my case for Lisp machine:
Even if the threading or sockets or streams foobars on you, there are
still plenty of options. Genera does not core dump running on an good
ivory[or 36xx] processor with good memory and good disk. The processor is
intelligent and knows how to handle various conditions, traps, flags or
signals and deal. You also have the FEP, which can warm boot the world in
memory. You can probably reboot the machine while running a program on a
Photon board.
Lisp is in a sad, sad state. Even though everybody in the lisp community
knows that this is technology for the worlds salvation, no one will
cooperate. Its all about the cash.
Sun is going to reinvent object-oriented intelligent hardware ignoring the
fact that it has already existed by at least four companies ten years ago.
They are going to reinvent all of Lisp only to years later, discover
through archaeology, that Lisp was there all along.
I think as human beings we do this over and over again. We are so
arrogant as to believe that our ancestors were primitive and that dolphins
are fish.
Seti beams radio signals in to deep space ignoring the fact that the
universe itself is an intelligent signal.
We treat third world countries and people like lesser important human
beings by enforcing export controls.
We force pedestrians into smaller spaces like cattle and make them breathe
carbon monoxide.
The lisp industry waned in the early nineties for no other reason than
humans are stuck in their material existance.
Erik is a stupid viking.
I am going to spend more time learning agrarian practices and less time
trying to convince people in the lisp community that we should be
cooperating.
Thanks,
Andrew K. Wolven
"R. Matthew Emerson" wrote:
> "Andrew K. Wolven" <·······@redfernlane.org> writes:
>
> > CL-HTTP promptly crashes ACL within 24 hours.
>
> there is no question that CL-HTTP in ACL isn't as polished as it is on
> MCL, but it doesn't crash within 24 hours, either.
>
> ACL running CL-HTTP is serving www.alu.org, and it's been running
> continuously since january 8th. (but nothing very complicated is
> going on).
>
> -matt
In article <·················@redfernlane.org>, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
> Lisp is in a sad, sad state. Even though everybody in the lisp community
> knows that this is technology for the worlds salvation, no one will
> cooperate. Its all about the cash.
Maybe when browsing your CL-HTTP documentation you see that
a lot of people have and are collaborating to make it
work. There are various contributed subsystems (from
HTML to XML parsing) and several ports have been done
by users. I personally are very thankful to these people.
Many man years have been gone into it.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.netWWW: http://www.ision.net/
Rainer Joswig wrote:
> In article <·················@redfernlane.org>, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> > Lisp is in a sad, sad state. Even though everybody in the lisp community
> > knows that this is technology for the worlds salvation, no one will
> > cooperate. Its all about the cash.
>
> Maybe when browsing your CL-HTTP documentation you see that
> a lot of people have and are collaborating to make it
> work. There are various contributed subsystems (from
> HTML to XML parsing) and several ports have been done
> by users. I personally are very thankful to these people.
> Many man years have been gone into it.
It referring to JCMA's pocket?
Background info: I am just now getting up to speed with application
development. I am not ashamed of this.
What this means is that if I contributed CL-HTTP, I still could not justify the
cost of the commercial license to my customer, because they expect Dreamweaver
and Apache.
You can chant Lisp all day and the customer doesn't know that you are not using
Perl CGI for their site.
You can't enter competition in an open source market with a price in the
thousands of dollars!
Thats ridiculous! Allegroserve is LGPL!
I am pretty sure that means I can port it to CMUCL if I wish. Thats
competative with Apache.
I don't need ultra high performance streams, because I am doing sites for Joe
Schmo not the prestigious Whitehouse.
Excuse the hell out of me for being confused into thinking that I could
actually contribute to the CL-HTTP project and still be able to afford to run
my contribution!
I understand that you guys are Mallery's friends but please get a clue.
>
>
> --
> Rainer Joswig, Technical Director, BU Partner,
> ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
> Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
> Email: ····················@ision.netWWW: http://www.ision.net/
Andrew K. Wolven wrote in message <·················@redfernlane.org>...
>What this means is that if I contributed CL-HTTP, I still could not justify
the
>cost of the commercial license to my customer, because they expect
Dreamweaver
>and Apache.
>You can chant Lisp all day and the customer doesn't know that you are not
using
>Perl CGI for their site.
>You can't enter competition in an open source market with a price in the
>thousands of dollars!
>Thats ridiculous! Allegroserve is LGPL!
>I am pretty sure that means I can port it to CMUCL if I wish. Thats
>competative with Apache.
>I don't need ultra high performance streams, because I am doing sites for
Joe
>Schmo not the prestigious Whitehouse.
>Excuse the hell out of me for being confused into thinking that I could
>actually contribute to the CL-HTTP project and still be able to afford to
run
>my contribution!
>I understand that you guys are Mallery's friends but please get a clue.
Preface: This isn't a defense of CL-HTTP, nor an attack of any kind on
Franz. This is simple a reality check for Andrew.
The Allegroserve may perhaps be an LGPL software package, but I can assure
that the ACL core that it relies upon is most certainly not. You can not use
the "free" ACL version for anything beyond simple development. To use it for
anything realistic, anything "commercial", means you need to license ACL
from Franz.
And Franz will want a large amount of money to serve up a commercial web
application. You cannot simply buy a copy and throw it on to your servers
(unless they've changed their terms recently, mind you). In the commercial
world, their prices are NOT unreasonable for Internet servers today, but
they are pretty darn high to Human Beings. (The dark side of their prices
isn't that they're high, it's that they're competitive!)
If the Allegroserve happens to work on other Lisps that may have cheaper
prices, then so be it. However, I'm sure that Franz's focus is in enhancing
their product to make it more enticing to a wider range of people, and the
server infrastructure is a critical piece. Most servers have a very similar
core, and by making that core robust and extensible, Franz adds value for
customers who wish to focus on the application rather than the plumbing.
If Allegroserve is the start of that sort of foundation, then more power to
them. By working on Allegroserve, Franz gets to work on both sides of the
server equation. The application side of interfaces, APIs and services as
well as the "OS" side of the core Lisp image that can be tweaked by Franz to
better support these kinds of servers. Since a lot of server functionality
is in the gray nether regiions not specified in the ANSI standard, Franz
will probably show little restraint in extending their system to meet these
kinds of needs.
I read the license agreement with CL-HTTP, and I didn't see any mention of
$200 anywhere. Seems similar to the muddled license of the past that many
users seem to have issue with.
What's interesting is that probably the worst news to hit something like
Allegroserve and Franz (along with most any other commercial Internet App
server) is Apples recent announcement dropping their WebObjects "unlimited"
deployment version from $50,000 down to $700. In the market, $50,000 was a
competitive (if a bit high) number. Many app server vendors are probably a
bit grumpy right now. WebObjects and Franz would probably hit completely
different markets, but regardless, folks are going to ask "How come theirs
is $700 for development and/or unlimited runtime...and yours isn't?".
Will Hartung
(······@home.com)
Will Hartung wrote:
> Andrew K. Wolven wrote in message <·················@redfernlane.org>...
> >What this means is that if I contributed CL-HTTP, I still could not justify
> the
> >cost of the commercial license to my customer, because they expect
> Dreamweaver
> >and Apache.
> >You can chant Lisp all day and the customer doesn't know that you are not
> using
> >Perl CGI for their site.
> >You can't enter competition in an open source market with a price in the
> >thousands of dollars!
> >Thats ridiculous! Allegroserve is LGPL!
> >I am pretty sure that means I can port it to CMUCL if I wish. Thats
> >competative with Apache.
> >I don't need ultra high performance streams, because I am doing sites for
> Joe
> >Schmo not the prestigious Whitehouse.
> >Excuse the hell out of me for being confused into thinking that I could
> >actually contribute to the CL-HTTP project and still be able to afford to
> run
> >my contribution!
> >I understand that you guys are Mallery's friends but please get a clue.
>
> Preface: This isn't a defense of CL-HTTP, nor an attack of any kind on
> Franz. This is simple a reality check for Andrew.
>
> The Allegroserve may perhaps be an LGPL software package, but I can assure
> that the ACL core that it relies upon is most certainly not. You can not use
> the "free" ACL version for anything beyond simple development. To use it for
> anything realistic, anything "commercial", means you need to license ACL
> from Franz.
Complete bullshit. I am just going to do it.
Franz gets their cash by legitimate means. They made me pay for an ACL license,
and now I am looking at the source code to a tiny fast web server (minus
uri.cl). I wouldn't be suprised If I could get this working under CLISP
[without threads].
>
>
> And Franz will want a large amount of money to serve up a commercial web
> application. You cannot simply buy a copy and throw it on to your servers
> (unless they've changed their terms recently, mind you). In the commercial
> world, their prices are NOT unreasonable for Internet servers today, but
> they are pretty darn high to Human Beings. (The dark side of their prices
> isn't that they're high, it's that they're competitive!)
>
I am looking at an email from my Franz salesperson [VAR, runtimes, etc.], and it
seems more reasonable than the rumors I get about CL-HTTP. But that is only if
I choose to use ACL instead of CMUCL or Open Genera for deployment.
>
> If the Allegroserve happens to work on other Lisps that may have cheaper
> prices, then so be it. However, I'm sure that Franz's focus is in enhancing
> their product to make it more enticing to a wider range of people, and the
> server infrastructure is a critical piece. Most servers have a very similar
> core, and by making that core robust and extensible, Franz adds value for
> customers who wish to focus on the application rather than the plumbing.
>
> If Allegroserve is the start of that sort of foundation, then more power to
> them. By working on Allegroserve, Franz gets to work on both sides of the
> server equation. The application side of interfaces, APIs and services as
> well as the "OS" side of the core Lisp image that can be tweaked by Franz to
> better support these kinds of servers. Since a lot of server functionality
> is in the gray nether regiions not specified in the ANSI standard, Franz
> will probably show little restraint in extending their system to meet these
> kinds of needs.
Franz damn well knows that I am going to be knocking at their door as soon as I
get a customer flow. We have maybe two more things to nail down in GWL before
we are going after the Lotus Notes market.
>
>
> I read the license agreement with CL-HTTP, and I didn't see any mention of
> $200 anywhere. Seems similar to the muddled license of the past that many
> users seem to have issue with.
>
> What's interesting is that probably the worst news to hit something like
> Allegroserve and Franz (along with most any other commercial Internet App
> server) is Apples recent announcement dropping their WebObjects "unlimited"
> deployment version from $50,000 down to $700. In the market, $50,000 was a
> competitive (if a bit high) number. Many app server vendors are probably a
> bit grumpy right now. WebObjects and Franz would probably hit completely
> different markets, but regardless, folks are going to ask "How come theirs
> is $700 for development and/or unlimited runtime...and yours isn't?".
>
What do you mean?? GWL is open source. CMUCL will see it for the misers.
They don't call it domino for nothin, and were going to knock it over.
AKW
>
> Will Hartung
> (······@home.com)
"Andrew K. Wolven" wrote:
> Will Hartung wrote:
>
> > Andrew K. Wolven wrote in message <·················@redfernlane.org>...
> > >What this means is that if I contributed CL-HTTP, I still could not justify
> > the
> > >cost of the commercial license to my customer, because they expect
> > Dreamweaver
> > >and Apache.
> > >You can chant Lisp all day and the customer doesn't know that you are not
> > using
> > >Perl CGI for their site.
> > >You can't enter competition in an open source market with a price in the
> > >thousands of dollars!
> > >Thats ridiculous! Allegroserve is LGPL!
> > >I am pretty sure that means I can port it to CMUCL if I wish. Thats
> > >competative with Apache.
> > >I don't need ultra high performance streams, because I am doing sites for
> > Joe
> > >Schmo not the prestigious Whitehouse.
> > >Excuse the hell out of me for being confused into thinking that I could
> > >actually contribute to the CL-HTTP project and still be able to afford to
> > run
> > >my contribution!
> > >I understand that you guys are Mallery's friends but please get a clue.
> >
> > Preface: This isn't a defense of CL-HTTP, nor an attack of any kind on
> > Franz. This is simple a reality check for Andrew.
> >
> > The Allegroserve may perhaps be an LGPL software package, but I can assure
> > that the ACL core that it relies upon is most certainly not. You can not use
> > the "free" ACL version for anything beyond simple development. To use it for
> > anything realistic, anything "commercial", means you need to license ACL
> > from Franz.
>
> Complete bullshit. I am just going to do it. [CMUCL]
> Franz gets their cash by legitimate means. They made me pay for an ACL license,
> and now I am looking at the source code to a tiny fast web server (minus
> uri.cl). I wouldn't be suprised If I could get this working under CLISP
> [without threads].
>
> >
> >
> > And Franz will want a large amount of money to serve up a commercial web
> > application. You cannot simply buy a copy and throw it on to your servers
> > (unless they've changed their terms recently, mind you). In the commercial
> > world, their prices are NOT unreasonable for Internet servers today, but
> > they are pretty darn high to Human Beings. (The dark side of their prices
> > isn't that they're high, it's that they're competitive!)
> >
>
> I am looking at an email from my Franz salesperson [VAR, runtimes, etc.], and it
> seems more reasonable than the rumors I get about CL-HTTP. But that is only if
> I choose to use ACL instead of CMUCL or Open Genera for deployment.
>
> >
> > If the Allegroserve happens to work on other Lisps that may have cheaper
> > prices, then so be it. However, I'm sure that Franz's focus is in enhancing
> > their product to make it more enticing to a wider range of people, and the
> > server infrastructure is a critical piece. Most servers have a very similar
> > core, and by making that core robust and extensible, Franz adds value for
> > customers who wish to focus on the application rather than the plumbing.
> >
> > If Allegroserve is the start of that sort of foundation, then more power to
> > them. By working on Allegroserve, Franz gets to work on both sides of the
> > server equation. The application side of interfaces, APIs and services as
> > well as the "OS" side of the core Lisp image that can be tweaked by Franz to
> > better support these kinds of servers. Since a lot of server functionality
> > is in the gray nether regiions not specified in the ANSI standard, Franz
> > will probably show little restraint in extending their system to meet these
> > kinds of needs.
>
> Franz damn well knows that I am going to be knocking at their door as soon as I
> get a customer flow. We have maybe two more things to nail down in GWL before
> we are going after the Lotus Notes market.
>
In case you want to know about GWL:
Those two things two nail down are:
Persistence methodology atop an RDMS, and
Use of templates for drag and drop interface building. (may just ignore and let
someone else do)
Of course we are going to have to always perfect:
url->lisp translation (well under control)
GC and session management (dependent upon server load)
and add:
online source code editing tools (likely to be java enhanced)
(just in case you don't have emacs)
I think we will be able to iron out this stuff in six months or less, and will very
likely be entirely free except for ACL. We are making it a point ot not be too ACL
dependent. Or server dependent (we plan on supporting CL-HTTP).
Thanks to Franz we have a lot more options to present to out potential customers.
We're going to be recommending things like Allegroserve for Franz and CL-HTTP for
Mac.
(reasonable??)
We still need to be able to have a CL solution to Apache however: LGPL Allegroserve
or custom.
There never would have been an Allegroserve if CL-HTTP hadn't been so aloof.
>
> >
> >
> > I read the license agreement with CL-HTTP, and I didn't see any mention of
> > $200 anywhere. Seems similar to the muddled license of the past that many
> > users seem to have issue with.
> >
> > What's interesting is that probably the worst news to hit something like
> > Allegroserve and Franz (along with most any other commercial Internet App
> > server) is Apples recent announcement dropping their WebObjects "unlimited"
> > deployment version from $50,000 down to $700. In the market, $50,000 was a
> > competitive (if a bit high) number. Many app server vendors are probably a
> > bit grumpy right now. WebObjects and Franz would probably hit completely
> > different markets, but regardless, folks are going to ask "How come theirs
> > is $700 for development and/or unlimited runtime...and yours isn't?".
> >
>
> What do you mean?? GWL is open source. CMUCL will see it for the misers.
>
> They don't call it domino for nothin, and were going to knock it over.
>
> AKW
>
> >
> > Will Hartung
> > (······@home.com)
On Wed, 17 May 2000 01:33:35 -0500, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
>Franz damn well knows that I am going to be knocking at their door as soon as I
>get a customer flow. We have maybe two more things to nail down in GWL before
>we are going after the Lotus Notes market.
Please forgive my ignorance, but...GWL??? :-?
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
On Wed, 17 May 2000 08:51:43 GMT, Fernando <·······@must.die> wrote:
>On Wed, 17 May 2000 01:33:35 -0500, "Andrew K. Wolven"
><·······@redfernlane.org> wrote:
>
>
>>Franz damn well knows that I am going to be knocking at their door as soon as I
>>get a customer flow. We have maybe two more things to nail down in GWL before
>>we are going after the Lotus Notes market.
>
>Please forgive my ignorance, but...GWL??? :-?
Forget the silly question. O:-)
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
"Will Hartung" <······@home.com> writes:
> What's interesting is that probably the worst news to hit something
> like Allegroserve and Franz (along with most any other commercial
> Internet App server) is Apples recent announcement dropping their
> WebObjects "unlimited" deployment version from $50,000 down to
> $700. In the market, $50,000 was a competitive (if a bit high)
> number. Many app server vendors are probably a bit grumpy right
> now. WebObjects and Franz would probably hit completely different
> markets, but regardless, folks are going to ask "How come theirs is
> $700 for development and/or unlimited runtime...and yours isn't?".
The more remarkable thing is that WO is way ahead of any of it's
competitors, and now it's going to be cheaper by an order of
magnitude. Not only is the web app development toolkit, the WO
Framework, much more cohesive, flexible, and scalable than any of the
competing frameworks, but the database interface, the EOF, is a
friggin' gem as well.
Despite the fact that WO provides solutions for problems that are
several levels above what the competitors are just now addressing (and
failing miserably with the EJB standard), this is still basic
infrastructure for building web apps. I believe it would benefit
Common Lisp quite a bit for this infrastructure to be Free Software,
available to all commercial and non-commercial entities in the lisp
community. This is neccesary to provide the basic level of
functionality needed to deliver internet applications, letting
developers concentrate on solving the problems that CL excels at.
The groundwork has already been done, but there is more work to do,
porting to the various CL runtimes, documenting, and adding
collections of commonly used interface components.
--
Craig Brozefsky <·····@red-bean.com>
Free Scheme/Lisp Software http://www.red-bean.com/~craig
"Hiding like thieves in the night from life, illusions of
oasis making you look twice. -- Mos Def and Talib Kweli
"Andrew K. Wolven" wrote:
> Mallery isn't doing anything for the benefit of the Lisp Community itself,
> and
> is only alienating supporters if he thinks I am going to make any customer
> cough up more than two hundred bucks for CL-HTTP when Apache is free.
If $200 is big money for a web server for your customers, then I don't
envy you, but supposedly your talent can add a lot of value to it (embed
the web server in an application, for example), and help you make decent
money. Or is it that you can't get the clients to pay for anything?
As for the CL-HTTP license, I think that a meaningful commercial license
(maybe along with an alternative, unsupported OS license) would be
preferable, but I don't think it would make a difference for you, as you
are clearly out of your medication.
Robert
Centuries ago, Nostradamus foresaw a time when Robert Monfera would say:
>"Andrew K. Wolven" wrote:
>> Mallery isn't doing anything for the benefit of the Lisp Community itself,
>> and
>> is only alienating supporters if he thinks I am going to make any customer
>> cough up more than two hundred bucks for CL-HTTP when Apache is free.
>
>If $200 is big money for a web server for your customers, then I don't
>envy you, but supposedly your talent can add a lot of value to it (embed
>the web server in an application, for example), and help you make decent
>money. Or is it that you can't get the clients to pay for anything?
Sometimes it's appallingly remarkable what clients are unwilling to
pay for...
The _problem_ is that with the availability of Apache, for free, and
Boa, for free, and Roxen Challenger, for free, and AOLserver, for
free, Zope, for free, IIS :-), for free, and probably a bunch of other
options, _for free_, a lot of people look pretty funny at the notion
of charging anything much for a web server.
Sure, Apple's WebObjects is pretty expensive; of course, it appears to
have dropped from $50K to $700, which is suggestive that declining web
server prices are the norm rather than than the abnorm.
I agree that $200 isn't all that much, but it's sure not unusual for
web servers to be priced vastly lower than that.
>As for the CL-HTTP license, I think that a meaningful commercial license
>(maybe along with an alternative, unsupported OS license) would be
>preferable, but I don't think it would make a difference for you, as you
>are clearly out of your medication.
The problem with the CL-HTTP license doesn't seem to me to be the
price, but rather the fact that the licensing is fairly vague.
If I wanted to build a derivative of CL-HTTP, and redistribute it, it
may be reasonably likely that it will be fine with the authors for me
to do so, perhaps at some fee, but it seems evident that there's not a
"pricing sheet" on how to do it. I'd essentially have to pre-arrange
terms individually with the authors.
The uncertainty that is entailed is discouraging.
I can't speak for Wolven's meds :-).
--
Rules of the Evil Overlord #140. "I will instruct my guards when
checking a cell that appears empty to look for the chamber pot. If the
chamber pot is still there, then the prisoner has escaped and they may
enter and search for clues. If the chamber pot is not there, then
either the prisoner is perched above the lintel waiting to strike them
with it or else he decided to take it as a souvenir (in which case he
is obviously deeply disturbed and poses no threat). Either way,
there's no point in entering." <http://www.eviloverlord.com/>
········@ntlug.org- <http://www.hex.net/~cbbrowne/lsf.html>
Christopher Browne wrote:
> >If $200 is big money for a web server [...]
> Sometimes it's appallingly remarkable what clients are unwilling to
> pay for...
A tested approach (with software definitely more expensive than $200) is
to build the costs into your pricing - whether you are hired to roll out
a product or are a consultant with an hourly rate. The point I made is
that if there is satisfactory cash flow to keep you happy, finding the
money for tools should not be a problem. in what commercial scenario do
you think there is an _unwillingness_ to pay? Who are those _clients_?
> The _problem_ is that with the availability of Apache, for free, and
> Boa, for free, and Roxen Challenger, for free, and AOLserver, for
> free, Zope, for free, IIS :-), for free, and probably a bunch of other
> options, _for free_, a lot of people look pretty funny at the notion
> of charging anything much for a web server.
You can get commercial support for Apache from IBM and maybe others, and
I don't think it's free.
> I agree that $200 isn't all that much, but it's sure not unusual for
> web servers to be priced vastly lower than that.
$200 is the cost of 1..5 developer hours in the USA. Think of how much
is spent on other unix Software or Microsoft-ware, or on those
queen-of-the-self CASE tools.
> The problem with the CL-HTTP license doesn't seem to me to be the
> price, but rather the fact that the licensing is fairly vague.
Yes, this is what I meant. As someone described it here: you would have
to work with a lawyer to be able to use it (at a time I guessed it's 2
mandays of work with the lawyer and the author(s)).
Robert
Centuries ago, Nostradamus foresaw a time when Robert Monfera would say:
>"Andrew K. Wolven" wrote:
>> Mallery isn't doing anything for the benefit of the Lisp Community itself,
>> and
>> is only alienating supporters if he thinks I am going to make any customer
>> cough up more than two hundred bucks for CL-HTTP when Apache is free.
>
>If $200 is big money for a web server for your customers, then I don't
>envy you, but supposedly your talent can add a lot of value to it (embed
>the web server in an application, for example), and help you make decent
>money. Or is it that you can't get the clients to pay for anything?
Sometimes it's appallingly remarkable what clients are unwilling to
pay for...
The _problem_ is that with the availability of Apache, for free, and
Boa, for free, and Roxen Challenger, for free, and AOLserver, for
free, Zope, for free, IIS :-), for free, and probably a bunch of other
options, _for free_, a lot of people look pretty funny at the notion
of charging anything much for a web server.
Sure, Apple's WebObjects is pretty expensive; of course, it appears to
have dropped from $50K to $700, which is suggestive that declining web
server prices are the norm rather than than the abnorm.
I agree that $200 isn't all that much, but it's sure not unusual for
web servers to be priced vastly lower than that.
>As for the CL-HTTP license, I think that a meaningful commercial license
>(maybe along with an alternative, unsupported OS license) would be
>preferable, but I don't think it would make a difference for you, as you
>are clearly out of your medication.
The problem with the CL-HTTP license doesn't seem to me to be the
price, but rather the fact that the licensing is fairly vague.
If I wanted to build a derivative of CL-HTTP, and redistribute it, it
may be reasonably likely that it will be fine with the authors for me
to do so, perhaps at some fee, but it seems evident that there's not a
"pricing sheet" on how to do it. I'd essentially have to pre-arrange
terms individually with the authors.
The uncertainty that is entailed is discouraging.
I can't speak for Wolven's meds :-).
--
Rules of the Evil Overlord #140. "I will instruct my guards when
checking a cell that appears empty to look for the chamber pot. If the
chamber pot is still there, then the prisoner has escaped and they may
enter and search for clues. If the chamber pot is not there, then
either the prisoner is perched above the lintel waiting to strike them
with it or else he decided to take it as a souvenir (in which case he
is obviously deeply disturbed and poses no threat). Either way,
there's no point in entering." <http://www.eviloverlord.com/>
········@ntlug.org- <http://www.hex.net/~cbbrowne/lsf.html>
In article <·················@redfernlane.org>, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
> To compare anything to they way it works on Genera is cheating.
I was talking about MCL and LW also, wasn't I?
> No there is no freaking defsystem to load CL-HTTP on Franz.
On every platform where possible CL-HTTP uses the native
defsystem. Ask Franz for support, since they are maintaining
the port to the latest ACL version.
Xanalys is maintaining the LispWorks port. It works there.
Digitool is maintaining parts of the MCL port it works
there, too. It should not be to complicated.
> CL-HTTP promptly crashes ACL within 24 hours.
Ask Franz, they are maintaining the port.
> using one command, It also loads a lot of stuff you don't need.
> (Therefore you
> must write your own build-cl-http command)
Ask Franz.
> wants to have bad blood with Franz, then I am sorry, because I am going
> to side
> with Franz.
See above.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.netWWW: http://www.ision.net/
Rainer Joswig wrote:
> In article <·················@redfernlane.org>, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> > To compare anything to they way it works on Genera is cheating.
>
> I was talking about MCL and LW also, wasn't I?
>
> > No there is no freaking defsystem to load CL-HTTP on Franz.
>
> On every platform where possible CL-HTTP uses the native
> defsystem. Ask Franz for support, since they are maintaining
> the port to the latest ACL version.
>
> Xanalys is maintaining the LispWorks port. It works there.
> Digitool is maintaining parts of the MCL port it works
> there, too. It should not be to complicated.
>
> > CL-HTTP promptly crashes ACL within 24 hours.
>
> Ask Franz, they are maintaining the port.
>
> > using one command, It also loads a lot of stuff you don't need.
> > (Therefore you
> > must write your own build-cl-http command)
>
> Ask Franz.
>
> > wants to have bad blood with Franz, then I am sorry, because I am going
> > to side
> > with Franz.
>
> See above.
>
> --
> Rainer Joswig, Technical Director, BU Partner,
> ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
> Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
> Email: ····················@ision.netWWW: http://www.ision.net/
Thanks for your advice. I asked Franz. They gave me Allegroserve for free.
In article <·················@redfernlane.org>, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
> Thanks for your advice. I asked Franz. They gave me Allegroserve for
> free.
Minus the price of ACL to be able to run it.
Hey, they even now are using hierarchical packages.
Seems like they are going to have a Lisp dialect
of there own.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.netWWW: http://www.ision.net/
I better submit my resume.
Rainer Joswig wrote:
> In article <·················@redfernlane.org>, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> > Thanks for your advice. I asked Franz. They gave me Allegroserve for
> > free.
>
> Minus the price of ACL to be able to run it.
>
> Hey, they even now are using hierarchical packages.
> Seems like they are going to have a Lisp dialect
> of there own.
>
> --
> Rainer Joswig, Technical Director, BU Partner,
> ISION Internet AG, Steinhoft 9, 20459 Hamburg, Germany
> Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
> Email: ····················@ision.netWWW: http://www.ision.net/
* Rainer Joswig wrote:
> Hey, they even now are using hierarchical packages.
> Seems like they are going to have a Lisp dialect
> of there own.
Well, I think it's kind of the idea that implementors should get to
experiment with extensions so that in due course some consensus can
emerge about what the right way of doing things is and a standard can
be evolved. I don't know about the hierarchical package stuff, and I
do know that I don't desperately like some of the other things they're
doing, but I'm definitely glad that they're doing *something* to
evolve CL, while making information available about what they're
implementing.
--tim
* Rainer Joswig <·············@ision.net>
| Minus the price of ACL to be able to run it.
Geez, _that's_ an intelligent counter-argument. Can you please be
explicit and explain your real beef with Franz Inc?
| Hey, they even now are using hierarchical packages.
| Seems like they are going to have a Lisp dialect
| of there own.
Although I don't necessarily think hierarchical packages is a good
move, they have published the source code necessary to make _any_
Common Lisp get hierarchical packages. It affects a single function
in the Common Lisp package. I'm impressed with your ability to
label this a "dialect", because it clearly suggests supernatural
mental powers.
#:Erik
--
If this is not what you expected, please alter your expectations.
In article <················@naggum.no>, Erik Naggum <····@naggum.no>
wrote:
> * Rainer Joswig <·············@ision.net>
> | Minus the price of ACL to be able to run it.
>
> Geez, _that's_ an intelligent counter-argument. Can you please be
> explicit and explain your real beef with Franz Inc?
It's hopeless Mr. Naggum - I'm not going to argue
with you.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinh�ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.netWWW: http://www.ision.net/
* Rainer Joswig <·············@ision.net>
| It's hopeless Mr. Naggum - I'm not going to argue with you.
Thank you for confirming that you have a _personal_ problem with
Franz Inc, which means we can expect more of the same and none of
the anticipated, but not expected, reasoning that would limit the
future amount of slander and negative propaganda against them.
#:Erik
--
If this is not what you expected, please alter your expectations.
In article <················@naggum.no>, Erik Naggum <····@naggum.no>
wrote:
> * Rainer Joswig <·············@ision.net>
> | It's hopeless Mr. Naggum - I'm not going to argue with you.
>
> Thank you for confirming that you have a _personal_ problem with
> Franz Inc, which means we can expect more of the same and none of
> the anticipated, but not expected, reasoning that would limit the
> future amount of slander and negative propaganda against them.
It's hopeless Mr. Naggum - I'm not going to argue with you.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinh�ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.net WWW: http://www.ision.net/
* Rainer Joswig <·············@ision.net>
| It's hopeless Mr. Naggum - I'm not going to argue with you.
But you are. If you didn't want to argue, you could just be
so kind as to refrain from slander and negative propaganda.
#:erik
--
If this is not what you expected, please alter your expectations.
In article <················@naggum.no>, Erik Naggum <····@naggum.no>
wrote:
> * Rainer Joswig <·············@ision.net>
> | It's hopeless Mr. Naggum - I'm not going to argue with you.
>
> But you are. If you didn't want to argue, you could just be
> so kind as to refrain from slander and negative propaganda.
Come on Mr. Naggum, waste others people time.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinh�ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.net WWW: http://www.ision.net/
* Rainer Joswig <·············@ision.net>
| Come on Mr. Naggum, waste others people time.
Wouldn't you love that? But it's quite simple: You explain what
your _real_ problem with Franz Inc is, instead of continuing to
slander them and hurting the entire Lisp community by causing
unnecessary animosity between developer groups, and we don't need
that. You're _really_ wasting people's time when you do it.
#;Erik
--
If this is not what you expected, please alter your expectations.
In article <················@naggum.no>, Erik Naggum <····@naggum.no>
wrote:
> Wouldn't you love that? But it's quite simple: You explain what
> your _real_ problem with Franz Inc is, instead of continuing to
> slander them and hurting the entire Lisp community by causing
> unnecessary animosity between developer groups, and we don't need
> that. You're _really_ wasting people's time when you do it.
You are obviously dreaming. Please waste other people's time.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinh�ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.net WWW: http://www.ision.net/
* Rainer Joswig <·············@ision.net>
| Please waste other people's time.
If I can waste your time so you don't continue to slander people and
companies and create unnecessary animosity between developer groups,
that's a _very_ good thing in my view. So, I'm glad your time is
wastable with such simple means.
#;Erik
--
If this is not what you expected, please alter your expectations.
In article <················@naggum.no>, Erik Naggum <····@naggum.no>
wrote:
> * Rainer Joswig <·············@ision.net>
> | Please waste other people's time.
>
> If I can waste your time so you don't continue to slander people and
> companies and create unnecessary animosity between developer groups,
> that's a _very_ good thing in my view. So, I'm glad your time is
> wastable with such simple means.
Sure, Sure, Mr.Naggum, tell me more about your mother.
--
Rainer Joswig, Technical Director, BU Partner,
ISION Internet AG, Steinh�ft 9, 20459 Hamburg, Germany
Tel: +49 40 3070 2950, Fax: +49 40 3070 2999
Email: ····················@ision.net WWW: http://www.ision.net/
Centuries ago, Nostradamus foresaw a time when Erik Naggum would say:
>* Rainer Joswig <·············@ision.net>
>| Please waste other people's time.
>
> If I can waste your time so you don't continue to slander people and
> companies and create unnecessary animosity between developer groups,
> that's a _very_ good thing in my view. So, I'm glad your time is
> wastable with such simple means.
Could you "squalling brats" _PLEASE_ take your "waste of time" to
email, where it won't trouble the rest of us?
Sometimes, my .sigmonster picks something incredibly more appropriate
than anything I would have _tried_ to choose myself; this is one of
those occasions...
--
This program posts news to billions of machines throughout the galaxy.
Your message will cost the net enough to bankrupt your entire planet.
As a result your species will be sold into slavery. Be sure you know
what you are doing. Are you absolutely sure you want to do this? [yn]
y
········@hex.net- <http://www.ntlug.org/~cbbrowne/lsf.html>
From: Erik Naggum
Subject: Re: Squalling Brats Flinging Dung At One Another Re: Servlets in CL
Date:
Message-ID: <3167704623050405@naggum.no>
* Christopher Browne
| Could you "squalling brats" _PLEASE_ take your "waste of time" to
| email, where it won't trouble the rest of us?
Why didn't you mail us your helpful suggestion?
You don't have to answer, just _think_ about it.
#:Erik
--
If this is not what you expected, please alter your expectations.
Centuries ago, Nostradamus foresaw a time when Erik Naggum would say:
>* Rainer Joswig <·············@ision.net>
>| Please waste other people's time.
>
> If I can waste your time so you don't continue to slander people and
> companies and create unnecessary animosity between developer groups,
> that's a _very_ good thing in my view. So, I'm glad your time is
> wastable with such simple means.
Could you "squalling brats" _PLEASE_ take your "waste of time" to
email, where it won't trouble the rest of us?
Sometimes, my .sigmonster picks something incredibly more appropriate
than anything I would have _tried_ to choose myself; this is one of
those occasions...
--
This program posts news to billions of machines throughout the galaxy.
Your message will cost the net enough to bankrupt your entire planet.
As a result your species will be sold into slavery. Be sure you know
what you are doing. Are you absolutely sure you want to do this? [yn]
y
········@hex.net- <http://www.ntlug.org/~cbbrowne/lsf.html>
From: Erik Naggum
Subject: Re: Squalling Brats Flinging Dung At One Another Re: Servlets in CL
Date:
Message-ID: <3168927323561701@naggum.no>
* Christopher Browne
| Could you "squalling brats" _PLEASE_ take your "waste of time" to
| email, where it won't trouble the rest of us?
You're responding to a message posted 2000-05-18 on 2000-06-02 with
a complaint like that? _Why_? I think you're the squalling brat
flinging dung at others when you take that long after THINGS HAVE
ALREADY QUIETED DOWN to complain. Just quit your pointless and
counterproductive complaining, will you? Sheesh!
#:Erik
--
If this is not what you expected, please alter your expectations.
From: Tim Bradshaw
Subject: Re: Squalling Brats Flinging Dung At One Another Re: Servlets in CL
Date:
Message-ID: <ey3k8g849wf.fsf@cley.com>
* Erik Naggum wrote:
* Christopher Browne
> | Could you "squalling brats" _PLEASE_ take your "waste of time" to
> | email, where it won't trouble the rest of us?
> You're responding to a message posted 2000-05-18 on 2000-06-02 with
> a complaint like that? _Why_? I think you're the squalling brat
> flinging dung at others when you take that long after THINGS HAVE
> ALREADY QUIETED DOWN to complain. Just quit your pointless and
> counterproductive complaining, will you? Sheesh!
I think the article got injected twice -- there's another copy in my
spool dated 19-may: I guess his newsreader fouled up?
--tim
Centuries ago, Nostradamus foresaw a time when Tim Bradshaw would say:
>* Erik Naggum wrote:
>* Christopher Browne
>> | Could you "squalling brats" _PLEASE_ take your "waste of time" to
>> | email, where it won't trouble the rest of us?
>
>> You're responding to a message posted 2000-05-18 on 2000-06-02 with
>> a complaint like that? _Why_? I think you're the squalling brat
>> flinging dung at others when you take that long after THINGS HAVE
>> ALREADY QUIETED DOWN to complain. Just quit your pointless and
>> counterproductive complaining, will you? Sheesh!
>
>I think the article got injected twice -- there's another copy in my
>spool dated 19-may: I guess his newsreader fouled up?
Reinjection of some spool; oops... Sorry about that...
--
·····@freenet.carleton.ca - <http://www.ntlug.org/~cbbrowne/>
Why does the word "lisp" have an "s" in it?
Erik Naggum wrote:
> * Rainer Joswig <·············@ision.net>
> | Come on Mr. Naggum, waste others people time.
>
> Wouldn't you love that? But it's quite simple: You explain what
> your _real_ problem with Franz Inc is, instead of continuing to
> slander them and hurting the entire Lisp community by causing
> unnecessary animosity between developer groups,
Oh, yeah, I couldn't imagine a lisp developer causing unneeded
animosity.
dave
Rainer Joswig wrote:
> Hey, they even now are using hierarchical packages.
> Seems like they are going to have a Lisp dialect
> of there own.
Being hard of thinking: what is up with this? Unless this breaks existing
code (which I assume it doesn't) and is not enforced in such a way that have
rather than choose to use hierarchical packages (which again is another
assumption, a something I would probably choose not to use because of
portability issues)...
Best Regards,
:) will
* Rainer Joswig wrote:
| Hey, they even now are using hierarchical packages. Seems like they
| are going to have a Lisp dialect of there own.
* William Deakin <·····@pindar.com>
| Being hard of thinking: what is up with this?
It's Rainer Joswig having a personal gripe with Franz Inc, but he
isn't honest enough to come out and state his real issue. From what
I have seen in the past, there isn't _anything_ Franz Inc can do to
appease him, either, so an informed choice is to ignore his noise.
| Unless this breaks existing code (which I assume it doesn't) and is
| not enforced in such a way that have rather than choose to use
| hierarchical packages (which again is another assumption, a
| something I would probably choose not to use because of portability
| issues)...
If you call find-package (or cause it to be called) with a string
designator that starts with a dot, there is no package so named, the
current package contains a dot in its name, there is a package with
the same name up to the dot and the string you gave find-package
from the dot, you will not get nil, but that package. I'm sure this
affects _some_ existing code.
Hierarchical packages have been added to aid in portability. It's
probably more an attempt to cater to the Java crowd, however. I'm
not sure what this will buy the Lisp community, but this is one of
the most innocuous changes I have seen anyone suggest.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * Rainer Joswig wrote:
> | Hey, they even now are using hierarchical packages. Seems like they
> | are going to have a Lisp dialect of there own.
>
> * William Deakin <·····@pindar.com>
> | Being hard of thinking: what is up with this?
>
> It's Rainer Joswig having a personal gripe with Franz Inc, but he
> isn't honest enough to come out and state his real issue. From what
> I have seen in the past, there isn't _anything_ Franz Inc can do to
> appease him, either, so an informed choice is to ignore his noise.
I don't there's _nothing_ they could do. I think they have a
decent product, but I consider their pricing structure unworkable.
If I was more committed to lisp, I might be annoyed, I guess, as
is it's the choice between the lesser of two evils. Although lispworks
is not as mature, they do provide *tons* of help.
dave
* David Hanley <···@ncgr.org>
| I don't there's _nothing_ they could do. I think they have a
| decent product, but I consider their pricing structure unworkable.
That's not it. You just _don't_ slam someone's technical or
professional _ethics_ because they charge a price _you_ don't want
to pay, when others clearly think it's OK to pay what they want.
The cost of an Allegro CL license should be nearly irrelevant to
most software projects, apart from the ones that are losing money by
design -- i.e., open source and hobbyist projects. These have the
ability to use the free Allegro CL Trial edition until they figure
out that they need to make money. Programmer time _really_ costs
money. Hardware and software does not.
Wanting more for less is quite universal, but thinking that it gives
those who want more than they get for less than they like a right to
call someone else's desire for same "unworkable" _unilaterally_ is
just plain stupid: Your demands as a cheap customer is just as
"unworkable" to a vendor as you think their pricing is to you. This
goes for _everything_ in _any_ market.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * David Hanley <···@ncgr.org>
> | I don't there's _nothing_ they could do. I think they have a
> | decent product, but I consider their pricing structure unworkable.
>
> That's not it. You just _don't_ slam someone's technical or
> professional _ethics_ because they charge a price _you_ don't want
> to pay, when others clearly think it's OK to pay what they want.
Well, there's 2 issues. I'm not slamming their ethics personally,
I'm just saying their pricing structure is not good. And, while some
may be willing to pay their fees, other clearly are not. Do you
see lisp dominating the PC market? Lisp hasn't taken over, and
there's a short list of potential reasons.
> The cost of an Allegro CL license should be nearly irrelevant to
> most software projects, apart from the ones that are losing money by
> design -- i.e., open source and hobbyist projects.
No. I know you like them and are unable to be objective on this
subject, but despite the very high startup costs to get the environment,
you must also pay a substantial percentage of your gross profits when
you distribute software written with their product. Also, there will
be high fees if you run a server written with their product. For the
kinds of software it seems you do, these may not be major roadblocks.
For what I want to do, it's basically unacceptable.
> These have the
> ability to use the free Allegro CL Trial edition until they figure
> out that they need to make money.
No. You may not be aware, but that is a limited version, unless'
you are not interested in targeting windows, and these differ
anyways.
> Wanting more for less is quite universal,
yep.
> but thinking that it gives
> those who want more than they get for less than they like a right to
> call someone else's desire for same "unworkable" _unilaterally_ is
> just plain stupid:
No, eric, what's stupid is your desire to defend something which is
working for you in particular, while it may be good for you, and
not good for someone else.
I have as much right to point out problems with pricing and distribution
fees as you have to insult people for asking reasonable questions.
Another fact to consider for perspective is that there are other options;
LWW is *vastly* cheaper and doesn't charge distribution fees. Maybe
I would consider ACL more workable if it was the only option.
However, it is not.
> Your demands as a cheap customer is just as
> "unworkable" to a vendor as you think their pricing is to you. This
> goes for _everything_ in _any_ market.
That's a pretty stupid statement ( again ). LWW uses a perfectly
reasonable pricing structure, and I have paid for it. Apparently
they feel that non draconian pricing isn't "unreasonable." In fact,
franz is the only compiler vendor which I can think of off hand which
requires distribution fees for apps compiled with their product.
I didn't get their stuff largely because of their pricing structure, which
is my right for anything_in_any_market. :)
dave
* David Hanley <···@ncgr.org>
| And, while some may be willing to pay their fees, other clearly are
| not. Do you see lisp dominating the PC market? Lisp hasn't taken
| over, and there's a short list of potential reasons.
What else do you know about that is not dominating the PC market?
| I know you like them and are unable to be objective on this subject,
Wrong. I'm not you, so you should be careful about extrapolating
your own problems with being objective to me and others.
| but despite the very high startup costs to get the environment, you
| must also pay a substantial percentage of your gross profits when
| you distribute software written with their product.
I know their policies better than you do (partly because I argue
with Franz Inc over them all the time :) and I can tell you this:
You do not get a right to post hurtful lies about people just
because you disagree with them. If you had to pay for the cost of
recovering from people believing your lies, you would be run into
the ground financially. Relax and make the effort to engage your
_brain_ when you read the license. Above all, don't let emotional
responses like "they're taking my food money!" overpower your
reading skills and comprehension. Franz Inc is _not_ the IRS.
| Also, there will be high fees if you run a server written with their
| product. For the kinds of software it seems you do, these may not
| be major roadblocks. For what I want to do, it's basically
| unacceptable.
Your behavior towards me now and towards Franz Inc is unacceptable
no matter how free it is, so clearly, money is _not_ an issue for
you to behave morally reprehensibly. Guess what: Other people don't
change their behavior when they start to charge money for it, either.
| No, eric, what's stupid is your desire to defend something which is
| working for you in particular, while it may be good for you, and not
| good for someone else.
That's "Erik". I'm not _defending_ anything, I'm deflecting unfair
criticism. These are two very different tasks, but only people who
care about fairness realize the fundamental difference. I was
showing you that being unilateral in such an issue is bad, but that
was also before I knew what kind of person you are: You're the kind
of person who thinks that when you work with or like someone, you
lose your objectivity and you start to defend them as people, no
matter what they _do_. In my eyes, that means you are not the kind
of person _anyone_ should be working with. (Which also explains why
you can't afford to pay for a license, but I digress.)
| I have as much right to point out problems with pricing and
| distribution fees as you have to insult people for asking reasonable
| questions.
But you think you have much wider rights to insult people than I
ever _imagined_ I had! I don't protest your right to point out
problems. I point out a problem with the way you did it: That when
you do so in unilateral ways, you're setting yourself up to be shot
down because you're _unfair_ in your criticism. I crack down on
unfairness wherever and whenever I see it and I think I can make a
difference and whoever the perpetrator is, even when it's our
beloved government. However, I think it's crucial to be fair when
cracking down on unfairness and so I crack down publicly when the
unfairness is public and privately when it is unfairness is private.
Being unfair in public because of private gripes about pricing or
policy is simply _unacceptable_. (No, it doesn't matter what you
think _I'm_ doing. _You_ don't get better by finding somebody else
you can throw dirt at.)
| > Your demands as a cheap customer is just as
| > "unworkable" to a vendor as you think their pricing is to you. This
| > goes for _everything_ in _any_ market.
|
| That's a pretty stupid statement ( again ).
You're proving to me that you're seriously lacking in objectivity,
which is probably why you think others have to be similarly
crippled. Why are you hurting yourself by proving this?
| In fact, franz is the only compiler vendor which I can think of off
| hand which requires distribution fees for apps compiled with their
| product.
What you can think of off hand or not is really nobody's business
when you are as unable to think as you have demonstrated so far.
| I didn't get their stuff largely because of their pricing structure,
| which is my right for anything_in_any_market. :)
Of course it is. But you're doing the sour grapes thing by slamming
the people whose products you didn't want to buy, and that is _not_
a right you have or can assume will be defended. If you don't want
a product, actually _go_ elsewhere, don't linger and pout, OK?
FWIW, I'm in constant talks with Franz Inc about ways to minimize
the cost of licensing for small clients until they make real money.
I have one customer who thinks just having to _discuss_ licensing
terms is a serious turn-off, while money is not: They're paying
_much_ more to Oracle for their database application and the ODBC
drivers cost an arm and a leg, too. My biggest client is a serious
Sun and Oracle outfit and the Franz Inc license costs are not even a
blip in the budget. And we're already paying 50 times these costs a
year to Microsoft for _their_ software and various development tools.
So I wonder: Who do you people think you're kidding? Business costs
are supposed to be passed on to the _custumers_, and if not, to be
assumed by your shareholders or other investors when you fold or get
acquired by (more) competent people. You're supposed to make money
by being cleverer than the average bear, not lose it. If you can't
make enough money to pay a license fee, that means not even your
_investors_ believe in you! Don't take that out on Franz Inc, for
crying out loud.
What the software industry lacks is an easy way to go from hobbyist
to real software developer in outfits that have real money to spend.
As it is, we're looking at a very serious change of mindset and that
will necessarily create much needless hostility, but when people who
tend towards unfairness as a solution to their problems think they
have a _right_ to engage in it because they disagree over pricing or
licensing terms, there is no hope that they will grow a clue in the
foreseeable future, either.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> | but despite the very high startup costs to get the environment, you
> | must also pay a substantial percentage of your gross profits when
> | you distribute software written with their product.
>
> I know their policies better than you do (partly because I argue
> with Franz Inc over them all the time :) and I can tell you this:
> You do not get a right to post hurtful lies about people just
> because you disagree with them.
Erik, since you are accusing me of posting hurtful lies, please either:
1) Point them out, or
2) Admit you don't know what you're talking about.
Since all the rest of your post is your usual uninformed drool,
and not at all responding to what I said, I'm just ignoring it, and
waiting for someone to respond to what I actually did say.
dave
* David Hanley <···@ncgr.org>
| Erik Naggum wrote:
|
|
| > | but despite the very high startup costs to get the environment, you
| > | must also pay a substantial percentage of your gross profits when
| > | you distribute software written with their product.
| >
| > I know their policies better than you do (partly because I argue
| > with Franz Inc over them all the time :) and I can tell you this:
| > You do not get a right to post hurtful lies about people just
| > because you disagree with them.
|
| Erik, since you are accusing me of posting hurtful lies, please either:
|
| 1) Point them out, or
| 2) Admit you don't know what you're talking about.
And what will you do when it has been pointed out? Will you shut up
and actually _stop_ lying and slandering people? I didn't think so.
"You must pay a substantial percentage of your gross profits" is a
lie, pure and simple. It would be a lie without "substantial", and
it would be a lie without "gross". It is of course a lie with the
"must", as there are (a little too many) options on how to become a
value added reseller or purchase an unlimited or number of runtime
licenses with or without a compiler and/or development environment.
As stated, the unequivocal
Furthermore, "when you distribute software written with their
product" is a lie. It's not the software _you_ have _written_ with
their product that they charge a license for, it's their own, which
_you_ ship with _your_ product. There is nothing at all in the
license which prohibits writing the software with Allegro CL and
shipping with something else. There couldn't be. Theirs is a
commercial software license, not the GNU Public License, which
_does_ incur such requirements as you seem to find "unworkable".
You are a despicable liar, David Henley, but that's no surprise.
| Since all the rest of your post is your usual uninformed drool, and
| not at all responding to what I said, I'm just ignoring it, and
| waiting for someone to respond to what I actually did say.
I see. Your intellectual policy is to pretend that you have really
awful enemies so you don't have to think, care, or be fair to them.
That's why you have to make Franz Inc much worse than they really
are, too. This is called "demonization" and is a common "problem"
with certain mental illnesses, such as rabid paranoia, "rabid" not
being a clinical term, but applicable nonetheless. Pleas find a
mental health professional willing to lsiten and just get yourself
cured, David -- this has nothing to do with Lisp, me, Franz Inc, or
anything else even remotely connected to this newsgroup, anymore.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * David Hanley <···@ncgr.org>
> | Erik Naggum wrote:
> |
> |
> | > | but despite the very high startup costs to get the environment, you
> | > | must also pay a substantial percentage of your gross profits when
> | > | you distribute software written with their product.
> | >
> | > I know their policies better than you do (partly because I argue
> | > with Franz Inc over them all the time :) and I can tell you this:
> | > You do not get a right to post hurtful lies about people just
> | > because you disagree with them.
> |
> | Erik, since you are accusing me of posting hurtful lies, please either:
> |
> | 1) Point them out, or
> | 2) Admit you don't know what you're talking about.
>
> And what will you do when it has been pointed out? Will you shut up
> and actually _stop_ lying and slandering people? I didn't think so.
Well, if you were smart enough to make a point, yes I would.
However, I am still waiting.
>
>
> "You must pay a substantial percentage of your gross profits" is a
> lie, pure and simple. It would be a lie without "substantial", and
> it would be a lie without "gross".
Um, fine, but they're both in there.
still waiting for a response from the mental midget(you).
dave
David Hanley <···@ncgr.org> writes:
> Erik Naggum wrote:
>
> > "You must pay a substantial percentage of your gross profits" is a
> > lie, pure and simple. It would be a lie without "substantial", and
> > it would be a lie without "gross".
>
> Um, fine, but they're both in there.
Let me paraphrase.
He said the quoted statement "is a lie, pure and simple" with both of
them "in there", and then he added that it would remain a lie if
"substantial" (x?)or "gross" were taken out.
- j
Jeff Dalton wrote:
> David Hanley <···@ncgr.org> writes:
>
> > Erik Naggum wrote:
> >
> > > "You must pay a substantial percentage of your gross profits" is a
> > > lie, pure and simple. It would be a lie without "substantial", and
> > > it would be a lie without "gross".
> >
> > Um, fine, but they're both in there.
>
> Let me paraphrase.
>
> He said the quoted statement "is a lie, pure and simple" with both of
> them "in there", and then he added that it would remain a lie if
> "substantial" (x?)or "gross" were taken out.
Then failed to back it up. Typical.
dave
David Hanley wrote:
>
> Jeff Dalton wrote:
>
> > David Hanley <···@ncgr.org> writes:
> >
> > > Erik Naggum wrote:
> > >
> > > > "You must pay a substantial percentage of your gross profits" is a
> > > > lie, pure and simple. It would be a lie without "substantial", and
> > > > it would be a lie without "gross".
> > >
> > > Um, fine, but they're both in there.
> >
> > Let me paraphrase.
> >
> > He said the quoted statement "is a lie, pure and simple" with both of
> > them "in there", and then he added that it would remain a lie if
> > "substantial" (x?)or "gross" were taken out.
>
> Then failed to back it up. Typical.
Well, being another Franz customer and having had many discussions with
them on this point as well, I will back him up: you are just plain
wrong.
/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
David Hanley <···@ncgr.org> writes:
> Jeff Dalton wrote:
>
> > David Hanley <···@ncgr.org> writes:
> >
> > > Erik Naggum wrote:
> > >
> > > > "You must pay a substantial percentage of your gross profits" is a
> > > > lie, pure and simple. It would be a lie without "substantial", and
> > > > it would be a lie without "gross".
> > >
> > > Um, fine, but they're both in there.
> >
> > Let me paraphrase.
> >
> > He said the quoted statement "is a lie, pure and simple" with both of
> > them "in there", and then he added that it would remain a lie if
> > "substantial" (x?)or "gross" were taken out.
>
> Then failed to back it up. Typical.
It depends a bit on how you sell the resulting code. Sell it as source
code, for further customer development (a nice, but perhaps unworkable
commercial model?), with or without strings attached: Franz'll say
nothing.
Sell it compiled, with a Franz runtime included: cheaper than selling
it with at least one commercial transaction monitor.
Sell it as FASL files: Ask Franz.
With at least one sales model including "no payment to Franz" and at
least one being no more expensive than other, comparable,
situations...
//Ingvar (aigh! I'm defending Erik!)
--
It is by caffeine alone I set my mind in motion, it is by
the beans of Java that thoughts acquire speed, the hands
acquire shaking, the shaking becomes a warning, it is by
caffeine alone I set my mind in motion. <·····@aber.ac.uk?>
* Ingvar Mattsson <····@algonet.se>
| //Ingvar (aigh! I'm defending Erik!)
Wrong. You're not defending _me_. You're not agreeing with _me_,
either. You're defending something you, too, believe to be true,
quite independently of me or that I agree with _you_. If you have
to subordinate your defense of truth or what you believe in to who
else believes it, I seriously suggest you rethink your value system.
It is as bogus to defend _people_ because of what they say or do as
it is to attack _people_ because of what they say or do. Those who
confuse them naturally feel attacked or defended personally because
of what they say or do, and that complicates things immensely when
you want to praise or criticize opinions or actions. Let's mature
beyond such _unwarranted_ feelings of personal closeness and instead
concentrate on ideas and actions -- that way, people who have said
or done something can be praised or criticized without affecting
their ego, and they can remain free to say or do something else. If
we focus on people, we lock them into what they have said and have
done, and restrict their freedom to change their mind and their ways.
This is not to say that personal like or dislike is immaterial, but
there is much evidence to suggest that human psyche is unable to
deal with more than a handful of _personal_ relationships, and that
there is a good reason for professional conduct instead of personal
conduct, and thus of limiting one's like or dislike to the relevant
aspects of whoever you're dealing with. Those who are forced beyond
their limits get into serious trouble, such as nurses who "bond"
with patients or prostitutes, to take but two well-known examples.
Personal relationships are _very_ valuable, and they create a very
valuable emotion: that of personal obligations. Do not waste them
on people you don't know -- it _will_ hurt you and others may not
appreciate the implicit obligations, either.
With that serious matter out of the way, I appreciate that you took
the time to help me set the record straight. You have no obligation
to continue to help me, and I hope nobody attacks you for anything
you could only be assumed to agree with, but on USENET, people who
fail completely to distinguish between the action and whoever acts
_have_ been known to fail in other ways, as well.
#:Erik
--
If this is not what you expected, please alter your expectations.
* David Hanley <···@ncgr.org>
| Well, if you were smart enough to make a point, yes I would.
| However, I am still waiting.
Are you holding your breath? That would help.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * David Hanley <···@ncgr.org>
> | Well, if you were smart enough to make a point, yes I would.
> | However, I am still waiting.
>
> Are you holding your breath? That would help.
No, i'll let the moronic childishness be purely your tool.
dave
* David Hanley <···@ncgr.org>
| No, i'll let the moronic childishness be purely your tool.
I'm sometimes impressed with the ability of some people to work
themselves into a frenzy over fairly simple questions and arguments.
And sometimes I'm not. The inability of some people to focus on
anything but _persons_ is fairly unimpressive, for instance, and
when it's people they don't know, but instead constantly demonize,
it's also fairly uninspiring. I'm glad David does this to himself,
however, and there is very little point in helping him, now.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * David Hanley <···@ncgr.org>
> | No, i'll let the moronic childishness be purely your tool.
>
> I'm sometimes impressed with the ability of some people to work
> themselves into a frenzy over fairly simple questions and arguments.
That's ironic, considering that you have been been making long
angry posts based on my dislike of a pricing strunctutre, and
I have been stringing you along with tiny replies.
dave
* David Hanley <···@ncgr.org>
| That's ironic, considering that you have been been making long
| angry posts based on my dislike of a pricing strunctutre, and
| I have been stringing you along with tiny replies.
If you weren't so self-servingly delusional, you'd be amusing.
Go see a doctor, now.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote to David Hanley:
> | but despite the very high startup costs to get the environment, you
> | must also pay a substantial percentage of your gross profits when
> | you distribute software written with their product.
> [...] You do not get a right to post hurtful lies about people just
> because you disagree with them. If you had to pay for the cost of
> recovering from people believing your lies, you would be run into
> the ground financially.
I don't think David Hanley's comments are hurtful to anybody other than
himself. There is an implied "IMHO" before all subjective, relativistic
opinions (very high costs etc.). Customers and potential buyers just do
not pay attention to agenda-loaded blanket statements about vendors or
products. There is no need to lend credibility to otherwise ignored
utterance.
Robert
* David Hanley wrote:
> Well, there's 2 issues. I'm not slamming their ethics personally,
> I'm just saying their pricing structure is not good. And, while some
> may be willing to pay their fees, other clearly are not. Do you
> see lisp dominating the PC market? Lisp hasn't taken over, and
> there's a short list of potential reasons.
Well, I hate to point out flaws in your logic, but since (you imply I
think, I don't actually know the prices of either product) LispWorks
is much cheaper than Franz's offering, and Lisp still hasn't taken
over, it can't be to do with pricing as far as I can see.
I think a much more plausible reason is that Lisp is a language which
requires a good deal of sophistication to appreciate, and that the
educational system is just dismally failing to produce people with the
sophistication to get it. So let's do something positive and try and
fix *that* problem: go teach someone Lisp today!
--tim
Tim Bradshaw <···@cley.com> wrote in message
····················@cley.com...
> I think a much more plausible reason is that Lisp is a language which
> requires a good deal of sophistication to appreciate, and that the
> educational system is just dismally failing to produce people with the
> sophistication to get it. So let's do something positive and try and
> fix *that* problem: go teach someone Lisp today!
Yes there is a problem with how universities teach Lisp.
I recently converted to lisp a friend that had been teached lisp in
university and was convinced, like almost everybody, that lisp were only
good to write toy software like elisa or research software.
Now he is throwing out his C++ to write industrial software in lisp.
So C++-- (who said infix syntax is pretty ?) and (incf lisp) ; nicer syntax
Marc Battyani
Centuries ago, Nostradamus foresaw a time when Marc Battyani would say:
>Tim Bradshaw <···@cley.com> wrote in message
>····················@cley.com...
>
>> I think a much more plausible reason is that Lisp is a language which
>> requires a good deal of sophistication to appreciate, and that the
>> educational system is just dismally failing to produce people with the
>> sophistication to get it. So let's do something positive and try and
>> fix *that* problem: go teach someone Lisp today!
>
>Yes there is a problem with how universities teach Lisp.
>I recently converted to lisp a friend that had been teached lisp in
>university and was convinced, like almost everybody, that lisp were only
>good to write toy software like elisa or research software.
>Now he is throwing out his C++ to write industrial software in lisp.
>So C++-- (who said infix syntax is pretty ?) and (incf lisp) ; nicer syntax
The problem is that the typical way of teaching Lisp (or Scheme)
seems to be as a _part_ of a course, often on comparative programming
languages.
If all you spend on the course is three weeks, there's hardly room
to touch on anything other than LET, LAMBDA, DO, CAR, and CDR. With
a few arithmetic operations thrown in to make it look as if you're
doing something useful. And probably ASSOC.
The obvious conclusion from that sort of introduction is that:
--> Even BASIC has arrays. Much better than lists for any nontrivial
work.
--> C has structs; the only way to simulate that in Lisp is via
those horribly inefficient association lists.
--> I/O? What's I/O?
You _don't_ get around to VECTORs, DEFINE-STRUCTURE, or LOOP in the
typical course.
Students tend to get the gist of what they're taught; when they're
taught that the only point to Lisp is to have a section in a course
where you do useless things with it, it _must_ be pretty useless.
This is _exactly_ where all the "driveby flamings" tend to come from.
[There haven't been any lately; it appears that the masochists, missing
their floggings, have taken to pointless verbal abuse usually involving
someone and Erik Naggum, flaming one another. Perhaps comp.lang.lisp
needs a good attack by some really clueless proponent of C++...]
--
"No matter how much money you spend, you can't make a racehorse out of
a pig. You can, however, make an awfully fast pig."
-- An old saying about program efficiency
········@hex.net - - <http://www.ntlug.org/~cbbrowne/lsf.html>
········@knuth.brownes.org (Christopher Browne) writes:
....
> The obvious conclusion from that sort of introduction is that:
> --> Even BASIC has arrays. Much better than lists for any nontrivial
> work.
> --> C has structs; the only way to simulate that in Lisp is via
> those horribly inefficient association lists.
> --> I/O? What's I/O?
>
> You _don't_ get around to VECTORs, DEFINE-STRUCTURE, or LOOP in the
^^^^^^^^^^^^^^^^
#+noise-on
DEFINE-STRUCTURE does not quite exist! (I mean, in the "standard"
documents)
CL has DEFSTRUCT, Scheme has ....... what does Scheme have? :)
Cheers
--
Marco Antoniotti ===========================================
>>>>> "Christopher" == Christopher Browne <········@knuth.brownes.org> writes:
Christopher> CL may have DEFSTRUCT, but the subset that instructors
Christopher> present to students is highly unlikely to include that
Christopher> portion of the vocabulary, just as it is highly
Christopher> unlikely to include arrays, LOOP, or, for that matter,
Christopher> hash tables.
I teach Lisp (CIS459.31) at Ohio State University. It's a one
credit-hour course that lasts one quarter (10 weeks). Just this year,
I switched primary texts to Graham's _ANSI_Common_Lisp_.
We cover structures, arrays, and hash tables. We discuss LOOP but not
in detail.
The class has five labs. Lab 4 requires students to write their own
program that will demonstrate use of arrays, structures, or hash
tables. I've seen folks implement things from Tic-Tac-Toe games to
databases to keep track of wines in a collection.
One quarter isn't long at all, and being only one credit-hour doesn't
help for the presentation of such a big language. But clearly,
(most) students _can_ and _do_ get it.
Perhaps proper instruction of Lisp is lacking in many places or in
general. I do think that by painting so broad a stroke, you're doing
those of us who teach Lisp and care about doing it well a great
disservice.
--
Matt Curtin ········@interhack.net http://www.interhack.net/people/cmcurtin/
Centuries ago, Nostradamus foresaw a time when Matt Curtin would say:
>>>>>> "Christopher" == Christopher Browne <········@knuth.brownes.org> writes:
>
> Christopher> CL may have DEFSTRUCT, but the subset that instructors
> Christopher> present to students is highly unlikely to include that
> Christopher> portion of the vocabulary, just as it is highly
> Christopher> unlikely to include arrays, LOOP, or, for that matter,
> Christopher> hash tables.
>
>I teach Lisp (CIS459.31) at Ohio State University. It's a one
>credit-hour course that lasts one quarter (10 weeks). Just this year,
>I switched primary texts to Graham's _ANSI_Common_Lisp_.
>
>We cover structures, arrays, and hash tables. We discuss LOOP but not
>in detail.
Sounds rather good.
>The class has five labs. Lab 4 requires students to write their own
>program that will demonstrate use of arrays, structures, or hash
>tables. I've seen folks implement things from Tic-Tac-Toe games to
>databases to keep track of wines in a collection.
>
>One quarter isn't long at all, and being only one credit-hour doesn't
>help for the presentation of such a big language. But clearly,
>(most) students _can_ and _do_ get it.
>
>Perhaps proper instruction of Lisp is lacking in many places or in
>general. I do think that by painting so broad a stroke, you're doing
>those of us who teach Lisp and care about doing it well a great
>disservice.
I'd tend to think that if Graham's book is used as the text,
it would be fairly difficult for an instructor to simultaneously
contend the relative "obsolescence" of Lisp.
I'm certainly not trying to preclude the thought that there are
indeed good presentations out there; good teaching is surely a
laudable thing.
--
········@hex.net - <http://www.ntlug.org/~cbbrowne/linux.html>
Microsoft: The Scientology of Computing
>>>>> "Christopher" == Christopher Browne <········@knuth.brownes.org> writes:
Christopher> I'm certainly not trying to preclude the thought that
Christopher> there are indeed good presentations out there; good
Christopher> teaching is surely a laudable thing.
At the same time, I readily recognize that there are serious problems
in many cases with how Lisp is taught. I don't think this is as much
an issue with Lisp as it is with teaching in general. Some folks are
just _boring_. Others don't really know their topic very well.
The big problem that Lisp does face in education is that there is the
possibility that what little folks do know about Lisp isn't highly
relevant to anything, or they'll have some general prejudice about
something that's so "old". (Strange how in any other context, that
sort of thing isn't called "old" or "obsolete", it's called "history"
and "culture", both generally considered Good Things.)
Combine that with the relatively little amount of noise that is made
about Lisp and the need for Lisp programmers in the industry and there
just isn't much in the way of incentive for folks to learn the
language well where there isn't someone there to give students a good
introduction. Anyone can look at the amount of hype that surrounds
Java and see that even where there isn't lots in the way of formal
education about a language, people can see that it's big enough to be
motivated to do what's necessary to learn it.
People sometimes ask me why there aren't billions of ads for Lisp
programmers, why they don't hear a lot about the need for there to be
more Lisp programmers, etc. I like to answer thusly:
o How much of all software development is maintenance work?
(Answers vary, but it's always a LOT, typically in the 80-95%
range.)
o If programs are written correctly the first time, there aren't
going to be lots of bugs that need to be fixed.
o If programs are designed well from the beginning, they should be
able to do quite a lot without needing to have lots of code changes
made.
o A language that helps programmers write good programs will
therefore require much less maintenance work.
o If the vast majority of programming is maintenance work, the need
for maintenance programmers will largely be dictated by how badly
the programs are being written. That can be heavily influenced by
the language choice.
So the answer, of course, is that there is a need for Lisp programmers
to build new systems. But there isn't much need for Lisp programmers
to fix all the junk that everyone else has already written four times.
As a result, there will be less "visibility" of the language.
Am I exaggerating? Probably. Is there an element of truth in what
I'm saying when I answer this way? I think so. But in what other way
could a genuine Lisp programmer answer? I think "getting it right
early on" is an important element of Lisp itself and its culture. I
like to convey that to students.
--
Matt Curtin ········@interhack.net http://www.interhack.net/people/cmcurtin/
········@knuth.brownes.org (Christopher Browne) writes:
> >Yes there is a problem with how universities teach Lisp. [...]
> The problem is that the typical way of teaching Lisp (or Scheme)
> seems to be as a _part_ of a course, often on comparative programming
> languages.
True, but even courses devoted to Lisp can give students a misleading
impression of the language.
Some trivial signs that Lisp is likely to end up looking bad: "Lisp" is
written as "LISP"; RPLACA and RPLACD get an earlier or more prominent
mention than SETF; you're told that EQ is only for comparing atoms
(or symbols); EQ is first introduced in a discussion of destructive
modification of lists; you're told that in Lisp "everything is a
list"; Lisp is called "an interpreted language"; Lisp is called "an AI
language"; you're given the impression that no one likes Lisp's syntax
but they put up with it in order to get the programming environment or
because it makes macros easy; the instructor keep apologising for
Lisp's syntax or for its use of names such as LAMBDA, COND, CAR, and
CDR; you're told that "CAR" stands for (rather than originally stood
for) "Contents of Address Register".
One of the worst problems, I think, is that courses often fail to get
students to understand Lisp data as objects, "objects" meaning that
they have an identity, live in their own storage, and so on (rather
than that they're instances of classes). From that, it will make
sense - in the sense that it forms a coherent picture - that procedure
calls and assignments don't copy, that EQ and EQL can test object
identity, that it's possible to determine the type of an object at
run-time, that types are primarily associated with objects rather
than variables or other places where (refs to) objects are stored;
that hash tables can make connections between particular objects;
that lists, arrays, etc, can contain arbitrary objects rather than
being specialised to some particular element type; that it's easy
to implement simple O-O "extensions"; and so on.
But courses that just do list processing in Lisp never get there.
> Students tend to get the gist of what they're taught; when they're
> taught that the only point to Lisp is to have a section in a course
> where you do useless things with it, it _must_ be pretty useless.
Yes, and when students are required to use Lisp and find it an
unpleasant experience, they tend to resent it, and this can turn
into an active dislike of the language.
> This is _exactly_ where all the "driveby flamings" tend to come from.
I think so too, in many cases at least.
Jeff Dalton <····@todday.aiai.ed.ac.uk> writes:
> ········@knuth.brownes.org (Christopher Browne) writes:
>
> > >Yes there is a problem with how universities teach Lisp. [...]
>
> > The problem is that the typical way of teaching Lisp (or Scheme)
> > seems to be as a _part_ of a course, often on comparative programming
> > languages.
>
> True, but even courses devoted to Lisp can give students a misleading
> impression of the language.
Forgive my post if I tend to ramble as I am still basking in the glow of
my newfound Lisp awareness and slowly catching up on news, but luckily I
have a counter example to share. I agree that in many instances what
you describe is the case, but I don't see why it has to be, or rather
why it tends to be as such. One of my co-workers teaches a class on
Lisp. It is a one credit hour technical elective, but the amount of
information and the impression students are left with is not even close
to what you have described in most cases, not that what you said
incorrect.
> Some trivial signs that Lisp is likely to end up looking bad: "Lisp" is
> written as "LISP"; RPLACA and RPLACD get an earlier or more prominent
> mention than SETF; you're told that EQ is only for comparing atoms
> (or symbols); EQ is first introduced in a discussion of destructive
> modification of lists; you're told that in Lisp "everything is a
> list"; Lisp is called "an interpreted language"; Lisp is called "an AI
> language"; you're given the impression that no one likes Lisp's syntax
> but they put up with it in order to get the programming environment or
> because it makes macros easy; the instructor keep apologising for
> Lisp's syntax or for its use of names such as LAMBDA, COND, CAR, and
> CDR; you're told that "CAR" stands for (rather than originally stood
> for) "Contents of Address Register".
It seems to me that if an instructor is always apologizing for Lisp's
syntax they just don't get it or they don't know how to teach it.
Perhaps that is the case many times. In the Lisp class I am referring to
other languages are poked fun of a bit in fact :-) The text for this
class is Graham's ANSI Common Lisp. If you stretch it a bit, you can go
all the way to CLOS (Chapter 11 non-inclusive) in just one quarter.
That may seem like a lot to do in one quarter, but I have been going
through the book on my own doing _all_ the exercises and I know that it
can be done (sans doing all the exercises). Most of the students keep
up, but a lot of that has to do with the quality of the instructor and
the amount of clues he can lay claim to. In the least they are left
with quite a different impression of Lisp than described here. Perhaps
we are just lucky!
> One of the worst problems, I think, is that courses often fail to get
> students to understand Lisp data as objects, "objects" meaning that
> they have an identity, live in their own storage, and so on (rather
> than that they're instances of classes). From that, it will make
> sense - in the sense that it forms a coherent picture - that procedure
> calls and assignments don't copy, that EQ and EQL can test object
> identity, that it's possible to determine the type of an object at
> run-time, that types are primarily associated with objects rather
> than variables or other places where (refs to) objects are stored;
> that hash tables can make connections between particular objects;
> that lists, arrays, etc, can contain arbitrary objects rather than
> being specialised to some particular element type; that it's easy
> to implement simple O-O "extensions"; and so on.
I feel that there is a fundamental flaw with how CS students are taught
in general. I see some courses and am left just asking myself "Why?" I
don't want to get into specifics, unfortunately, but I am pretty
discontent with the whole thing really. I see a desire to teach a
paradigm where many limitations of certain languages are worked around
and ideas like software reuse and OO techniques are employed, and I am
left wondering why the implementation language used to teach these
concepts is not Lisp or something more along those lines. Why?
How hard are these concepts to grasp really? Is it that difficult to
teach functional programming? I just don't get it. I don't see how
many people just cannot put their heads around the concepts you just
listed. I think that the fact that one paradigm is drilled into their
heads from day one while everything else is left as an after thought
might have something to do with it. To me this is a serious flaw.
> But courses that just do list processing in Lisp never get there.
Very true, but this is not how it has to be, even for a one credit hour
class.
> > Students tend to get the gist of what they're taught; when they're
> > taught that the only point to Lisp is to have a section in a course
> > where you do useless things with it, it _must_ be pretty useless.
>
> Yes, and when students are required to use Lisp and find it an
> unpleasant experience, they tend to resent it, and this can turn
> into an active dislike of the language.
Does anyone feel that Lisp or scheme should be taught first in a CS
program? It seems to me that this would give people a real chance to
learn something. My academic experience, which is highly unusual to say
the, involved learning C++ first, then C. I found these languages
fairly easy to grasp even if they were taught haphazardly. Then
outside of the academic environment I discover things like Lisp where
there are greater possibilities. Now, I am not yet an expert on Lisp,
but I can feel it whenever I read a new chapter out of Graham's ANSI
Common Lisp or work on the chapter questions. I can catch a glimpse of
something greater, where I am not limited in many ways. It's like I
have a whole new way of approaching problems that I would not have had
otherwise. It seems to me that I would have had a much more productive
time if I would have been introduced to Lisp first.
> > This is _exactly_ where all the "driveby flamings" tend to come from.
>
> I think so too, in many cases at least.
Those who perpetrate "driveby flamings" scare me. These are most likely
the same people who write the following code:
while (1) {
fork();
}
printf("Got here.\n");
Sorry about the C there, but I didn't make that up. I see this all the
time, and since I am the system administrator I know when the labs that
deal with these basic concepts are due, and in classes where these
concepts should already have been learned. It may seem like people
writing code like this is a fluke, but let me tell you, a lot of people
who are moderately advanced in CS programs do it, and not just here.
Much to my dismay I am left wondering if most of the people who get CS
degrees really know anything at all.
--
Shaun Rowland ·······@cis.ohio-state.edu
http://www.cis.ohio-state.edu/~rowland/
So after reading random stuff from this thread, here's my $0.02 about
what should be taught first:
> Does anyone feel that Lisp or scheme should be taught first in a CS
> program? It seems to me that this would give people a real chance
> to learn something. My academic experience, which is highly unusual
> to say the, involved learning C++ first, then C.
here's what I believe to be true about human beings. Let's first make
the generalization that the "city" is better than the "country" in the
same way that "Common Lisp" is better than "C" (note: not Lisp, but
particularly CL; I am not talking about Scheme here). This is just an
assumption, and I'm trying to make a point.
It seems, from my experience, that if you take a country boy, born and
raised in the country, and then move him to the city, and he
assimilates, and gets used to it, then it's very hard for him to go
back; much harder, in fact, than for a city boy, born and raised in
the city, to go to the country, get used to it out there (farming or
whatever) and then return to the city.
Applying this [mere generalization] to the programming question of
which language to teach first, it would make sense to teach C first
and CL second. Because then programmers "know" C, but once they've
"seen the better way" it's hard to go back. Given the choice, they'd
rather "live in the city" now.
This is basically already what is happening. _very_ few
students/programmers see CL before either C, C++, Pascal, Basic, or
Java. And this is kinda the way I like it. It means that nearly ALL
CL programmers know another language, and that makes CL prorammers
wiser, more knowledgable, and more versatile. I'd rather that the CL
community remain small, but that its members be, on average, "wiser"
than other programmers. Does this sound elitist? Maybe. Probably.
But I think it's the right way. I don't want to see high school kids
and your typical Perl scripting hackers having a serious impact on CL.
The other side of the coin here is that with the influence of features
in other languages, people will move over to CL and then miss things
not present in CL, but in other languages, and so the trend will be to
wash CL into the other languages. But it's good for CL to have users
who know about features in other languages.
My only gripe is that Common Lisp is not marketed properly to many
people who would really benefit from it. I'll just recount a story.
I met a guy a year or so ago who graduated CMU with an advanced degree
in computer science, and then came to MIT to do his PhD. He was doing
all sorts of complex analyses, proofs, etc. We were talking, and he
said he used C. I said something like "coming from CMU, I'd have
figured that you, especially, would use CL for your work..." He then
said something like "Are you crazy? that language sucks! it's not
even compiled!". And this is someone who got his B.S. and M.S. at
CMU. So go figure. people just arn't getting the right information.
they're also not getting the opportunities to use CL with the right
level of instruction.
dave
David Bakhash <·····@alum.mit.edu> writes:
> But I think it's the right way. I don't want to see high school kids
> and your typical Perl scripting hackers having a serious impact on CL.
Um, since I just now finished 11th grade, does that mean I should go
away now?
The REST of it was quite well written... =)
--
vsync
http://quadium.net/ - last updated Thu Jun 1 23:57:30 MDT 2000
Orjner.
vsync <·····@quadium.net> writes:
> David Bakhash <·····@alum.mit.edu> writes:
>
> > But I think it's the right way. I don't want to see high school kids
> > and your typical Perl scripting hackers having a serious impact on CL.
>
> Um, since I just now finished 11th grade, does that mean I should go
> away now?
This is why I hate generalizations. You must consider what I'm saying
in context.
Trying to standardize and perfect a programming language comes from a
maturity in programming that tends to come over time. I'm 26, and still
find myself learning tons, both about CL and about programming in
general. that's all I was trying to say.
dave
David Bakhash <·····@alum.mit.edu> writes:
> vsync <·····@quadium.net> writes:
>
> > David Bakhash <·····@alum.mit.edu> writes:
> >
> > > But I think it's the right way. I don't want to see high school kids
> > > and your typical Perl scripting hackers having a serious impact on CL.
> >
> > Um, since I just now finished 11th grade, does that mean I should go
> > away now?
>
> This is why I hate generalizations. You must consider what I'm saying
> in context.
Sorry, humor's kind of hard to convey on Usenet... I think I had a
smiley in there somewhere. Anyway, I did get your point and I
agree with you.
Unfortunately, the way society is, a more "elite" language will just
be regarded as somehow bad. This happens at work all the time. If I
need to implement something, some of the managers (not all, thank
goodness) request details about how I'm doing it. Most of the time,
these managers don't understand the information they're given, and it
ends up being that if they've heard of the tool I'm using, they like
it. If they haven't, they don't, and no additional information will
change their minds.
If programmers themselves often don't understand what Lisp is, as in
your example of the CMU student, then this will translate to even less
acceptance among those who make the decisions. This isn't to say that
you aren't right, but I don't think there really is a good solution.
Personally, I wish they'd trust those of us who do the actual work to
be able pick the right tool for the job, but... =)
> Trying to standardize and perfect a programming language comes from a
> maturity in programming that tends to come over time. I'm 26, and still
> find myself learning tons, both about CL and about programming in
> general. that's all I was trying to say.
Very well stated. I think if a person isn't willing to learn, they
shouldn't be allowed anywhere near a computer (but that's another
topic =), much less programming, because there's so much exploration
involved. I think one of my favorite things about Lisp is the way it
encourages such exploration. Instead of having to lay out every
detail of a program beforehand, with a huge cost if I change my mind
later, I can just let it grow organically as I add features. Lisp
really puts fun and adventure into programming, which is as it should
be.
--
vsync
http://quadium.net/ - last updated Thu Jun 1 23:57:30 MDT 2000
Orjner.
David Bakhash <·····@alum.mit.edu> writes:
> Applying this [mere generalization] to the programming question of
> which language to teach first, it would make sense to teach C first
> and CL second. Because then programmers "know" C, but once they've
> "seen the better way" it's hard to go back. Given the choice, they'd
> rather "live in the city" now.
I can follow your logic, and it seems sound to me. I believe that I
wrote more of my own experiences in there than I wanted to. The reason
that I was thinking of teaching Lisp first is because it would give
students the exposure to a new language they might not otherwise get.
The problem in the one credit hour technical elective example is the
class is optional. There are a few different technical electives that a
student can take in fact, and the Lisp class is only offered one quarter
of the year. That's very unfortunate in my eyes. I could go on for
hours about how I think what is actually learned by most students hurts
them in ways, but I won't. I just wish that the Lisp class was
required. I think that everyone would benefit from the experience.
What is even worse in this situation is that one must know what Lisp is
already. How else would one know to take the class in the first place?
There is not much motivation given the C/C++ "focus" in other classes
unless clues about what Lisp is and why it might be important to learn
are already present. If Lisp were taught first it would still be
necessary to learn C or C++. This would accomplish my goal, but the
reasons for my general question are too specific.
> This is basically already what is happening. _very_ few
> students/programmers see CL before either C, C++, Pascal, Basic, or
> Java. And this is kinda the way I like it. It means that nearly ALL
> CL programmers know another language, and that makes CL prorammers
> wiser, more knowledgable, and more versatile. I'd rather that the CL
> community remain small, but that its members be, on average, "wiser"
> than other programmers. Does this sound elitist? Maybe. Probably.
> But I think it's the right way. I don't want to see high school kids
> and your typical Perl scripting hackers having a serious impact on CL.
I can see the benefits you state here, but I don't agree with the last
sentence in this context. The students in a CS program are not supposed
to be just typical Perl scripting hackers. If you want that go to a
technical college or learn it on your own.
> My only gripe is that Common Lisp is not marketed properly to many
> people who would really benefit from it. I'll just recount a story.
> I met a guy a year or so ago who graduated CMU with an advanced degree
> in computer science, and then came to MIT to do his PhD. He was doing
> all sorts of complex analyses, proofs, etc. We were talking, and he
> said he used C. I said something like "coming from CMU, I'd have
> figured that you, especially, would use CL for your work..." He then
> said something like "Are you crazy? that language sucks! it's not
> even compiled!". And this is someone who got his B.S. and M.S. at
> CMU. So go figure. people just arn't getting the right information.
> they're also not getting the opportunities to use CL with the right
> level of instruction.
Right, but I don't take too much stock in what people with B.S. and
M.S. degrees tell me. I have too much experience dealing with them :-)
Unfortunately people with B.S. and M.S. degrees often do a good job
representing their own population, just not a good job representing the
real truth that can be found in Lisp. Now that probably sounds elitist!
None of what I am saying here means that Lisp is the one true way for
everyone. I would rather see students receive good exposure to Lisp
along with other languages. If that were happening I don't believe we
would see answers like "that language sucks! It's not even compiled!"
Any objections would be more along the lines of what I expect from
someone with an advanced degree in computer science and not someone who
has no real idea of the benefits of one language over another for
certain applications. Often times I believe I expect too much.
--
Shaun Rowland ·······@cis.ohio-state.edu
http://www.cis.ohio-state.edu/~rowland/
> CL programmers know another language, and that makes CL prorammers
> wiser, more knowledgable, and more versatile. I'd rather that the
CL
> community remain small, but that its members be, on average, "wiser"
> than other programmers. Does this sound elitist? Maybe. Probably.
> But I think it's the right way. I don't want to see high school
kids
> and your typical Perl scripting hackers having a serious impact on
CL.
Umm, I'm confused. How does using have an impact, particularly
a negative one? (Perl went non-linear because Larry Wall is
non-linear.)
Precisely what is the virtue in being a back-water language?
A lisp could, and arguably should, have had the position that
Python is getting. Yet, due to the lisp communities insistence
on fragmentation and not acknowledging the importance of interaction
with the outside world, Python is going to be important, and lisp
is, well, lisp.... (Python is interpreted as bytecodes.)
This story has been repeated a number of times. Why is this
a good thing?
-andy
Sent via Deja.com http://www.deja.com/
Before you buy.
In article <············@nnrp1.deja.com>,
Andy Freeman <······@earthlink.net> wrote:
> This story has been repeated a number of times. Why is this
> a good thing?
I mentioned Python because that's the best choice for what
I'm doing now.
After I hit "send", I remembered a dinner that I had with a
very senior executive at an e-commerce company. I asked a
meaningless question and was rather floored by the response
"you know, XML could have been lisp, but ...." That wasn't
an answer to my question, and I doubt that he knew that I
knew lisp, let alone preferred it, but it's still interesting.
When I have to start processing XML with Python, I'm not
going to think fondly of the people who find virtue in
keeping lisp in the back-water. It's not pure, it's stagnant.
-andy
Sent via Deja.com http://www.deja.com/
Before you buy.
David Bakhash <·····@alum.mit.edu> writes:
> So after reading random stuff from this thread, here's my $0.02 about
> what should be taught first:
>
> > Does anyone feel that Lisp or scheme should be taught first in a CS
> > program? It seems to me that this would give people a real chance
> > to learn something. My academic experience, which is highly unusual
> > to say the, involved learning C++ first, then C.
>
> here's what I believe to be true about human beings. Let's first make
> the generalization that the "city" is better than the "country" in the
> same way that "Common Lisp" is better than "C" (note: not Lisp, but
> particularly CL; I am not talking about Scheme here). This is just an
> assumption, and I'm trying to make a point.
>
> It seems, from my experience, that if you take a country boy, born and
> raised in the country, and then move him to the city, and he
> assimilates, and gets used to it, then it's very hard for him to go
> back; much harder, in fact, than for a city boy, born and raised in
> the city, to go to the country, get used to it out there (farming or
> whatever) and then return to the city.
>
> Applying this [mere generalization] to the programming question of
> which language to teach first, it would make sense to teach C first
> and CL second. Because then programmers "know" C, but once they've
> "seen the better way" it's hard to go back. Given the choice, they'd
> rather "live in the city" now.
>
> This is basically already what is happening. _very_ few
> students/programmers see CL before either C, C++, Pascal, Basic, or
> Java.
the problem, as i see it, is that very few students/programmers see CL
at all. as an electrical engineer, i can tell you i was never exposed
to it in any class.
if they do, the don't see it before learning *all of* C, C++, pascal,
basic *and* java. then they see lisp. they think -- lisp looks
weird. it's got all these parenthesis and the professor gives three
minor projects having to do with flattening lists. student gets the
idea that
1) flattening lists is a pointless exercise.
2) it was harder than it ought to be.
3) dealing with lists was supposed to be lisp's strength.
4) lisp sucks.
this is *not* the impression i want to have lisp give.
if you have people fluent in lisp before seeing C++, maybe they will
react in appropriate horror to that. as it is, they put up with C++
because it's more like things they know. once someone has invested
the effort to be good at C++ it will be *very* hard to expand their
minds. no one wants to admit they are wrong.
> And this is kinda the way I like it. It means that nearly ALL
> CL programmers know another language, and that makes CL prorammers
> wiser, more knowledgable, and more versatile.
it also means there are too damn few lisp programmers. this means
that
1) lisp compilers and tools are expensive
2) no more lisp machines
3) significant bit rot on useful things like window systems, number
crunching &c.
i mean, lisp would be *the* natural language for window based
programming. look how well emacs does with reacting to keystrokes.
C++ limps along with class inheritance, but it doesn't have lambda.
CL has functional *and* OOP style. it's a natural for windows (not
microsoft but in general). but does everyone have nice lisp based
windowing systems? noooooo we are struggling with monstrosities
constructed in C++. why? because lisp isn't near popular enough.
> I'd rather that the CL
> community remain small, but that its members be, on average, "wiser"
> than other programmers. Does this sound elitist? Maybe. Probably.
> But I think it's the right way. I don't want to see high school kids
> and your typical Perl scripting hackers having a serious impact on
> CL.
what i am afraid of is worse is better and we will, in the end, all be
reduced to windows visual basic.
lots of idiots knowing lisp doesn't bother me in the least. it will
mean lots of smart people knowing lisp too. a language isn't good or
bad just on popularity alone, but being popular helps. sometimes,
more really *is* more.
> The other side of the coin here is that with the influence of features
> in other languages, people will move over to CL and then miss things
> not present in CL, but in other languages, and so the trend will be to
> wash CL into the other languages. But it's good for CL to have users
> who know about features in other languages.
i'd be happy if CL just had users period. (yes, i know CL does have
users, but i want more of them!)
> My only gripe is that Common Lisp is not marketed properly to many
> people who would really benefit from it. I'll just recount a story.
> I met a guy a year or so ago who graduated CMU with an advanced degree
> in computer science, and then came to MIT to do his PhD. He was doing
> all sorts of complex analyses, proofs, etc. We were talking, and he
> said he used C. I said something like "coming from CMU, I'd have
> figured that you, especially, would use CL for your work..." He then
> said something like "Are you crazy? that language sucks! it's not
> even compiled!". And this is someone who got his B.S. and M.S. at
> CMU. So go figure. people just arn't getting the right information.
> they're also not getting the opportunities to use CL with the right
> level of instruction.
i agree totally with this last paragraph.
--
johan kullstam l72t00052
Johan Kullstam writes:
> i mean, lisp would be *the* natural language for window based
> programming. look how well emacs does with reacting to keystrokes.
> C++ limps along with class inheritance, but it doesn't have lambda.
> CL has functional *and* OOP style. it's a natural for windows (not
> microsoft but in general). but does everyone have nice lisp based
> windowing systems? noooooo we are struggling with monstrosities
> constructed in C++. why? because lisp isn't near popular enough.
I have to agree with this, but I don't know what to do about it.
(Write software, I suppose.) I'm a CL dabbler; I like it, it's fun,
and it's a good source of ideas. Over the weekend, though, I found
myself re-implementing a miniature version of CLOS generic functions
in Python, because that style of modularity fit what I was doing
better. I was just doing simple GUI development, telling objects how
to use the GUI to display themselves, edit themselves, and so on.
It's a beautiful fit; the GUI stuff stays in the GUI methods, while
everything else stays in the more basic methods. Wow. A light goes
on.
I don't quite know where to go from here, but I'm back to watching CL
very carefully. I like the feeling.
--
Johann Hibschman ······@physics.berkeley.edu
Jeff Dalton <····@todday.aiai.ed.ac.uk> writes:
> ········@knuth.brownes.org (Christopher Browne) writes:
>
> > >Yes there is a problem with how universities teach Lisp. [...]
>
> > The problem is that the typical way of teaching Lisp (or Scheme)
> > seems to be as a _part_ of a course, often on comparative programming
> > languages.
>
> True, but even courses devoted to Lisp can give students a misleading
> impression of the language.
Forgive my post if I tend to ramble as I am still basking in the glow of
my newfound Lisp awareness and slowly catching up on news, but luckily I
have a counter example to share. I agree that in many instances what
you describe is the case, but I don't see why it has to be, or rather
why it tends to be as such. One of my co-workers teaches a class on
Lisp. It is a one credit hour technical elective, but the amount of
information and the impression students are left with is not even close
to what you have described in most cases, not that what you said is
incorrect.
> Some trivial signs that Lisp is likely to end up looking bad: "Lisp" is
> written as "LISP"; RPLACA and RPLACD get an earlier or more prominent
> mention than SETF; you're told that EQ is only for comparing atoms
> (or symbols); EQ is first introduced in a discussion of destructive
> modification of lists; you're told that in Lisp "everything is a
> list"; Lisp is called "an interpreted language"; Lisp is called "an AI
> language"; you're given the impression that no one likes Lisp's syntax
> but they put up with it in order to get the programming environment or
> because it makes macros easy; the instructor keep apologising for
> Lisp's syntax or for its use of names such as LAMBDA, COND, CAR, and
> CDR; you're told that "CAR" stands for (rather than originally stood
> for) "Contents of Address Register".
It seems to me that if an instructor is always apologizing for Lisp's
syntax they just don't get it or they don't know how to teach it.
Perhaps that is the case many times. In the Lisp class I am referring to
other languages are poked fun of a bit in fact :-) The text for this
class is Graham's ANSI Common Lisp. If you stretch it a bit, you can go
all the way to CLOS (Chapter 11 non-inclusive) in just one quarter.
That may seem like a lot to do in one quarter, but I have been going
through the book on my own doing _all_ the exercises and I know that it
can be done (sans doing all the exercises). Most of the students keep
up, but a lot of that has to do with the quality of the instructor and
the amount of clues he can lay claim to. In the least they are left
with quite a different impression of Lisp than described here. Perhaps
we are just lucky!
> One of the worst problems, I think, is that courses often fail to get
> students to understand Lisp data as objects, "objects" meaning that
> they have an identity, live in their own storage, and so on (rather
> than that they're instances of classes). From that, it will make
> sense - in the sense that it forms a coherent picture - that procedure
> calls and assignments don't copy, that EQ and EQL can test object
> identity, that it's possible to determine the type of an object at
> run-time, that types are primarily associated with objects rather
> than variables or other places where (refs to) objects are stored;
> that hash tables can make connections between particular objects;
> that lists, arrays, etc, can contain arbitrary objects rather than
> being specialised to some particular element type; that it's easy
> to implement simple O-O "extensions"; and so on.
I feel that there is a fundamental flaw with how CS students are taught
in general. I see some courses and am left just asking myself "Why?" I
don't want to get into specifics, unfortunately, but I am pretty
discontent with the whole thing really. I see a desire to teach a
paradigm where many limitations of certain languages are worked around
and ideas like software reuse and OO techniques are employed, and I am
left wondering why the implementation language used to teach these
concepts is not Lisp or something more along those lines. Why?
How hard are these concepts to grasp really? Is it that difficult to
teach functional programming? I just don't get it. I don't see how
many people just cannot put their heads around the concepts you just
listed. I think that the fact that one paradigm is drilled into their
heads from day one while everything else is left as an after thought
might have something to do with it. To me this is a serious flaw.
> But courses that just do list processing in Lisp never get there.
Very true, but this is not how it has to be, even for a one credit hour
class.
> > Students tend to get the gist of what they're taught; when they're
> > taught that the only point to Lisp is to have a section in a course
> > where you do useless things with it, it _must_ be pretty useless.
>
> Yes, and when students are required to use Lisp and find it an
> unpleasant experience, they tend to resent it, and this can turn
> into an active dislike of the language.
Does anyone feel that Lisp or scheme should be taught first in a CS
program? It seems to me that this would give people a real chance to
learn something. My academic experience, which is highly unusual to say
the least, involved learning C++ first then just working with C. I
found these languages fairly easy to grasp even if they were taught
haphazardly. Then outside of the academic environment I discover things
like Lisp where there are greater possibilities. I am not an expert on
Lisp, but I can feel it whenever I read a new chapter out of Graham's
ANSI Common Lisp or work on the chapter questions. I can catch a
glimpse of something greater, where I am not limited in many ways. It's
like I have a whole new way of approaching problems that I would not
have had otherwise. It seems to me that I would have had a much more
productive time if I would have been introduced to Lisp first.
> > This is _exactly_ where all the "driveby flamings" tend to come from.
>
> I think so too, in many cases at least.
Those who perpetrate "driveby flamings" scare me. These are most likely
the same people who write the following code:
while (1) {
fork();
}
printf("Got here.\n");
Sorry about the C there, but I didn't make that up. I see this all the
time, and since I am the system administrator I know when the labs that
deal with these basic concepts are due, and in classes where these
concepts should already have been learned. It may seem like people
writing code like this is a fluke, but let me tell you a lot of people
who are moderately advanced in CS programs do it, and not just here.
Much to my dismay I am left wondering if many who get CS degrees really
know anything at all.
--
Shaun Rowland ·······@cis.ohio-state.edu
http://www.cis.ohio-state.edu/~rowland/
········@knuth.brownes.org (Christopher Browne) writes:
> --> C has structs; the only way to simulate that in Lisp is via
> those horribly inefficient association lists.
Alists? What are those? My first lisp course covered propertly lists
on symbols.....
--
~jrm
Joe Marshall <·········@alum.mit.edu> writes:
> ········@knuth.brownes.org (Christopher Browne) writes:
>
> > --> C has structs; the only way to simulate that in Lisp is via
> > those horribly inefficient association lists.
>
> Alists? What are those? My first lisp course covered propertly lists
> on symbols.....
we'll, I'd have retored by saying that CL has well-capable and efficient
structs in most implementations. But symbol plists arn't considered to
be extremely efficient, and I'd guess that alists and plists are on the
same order of inefficiency wrt access times in most implementations.
One reason I don't like using plists is that I feel that the data is
somehow hidden. Of course, it's not, but it always feels like it's one
level underneath the surface. I feel that while compile-time systems
should make full use of plists, your runtime systems should not use them
heavily. But I'd like to know more about how people use them, since I
very seldom use them.
dave
David> I feel that while compile-time systems should make full use of plists,
David> your runtime systems should not use them heavily. But I'd like to
David> know more about how people use them, since I very seldom use them.
Well, I know lisp mostly from emacs-hacking, but rarely come across it.
Based on the emacs lisp corpus I find the following stats:
emacs/20.5/lisp/*.el has 236 uses of GET in 191995 lines of code (0.1 %)
emacs/20.5/lisp/emacs-lisp/*.el has 116 uses of GET in 35448 lines (0.3 %)
supporting the feeling that plists aren't a main stream thing. However, they
do seem to be a very useful construct for clever things like compilers,
debuggers, profilers and code-instrumenters such as advice.el. I expect there
to be correlation between the occurrences of EVAL and of p-lists :-)
Philip
--
Is your dog vegetarian too?
-----------------------------------------------------------------------------
Philip Lijnzaad, ········@ebi.ac.uk \ European Bioinformatics Institute,rm A2-24
+44 (0)1223 49 4639 / Wellcome Trust Genome Campus, Hinxton
+44 (0)1223 49 4468 (fax) \ Cambridgeshire CB10 1SD, GREAT BRITAIN
PGP fingerprint: E1 03 BF 80 94 61 B6 FC 50 3D 1F 64 40 75 FB 53
David Bakhash <·····@alum.mit.edu> writes:
> But I'd like to know more about how people use them, since I
> very seldom use them.
Nor should you. Plists are a `poor man's 2-d table'. The problem is
that this table is not first class, the key pairs come from a global
namespace, and the table shared with every function that uses a 2-d
table.
The latter two problems are ameliorated somewhat by packages, but in
most cases where a plist is being used, a hash table, alist, btree,
etc. would be a better choice.
* David Bakhash
| But I'd like to know more about how people use them, since I
| very seldom use them.
* Joe Marshall <·········@alum.mit.edu>
| Nor should you.
What nonsense!
| Plists are a `poor man's 2-d table'.
Well, that's _one_ view.
| The problem is that this table is not first class, the key pairs
| come from a global namespace, and the table shared with every
| function that uses a 2-d table.
Firstclassitude is _so_ overrated. Don't listen to Schemers!
Property lists exist in many objects, not just symbols. They are
very convenient ways to store _incidental_ information with the
object when the object itself is vastly more important than the
_property_ you want to associate with it. Take symbols. If the
symbol has it's very own, very good raison d'�tre, you'd be a fool
to set up an eq hash table to store some particular properties on
them, and particular to set up a whole bunch of them, probably named
by their own symbols, to hold on to each one. You already have the
symbol in hand, so it's a wanton waste of resources to use an
expensive external mechanism to avoid being criticized by Schemers.
However, if you don't use symbols for the many good reasons for
which you _should_ use symbols, but for some of the more numerous
good reasons to _avoid_ using symbols, it's just stupid to waste
resources on symbols only for their property lists. Remember,
symbols are _expensive_ little things. If you gotta have them, you
might as well exploit that fact, but if don't have to for some other
reason, property lists is not a reason to use them.
Scheme exposure damages your brain if you want to program in Common
Lisp in your real life. Scheme doesn't have symbols, so they're
_envious_ of our very useful property lists! (Or whatever, they
simply don't _get_� the symbol as a first-class object, no matter
how much they otherwise rant and rave about firstclassitude�.)
#:Erik
-------
� pun intended
� this really should be a word
--
If this is not what you expected, please alter your expectations.
Erik Naggum <····@naggum.no> writes:
> Property lists exist in many objects, not just symbols. They are
> very convenient ways to store _incidental_ information with the
> object when the object itself is vastly more important than the
> _property_ you want to associate with it.
I was objecting to the `globalness' of property lists. `Naked'
property lists don't have this problem.
> Take symbols. If the symbol has it's very own, very good raison
> d'�tre, you'd be a fool to set up an eq hash table to store some
> particular properties on them, and particular to set up a whole
> bunch of them, probably named by their own symbols, to hold on to
> each one. You already have the symbol in hand, so it's a wanton
> waste of resources to use an expensive external mechanism to avoid
> being criticized by Schemers.
Erik raises a good point, but there are drawbacks to using plists even
in this situation. Since symbol plists are a global structure, they
share the same drawbacks as any other global structure, like a global
variable.
There is no way to manipulate properties as a collection, however.
It is difficult, for example, to enumerate all symbols with a
particular property, or to maintain two sets of properties with the
same name.
A particular example is in a compiler. Many compilers put properties
on symbols indicating how to process forms with these symbols.
However, when you want to retarget the compiler, you have problems.
You must replace the properties that the compiler is looking for with
a new set, but doing that makes the compiler useless for the current
machine. There is no easy way to say `ok, do the compile, but use a
non-standard set of properties'. If there were instead an explicit
table rather than an *implicit* one scattered about the environment,
it would be trivial to hand the compiler a different set of special
symbols.
Perhaps I should say: `it's a wanton violation of abstraction to use
a global mechanism to avoid being criticized by Erik.'
> Scheme exposure damages your brain if you want to program in Common
> Lisp in your real life. Scheme doesn't have symbols, so they're
> _envious_ of our very useful property lists! (Or whatever, they
> simply don't _get_� the symbol as a first-class object, no matter
> how much they otherwise rant and rave about firstclassitude�.)
MIT Scheme has both symbols and propertly lists, but they are not
advertised.
* Joe Marshall <·········@alum.mit.edu>
| Erik raises a good point, but there are drawbacks to using plists
| even in this situation. Since symbol plists are a global structure,
| they share the same drawbacks as any other global structure, like a
| global variable.
I'm sure it's possible to view them this way, but _why_? This looks
like unfounded fear to me, and I'm not at all sure how to respond
when the technical issues at hand seem completely irrelevant. Are
global variables unilaterally and universally _bad_, now? Furrfu!
| There is no way to manipulate properties as a collection, however.
| It is difficult, for example, to enumerate all symbols with a
| particular property, or to maintain two sets of properties with the
| same name.
The latter is obviously solved by the package system. The former is
the same problem as finding all the keys that have values in a hash
table which match certain criteria. If you have such a need, you
cater to it by using something else. If you don't have such a need,
you don't need to do something else. Contrary to what nearly every
Scheme proponent will react as if he meant, we do not deal with the
fully general abstract notion of a program, but an actual program
that has a number of real-world constraints. If using property
lists would lead to problems, smart people use something else, but
Scheme proponents conclude that it is problem-laden and will never
use it even when there are no problems.
| A particular example is in a compiler. Many compilers put
| properties on symbols indicating how to process forms with these
| symbols. However, when you want to retarget the compiler, you have
| problems. You must replace the properties that the compiler is
| looking for with a new set, but doing that makes the compiler
| useless for the current machine. There is no easy way to say `ok,
| do the compile, but use a non-standard set of properties'. If there
| were instead an explicit table rather than an *implicit* one
| scattered about the environment, it would be trivial to hand the
| compiler a different set of special symbols.
Or you could be sufficiently smart and use one level of indirection
in the properties: The code does not reference properties by a
constant value (symbol), but by the value of a global symbol.
I'm sorry, but I"m not into this "let's find one more ridiculous
problem with something I don't like" game. If you enjoy such games
and impractical purity, Scheme is for you. If you are smart enough
to deal with the real world in its sometimes messy complexity, stay
with Common Lisp and observe that unlike static typing and Scheme
proponents, other programmers trust that you have a working brain
and know when using a practical, useful mechanism is a good idea and
when it is not. Smart people never listen to those who claim that
something is always bad or always good, anyway.
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum <····@naggum.no> writes:
>
> * Joe Marshall <·········@alum.mit.edu>
> | Erik raises a good point, but there are drawbacks to using plists
> | even in this situation. Since symbol plists are a global structure,
> | they share the same drawbacks as any other global structure, like a
> | global variable.
>
> I'm sure it's possible to view them this way, but _why_? This looks
> like unfounded fear to me, and I'm not at all sure how to respond
> when the technical issues at hand seem completely irrelevant. Are
> global variables unilaterally and universally _bad_, now? Furrfu!
Symbol plists are global only because symbols are global.
Actually, they're not really global - they're just in tables
(packages) so you can look them up, and the packages are global
(because of find-package).
Anyway, for many purposes, that degree of globalness is not a very
serious problem. The existence of cases in which it is a problem
(e.g. certain uses of symbol plists by compilers) does not change
that.
Erik Naggum <····@naggum.no> writes:
> The cost of [a commercial Lisp implementation] license should be
> nearly irrelevant to most software projects, apart from the ones
> that are losing money by design -- i.e., open source and hobbyist
> projects.
In my experience, the cost of language implementations is very
important for their adoption in industry. Most new tools are tried
out on small side-projects first, and usually by R&D organizations (as
opposed to product organizations) on fixed budgets. If you can only
get a full version for several thousand dollars per developer, the
system is usually not even going to make it in the door, and any form
of runtime license usually just kills it.
Also keep in mind that the competition from free and cheap software is
really tough. Python is an excellent dynamic language with lots of
libraries, and it's completely free. Java comes with one of the most
extensive standard libraries, good reflection capabilities, and good
enterprise support, and it's also essentially free. And while VB is
iffy as a language, it comes with a good environment and does have
most of the dynamic language features most normal programmers will
ever need, and VB is also quite cheap.
Tom.
tom wrote:
> ...If you can only get a full version for several thousand dollars per
> developer, the system is usually not even going to make it in the door,
> and any form of runtime license usually just kills it.
Do you know this to be the actual cost of a `full version' of the software
(particularly if you have more a single user licence) or is this a bit
more of `lisp is dead' sinmply packaged in a new more subtle way?
> Also keep in mind that the competition from free and cheap software is
> really tough. [...elided an excellent and representive list of other
> software...]
Is this the market that people who use cl are in? I'm not sure that it
is. I think cl is *not* in direct competition with these mass market
language. However, just because every man and his dog is not using cl[1],
it is still (IMHO) thriving in the niche where quality, reliablity,
elegance and intelligence are highly regarded,
Best Regards,
:) will
[1] although *maybe* they ought.
William Deakin <·····@pindar.com> writes:
> tom wrote:
> > ...If you can only get a full version for several thousand dollars per
> > developer, the system is usually not even going to make it in the door,
> > and any form of runtime license usually just kills it.
>
> Do you know this to be the actual cost of a `full version' of the software
> (particularly if you have more a single user licence) or is this a bit
> more of `lisp is dead' sinmply packaged in a new more subtle way?
Well, it's sure what killed Smalltalk. $4000/seat dev license +
non-trivial runtimes made it very clear that the only adopters were
religious ones.
> > Also keep in mind that the competition from free and cheap software is
> > really tough. [...elided an excellent and representive list of other
> > software...]
Java nailed the coffin shut on Smalltalk just as people were getting
fed up enough with the-language-that-shall-not-be-named to start
thinking about switching. I was there. It was a sorry sight.
> However, just because every man and his dog is not using cl[1],
> it is still (IMHO) thriving in the niche where quality, reliablity,
> elegance and intelligence are highly regarded,
In short, a rapidly shrinking one.
david rush
--
And Visual Basic programmers should be paid minimum wage :)
-- Jeffrey Straszheim (on comp.lang.functional)
* David Rush wrote:
> William Deakin <·····@pindar.com> writes:
>> However, just because every man and his dog is not using cl[1],
>> it is still (IMHO) thriving in the niche where quality, reliablity,
>> elegance and intelligence are highly regarded,
> In short, a rapidly shrinking one.
I don't think it's shrinking, why do you?
--tim
David Rush wrote:
> > However, just because every man and his dog is not using cl[1],
> > it is still (IMHO) thriving in the niche where quality, reliablity,
> > elegance and intelligence are highly regarded,
>
> In short, a rapidly shrinking one.
You are so vague it's almost like you did not say anything. For
example, are you thinking in absolute or relative terms?
Absolute: I can't speak for everybody, but my experiences tell it's
modestly growing. Choice of language does not matter as much now that
there are interoperability standards like Corba, HTTP and SQL, and that
the competition among C(++), Java, Visual Basic, scripting and "4GL"
languages erodes the notion of which language is the only politically
correct choice.
>
> david rush
> --
> And Visual Basic programmers should be paid minimum wage :)
> -- Jeffrey Straszheim (on comp.lang.functional)
I didn't know they don't make at least that much. Does it mean it's
illegal to program in Visual Basic? Maybe it's like farming - the less
you produce, the more you are paid.
Robert
William Deakin <·····@pindar.com> writes:
> tom wrote:
> > ...If you can only get a full version for several thousand dollars per
> > developer, the system is usually not even going to make it in the door,
> > and any form of runtime license usually just kills it.
>
> Do you know this to be the actual cost of a `full version' of the software
> (particularly if you have more a single user licence) or is this a bit
> more of `lisp is dead' sinmply packaged in a new more subtle way?
Well, it's sure what killed Smalltalk. $4000/seat dev license +
non-trivial runtimes made it very clear that the only adopters were
religious ones.
> > Also keep in mind that the competition from free and cheap software is
> > really tough. [...elided an excellent and representive list of other
> > software...]
Java nailed the coffin shut on Smalltalk just as people were getting
fed up enough with the-language-that-shall-not-be-named to start
thinking about switching. I was there. It was a sorry sight.
> However, just because every man and his dog is not using cl[1],
> it is still (IMHO) thriving in the niche where quality, reliablity,
> elegance and intelligence are highly regarded,
In short, a rapidly shrinking one.
david rush
--
And Visual Basic programmers should be paid minimum wage :)
-- Jeffrey Straszheim (on comp.lang.functional)
Erik Naggum wrote:
> ...[elided a description of ACL find-package]... I'm sure this affects
> _some_ existing code.
Agreed. But as you say, in a quite innocuous fashion.
> Hierarchical packages have been added to aid in portability. It's
> probably more an attempt to cater to the Java crowd, however.
This is my suspicion also.
> I'm not sure what this will buy the Lisp community, but this is one of
> the most innocuous changes I have seen anyone suggest.
Then again (I suspect) that if I were going to give a �10-note to each
member of the said community, there would be somebody that would
grumble...
;) will
* Erik Naggum
| ...[elided a description of ACL find-package]... I'm sure this
| affects _some_ existing code.
* William Deakin
| Agreed. But as you say, in a quite innocuous fashion.
I _was_ being facetious with that last comment. I don't actually
think _any_ existing code will notice anything. (Rainer Joswig may
now helpfully provide a contrived counter-example, of course.)
#:Erik
--
If this is not what you expected, please alter your expectations.
Erik Naggum wrote:
> * Erik Naggum
> I _was_ being facetious with that last comment. I don't actually
> think _any_ existing code will notice anything.
And I was (clearly) being terminally stupid.
:( will
While the hierarchial package specification does allow for
naming packages with relative package names, that's not
used in the AllegroServe source. Thus all AllegroServe
package references can be understood by any Common Lisp.
Hierarchial package names are for the most part a
naming convention designed to prevent package names
we choose from conflicting with package names
other users choose.
Fernando wrote:
> On Mon, 15 May 2000 11:57:35 -0500, "Andrew K. Wolven"
> <·······@redfernlane.org> wrote:
>
> >P.S.: Saving state information of clients in a Lisp web server is a
> >no-brainer. (when to GC is the hard part) I believe you already proposed
>
> Please, enlighten us. O:-) I also have to save some client
> related stuff into a database (results of a search, so I can show the
> first 10 matches and latter show the rest without repeating the
> query), and I'm not sure when to "GC" the temporary tables holding
> that info. I was planning to use a chron process to cleanup every
> entry older than X minutes... :-?
>
It would be useful to know what Lisp platforms you intend to use.
If you are dealing with a production web server or a production web-application
server, you are going to have to deal with sockets and threads, which are
platform dependent.
The questions I have seen relate to emission of client [session] specific
content.
To do this easily, you will want a long-lived lisp process that will maintain
state for you. (not CGI)
What is the OS?
What is the Lisp?
Are you wanting to set up a Lisp web server, or are you wanting to set up a
web-application server to be run under a conventional server? (such as under
Apache)
Less critical: what DB do you want to use, and what do you plan to store in it?
Some of the things that Perl/CGI hackers put in a database can actually live as
lisp objects in memory.
You don't need a database unless you really need to store things in tables.
For now go to allegroserve.sourceforge.net and download allegroserve. There is
a file you will find very useful (htmlgen.cl). Htmlgen is the simplest of the
simple in terms of dynamic html generation with common lisp. Even if you plan
to use templates, you will find this macro very useful.
AKW
>
> >the solution: Simply stuff a client-id into emitted anchors and forms.
> >Don't call it a session, however, or you might be violating some
> >dickhead's patent. (On that logic, you can't use lisp, because people
>
> Is this what Jeff "Dickhead" Bezos patented (saving client-ids
> into forms and anchors)??!!! =:-O I guess the real dickheads are at
> the patents office then...
>
> >patent CL facilities all the time [I am thinking about starting a
> >bogus-patent web ring for fun, please send your bogus patents to me])
>
> Fine, I'll patent the use of ">" for quoting text.
>
> BTW, is there a more appropriate NG for discussing web server
> programming? Most of this isn't really Lisp specific, and I'm afraid
> I might be boring other readers... O:-)
>
> //-----------------------------------------------
> // Fernando Rodriguez Romero
> //
> // frr at mindless dot com
> //------------------------------------------------
"Andrew K. Wolven" wrote:
> Fernando wrote:
>
> > On Mon, 15 May 2000 11:57:35 -0500, "Andrew K. Wolven"
> > <·······@redfernlane.org> wrote:
> >
> > >P.S.: Saving state information of clients in a Lisp web server is a
> > >no-brainer. (when to GC is the hard part) I believe you already proposed
> >
> > Please, enlighten us. O:-) I also have to save some client
> > related stuff into a database (results of a search, so I can show the
> > first 10 matches and latter show the rest without repeating the
> > query), and I'm not sure when to "GC" the temporary tables holding
> > that info. I was planning to use a chron process to cleanup every
> > entry older than X minutes... :-?
> >
>
> It would be useful to know what Lisp platforms you intend to use.
>
> If you are dealing with a production web server or a production web-application
> server, you are going to have to deal with sockets and threads, which are
> platform dependent.
>
> The questions I have seen relate to emission of client [session] specific
> content.
>
> To do this easily, you will want a long-lived lisp process that will maintain
> state for you. (not CGI)
>
> What is the OS?
> What is the Lisp?
> Are you wanting to set up a Lisp web server, or are you wanting to set up a
> web-application server to be run under a conventional server? (such as under
> Apache)
>
> Less critical: what DB do you want to use, and what do you plan to store in it?
>
> Some of the things that Perl/CGI hackers put in a database can actually live as
> lisp objects in memory.
> You don't need a database unless you really need to store things in tables.
>
> For now go to allegroserve.sourceforge.net and download allegroserve. There is
> a file you will find very useful (htmlgen.cl). Htmlgen is the simplest of the
> simple in terms of dynamic html generation with common lisp. Even if you plan
> to use templates, you will find this macro very useful.
>
> AKW
HTMLGEN will show you that HTML should have just been lisp in the first place. If
you use emacs or have paren matching turned on in vi, you will never miss an end
tag.
>
>
> >
> > >the solution: Simply stuff a client-id into emitted anchors and forms.
> > >Don't call it a session, however, or you might be violating some
> > >dickhead's patent. (On that logic, you can't use lisp, because people
> >
> > Is this what Jeff "Dickhead" Bezos patented (saving client-ids
> > into forms and anchors)??!!! =:-O I guess the real dickheads are at
> > the patents office then...
> >
> > >patent CL facilities all the time [I am thinking about starting a
> > >bogus-patent web ring for fun, please send your bogus patents to me])
> >
> > Fine, I'll patent the use of ">" for quoting text.
> >
> > BTW, is there a more appropriate NG for discussing web server
> > programming? Most of this isn't really Lisp specific, and I'm afraid
> > I might be boring other readers... O:-)
> >
> > //-----------------------------------------------
> > // Fernando Rodriguez Romero
> > //
> > // frr at mindless dot com
> > //------------------------------------------------
On Tue, 16 May 2000 12:47:52 -0500, "Andrew K. Wolven"
<·······@redfernlane.org> wrote:
>> Please, enlighten us. O:-) I also have to save some client
>> related stuff into a database (results of a search, so I can show the
>> first 10 matches and latter show the rest without repeating the
>> query), and I'm not sure when to "GC" the temporary tables holding
>> that info. I was planning to use a chron process to cleanup every
>> entry older than X minutes... :-?
>>
>
>It would be useful to know what Lisp platforms you intend to use.
LispWorks
>What is the OS?
Linux and Apache. I probably won't be able to change this
(Apache).
>Are you wanting to set up a Lisp web server, or are you wanting to set up a
>web-application server to be run under a conventional server? (such as under
>Apache)
The latter.
>Less critical: what DB do you want to use, and what do you plan to store in it?
I'm considering interbase.
>For now go to allegroserve.sourceforge.net and download allegroserve. There is
>a file you will find very useful (htmlgen.cl). Htmlgen is the simplest of the
>simple in terms of dynamic html generation with common lisp. Even if you plan
>to use templates, you will find this macro very useful.
OK, bt is this a http server? What would really interesting
(IMHO) would be an application server for CL....
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------
Fernando <·······@must.die> writes:
> Hi!
>
> I have to write a few servlets and I'm planning to use CL. I
> was wondering if there's some library that encapsulates state
> retention, something like Java's HttpSession, available? O:-)
>
> TIA
>
You might want to have a look at ``gwl'' at http://gwl.sourceforge.net.
-dave
--
David J. Cooper Jr, Chief Engineer Genworks International
·······@genworks.com 5777 West Maple, Suite 130
(248) 932-2512 (Genworks HQ/voicemail) West Bloomfield, MI 48322-2268
(248) 407-0633 (pager) http://www.genworks.com
On 10 May 2000 08:00:27 -0400, ·······@genworks.com (David J. Cooper)
wrote:
>You might want to have a look at ``gwl'' at http://gwl.sourceforge.net.
OK :-)
//-----------------------------------------------
// Fernando Rodriguez Romero
//
// frr at mindless dot com
//------------------------------------------------