From: Peter.VanEynde
Subject: One the Mark Kantrowitz copyright
Date:
Message-ID: <EsL6nt.5tx@uia.ua.ac.be>
Hi people,
I've been busy lately packaging CMUCL and other Lisp stuff
in debian packages. (a debian package is a binary distribution
of a program or library for debian. It makes installing, configuring and
de-installing software easy and user-friendly)
To this end I have to investigate the copyright of the stuff
I package quite attentively, in doing this I noticed the
copyright on the programs from Mark Kantrowitz, which is
also used by other authors. It states:
;;; Copyright (c) 1989-95 by Mark Kantrowitz. All rights reserved.
;;; Use and copying of this software and preparation of derivative works
;;; based upon this software are permitted, so long as the following
;;; conditions are met:
;;; o no fees or compensation are charged for use, copies, or
;;; access to this software
;;; o this copyright notice is included intact.
;;; This software is made available AS IS, and no warranty is made about
;;; the software or its performance.
This is a problem. (This piece is from "defsystem" in the CMUCL contrib
directory.) The problem is the first "o": no fees means that I can't upload
this to the main or contrib distribution, because people make CD-ROM's from
those parts and _sell_ them. This is charging a fee for copies. So I can't
use it :-(.
Note that it is not only debian who is hurt by this: also FreeBSD shouldn't
include defsystem... And I would hate to have to re-write defsystem just
because of this one clause.
Is Mr. Kantrowitz still out there? (I tried to contact him 2 weeks ago)
Also, why are copyrights like this so common in "Free" Lisp code? I'm not
even talking about the cl-http license... but it seems that they don't want
more users anyway :-(.
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
"Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
> This is a problem. (This piece is from "defsystem" in the CMUCL contrib
> directory.) The problem is the first "o": no fees means that I can't upload
> this to the main or contrib distribution, because people make CD-ROM's from
> those parts and _sell_ them. This is charging a fee for copies. So I can't
> use it :-(.
Now I can't speak for Mr Kantrowitz, but in general the "no fees"
restriction above means, "you can't charge extra for a system
that ships with this." That is, you can't treat these pieces as
an "extra-cost option", but you can charge money for the package
containing them.
- David
"Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
> Also, why are copyrights like this so common in "Free" Lisp code? I'm not
> even talking about the cl-http license... but it seems that they don't want
> more users anyway :-(.
What is the particular problem? I have been using CL-HTTP without any
licensing problems since some time. Making sure that the
improvements will go back into the original seems to be a worthwile goal.
CL-HTTP should not have the fate of CLIM:
- no free version
- no source code
- commercial version lacking support
- unclear future
- commercial versions not based on a single source code anymore
- no active development
People are developing their extensions to CL-HTTP.
I know that some (not many) people have complained about
the CL-HTTP license. Others are selling commercial software
based on CL-HTTP already.
Actually, we should ask "What can we do for CL-HTTP?" and
not "What will I get for free?". ;-)
Greetings from Hamburg,
Rainer Joswig
From: Hartmann Schaffer
Subject: Re: One the Mark Kantrowitz copyright
Date:
Message-ID: <3552153C.2A0CA799@netcom.ca>
Rainer Joswig wrote:
> ...
> People are developing their extensions to CL-HTTP.
> I know that some (not many) people have complained about
> the CL-HTTP license. Others are selling commercial software
> based on CL-HTTP already.
>
> Actually, we should ask "What can we do for CL-HTTP?" and
> not "What will I get for free?". ;-)
The problem is: I want to use x, I certainly don't mind putting any
modifications / improvements / ... back, but what if the license
prevents me from shipping any products I want to sell (do for a
customer)?
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
From: Peter.VanEynde
Subject: Re: One the Mark Kantrowitz copyright
Date:
Message-ID: <Essp64.7nL@uia.ua.ac.be>
Rainer Joswig <······@lise.lavielle.com> wrote:
: "Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
:> Also, why are copyrights like this so common in "Free" Lisp code? I'm not
:> even talking about the cl-http license... but it seems that they don't want
:> more users anyway :-(.
: What is the particular problem? I have been using CL-HTTP without any
People tell me that things like this:
>Additionally, all rights are reserved to any derivative works based on the
>Lisp source code in the CL-HTTP distribution, in particular but not limited to
>any automatic or manual translation of the source code into other computer
>languages or executables.
In fact say that compiling the code is forbidden. As English is not my
native language I can only admit that I'm confused by all the "all rights
reserved" in there.
[snip]
: Actually, we should ask "What can we do for CL-HTTP?" and
: not "What will I get for free?". ;-)
I assumed that making cl-http easily installable for novices,
with a manual that explains how to program simple web-based applications
would pull Lisp out from the mists of a (perceived) 1960's based interface
to the bright new world of web-based applications. I wanted to give
the people at the university of Naples a debian distribution with the words:
install debian, add CMUCL and cl-http and write the class assignments as
web-based applications. People would stop complaining about the brackets
and the tty-based interface and would marvel at the speed of the
computed response functions. I've stacked cl-http up against apache, and
for large files it loses badly, due to the not-quite-good enough multitasking of
CMUCL. But for small files (I'm not even talking about cgi here) it's
twice as fast. (measurements with wget, so with http 1.0)
But it seems people don't want it... :-(
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
"Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
> People tell me that things like this:
>
> >Additionally, all rights are reserved to any derivative works based on the
> >Lisp source code in the CL-HTTP distribution, in particular but not limited to
> >any automatic or manual translation of the source code into other computer
> >languages or executables.
>
> In fact say that compiling the code is forbidden. As English is not my
> native language I can only admit that I'm confused by all the "all rights
> reserved" in there.
If you are trying to automatically convert CL-HTTP to Java, it is still CL-HTTP.
"Similar" commercial software (WebObjects, VisualWave, ...) is *extremely*
expensive. CL-HTTP is free and it is being used in mission critical
applications (hey, it is used in the White House), so people
are trying to make sure that it works. It is also
available for very diverse systems (from a Lisp machine to
MCL). So I don't complain that a complex software I'd otherwise
had to program and maintain myself is available. Giving
improvements of the core system (bug fixes are always welcome)
back to the community seems to be a small price.
You just can not expect that people will prepare
the system in a way that it is immediately useful for yourself.
One of the main purposes of the license of CL-HTTP is to make
sure that it can be used by developers without other people trying
to get hold of CL-HTTP and lock it inside their companies.
Just avoid the fate of CLIM where the companies did have a less than
optimal marketing and development strategy.
> I assumed that making cl-http easily installable for novices,
> with a manual that explains how to program simple web-based applications
> would pull Lisp out from the mists of a (perceived) 1960's based interface
> to the bright new world of web-based applications.
In principle a worthwile task. Actually there is no way for developing
simple web-based applications in Common Lisp, CL-HTTP and
HTTP for novices. The domain itself is already very complicated.
Common Lisp itself is already a big hurdle.
An introductory text which would show up the way
would be welcome, though. Any takers?
Who writes a text? There is a nice CL-HTTP software that serves
interactive books (Interbook). This would be a good base for
documentation. As long there is nobody doing this, there
will be no documentation. Maybe there are people/companies
who would like to sponsor such an effort?
> I wanted to give
> the people at the university of Naples a debian distribution with the words:
> install debian, add CMUCL and cl-http and write the class assignments as
> web-based applications.
Actually the University of Hamburg has done this. Students have been
programming a web-walker-based broker. This time there is a project about
with functional VRML programming, which eventually will may
be written with the CL-HTTP library. Well, and they have been starting
their computer science course with Common Lisp.
The computer science department has a site license of
ACL (with composer and CLIM) on SPARCs. Students and projects can use
this stuff from everywhere (some actually are doing this ;-) ).
> But it seems people don't want it... :-(
Which people?
CL-HTTP developers are doing something different, AFAIK. They are doing
research and are maintaining some systems based on CL-HTTP. It is just
not on their list of tasks. They don't have time
to do that. If somebody would like to hack on that and
provide a way to make CL-HTTP more accessible for students
he or she would be welcome.
Currently CL-HTTP is a big library of web-related Common Lisp
code with a lot of example code. Use it, make sure
that you don't reinvent the wheel and that the library
is in a usable state. Still a lot of different concepts are
involved. Especially when it comes to developing an
efficient piece of software that works on a lot of platforms and
inside much bigger systems, you will be confronted with
coding techniques that may not be common.
Summary:
** Using CL-HTTP and providing contributions is very welcome. **
Rainer Joswig
In article <············@lise.lavielle.com>, Rainer Joswig
<······@lise.lavielle.com> wrote:
>One of the main purposes of the license of CL-HTTPD is to make
>sure that it can be used by developers without other people trying
>to get hold of CL-HTTP and lock it inside their companies.
That's exactly what GNU Copyleft was invented for. Use the LGPL for
CL-HTTP and there wouldn't be a problem. I really don't understand why
there are so much programmers/projects out there that want to provide some
"free software" and only get to the point of "software for free" by
installing their own proprietary license. Hey, it's the century of the
fruit bat^w^wcode reuse, so why not reuse a perfectly good license, too?
bye, Georg
--
#!/usr/bin/perl -0777pi
s/TS:.*?\0/$_=$&;y,a-z, ,;s, $,true,gm;s, 512,2048,;$_/es
··@hugo.westfalen.de (Georg Bauer) writes:
> In article <············@lise.lavielle.com>, Rainer Joswig
> <······@lise.lavielle.com> wrote:
>
> >One of the main purposes of the license of CL-HTTPD is to make
> >sure that it can be used by developers without other people trying
> >to get hold of CL-HTTP and lock it inside their companies.
>
> That's exactly what GNU Copyleft was invented for. Use the LGPL for
> CL-HTTP and there wouldn't be a problem.
Maybe not every software author agrees with this license.
> I really don't understand why
> there are so much programmers/projects out there that want to provide some
> "free software" and only get to the point of "software for free" by
> installing their own proprietary license. Hey, it's the century of the
> fruit bat^w^wcode reuse, so why not reuse a perfectly good license, too?
Because of different goals?
In article <············@lise.lavielle.com>, Rainer Joswig
<······@lise.lavielle.com> wrote:
>Because of different goals?
Actually there are so much different licenses available, that I think it
should be possible to use one of them instead of inventing another one
(with the problems _any_ new thing has - it will be buggy). Ever read the
Debian Free Software Guidelines? They give a good summary of what can be
seen as really "free software". It should be recommended reading for any
author of free software. Helps a lot on deciding what to put in a license
and what not. (I still prefer going with some of the established licenses
- GPL, LGPL and Artistic are my favorites, BSD can be ok, if done right).
bye, Georg
--
#!/usr/bin/perl -0777pi
s/TS:.*?\0/$_=$&;y,a-z, ,;s, $,true,gm;s, 512,2048,;$_/es
From: Peter.VanEynde
Subject: Re: On the Mark Kantrowitz copyright
Date:
Message-ID: <Esvx66.8DM@uia.ua.ac.be>
Rainer Joswig <······@lise.lavielle.com> wrote:
: One of the main purposes of the license of CL-HTTP is to make
: sure that it can be used by developers without other people trying
: to get hold of CL-HTTP and lock it inside their companies.
: Just avoid the fate of CLIM where the companies did have a less than
: optimal marketing and development strategy.
Why oh why do I need to be stumped by several pages of
legaleese? I asked for an explenation and all I got was "we have
enough users allready". This is not the attitude of a system
striving for world-domination :-).
: In principle a worthwile task. Actually there is no way for developing
: simple web-based applications in Common Lisp, CL-HTTP and
: HTTP for novices. The domain itself is already very complicated.
: Common Lisp itself is already a big hurdle.
Lisp they _have_ to learn. And replacing "read" and "princ" isn't
that dificult. I'm not talking about virtual worlds here, I mean
simple programs. Examples from text-books: (fact x), not VRML.
...
: Summary:
:
: ** Using CL-HTTP and providing contributions is very welcome. **
I will probably do that, but being in non-free will limit the
distribution.
PS: I just want to clarify that I am _very_ gratefull for programs like
cl-http and defsystem. I'm just annoyed that they limit themselves by
having strange and confusing licenses. Making a free unix clone is a lot
easier then making a free lisp machine, if only because in the unix world
almost all software is either GPL/FreeBSD or Artistic licenses. For Lisp
programs I lose more time searching for and reading the license then I do
debianizing them :-(.
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
In article <··········@uia.ua.ac.be>, "Peter.VanEynde"
<·······@hhipe.uia.ac.be> wrote:
>But it seems people don't want it... :-(
Hey, don't you dare and stop your work: _I_ would like it. Makes
installing _much_ easier, if you can get standard .deb's. Even if the
result would go into non-free, it would be worth the work.
bye, Georg
--
#!/usr/bin/perl -0777pi
s/TS:.*?\0/$_=$&;y,a-z, ,;s, $,true,gm;s, 512,2048,;$_/es
From: Peter.VanEynde
Subject: Re: One the Mark Kantrowitz copyright
Date:
Message-ID: <Esvwov.6oA@uia.ua.ac.be>
Georg Bauer <··@hugo.westfalen.de> wrote:
: In article <··········@uia.ua.ac.be>, "Peter.VanEynde"
: <·······@hhipe.uia.ac.be> wrote:
:>But it seems people don't want it... :-(
: Hey, don't you dare and stop your work: _I_ would like it. Makes
: installing _much_ easier, if you can get standard .deb's. Even if the
: result would go into non-free, it would be worth the work.
cmucl, series and clio/clue should appear in unstable in the next
few days. defsystem in unstable/non-free :-(.
I'm having problems with a lot of other software mainly because
the licenses are so non-standard. I would like to include f2cl, but
there _is_ no license, so I must assume that I can't even upload it
to non-free :-(. cl-http was almost done, but I have my hands
full in RL now. The HyperSpec is also packaged, but I need to check the
license with the mentors. Plob doesn't build yet :-(. Oh, and
I'm also working on the completion of the glibc2 port and a MySql library.
That and my thesis makes me wait for the next version of cl-http...
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
"Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
> the licenses are so non-standard. I would like to include f2cl, but
> there _is_ no license, so I must assume that I can't even upload it
> to non-free :-(. cl-http was almost done, but I have my hands
That's probably because you obtained my modified copy of f2cl, which
didn't contain all of the original f2cl code. I should have left a
pointer on where to get the original.
In any case, the original distribution gives a copyright statement
saying:
Summary:
Permission is granted to use, copy, modify this program, f2cl,
EXCEPT that the copyright notice must be reproduced on copies, and
credit should be given to the authors where it is due.
WE MAKE NO WARRANTY AND ACCEPT NO LIABILITY FOR THIS PROGRAM.
In detail:
1. Permission to use, copy, modify this software and its documentation
for any purpose is hereby granted without fee, provided that
- the above copyright notice appear in all copies,
- both that copyright notice and this permission notice appear in
supporting documentation, and that
- you cause modified files to carry prominent notices stating that
you changed the files and the date of any change.
But then later says it is under GNU GPL too.
I obviously need to update my modified copy of f2cl to comply with
these requirements and will do so shortly. (And add one additional
bug fix.)
Ray
In article <··········@uia.ua.ac.be>, "Peter.VanEynde"
<·······@hhipe.uia.ac.be> wrote:
>cmucl, series and clio/clue should appear in unstable in the next
>few days. defsystem in unstable/non-free :-(.
No chance that it appears for bo-unstable? Okok, stop laughing, it just
was a question ... ;-)
bye, Georg
--
#!/usr/bin/perl -0777pi
s/TS:.*?\0/$_=$&;y,a-z, ,;s, $,true,gm;s, 512,2048,;$_/es
From: Lieven Marchand
Subject: Re: One the Mark Kantrowitz copyright
Date:
Message-ID: <6j24ok$drb$1@xenon.inbe.net>
"Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
> To this end I have to investigate the copyright of the stuff
> I package quite attentively, in doing this I noticed the
> copyright on the programs from Mark Kantrowitz, which is
> also used by other authors. It states:
>
> ;;; Copyright (c) 1989-95 by Mark Kantrowitz. All rights reserved.
>
> ;;; Use and copying of this software and preparation of derivative works
> ;;; based upon this software are permitted, so long as the following
> ;;; conditions are met:
> ;;; o no fees or compensation are charged for use, copies, or
> ;;; access to this software
> ;;; o this copyright notice is included intact.
> ;;; This software is made available AS IS, and no warranty is made about
> ;;; the software or its performance.
>
> This is a problem. (This piece is from "defsystem" in the CMUCL contrib
> directory.) The problem is the first "o": no fees means that I can't upload
> this to the main or contrib distribution, because people make CD-ROM's from
> those parts and _sell_ them. This is charging a fee for copies. So I can't
> use it :-(.
Franz bundles a few programs by Mr. Kantrowitz in their
contrib/mkant-cmu directory so I suppose it is ok to add them to a
contrib style directory. Actually looking at the copyright notices
some are even inconsistent.
From logical-pathnames.lisp:
;;; This code is in the public domain and is distributed without warranty
;;; of any kind.
followed by the conditions you quoted. But once you've put something
in the public domain you have given up every right to put constraints
on the copying or use.
In the same directory they also put this README:
Return-Path: layer
Return-Path: <layer>
Received: by akbar.Franz.COM (3.2/FI-1.0)
id AA29724; Mon, 6 Aug 90 16:00:26 PDT
From: layer (Kevin Layer)
Message-Id: <··················@akbar.Franz.COM>
To: cxh
Cc: local-tech, layer
Subject: Re: ·····@a.gp.cs.cmu.edu: Portable Common Lisp Utilities
In-Reply-To: Your message of Tue, 31 Jul 90 00:32:59 GMT.
<··················@fiolin.Franz.COM>
Date: Mon, 06 Aug 90 16:00:24 -0700
These can be found in /cl/master/share/mkant-cmu/.
>> From: ·····@a.gp.cs.cmu.edu (Mark Kantrowitz)
>> Newsgroups: comp.lang.lisp
>> Subject: Portable Common Lisp Utilities
>> Keywords: portable common-lisp tools
>> Date: 18 Jul 90 18:49:07 GMT
>> Organization: Carnegie-Mellon University, CS/RI
>>
>>
>> This post describes four portable utilities for Common Lisp which are
>> available by anonymous ftp from a.gp.cs.cmu.edu in the directory
>> /usr/mkant/Public.
>>
>> All have been placed in the public domain and are distributed without
>> warranty of any kind.
>>
>> defsystem.lisp A portable system definition facility, similar
>> to the Symbolics system construction tool.
>> Documentation is in defsystem.ps.
>>
>> logical-pathnames.lisp A portable implementation of the X3J13 June
>> 1989 specification for logical pathnames, as
>> documented in Guy Steele CLtL2, section 23.1.5.
>>
>> metering.lisp A portable code profiling tool, for gathering
>> timing and consing statistics while a program
>> is running. This is a combination of the Monitor
>> package written by Chris McConnell and the
>> Profile package written by Skef Wholey and
>> Rob MacLachlan, with a few extensions.
>> Documentation is in the first few pages.
>>
>> xref.lisp A portable cross referencing tool for determining
>> callers of functions and variables, and for
>> maping out the structure of a program. Similar
>> to Symbolics Who-Calls and Xerox Masterscope.
>> Includes an interface to Joe Bates' Postscript
>> graphing tool (psgraph.lisp).
>>
>>
>> Please send bug reports, comments, and suggestions to ·····@cs.cmu.edu.
>>
>> Also, if you'd like to receive future notification of updates,
>> bug-fixes, and new utilities, send email to ·····@cs.cmu.edu.
>>
>> --mark
I suppose this is a good example why it is smarter to use BSD/GPL/MPL
(in no particular order) like licenses that have been carefully
drafted by qualified lawyers.
If you want to replace it, there is a defsystem in the contrib/clx
directory that has a BSD style copyright.
--
Lieven Marchand <···@bewoner.dma.be>
------------------------------------------------------------------------------
Few people have a talent for constructive laziness. -- Lazarus Long
Fodder to keep nntp server happy
Lieven Marchand <···@bewoner.dma.be> writes:
> "Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
>
...
>
> I suppose this is a good example why it is smarter to use BSD/GPL/MPL
> (in no particular order) like licenses that have been carefully
> drafted by qualified lawyers.
>
> If you want to replace it, there is a defsystem in the contrib/clx
> directory that has a BSD style copyright.
>
Unfortunately, IMHO, the DEFSYSTEM in the contrib/clx is not up to par
with Mark Kantrowitz's one.
It is unfortunate that he does not answer email so that this copyright
matter can be cleared. OTOH, it seems that "adding" the DEFSYSTEM code
to a "product" without asking for a fee specifically for this reason,
would not really infringe the copyright.
Cheers
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - (0)6 - 68 80 79 23, fax. +39 - (0)6 - 68 80 79 26
http://www.parades.rm.cnr.it
* Marco Antoniotti
| OTOH, it seems that "adding" the DEFSYSTEM code to a "product" without
| asking for a fee specifically for this reason, would not really infringe
| the copyright.
I have hesitated to bring up a legal point because I'm not at all certain
about its application in this case, but during a course in U.S. contract
law given by the U.S. Embassy in Oslo a few years back, a fairly strong
warning was issued to us that contracts that in any way would require one
party to lose money by performing the requirements of the contract would
be unenforceable at law, and could land us in real trouble defending the
_rest_ of the contract in a hostile court. several obvious examples were
given that were tantamount to fraud.
if I understood this principle right, the "permission" to reproduce could
be construed to require that whoever is stupid enough to obey the letter
of the contract would run the high risk of having to pay for the whole
shebang at other people's unlimited request were the license enforceable,
but Mark Kantrowitz is in no legal position to require such a loss from
his grantees. my understanding from said course was that a friendly
court could rule "without fee or compensation" to mean "above and beyond
recovery of costs pertaining to execution of license" to avoid declaring
the restrictions invalid and essentially grant an _unlimited_ license to
use, copy, and access the material provided by Mark, which is an option.
my experience with this part of contract law is non-existing as I've done
my best to stay away from problematic areas, so I may have misapplied
this principle or not understood its ramifications. if anyone _knows_
what I'm hinting at, I'd appreciate corrections and comments. in any
case, were I to want to reproduce his works in part of a (physical)
package that I would charge some fee for producing, but _not_ require
secondary payment from those who would reproduce further copies (which
would clearly fall under the restrictions Mark wants exercised), I would
talk to a lawyer about it.
#:Erik
--
Support organized crime: use Microsoft products!
From: Peter.VanEynde
Subject: Re: One the Mark Kantrowitz copyright
Date:
Message-ID: <EssqD0.B7F@uia.ua.ac.be>
Marco Antoniotti <·······@galvani.parades.rm.cnr.it> wrote:
: Lieven Marchand <···@bewoner.dma.be> writes:
:> "Peter.VanEynde" <·······@hhipe.uia.ac.be> writes:
:>
: ...
:>
:> I suppose this is a good example why it is smarter to use BSD/GPL/MPL
:> (in no particular order) like licenses that have been carefully
:> drafted by qualified lawyers.
:>
:> If you want to replace it, there is a defsystem in the contrib/clx
:> directory that has a BSD style copyright.
:>
Please note that I didn't write this, but I agree with it mainly because
as a debian maintainer it's easier to include stuff if it has a "well-known"
license.
: Unfortunately, IMHO, the DEFSYSTEM in the contrib/clx is not up to par
: with Mark Kantrowitz's one.
: It is unfortunate that he does not answer email so that this copyright
: matter can be cleared. OTOH, it seems that "adding" the DEFSYSTEM code
: to a "product" without asking for a fee specifically for this reason,
: would not really infringe the copyright.
I got main from Mr. Kantrowitz, he says that he reads his CMU account only
rarely.
He also says that the copyright is written that way because "it is intended to
be free, not for others to sell as a source of profit.". And that he wants
people who want to redistribute DEFSYSTEM to contact him for an exception.
I fear that I will have to reply to him that this is not enough, the
debian rules prohibit discrimination, and only allowing redistribution with
debian is such a discrimination.
Note that I fully understand his position: he contributed something for
free and he doesn't want people to grow rich from his work. But the
present climate has changed and downloading software is not anymore for the
elitist professionals, I hope/fear that with Netscape going free the unwashed masses
will discover "free" software and that alternative licenses like for
DEFSYSTEM will get largely ignored and DEFSYSTEM will be treated like BSD software.
(even now people have a hard time grasping that MySql and qt are _not_ free)
Debian and other projects have a habit of absorbing software that is assumed to
be free, and all-ready it is difficult to add non-GPL/BSD/MPL software... It's
not worth the hassle of checking the copyright with a fine comb...
I started this with thinking of adding the option of "make a package out of
this module" to DEFSYSTEM... it now looks like I will have to reimplement it :-(
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
On a related point, I tried to send a bugfix to the author of the SOURCE-COMPARE
utility (by MK?, file sc.lsp), at the address listed in the code. While it
didn't bounce, neither did it get a reply, nor has it been incorporated in the
archive. This was nearly a year ago.
I've since found another bug too, where SC gives the wrong results.
Where should fixes to code in the CMU archives be sent?
(It may be published in some prominent area, but I haven't seen it).
Dave Tenny
In article <··················@news.ma.ultranet.com>,
············@ma.ultranet.com wrote:
>On a related point, I tried to send a bugfix to the author of the
SOURCE-COMPARE
>utility (by MK?, file sc.lsp), at the address listed in the code. While it
>didn't bounce, neither did it get a reply, nor has it been incorporated in the
>archive. This was nearly a year ago.
I had a similar non-response to some changes I submitted for Mark's
DEFSYSTEM well over a year ago. I ended up posting the (clearly marked)
revised file on Digitool's FTP site (since the changes are intended
primarily for compatibility with later versions of MCL.)
OTOH, I've submitted updates to the FAQ; these have been incorporated in a
timely manner.
I remember reading a while ago that Mark was getting back to finishing his
degree, or something. My guess is that he "got a life."
But what becomes of the AI Repository? I'd like to see it _not_
stagnate. Is there any talk of handing over the reins to a new
maintainer?
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
············@ma.ultranet.com writes:
> On a related point, I tried to send a bugfix to the author of the
> SOURCE-COMPARE utility (by MK?, file sc.lsp), at the address listed
> in the code. While it didn't bounce, neither did it get a reply,
> nor has it been incorporated in the archive. This was nearly a year
> ago.
>
> I've since found another bug too, where SC gives the wrong results.
> Where should fixes to code in the CMU archives be sent? (It may be
> published in some prominent area, but I haven't seen it).
What email address are people using? I just searched for 'Mark
Kantrowitz home page' using hotbot & found what looks to be his home
page (at
http://128.2.242.152/afs/cs.cmu.edu/user/mkant/www/home.html),
although it was last modified almost a year ago. It lists:
Mark Kantrowitz Office: Wean Hall 8104
School of Computer Science Phone: 412-268-2582
Carnegie Mellon University Fax: 412-268-5576
5000 Forbes Avenue E-mail: ·····@cs.cmu.edu
Pittsburgh, PA 15213-3891 Home: 412-422-6190
Interested parties might try this email address or calling him at his
office.
--
Harvey J. Stein
BFM Financial Research
·······@bfr.co.il