Path: gmd.de!Germany.EU.net!mcsun!uunet!europa.eng.gtefsd.com!
howland.reston.ans.net!usenet.ins.cwru.edu!
magnus.acs.ohio-state.edu!cis.ohio-state.edu!gnu.ai.mit.edu!rms
From: r...@gnu.ai.mit.edu (Richard Stallman)
Newsgroups: gnu.misc.discuss
Subject: GPL not bypassed
Date: 2 Jul 1993 15:32:27 -0400
Organization: GNUs Not Usenet
Lines: 4
Sender: dae...@cis.ohio-state.edu
Distribution: gnu
Message-ID: < 9307021931.AA12919@mole.gnu.ai.mit.edu>
It is not as easy to bypass the GPL as some people might think.
The GPL does not permit what is being done with RSAREF and GNU mp,
and as soon as I found out about it (a couple of hours ago)
I informed the distributors of this.
Xref: gmd.de gnu.misc.discuss:5315 alt.security.pgp:3295
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!
math.ohio-state.edu!cyber1.cyberstore.ca!vanbc.wimsey.com!
vanbc.wimsey.com!not-for-mail
From: ma...@vanbc.wimsey.com (Mark C. Henderson)
Newsgroups: gnu.misc.discuss,alt.security.pgp
Subject: Re: GPL not bypassed
Date: 2 Jul 1993 14:51:32 -0700
Organization: Wimsey
Lines: 62
Distribution: world
Message-ID: <212al4$o9t@vanbc.wimsey.com>
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
NNTP-Posting-Host: vanbc.wimsey.com
In article < 930702193...@mole.gnu.ai.mit.edu>
r...@gnu.ai.mit.edu (Richard Stallman) writes:
>It is not as easy to bypass the GPL as some people might think.
>The GPL does not permit what is being done with RSAREF and GNU mp,
>and as soon as I found out about it (a couple of hours ago)
>I informed the distributors of this.
And Mark Riordan has removed it from the distribution. I respect
his decision to do so. It is disappointing not to be able to hook
free software together with free software to make better free
software.
I'm certainly not convinced that what Richard Stallman says the
GPL implies is what the GPL implies. For instance, if I go out and
write my own library with calls like mpz_mul and such and then distribute
the hooks with that, am I still violating the GPL?
I'm going to offer some observations about how I did this, for
informational purposes only. It represents an hour or two of work.
This will only make sense, if you have both gmp documentation and
access to RSAREF (or, at least the portion of RSAREF published in
DDJ last year).
I wrote two routines to essentially perform the functions of NN_Encode
and NN_Decode to translate to and from the GNU mp format instead of
the format required by nn.c
Second I modified the routines in rsa.c and r_keygen.c to
translate call these new functions (lets call them MP_Encode and
MP_Decode) and the appropriate functions in gmp.
Some examples of the appropriate translation are as follows:
NN_Encode MP_Encode
NN_Decode MP_Decode
NN_ModExp mpz_powm
NN_Mod mpz_mmod
NN_Cmp mpz_cmp
NN_AssignZero mpz_set_ui (with zero as second parameter)
NN_ModMult mpz_mul followed by mpz_mmod
NN_Add mpz_add
NN_ModInv mpz_gcdext with appropriate parameters
NN_Assign mpz_set
NN_AssignDigit mpz_set_ui
The only part of this that required any thought is the writing of
MP_Encode and MP_Decode. The rest was purely mechanical. If any
of you have picked up my LUC package (called L3), functions very
similar to the MP_Encode and MP_Decode you'd need to write are used to
encode chunks of pseudo-random data into MP_INTs and reverse
this operation. These are called mptcs and cstmp, in the file lucas.c
(these stand for character string to mp and mp to character string).
The resulting RIPEM (TIS/PEM) is roughly twice as fast (better on some
platforms). Presumably if RSADSI and the PGP folks come to an agreement
the same technique could be applied to PGP.
The above is for informational purposes only. Consider it the report
of an interesting experiment or a testimony to how good gmp really is.
--
Mark Henderson
ma...@wimsey.bc.ca (personal account)
RIPEM key available by key server/finger/E-mail
MD5OfPublicKey: F1F5F0C3984CBEAF3889ADAFA2437433
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!uknet!bnr.co.uk!pipex!uunet!
europa.eng.gtefsd.com!darwin.sura.net!news-feed-2.peachnet.edu!
concert!uvaarpa!murdoch!fermi.clas.Virginia.EDU!gsh7w
From: gs...@fermi.clas.Virginia.EDU (Greg Hennessy)
Subject: Re: GPL not bypassed
Message-ID: < C9K1M0.9uM@murdoch.acc.Virginia.EDU>
Sender: use...@murdoch.acc.Virginia.EDU
Organization: University of Virginia
References: < 9307021931.AA12919@mole.gnu.ai.mit.edu>
Distribution: gnu
Date: Fri, 2 Jul 1993 20:44:24 GMT
Lines: 25
Richard Stallman writes:
#It is not as easy to bypass the GPL as some people might think.
#The GPL does not permit what is being done with RSAREF and GNU mp,
#and as soon as I found out about it (a couple of hours ago)
#I informed the distributors of this.
If writing code such as
main(){
call gpl_function();
}
is subject to the GNU copyright terms, isn't this the same argument
that USL is using to claim that the Net-2 distribution is a derived
work?
That sword cuts both ways.
--
-Greg Hennessy, University of Virginia
USPS Mail: Astronomy Department, Charlottesville, VA 22903-2475 USA
Internet: gs...@virginia.edu
UUCP: ...!uunet!virginia!gsh7w
Path: gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!
agate!foucault.eecs.berkeley.edu!jbuck
From: jb...@foucault.eecs.berkeley.edu (Joe Buck)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 2 Jul 1993 23:47:09 GMT
Organization: University of California at Berkeley
Lines: 31
Distribution: gnu
Message-ID: <212hdt$k8v@agate.berkeley.edu>
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
<C9K1M0.9uM@murdoch.acc.virginia.edu>
NNTP-Posting-Host: foucault.eecs.berkeley.edu
gs...@fermi.clas.Virginia.EDU (Greg Hennessy) writes:
>If writing code such as
>
>main(){
> call gpl_function();
>}
>
>is subject to the GNU copyright terms, isn't this the same argument
>that USL is using to claim that the Net-2 distribution is a derived
>work?
It's not clear to me that this is what RMS is claiming. With Net-2,
the analogy would be that gpl_function() has been cloned. If there were
another multiprecision math package in existence with the same interface
as GNU MP, then RMS would have no legitimate objection to it without
making a user interface claim. But what he is objecting to here is that
the system is designed to use GNU MP, the instructions say to use GNU MP
and that is what the makefiles are set up to do -- there is no alternative
provided. It seems reasonable to say that what is being distributed is a
system that includes GPL code, it's just being distributed in two pieces.
If it is indeed possible to bypass the GPL this way, the GPL is
effectively dead. Anyone can bypass it by shipping a system that contains
no GPL'd code, but has a script that opens an anonymous FTP connection,
downloads GPL'ed code, patches it and merges it into the proprietary code.
Compiler writers can build on gcc's back end in exactly the same way as
RSAREF proposes to use GNU MP.
--
Joe Buck jb...@ohm.EECS.Berkeley.EDU
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!uunet!europa.eng.gtefsd.com!
howland.reston.ans.net!newsserver.jvnc.net!louie!udel!news.udel.edu!
me.udel.edu!johnston
From: john...@me.udel.edu (Bill Johnston)
Subject: Re: GPL not bypassed
Message-ID: <C9KnqK.Mv3@news.udel.edu>
Sender: use...@news.udel.edu
Nntp-Posting-Host: me.udel.edu
Organization: University of Delaware
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
<C9K1M0.9uM@murdoch.acc.virginia.edu> <212hdt$k8v@agate.berkeley.edu>
Distribution: gnu
Date: Sat, 3 Jul 1993 04:42:20 GMT
Lines: 37
In article < 212hdt$k...@agate.berkeley.edu>
jb...@foucault.eecs.berkeley.edu (Joe Buck) writes:
>If it is indeed possible to bypass the GPL this way, the GPL is
>effectively dead.
I disagree. The GPL has other merits. Even if some people violate
its spirit by distributing software under any of the various proprietary
extension scenarios discussed periodically in gnu.misc.discuss, free
software will continue to advance simply because the GPL and FSF provides
a useful framework for people to make and contribute improvements.
Proprietary extensions to GPL'd software will have approximately the
same effect on this process that continued development of cheaper and
better commercial software will have. If someone chooses to develop
and distribute an inexpensive commercial alternative to gcc or emacs,
there will be people who choose to use it. When the temporary advantage
afforded by such proprietary software is implemented in GPL'd software,
these users may choose to switch back -- or not. It's their choice.
The GPL is `dead' in such a scenario only if one believes that the
GPL needs to succeed completely as an instrument to coerce people
into adopting a `purist' attitude to software development and use.
I think that FSF is correct in attempting to vigorously enforce
its license; such efforts may result in would-be GPL-circumventers
choosing better distribution policies for their work. If some attempts
to circumvent the GPL succeed, the free software community is still free
to make progress by working with from existing base of free software.
Competition from proprietary alternatives may slow the development
of free software, but it cannot stop it. That's the real strength
of the GPL -- no more orphaned products or ideas. Progress continues.
--
-- Bill Johnston (john...@me.udel.edu)
-- 38 Chambers Street; Newark, DE 19711; (302)368-1949
Xref: gmd.de gnu.misc.discuss:5333 alt.security.pgp:3297
Newsgroups: gnu.misc.discuss,alt.security.pgp
Path: gmd.de!xlink.net!howland.reston.ans.net!newsserver.jvnc.net!
princeton!hart!mg
From: m...@hart.Princeton.EDU (Michael Golan)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul3.072740.24772@Princeton.EDU>
Originator: news@nimaster
Sender: ne...@Princeton.EDU (USENET News System)
Nntp-Posting-Host: hart.princeton.edu
Organization: Princeton University
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
<212al4$o9t@vanbc.wimsey.com>
Date: Sat, 3 Jul 1993 07:27:40 GMT
Lines: 25
ma...@vanbc.wimsey.com (Mark C. Henderson) writes:
>In article < 930702193...@mole.gnu.ai.mit.edu> r...@gnu.ai.mit.edu
>(Richard Stallman) writes:
>>It is not as easy to bypass the GPL as some people might think.
>>The GPL does not permit what is being done with RSAREF and GNU mp,
>>and as soon as I found out about it (a couple of hours ago)
>>I informed the distributors of this.
>It is disappointing not to be able to hook
>free software together with free software to make better free
>software.
By definition of "free", it is possible to do what you would like.
Unfortunately, GPL'ed code is not free. It is nothing more than
FSF copyrighted code with limited free distribution and a license.
Maybe this will wake up some people to stop writing GPL code and
stop enhancing GNU tools, and write real free code: public domain.
See duel for an example.
-- Michael
--
Michael Golan | Duel, an addon to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!uunet!spool.mu.edu!wupost!udel!
princeton!hart!mg
From: m...@hart.Princeton.EDU (Michael Golan)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul3.083602.118@Princeton.EDU>
Originator: news@nimaster
Sender: ne...@Princeton.EDU (USENET News System)
Nntp-Posting-Host: hart.princeton.edu
Organization: Princeton University
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
<C9K1M0.9uM@murdoch.acc.virginia.edu> <212hdt$k8v@agate.berkeley.edu>
<C9KnqK.Mv3@news.udel.edu>
Distribution: gnu
Date: Sat, 3 Jul 1993 08:36:02 GMT
Lines: 70
john...@me.udel.edu (Bill Johnston) writes:
>In article < 212hdt$k...@agate.berkeley.edu>
jb...@foucault.eecs.berkeley.edu (Joe Buck) writes:
>>If it is indeed possible to bypass the GPL this way, the GPL is
>>effectively dead.
>I disagree. The GPL has other merits. Even if some people violate
>its spirit by distributing software under any of the various proprietary
>extension scenarios discussed periodically in gnu.misc.discuss, free
>software will continue to advance simply because the GPL and FSF provides
>a useful framework for people to make and contribute improvements.
And what exactly is this framework that can not be sustained without
a limiting and confuisng license? There is plenty of evidence of
PD and free software that has been a success and still succeed.
IMNHO, the GNU project has succeeded so well primarily due to the
efforts of a few very smart people who are also exceptionally good
programmers, and their dedication to a whole project of freely
available software.
>The GPL is `dead' in such a scenario only if one believes that the
>GPL needs to succeed completely as an instrument to coerce people
>into adopting a `purist' attitude to software development and use.
Well, why not do away with the GPL? make the software REALLY FREE!
I am cetrain, like you seems to be, that the project will continue
to evolve - maybe even at a greater speed - there are many people
out there who would like to contribute free software, but are not
willing to do so under a limited restricting license. I can assure
you that Duel would still be free code, but much better integrated
with GDB and much easier to write, if I didn't have to work around
the GPL restrictions and not copy a single line of gdb code. The
recent security software is just another case of free software being
limited in usefulness simply because the GNU code isn't really free.
>I think that FSF is correct in attempting to vigorously enforce
>its license; such efforts may result in would-be GPL-circumventers
>choosing better distribution policies for their work. If some attempts
>to circumvent the GPL succeed, the free software community is still free
>to make progress by working with from existing base of free software.
Now you start to sound like a lawyer. The FSF sure seems to be able
to catch the "bad guys", while the good guys simply avoid touching
or using GNU code. That's the way to go! make life harder for everyone
in order to catch a few bad guys. the US government seems to be on the
same track. cool.
>Competition from proprietary alternatives may slow the development
>of free software, but it cannot stop it. That's the real strength
>of the GPL -- no more orphaned products or ideas. Progress continues.
No, its the strength of free software. the GPL merely handicap it.
Sure, if there was no GPL, there would have been more GNU-based
commercial software. I don't think this is bad - if anything, it
will increase knowledge about the existence of the free software
and drive commercial software into more competition, not less.
You should also note that few commerical companies who could have
taken advantage of GNU code have done so - for example, GNU utils
could have been added to the Borland compiler. Borland had nothing
to loose, it only enhance the overall product, and having binaries
of GNU products (if the source is available someplace) within
the distribution is totally allowed. So, have they done so? nope!
Too much legal hassle, not enough utility, I suppose. my loss.
-- Michael
--
Michael Golan | Duel, an addon to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!
Xref: gmd.de gnu.misc.discuss:5340 alt.security.pgp:3299
Path: gmd.de!xlink.net!howland.reston.ans.net!spool.mu.edu!caen!
msuinfo!scss3.cl.msu.edu!mrr
From: m...@scss3.cl.msu.edu (Mark Riordan)
Newsgroups: gnu.misc.discuss,alt.security.pgp
Subject: Re: GPL not bypassed
Followup-To: gnu.misc.discuss,alt.security.pgp
Date: 3 Jul 1993 14:18:31 GMT
Organization: Michigan State University
Lines: 27
Message-ID: <2144fn$om8@msuinfo.cl.msu.edu>
References: <1993Jul3.072740.24772@Princeton.EDU>
NNTP-Posting-Host: scss3.cl.msu.edu
X-Newsreader: TIN [version 1.1 PL8]
Michael Golan (m...@hart.Princeton.EDU) wrote:
: >In article <930702193...@mole.gnu.ai.mit.edu> r...@gnu.ai.mit.edu
: >(Richard Stallman) writes:
: >>It is not as easy to bypass the GPL as some people might think.
: >>The GPL does not permit what is being done with RSAREF and GNU mp,
: >>and as soon as I found out about it (a couple of hours ago)
: >>I informed the distributors of this.
: By definition of "free", it is possible to do what you would like.
: Unfortunately, GPL'ed code is not free. It is nothing more than
: FSF copyrighted code with limited free distribution and a license.
: Maybe this will wake up some people to stop writing GPL code and
: stop enhancing GNU tools, and write real free code: public domain.
My feelings, too.
In fact, I do not think that distributing RIPEM with optional
GMP hooks violated the GNU copyleft.
However, I ceased distributing the hooks because, as nice as
GMP is, it's just not worth the hassle and ill will that would
be generated by its proprietors.
This quibbling is an example of why I never write GPL code;
if I want my code to be free, I simply place it in the public
domain.
Mark R.
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!dkuug!uts!iesd!news.iesd.auc.dk!abraham
From: abr...@iesd.auc.dk (Per Abrahamsen)
Subject: Re: GPL not bypassed
In-Reply-To: mg@hart.Princeton.EDU's message of Sat, 3 Jul 1993 08:36:02 GMT
Content-Type: text/plain; charset=iso-8859-1
Message-ID: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
Organizaion: AUC
Sender: ne...@iesd.auc.dk (UseNet News)
Content-Transfer-Encoding: 8bit
Organization: AT&T Bell Laboratories, Murray Hill
X-Newsreader: GNUS 3.15
References: <9307021931.AA12919@mole.gnu.ai.mit.edu>
<C9K1M0.9uM@murdoch.acc.virginia.edu>
<212hdt$k8v@agate.berkeley.edu> <C9KnqK.Mv3@news.udel.edu>
<1993Jul3.083602.118@Princeton.EDU>
Mime-Version: 1.0
Distribution: gnu
Date: 3 Jul 93 19:14:30
Lines: 44
>>>>> "Michael" == Michael Golan < m...@hart.Princeton.EDU> writes:
Michael> Well, why not do away with the GPL?
There would not have been a GNU Objective C front end, had NeXT not
been forced to donate it to FSF because of the GPL.
Michael> I am cetrain, like you seems to be, that the project will continue
Michael> to evolve - maybe even at a greater speed - there are many people
Michael> out there who would like to contribute free software, but are not
Michael> willing to do so under a limited restricting license.
It is interesting to compare how 386BSD (very generous copyright) and
Linux (GPL) evolve. 386BSD had a head start, but by now Linux seems
to be more popular and evolving faster.
Michael> I can assure
Michael> you that Duel would still be free code, but much better integrated
Michael> with GDB and much easier to write, if I didn't have to work around
Michael> the GPL restrictions and not copy a single line of gdb code.
If people insist on making life difficult for themselves, there are
little FSF can do to help them.
When I give my software away for free, I prefer to give it only to
people who in return will give me any enhancement or bug fixes they
make. The simplest way to ensure that, is to cover the software by
the GPL.
Michael> Sure, if there was no GPL, there would have been more GNU-based
Michael> commercial software.
There would also have been less free software. There have been large
contributions to GPL'ed software by companies who otherwise would have
kept it proprietary. The availability of commercial software would
most likely also delay the development of free equivalents.
Michael> for example, GNU utils could have been added to the Borland
Michael> compiler.
Luckily, this never happened. I doubt the DOS version of GCC would
have been so popular, had there existed a cheap 32-bit compiler from
Borland.
Path: gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!wupost!
kuhub.cc.ukans.edu!hubble.asymetrix.com!netnews.nwnet.net!
news.u.washington.edu!stein2.u.washington.edu!tzs
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Message-ID: <215at5$hnu@news.u.washington.edu>
From: t...@stein2.u.washington.edu (Tim Smith)
Date: 4 Jul 1993 01:14:13 GMT
References: <C9KnqK.Mv3@news.udel.edu> <1993Jul3.083602.118@Princeton.EDU>
<ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
Distribution: gnu
Organization: University of Washington School of Law, Class of '95
NNTP-Posting-Host: stein2.u.washington.edu
Lines: 12
abr...@iesd.auc.dk (Per Abrahamsen) writes:
>Michael> Well, why not do away with the GPL?
>
>There would not have been a GNU Objective C front end, had NeXT not
>been forced to donate it to FSF because of the GPL.
NeXT are the only people in the world capable of writing one?
--
"Pope moved that we strike from the State's brief and appendix a selection from
the Year Book of 1484 written in Medieval Latin and references thereto. The
State provided no translation and conceded a total lack of knowledge of what it
meant. The motion is granted" 396 A.2d 1054 --Tim Smith
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!sunic!uts!iesd!news.iesd.auc.dk!abraham
From: abr...@iesd.auc.dk (Per Abrahamsen)
Subject: Re: GPL not bypassed
In-Reply-To: tzs@stein2.u.washington.edu's message of 4 Jul 1993 01:14:13 GMT
Content-Type: text/plain; charset=iso-8859-1
Message-ID: <ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
Sender: ne...@iesd.auc.dk (UseNet News)
Content-Transfer-Encoding: 8bit
Organization: AUC
X-Newsreader: GNUS 3.15
References: <C9KnqK.Mv3@news.udel.edu> <1993Jul3.083602.118@Princeton.EDU>
<ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu>
Mime-Version: 1.0
Distribution: gnu
Date: 4 Jul 93 15:36:55
Lines: 12
abr...@iesd.auc.dk (Per Abrahamsen) writes:
>There would not have been a GNU Objective C front end, had NeXT not
>been forced to donate it to FSF because of the GPL.
>>>>> "Tim" == Tim Smith <t...@stein2.u.washington.edu> writes:
Tim> NeXT are the only people in the world capable of writing one?
They are the only one with a sufficient large financial interest in
Objective-C to write one.
Path: gmd.de!Germany.EU.net!mcsun!uunet!wupost!kuhub.cc.ukans.edu!
hubble.asymetrix.com!netnews.nwnet.net!news.u.washington.edu!
stein.u.washington.edu!kenney
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Message-ID: <215se3$juc@news.u.washington.edu>
From: ken...@stein.u.washington.edu (Michael Kenney)
Date: 4 Jul 1993 06:13:23 GMT
References: <C9KnqK.Mv3@news.udel.edu>
<1993Jul3.083602.118@Princeton.EDU> <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
Distribution: gnu
Organization: University of Washington, Seattle
NNTP-Posting-Host: stein.u.washington.edu
Lines: 35
In article < ABRAHAM.93...@loke.iesd.auc.dk> abr...@iesd.auc.dk
(Per Abrahamsen) writes:
>
>>>>>> "Michael" == Michael Golan < m...@hart.Princeton.EDU> writes:
>
>Michael> Well, why not do away with the GPL?
>
>There would not have been a GNU Objective C front end, had NeXT not
>been forced to donate it to FSF because of the GPL.
>
>Michael> I am cetrain, like you seems to be, that the project will continue
>Michael> to evolve - maybe even at a greater speed - there are many people
>Michael> out there who would like to contribute free software, but are not
>Michael> willing to do so under a limited restricting license.
>
>It is interesting to compare how 386BSD (very generous copyright) and
>Linux (GPL) evolve. 386BSD had a head start, but by now Linux seems
>to be more popular and evolving faster.
>
Weak argument.
The speed of evolution of Linux compared to 386BSD has, IMHO, nothing to
do with the copyright/copyleft differences. Both products are excellent
but are developed in completely different ways ... Linux's distributed
development encourages lots of new features also it can run on smaller
systems than 386BSD, this probably explains why it is more popular.
(claiming that one is more popular than the other is sure to draw some
flames :-).
----
Mike Kenney
UW Applied Physics Lab
mi...@apl.washington.edu
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!uunet!munnari.oz.au!cs.mu.OZ.AU!
munta.cs.mu.OZ.AU!fjh
From: f...@munta.cs.mu.OZ.AU (Fergus James HENDERSON)
Subject: 386BSD vs Linux (was: Re: GPL not bypassed)
Message-ID: < 9318600.1260@mulga.cs.mu.OZ.AU>
Sender: ne...@cs.mu.OZ.AU
Organization: Computer Science, University of Melbourne, Australia
References: <C9KnqK.Mv3@news.udel.edu>
<1993Jul3.083602.118@Princeton.EDU>
<ABRAHAM.93Jul3191430@loke.iesd.auc.dk> <215se3$juc@news.u.washington.edu>
Distribution: gnu
Date: Sun, 4 Jul 1993 14:24:28 GMT
Lines: 25
ken...@stein.u.washington.edu (Michael Kenney) writes:
>abr...@iesd.auc.dk (Per Abrahamsen) writes:
>>
>>It is interesting to compare how 386BSD (very generous copyright) and
>>Linux (GPL) evolve. 386BSD had a head start, but by now Linux seems
>>to be more popular and evolving faster.
>>
>
>Weak argument.
Indeed.
>The speed of evolution of Linux compared to 386BSD has, IMHO, nothing to
>do with the copyright/copyleft differences.
Yes, as Linus Torvalds himself has pointed out several times, the
reason why Linux is more popular than 386BSD is because of the
cool name :-)
--
Fergus Henderson This .signature virus might be
f...@munta.cs.mu.OZ.AU getting old, but you still can't
consistently believe it unless you
Linux: Choice of a GNU Generation copy it to your own .signature file!
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!news.funet.fi!funic!nntp.hut.fi!nntp!sja
From: s...@snakemail.hut.fi (Sakari Jalovaara)
Subject: Re: GPL not bypassed
In-Reply-To: abraham@iesd.auc.dk's message of 4 Jul 93 15:36:55
Message-ID: <SJA.93Jul4222642@gamma.hut.fi>
Sender: use...@nntp.hut.fi (Usenet pseudouser id)
Nntp-Posting-Host: gamma.hut.fi
Organization: Helsinki University of Technology, Finland
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu> <ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
Date: 04 Jul 93 20:26:42 GMT
Lines: 30
>>There would not have been a GNU Objective C front end, had NeXT not
>>been forced to donate it to FSF because of the GPL.
>Tim> NeXT are the only people in the world capable of writing one?
>They are the only one with a sufficient large financial interest in
>Objective-C to write one.
Is financial interest somehow relevant? Hint: "no".
Do you have first hand knowledge (or *any* likely indication) that
1) NeXT would not have made their compiler front-end available
without threat of legal persecution by the FSF?
2) that soneone else would not have thought, "heck; if those
guys can write an Objective-C compiler, so can I" and proceeded
to produce one?
Is there some real indication that if the GPL gave people the freedom
of choice, there would be less GNU code?
(Point to the contrary: I have written enhancements to GPL'd code but
can't give them away because I find the "*force* someone to donate
software to FSF" sentiment unethical. I can't put that requirement
on my code, sorry. Well, I'm re-writing some of the GPL'd base from
scratch now. If enough of the GNU code gets re-implemented without
the legal GPL baggage, maybe others won't have the trouble with FSF that
RSAREF is having.)
++sja
Newsgroups: gnu.misc.discuss
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!noc.near.net!
uunet!pipex!bnr.co.uk!bnrgate!bcars267!bcars267!sbo
From: s...@bcars656.bnr.ca (Stephane Boucher)
Subject: Re: GPL not bypassed
In-Reply-To: sja@snakemail.hut.fi's message of 04 Jul 93 20:26:42 GMT
Message-ID: <SBO.93Jul4165655@bcars656.bnr.ca>
Sender: ne...@bnr.ca (usenet)
Nntp-Posting-Host: bcars656
Organization: BNR
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu>
<ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>
Date: Sun, 4 Jul 1993 20:56:54 GMT
Lines: 27
>>>>> On 04 Jul 93 20:26:42 GMT, s...@snakemail.hut.fi (Sakari Jalovaara)
>>>>> said:
Sakari>(Point to the contrary: I have written enhancements to GPL'd
Sakari>code but can't give them away because I find the "*force*
Sakari>someone to donate software to FSF" sentiment unethical. I
Sakari>can't put that requirement on my code, sorry. Well, I'm
Sakari>re-writing some of the GPL'd base from scratch now. If enough
Sakari>of the GNU code gets re-implemented without the legal GPL
Sakari>baggage, maybe others won't have the trouble with FSF that
Sakari>RSAREF is having.)
Bingo!
If you don't like GPL, rewrite the code yourself, and let the FSF do
whatever it wants. The FSF is not forcing anybody to use its
softwares.
--
,
Stephane Boucher Recherche Bell-Northern Research
s...@bnr.ca
[Mes opinions sont les miennes -- My opinions are mine]
Path: gmd.de!Germany.EU.net!mcsun!news.funet.fi!hydra!klaava!klaava!
not-for-mail
From: torv...@klaava.Helsinki.FI (Linus Torvalds)
Newsgroups: gnu.misc.discuss
Subject: Re: 386BSD vs Linux (was: Re: GPL not bypassed)
Date: 5 Jul 1993 00:55:45 +0300
Organization: University of Helsinki
Lines: 27
Distribution: gnu
Message-ID: <217jl1$omv@klaava.Helsinki.FI>
References: <C9KnqK.Mv3@news.udel.edu>
<ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215se3$juc@news.u.washington.edu> <9318600.1260@mulga.cs.mu.oz.au>
NNTP-Posting-Host: klaava.helsinki.fi
In article < 931860...@mulga.cs.mu.oz.au> f...@munta.cs.mu.OZ.AU
(Fergus James HENDERSON) writes:
>ken...@stein.u.washington.edu (Michael Kenney) writes:
>
>>The speed of evolution of Linux compared to 386BSD has, IMHO, nothing to
>>do with the copyright/copyleft differences.
Actually, there was something of a "copyleft/copyright" argument in the
386bsd camp - William Jolitz and the CSGR had a disagreement about the
conditions of the first 386bsd version which resulted in the current
386bsd <-> bsdi split, and Bill had to start over "from scratch" for the
free 386bsd release (somewhat simplified, but..). So even though the
386bsd project got started before linux did, 386bsd was available for
the public later. Of course, this might be due to my incredible coding
skills (and modesty), but there are some aspects of copyright problems
within the bsd groups.
But I agree that the biggest reason is probably the open development
which turns out to be a very good idea (not that it's without its
problems, but I can generally warmly recommend it).
>Yes, as Linus Torvalds himself has pointed out several times, the
>reason why Linux is more popular than 386BSD is because of the
>cool name :-)
Tongue firmly in cheek,
Linus
Newsgroups: gnu.misc.discuss
Path: gmd.de!newsserver.jvnc.net!princeton!elan!mg
From: m...@elan.Princeton.EDU (Michael Golan)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul5.094336.4859@Princeton.EDU>
Originator: news@nimaster
Sender: ne...@Princeton.EDU (USENET News System)
Nntp-Posting-Host: elan.princeton.edu
Organization: Princeton University
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu> <ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>< SBO.93Jul4165655@bcars656.bnr.ca>
Date: Mon, 5 Jul 1993 09:43:36 GMT
Lines: 22
s...@bcars656.bnr.ca (Stephane Boucher) writes:
>Bingo!
>If you don't like GPL, rewrite the code yourself, and let the FSF do
>whatever it wants. The FSF is not forcing anybody to use its
>softwares.
The FSF doesn't force anyone to use GNU code. But it is engaged in
propaganda to fool people into thinking that GNU is the "real free
software" and that using & supporting GNU is better than using
and supporting PD. This is what I and others are fighting against.
Maybe it is time for the RFSF - the Real Free Software Foundation,
which will collect under one roof all the PD code and provide a
measure against the FSF propaganda machine.
-- Michael
--
Michael Golan | Duel, an addon to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!
Path: gmd.de!xlink.net!howland.reston.ans.net!spool.mu.edu!think.com!
grapevine.lcs.mit.edu!ai-lab!life.ai.mit.edu!burley
From: bur...@wookumz.gnu.ai.mit.edu (Craig Burley)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 5 Jul 93 14:06:36
Organization: Free Software Foundation 545 Tech Square Cambridge,
MA 02139
Lines: 110
Message-ID: <BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu>
<ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>
<SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@Princeton.EDU>
NNTP-Posting-Host: wookumz.gnu.ai.mit.edu
In-reply-to: mg@elan.Princeton.EDU's message of Mon,
5 Jul 1993 09:43:36 GMT
In article < 1993Jul5.0...@Princeton.EDU> m...@elan.Princeton.EDU
(Michael Golan) writes:
Maybe it is time for the RFSF - the Real Free Software Foundation,
which will collect under one roof all the PD code and provide a
measure against the FSF propaganda machine.
Where have you been? It has been "time" for the RFSF for at least 1.5
years, as I recall.
This subject has been beaten to death over and over again, and each
time it is (usually) a new person or persons who claims PD is
better than the GPL, and ends up saying "well, let's create a
RFSF that does PD code". Last I heard, there was a serious attempt
at writing a PD C compiler from scratch, to compete with GCC. I
haven't heard anything since, but writing a C compiler takes quite a
bit of time. (I should know, I'm writing GNU Fortran. :-)
So, please contact the other people so obviously enthusiastic about
creating a RFSF, and get going on writing PD software that proves
PD is better than the GPL! You'll only make things better, not
worse, since the cross-pollination that will result will be quite
beneficial -- and lots of VARs who currently avoid anything called
"free software" because it's generally either PD binary-only
(sourceless) or GNU (source but with GPL restrictions) will be
more than happy to swarm over all the new high-quality PD source-
available software you create, making copyrighted proprietary
executables out of it that you won't be able to maintain yourself.
(But, they'll have the freedom to do that, which is clearly what
you want, and is exactly what many GPL supporters would like
to prevent happening to their code.)
On a related topic, I hardly see it as a swipe against the GPL that
the Manifesto makes claims about "reinventing the wheel" not being
necessary and yet GPL supporters tell PD supporters to just rewrite
code. The whole point of the GPL is that _if_ you buy into the
idea of the GPL, you need _neither_ to reinvent the wheel _nor_ to
pay huge sums of money to make use of free (source-available) code
(or any of its derivatives, such as executables). People who'd
rather have everything be PD obviously haven't bought into the GPL
any more than people who want to have everything be proprietary (and,
interestingly, those two camps support each other somewhat), so
naturally at the "meta-level" of licensing issues, reinvention _is_
necessary.
Yes, the FSF has to use some anti-freedom tactics in their overall
strategy to promote (source-code) freedom. Inevitably, the FSF
might go too far in one direction or another. The cost of being
too permissive is that all the GPL code in existence will end up
being (essentially) PD, which will result in the very things
happening that not only are not desired by rms and his followers,
but which he has already _proven_ WILL happen (by his own
experience). The cost of being too aggressive is that it hurts the
FSF's image as promoting freedom -- but as long as that aggressiveness
is properly within (and understood as) the bounds of protecting
the FSF's own investment in the GPL and GPL'ed software as _source-free_
(and derivative-protected) software, the FSF can at least be said
to be acting within its own interests (including its desires for
making free software widely available, where "free" means "source
available for original and derivatives, including executables", among
other things).
But keep in mind the opposing tactics of those who are against free
(GPL-style) software, especially those who actually prefer totally
proprietary software. Remember, to create proprietary executables,
you can use PD code. Therefore, the availability of high-quality
PD source code can decrease your R&D cost without having to pass _any_
savings on to your consumers (and you don't have to tell them about
the PD origins of any of your products). So, you might be eyeing
GCC, but as its under the GPL, you can't (ab)use it as you wish.
However, any means by which you can convince those who write things
like GCC to instead make their code PD will help you, including such
means as claiming to be in favor of "really free" software and,
ironically, portraying the FSF as "against really free" software because
they attack your public attempts to make PD software available that
uses (or might use) GPL'ed software.
Now obviously there are many who truly, in their hearts, feel PD is
better than the GPL, and wouldn't _ever_ themselves release anything
_other_ than PD source code, probably including most, or all, of
those who in g.m.d object to the GPL.
However, given that once something is PD, there _is_ no author, in
terms of the resulting intellectual market, what difference _is_
there between someone who sincerely prefers PD over the GPL and
thus doesn't like the FSF, and someone who simply makes such a claim
but would turn around and take any and all PD source and make proprietary
executables out of it (plus mods that are hard to reverse-engineer,
of course)? The answer in this context is that there is no difference.
It does not matter whether contributors like myself are convinced to
release source code as PD instead of under the GPL by "true believers"
in PD or by wolves' in sheeps' clothing. If we are so convinced, our
source code _will_ be used (freely) by those who are totally _against_
free software in _any_ context, and are happy to have the free ammunition
(unless our code is next to useless). There is no debate on this point;
it has been proven true, and supporters of PD software clearly admit
it (that's their point of "truly free"; truly free to make totally
unfree derivatives).
So, to other contributors who, like myself, are intending to release
code under the GPL but see recurring threads in g.m.d and elsewhere
telling you to make it PD or you aren't really "freedom-minded", I
suggest you think long and hard about who will _really_ benefit _most_
from having your source code available in the public domain before
you take that step (something that cannot be untaken, unlike placing
it in the GPL and later making it PD, as long as you retain copyright).
--
James Craig Burley, Software Craftsperson bur...@gnu.ai.mit.edu
Member of the League for Programming Freedom (LPF) l...@uunet.uu.net
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!howland.reston.ans.net!torn!
newshub.ccs.yorku.ca!oz
From: o...@ursa.sis.yorku.ca (Ozan S. Yigit)
Subject: Re: GPL not bypassed
In-Reply-To: burley@wookumz.gnu.ai.mit.edu's message of
5 Jul 93 14:06:36
Message-ID: <OZ.93Jul5155018@ursa.sis.yorku.ca>
Sender: ne...@newshub.ccs.yorku.ca (USENET News System)
Organization: York U. Student Information Systems Project
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu>
<ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>
<SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@Princeton.EDU>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
Date: Mon, 5 Jul 1993 20:50:18 GMT
Lines: 19
Craig Burley writes:
So, please contact the other people so obviously enthusiastic about
creating a RFSF, and get going on writing PD software that proves
PD is better than the GPL!
There is no need for RFSF. All one needs to to is to contribute to
efforts similar to BSD, X or whatever else that comes up that does
not create a false dichotomy such as "the bad hoarders vs us". The
results may not be absolutely PD, but close enough while remaining
free and without ideological baggage.
... You'll only make things better, not
worse, since the cross-pollination that will result will be quite
beneficial --
This has been proven already.
oz
Path: gmd.de!xlink.net!howland.reston.ans.net!wupost!uunet!think.com!
grapevine.lcs.mit.edu!ai-lab!life.ai.mit.edu!burley
From: bur...@wookumz.gnu.ai.mit.edu (Craig Burley)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 5 Jul 93 21:09:16
Organization: Free Software Foundation 545 Tech Square Cambridge, MA 02139
Lines: 58
Message-ID: <BURLEY.93Jul5210916@wookumz.gnu.ai.mit.edu>
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<215at5$hnu@news.u.washington.edu>
<ABRAHAM.93Jul4153655@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>
<SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@Princeton.EDU>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<OZ.93Jul5155018@ursa.sis.yorku.ca>
NNTP-Posting-Host: wookumz.gnu.ai.mit.edu
In-reply-to: oz@ursa.sis.yorku.ca's message of Mon, 5 Jul 1993 20:50:18 GMT
In article < OZ.93Ju...@ursa.sis.yorku.ca> o...@ursa.sis.yorku.ca
(Ozan S. Yigit) writes:
Craig Burley writes:
So, please contact the other people so obviously enthusiastic about
creating a RFSF, and get going on writing PD software that proves
PD is better than the GPL!
There is no need for RFSF. All one needs to to is to contribute to
efforts similar to BSD, X or whatever else that comes up that does
not create a false dichotomy such as "the bad hoarders vs us". The
results may not be absolutely PD, but close enough while remaining
free and without ideological baggage.
[I skipped this post the first time, but realize I should have
responded to it:]
PLEASE understand that I have no desire to create a "false dichotomy",
especially between PD and GPL _contributors_. I do believe a
major dichotomy exists between those who freely distribute source code
and those who don't, but it is hardly a false one that needs creating
(simply ask Bill Gates or almost anyone earning $$ selling proprietary
software why they don't ship executables and loosen their copying
restrictions, and see how far you get across this dichotomy).
But if someone wants to put his/her _own_ code into PD, _fine_ by me
I say, though they should realize it'll (hopefully, if it's good
enough) get turned into proprietary derived versions they can't
maintain. Similarly, if someone wants to put his/her _own_ code under
the GPL, _fine_ as well, though obviously they have to understand
the implications (countless accusations that their code is "not really
free", therefore they didn't really intend to do any "public good",
blah blah blah).
What I happen to know for myself is that, RIGHT NOW, THIS MINUTE, there
are many corporations spending _real_ $$ doing development (bug-fixing
and enhancing) of GPL software whereby the results of their development
will be directly usable by _all_ of us. I don't know that any such
equivalent thing is going on for PD source-available software. So,
since I value the contributions (whether volunteered or funded) by
anyone, corporate, private, or whatever, I choose the GPL for my
own software.
This does _not_ mean that I'm _against_ people releasing PD software --
in fact, I'm all for it! -- just that I have a problem with people
calling _me_ all kinds of names (implicitly) in their tirades against
those of us who write GPL'ed software.
For the most part: IF YOU CAN'T USE GPL SOFTWARE, IT'S YOUR PROBLEM,
NOT THE AUTHOR'S. Stop blaming us -- we aren't blaming the authors
of PD software for the fact that, for most copies of
originals+derivatives of that software, the source code is
completely unavailable (at least, I'm not, and we shouldn't be in
general).
--
James Craig Burley, Software Craftsperson bur...@gnu.ai.mit.edu
Member of the League for Programming Freedom (LPF) l...@uunet.uu.net
Newsgroups: gnu.misc.discuss
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!wupost!uunet!
news.tele.fi!news.funet.fi!funic!nntp.hut.fi!nntp!sja
From: s...@snakemail.hut.fi (Sakari Jalovaara)
Subject: Re: GPL not bypassed
In-Reply-To: pgr@comlab.ox.ac.uk's message of Mon, 5 Jul 1993 14:14:06 GMT
Message-ID: <SJA.93Jul6004759@gamma.hut.fi>
Sender: use...@nntp.hut.fi (Usenet pseudouser id)
Nntp-Posting-Host: gamma.hut.fi
Organization: Helsinki University of Technology, Finland
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi> <SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@princeton.edu> <pgr.150051.05Jul1993@comlab.ox.ac.uk>
Date: 05 Jul 93 22:47:59 GMT
Lines: 62
> [if you write public domain code]
> And then a company will come along, take all your code, sell it, improve
> it, and bug fix it. And then you can do the same patches yourself
> (without their help, of course, so you'll have to reinvent them).
Sure, this happens. The trick is to recognize that a competitive
version will not make your version disappear (unless, of course,
your version is so bad it can't survive competition. Think of
it as evolution in action.)
I don't mind competition (and I don't mind someone else making money;
I hope they'll buy something really nice for themselves/their families.)
> (Perhaps they won't; until there is a group of people who are willing to
> try, we'll never know, will we?
Ooh, sure we know! The largest examples that come immediately to
mind are BSD UN*X, TeX and X. Two (if memory serves) of the three
are not PD but neither do they "infect" other programmers' code. All
exist in both "free" and "hoarded" (for a lack of a less silly word)
form.
Take your pick; exercise your freedom.
Do you need the enhanced versions? See the classified ads. Want the
free source? It's there. Want enhanced versions? Write them, ask for
$$ for your effort and compete with your excellent programming skills.
Or make them PD and *really* blow away the competition.
According to one theory, it is simply such a *good* thing to have
source code to the programs you use that free software will triumph
in the end. Strangely enough, there is another theory that states
that software needs restirctions of use and bogosities like proprietary
programming interfaces (gmp_mul() or whatever) to survive.
> Each to their own. You can call yours `really free' and people will say
> `but I have to pay for it! and I can't get the source from that company
> who is selling it!'
This doesn't follow. If you assume that a PD program is enhanced and
redistributed (or redistributed unchanged) for financial compensation,
surely that doesn't affect the original free version? Put it out for
FTP; as long as you don't "rm" it it'll be there, assuming it is worth
the disk space.
If someone wants the source, give it to them! If they absolutely need
commercially produced enhancements, tell them to either
- implement them themselves (which is what FSF proponents
sometimes tell people.)
- hire someone (e.g. you) to implement them (which is what FSF
proponents sometimes tell people.)
- accept that what they have is not the PD program (there are
commercial C compilers - whether they use a PD version as
an ancestor or not makes little difference at the end - except
that one using a PD base is likely to be less expensive
due to smaller development cost and increased competition.
And if you want the source, the PD base version will *still*
be there.)
++sja
Newsgroups: gnu.misc.discuss
Path: gmd.de!Germany.EU.net!mcsun!dkuug!uts!iesd!news.iesd.auc.dk!abraham
From: abr...@iesd.auc.dk (Per Abrahamsen)
Subject: Re: GPL not bypassed
In-Reply-To: sja@snakemail.hut.fi's message of 05 Jul 93 22:47:59 GMT
Content-Type: text/plain; charset=iso-8859-1
Message-ID: <ABRAHAM.93Jul6014503@loke.iesd.auc.dk>
Sender: ne...@iesd.auc.dk (UseNet News)
Content-Transfer-Encoding: 8bit
Organization: AUC
X-Newsreader: GNUS 3.15
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi>
<SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@princeton.edu>
<pgr.150051.05Jul1993@comlab.ox.ac.uk>
<SJA.93Jul6004759@gamma.hut.fi>
Mime-Version: 1.0
Date: 6 Jul 93 01:45:03
Lines: 23
>>>>> "Sakari" == Sakari Jalovaara < s...@snakemail.hut.fi> writes:
Sakari> I don't mind competition (and I don't mind someone else making
Sakari> money; I hope they'll buy something really nice for
Sakari> themselves/their families.)
I agree 100%, and this is the beauty of the GPL. Unlike proprietary
programs, the competition does not need to clone your work first. On
the other hand, unlike public domain software, you do not need to
clone the competitions work either. This makes the GPL the best
choice if you want a strongly competing market.
Sakari> This doesn't follow. If you assume that a PD program is enhanced and
Sakari> redistributed (or redistributed unchanged) for financial compensation,
Sakari> surely that doesn't affect the original free version?
It does negatively effect the development of free variant. A customer
who needs some feature only found in the proprietary variant can now
just buy it, instead of paying someone to enhance the free version.
I don't see any problems with redistributing it unchanged, in that
case you are just selling support.
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!howland.reston.ans.net!newsserver.jvnc.net!
princeton!tyrolia!mg
From: m...@tyrolia.Princeton.EDU (Michael Golan)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul6.085837.2043@Princeton.EDU>
Originator: news@nimaster
Sender: ne...@Princeton.EDU (USENET News System)
Nntp-Posting-Host: tyrolia.princeton.edu
Organization: Princeton University
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<SJA.93Jul4222642@gamma.hut.fi> <SBO.93Jul4165655@bcars656.bnr.ca>
<1993Jul5.094336.4859@princeton.edu>
<pgr.150051.05Jul1993@comlab.ox.ac.uk> <SJA.93Jul6004759@gamma.hut.fi>
<ABRAHAM.93Jul6014503@loke.iesd.auc.dk>
Date: Tue, 6 Jul 1993 08:58:37 GMT
Lines: 50
abr...@iesd.auc.dk (Per Abrahamsen) writes:
>>>>>> "Sakari" == Sakari Jalovaara < s...@snakemail.hut.fi> writes:
>Sakari> I don't mind competition (and I don't mind someone else making
>Sakari> money; I hope they'll buy something really nice for
>Sakari> themselves/their families.)
>I agree 100%, and this is the beauty of the GPL. Unlike proprietary
>programs, the competition does not need to clone your work first. On
>the other hand, unlike public domain software, you do not need to
>clone the competitions work either. This makes the GPL the best
>choice if you want a strongly competing market.
Your argument sounds nice until you get to the punch line. There is
no strong competing market for GPL-only software.
(before you give me some minor competition examples, check with
your local economist what a strong competing market usually means.
Show me sale prices, special deals, advertising, etc., for GPL code.)
While the GPL allows me to sell copies, it eliminates my ability
to protect my intellectual property by controlling copies. And
the existing competing market is based on such control, like it or not.
Maybe in the future GPL code could provide the best choice for
a stongly competing market. Right now it is clearly PD code that
does that.
-- Michael
>Sakari> This doesn't follow. If you assume that a PD program is enhanced and
>Sakari> redistributed (or redistributed unchanged) for financial compensation,
>Sakari> surely that doesn't affect the original free version?
>It does negatively effect the development of free variant. A customer
>who needs some feature only found in the proprietary variant can now
>just buy it, instead of paying someone to enhance the free version.
>I don't see any problems with redistributing it unchanged, in that
>case you are just selling support.
--
Michael Golan | Duel, an addon to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!howland.reston.ans.net!wupost!csus.edu!
netcom.com!yci
From: ad...@netcom.com (Adam J. Richter)
Subject: Re: GPL not bypassed
Message-ID: <yciC9qtDC.51n@netcom.com>
Sender: y...@netcom.com (Yggdrasil Computing)
Organization: Yggdrasil Computing, Incorporated
References: <SJA.93Jul6004759@gamma.hut.fi>
<ABRAHAM.93Jul6014503@loke.iesd.auc.dk>
<1993Jul6.085837.2043@Princeton.EDU>
Date: Tue, 6 Jul 1993 12:29:36 GMT
Lines: 85
In article < 1993Jul6.0...@Princeton.EDU> m...@tyrolia.Princeton.EDU
(Michael Golan) writes:
>abr...@iesd.auc.dk (Per Abrahamsen) writes:
>>I agree 100%, and this is the beauty of the GPL. Unlike proprietary
>>programs, the competition does not need to clone your work first. On
>>the other hand, unlike public domain software, you do not need to
>>clone the competitions work either. This makes the GPL the best
>>choice if you want a strongly competing market.
>
>Your argument sounds nice until you get to the punch line. There is
>no strong competing market for GPL-only software.
>(before you give me some minor competition examples, check with
>your local economist what a strong competing market usually means.
>Show me sale prices, special deals, advertising, etc., for GPL code.)
"Check with your local economist." Give me a break. If you
want to introduce a piece of terminology like "strong competing market",
then you should either post a definition or not use it. _Strong
competing market_ is not in _The MIT Dictionary of Modern Economics_
(4th ed.), although it defines _competitive market_ as "A market in
which a very large number of small buyers and sellers trade
independently, and as such no one trader can significantly influence
price," which is probably not exactly what Per Abrahamsen or you meant.
Regarding the PD vs. copyleft debate, here is a data point which
you might find relevant.
In July of 1992, I started my company (Yggdrasil Computing, Inc.)
to sell a free operating system on CDROM. At the time, I had a system
running Mach 3.0 the "BSD Net 2 single server" (unix emulator) up and
running. The problem was that the software wasn't protected with copyleft,
I wanted to copyleft the code, but at least one of the people at CMU didn't
want me to do that. Of course, I could have copylefted the code anyhow,
but I got the impression that that would have created a lot of ill will
with the folks at CMU. If I wanted to go with Mach, the only other
choices that I had were to get into the proprietary software business
or not to have a level playing field on which to compete with other
businesses that might make a proprietary version of my distribution.
I knew that I wanted to be in the free software business because
I personally couldn't do the engineering to maintain the whole operating
system myself, and even if I had had the money to pay people to do the
engineering, I would have needed sales like Microsoft's in order to pay
for an engineering effort like Microsoft's since engineering would have
to be paid for from sales. With free software, on the other hand, people
are quite interested in not only fixing bugs, but also in doing projects,
because they realize that they're contributing to a pool of free software
that everybody shared. Also, if their project is a derivative work such
as a device driver, then they aren't tempted by the possibilities of
making their project proprietary.
Fortunately, sometime around August, I learned about an unknown
operating system called "Linux" that at the time was covered by license
that forbade any distribution for profit. Some friends of mine and I
sent mail to Linus Torvalds explaining the situation and asking him
to switch to the GNU General Public License, which he did (I got
the impression that he was planning to anyway).
I still toyed with idea of doing a copylefted Mach/BSD
distribution for a while, but with the news of the USL-BSDI lawsuit,
I decided to delete the Mach/BSD-Net-2 system and switch to Linux.
In December 1992, I shipped an alpha release CDROM. In February, I
shipped the first beta release, complete with X windows and networking.
Today, according to the June 1993 USENET leadership report,
comp.os.linux and comp.os.linux.announce each have a higher estimated
readership than the largest newsgroup with "bsd" in its name. About
2,500 beta copies of the Yggdrasil Linux/GNU/X beta distribution have
been shipped, only about 100 of which were freebies. I also have two
competitors now (SLS and JANA), so there are competitive market forces
at work here.
I think the question of whether copyleft or public domain is
better for commercialization misses the point. The question should
be "Do you want the commercial software to be proprietary or do you
want the commercial software to be free?" If you don't establish
a legal framework to keep the commercial software free, then competitive
forces will encourage companies to make their distributions proprietary.
If you are actively hostile to copyleft, then people who want to do
commercial free software distributions may go elsewhere.
--
Adam J. Richter Yggdrasil Computing, Incorporated
409 Evelyn Ave., Apt. 312, Albany CA 94706 PO Box 8418, Berkeley CA 94707-8418
(510) 528-3209 (510) 526-7531, fax: (510) 528-8508
ad...@netcom.com yggd...@netcom.com
Another member of the League for Programming Freedom (l...@uunet.uu.net).
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!agate!
forney.eecs.berkeley.edu!jbuck
From: jb...@forney.eecs.berkeley.edu (Joe Buck)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 6 Jul 1993 18:15:39 GMT
Organization: University of California at Berkeley
Lines: 70
Message-ID: <21cfgb$ltk@agate.berkeley.edu>
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<OZ.93Jul5155018@ursa.sis.yorku.ca>
<BURLEY.93Jul5210916@wookumz.gnu.ai.mit.edu>
NNTP-Posting-Host: forney.eecs.berkeley.edu
bur...@wookumz.gnu.ai.mit.edu (Craig Burley) writes:
>PLEASE understand that I have no desire to create a "false dichotomy",
>especially between PD and GPL _contributors_.
The dichotomy is not between "PD" and GPL, since almost nothing is truly
"PD". The dichotomy is between the GPL and licensing terms using the
"Berkeley license" or the (almost identical) "MIT license".
>What I happen to know for myself is that, RIGHT NOW, THIS MINUTE, there
>are many corporations spending _real_ $$ doing development (bug-fixing
>and enhancing) of GPL software whereby the results of their development
>will be directly usable by _all_ of us. I don't know that any such
>equivalent thing is going on for PD source-available software.
There are several notable examples, the X Window system is the most
well-known example. But for the most part, free software on other terms
is developed in universities with substantial industry sponsorship, so
that the corporate money is paying for it. Yes, one of their angles is
that they wish to eventually "hoard" the resulting software. But in the
process, the rest of the world gets it for free, and the GNU project is
free to grab the most promising stuff, enhance it, and slap the GPL on it.
I think that the GPL has clearly made GCC much better than it otherwise
would have been. But I also think that the Berkeley/MIT license terms
have also contributed to the development of much valuable free software.
I want to see *both* types of licensing continue. I think that each type
of licensing can contribute to the development of free software, and I
think that the existence of Berkeley/MIT licensing has actually helped the
goals of the FSF as well. FSF has been able to grab code paid for by
corporate money that never would have appeared if the original development
had been under the GPL.
>For the most part: IF YOU CAN'T USE GPL SOFTWARE, IT'S YOUR PROBLEM,
>NOT THE AUTHOR'S. Stop blaming us --
Well, yes and no. To the extent that the GPL hampers the development of
free software on other terms (by forcing people who are creating free
software to waste their time re-implementing the code), it can backfire
and hurt the GPL authors themselves. For example, the GNU project is now
using Tcl and expect in testing (Berkeley-licensed free software), the X
Window system, as well as a number of other similar free software modules.
It is for this reason that I was very, very happy to see the LGPL
developed. The reason is that those of us who produce free software on
terms other than the GPL can use LGPL'ed code without having FSF dictate
the licensing terms we use on our code. Before the LGPL existed, the
Stanford InterViews project wasted a lot of time re-implementing the C++
streams classes because the (then GPL'ed) libg++ streams classes could not
be used with GPLing InterViews, which would have been politically and
financially impossible. Once it was LGPL'ed, the Ptolemy project at
Berkeley was able to use libg++ without a problem. I am sad to hear that
the FSF considers the LGPL a failure; the LGPL is one of the keys to the
success of the Linux system (it permits binaries to be distributed that
use shared libraries).
> -- we aren't blaming the authors
>of PD software for the fact that, for most copies of
>originals+derivatives of that software, the source code is
>completely unavailable (at least, I'm not, and we shouldn't be in
>general).
The FSF is free to grab any piece of "PD" or Berkeley/MIT licensed
software, slap a GPL on it, and prevent anyone from "hoarding" the
enhancements they add. This is basically what was done with bison,
remember?
--
Joe Buck jb...@ohm.EECS.Berkeley.EDU
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!spool.mu.edu!
olivea!grapevine.lcs.mit.edu!ai-lab!life.ai.mit.edu!mib
From: m...@geech.gnu.ai.mit.edu (Michael I Bushnell)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 06 Jul 1993 19:04:42 GMT
Organization: FOO
Lines: 24
Message-ID: <MIB.93Jul6150443@geech.gnu.ai.mit.edu>
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<OZ.93Jul5155018@ursa.sis.yorku.ca>
<BURLEY.93Jul5210916@wookumz.gnu.ai.mit.edu>
<21cfgb$ltk@agate.berkeley.edu>
NNTP-Posting-Host: geech.gnu.ai.mit.edu
In-reply-to: jbuck@forney.eecs.berkeley.edu's message of
6 Jul 1993 18:15:39 GMT
In article <21cfgb$l...@agate.berkeley.edu>
jb...@forney.eecs.berkeley.edu (Joe Buck) writes:
Once it was LGPL'ed, the Ptolemy project at
Berkeley was able to use libg++ without a problem. I am sad to hear that
the FSF considers the LGPL a failure; the LGPL is one of the keys to the
success of the Linux system (it permits binaries to be distributed that
use shared libraries).
The FSF considers the LGPL a failure because nobody has used it in the
way it was intended: nobody is making the libraries more available
through use of the LGPL. When rms asked "are you using the LGPL to
aid in the distribution of a product you otherwise wouldn't be able to
do"; nearly nobody responded. If it were actually being used, then
that would be great. But the evidence is that it isn't.
Now, you seem to be a counterexample. I would advise you to tell rms
(who does not read this newsgroup) why the LGPL is important to you.
This is an important thing to do.
--
+1 617 623 3248 (H) | The LORD is gracious and full of compassion,
+1 617 253 8568 (W) -+- slow to anger and of great kindness.
1105 Broadway | The LORD is loving to everyone
Somerville, MA 02144 | and his compassion is over all his works.
Newsgroups: gnu.misc.discuss
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!ux1.cso.uiuc.edu!
moe.ksu.ksu.edu!crcnis1.unl.edu!wupost!cs.utexas.edu!uunet!pipex!uknet!
cf-cm!cybaswan!iiitac
From: iiitac@swan.pyr (Alan Cox)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul7.093505.14742@swan.pyr>
Organization: Swansea University College
References: <BURLEY.93Jul5210916@wookumz.gnu.ai.mit.edu>
<1cfgb$ltk@agate.berkeley.edu> <MIB.93Jul6150443@geech.gnu.ai.mit.edu>
Date: Wed, 7 Jul 1993 09:35:05 GMT
Lines: 8
The LGPL is important because the gnu C library and Linux will be non
entities without it thanks to no non GPL software. As it is it appears that
people are going to have go through the library and dumb gdbm, termcap
and other bits and rewrite them because they are GPL not LGPL. Fortunately
there are as good if not better replacements for most of these.
Alan
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!agate!ucbvax!
silverton.berkeley.edu!djb
From: d...@silverton.berkeley.edu (D. J. Bernstein)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Message-ID: <17634.Jul1518.12.0993@silverton.berkeley.edu>
Date: 15 Jul 93 18:12:09 GMT
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<SBO.93Jul4165655@bcars656.bnr.ca> <1993Jul5.094336.4859@Princeton.EDU>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
Organization: IR
Lines: 18
In article <BURLEY.93...@wookumz.gnu.ai.mit.edu>
bur...@wookumz.gnu.ai.mit.edu (Craig Burley) writes:
> In article < 1993Jul5.0...@Princeton.EDU> m...@elan.Princeton.EDU
> (Michael Golan) writes:
> Maybe it is time for the RFSF - the Real Free Software Foundation,
> which will collect under one roof all the PD code and provide a
> measure against the FSF propaganda machine.
> Where have you been? It has been "time" for the RFSF for at least 1.5
> years, as I recall.
A question for the opponents of a Public-Domain Software Foundation:
How can you say that such a thing wouldn't be successful, given the
sheer volume of PD code available through services such as netlib,
comp.sources.unix, and government/industry groups?
A question for the proponents of a Public-Domain Software Foundation:
What would such an organization accomplish, beyond what has already been
done by the existing mechanisms for collecting and distributing PD code?
---Dan
Newsgroups: gnu.misc.discuss
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!
europa.eng.gtefsd.com!darwin.sura.net!udel!princeton!elan!mg
From: m...@elan.Princeton.EDU (Michael Golan)
Subject: Re: GPL not bypassed
Message-ID: <1993Jul15.202633.332@Princeton.EDU>
Originator: news@nimaster
Sender: ne...@Princeton.EDU (USENET News System)
Nntp-Posting-Host: elan.princeton.edu
Organization: Princeton University
References: <ABRAHAM.93Jul3191430@loke.iesd.auc.dk>
<SBO.93Jul4165655@bcars656.bnr.ca> <1993Jul5.094336.4859@Princeton.EDU>
<BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<17634.Jul1518.12.0993@silverton.berkeley.edu>
Date: Thu, 15 Jul 1993 20:26:33 GMT
Lines: 17
d...@silverton.berkeley.edu (D. J. Bernstein) writes:
>A question for the proponents of a Public-Domain Software Foundation:
>What would such an organization accomplish, beyond what has already been
>done by the existing mechanisms for collecting and distributing PD code?
Its ain't the PDSF, it is the SFS and the UNG project :-)
The Society for Free Software would do similar things that
the FSF does, but using PD code and w/o a political agenda.
That is, ask for donations and fund specific PD projects,
and act as a clearing house for all PD code/modifications, etc.
--
Michael Golan | Duel, PD add-on to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger, etc.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!howland.reston.ans.net!noc.near.net!uunet!
news.claremont.edu!jarthur.claremont.edu!dhosek
From: dho...@jarthur.claremont.edu (D Hosek)
Subject: Re: GPL not bypassed
Message-ID: <CA8I4D.1Dr@news.claremont.edu>
Sender: ne...@news.claremont.edu (The News System)
Organization: Quixote Digital Typography, Claremont, CA 91711
References: <BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<17634.Jul1518.12.0993@silverton.berkeley.edu>
<1993Jul15.202633.332@Princeton.EDU>
Date: Fri, 16 Jul 1993 01:43:25 GMT
Lines: 37
In article < 1993Jul15....@Princeton.EDU> m...@elan.Princeton.EDU
(Michael Golan) writes:
>d...@silverton.berkeley.edu (D. J. Bernstein) writes:
>>A question for the proponents of a Public-Domain Software Foundation:
>>What would such an organization accomplish, beyond what has already been
>>done by the existing mechanisms for collecting and distributing PD code?
>Its ain't the PDSF, it is the SFS and the UNG project :-)
>The Society for Free Software would do similar things that
>the FSF does, but using PD code and w/o a political agenda.
>That is, ask for donations and fund specific PD projects,
>and act as a clearing house for all PD code/modifications, etc.
Plus it would be producing code free of the silly GPL. As an example
of how the GPL would have hurt the development of a product, consider
TeX. This has the same license components that I described (less the
credit requirement even). The state of equation editors in general
has been improved by the availability of TeX source code with no
restrictions on its use. What's more by not requiring source disclosure,
the propogation of the program to different platforms has been aided.
Consider this: I can run a full implementation of TeX on an 8088 PC
running MS-DOS. I cannot do the same with GNU EMACS. My claim is that
no one has done the former because for the work involved there would be
no pay-off. Write the code and your potentially unscrupulous competitors
will just steal it from you since you have to make it freely available.
On the other hand, in the TeX world, because of the lack of such
requirements, it was possible for several competing TeX implementations
for DOS machines to exist fairly early on. In the meantime, public domain
products eventually appeared. At first they were uncompetitive, but
as time has gone on, emTeX (free) is now the fastest TeX for DOS
available. The competition as forced some commercial TeX vendors out
of business and others to improve the quality of their product. This has
enhanced the greater good of the whole TeX community and not just a
select few with access to the internet.
-dh
Path: gmd.de!xlink.net!scsing.switch.ch!news.univie.ac.at!
paladin.american.edu!howland.reston.ans.net!math.ohio-state.edu!
news.acns.nwu.edu!129.105.49.52!richter
From: ric...@kepler.math.nwu.edu (Bill Richter)
Newsgroups: gnu.misc.discuss
Subject: Re: GPL not bypassed
Date: 16 Jul 1993 19:42:53 GMT
Organization: Northwestern University, Evanston, IL USA
Lines: 25
Distribution: gnu
Message-ID: <RICHTER.93Jul16144253@kepler.math.nwu.edu>
References: <BURLEY.93Jul5140636@wookumz.gnu.ai.mit.edu>
<17634.Jul1518.12.0993@silverton.berkeley.edu>
<1993Jul15.202633.332@Princeton.EDU>
<CA8I4D.1Dr@news.claremont.edu>
NNTP-Posting-Host: kepler.math.nwu.edu
In-reply-to: dhosek@jarthur.claremont.edu's message of Fri,
16 Jul 1993 01:43:25 GMT
Plus it would be producing code free of the silly GPL. As an example
of how the GPL would have hurt the development of a product, consider
TeX ... I can run a full implementation of TeX on an 8088 PC
running MS-DOS. I cannot do the same with GNU EMACS.
Couldn't resist this one:-) There would have been be one big advantage
to TeX being part of the GNU project---better coding!
Emacs is really a beautiful piece of code compared to TeX. I'm a big
fan of TeX; the main thing I use the computer for is producing my
shiny dean fodder. But's it's not structured programming.
TeX would be very much improved if it directly created a postscript
file, for instance, rather than these dvi (device dependent) files.
As a result, it's really a headache to try to include a picture. You
can special in postscript of course, but it won't have nice TeX fonts.
Probably the reason that GNU emacs is slow to come to the PC world has
more to do with the limitations of the PCs and the operating systems.
But an 8088? What museum did you steal it from? With Linux/X386free
you can run emacs 19 on a a good PC, and on my 3b1 I've got an
excellent version of GNU emacs 18 running, nice mouse support.
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!
noc.near.net!uunet!news.claremont.edu!jarthur.claremont.edu!dhosek
From: dho...@jarthur.claremont.edu (D Hosek)
Subject: Re: GPL not bypassed
Message-ID: <CAFEzF.2y9@news.claremont.edu>
Sender: ne...@news.claremont.edu (The News System)
Organization: Quixote Digital Typography, Claremont, CA 91711
References: <1993Jul15.202633.332@Princeton.EDU>
<CA8I4D.1Dr@news.claremont.edu>
<RICHTER.93Jul16144253@kepler.math.nwu.edu>
Distribution: gnu
Date: Mon, 19 Jul 1993 19:18:42 GMT
Lines: 61
In article < RICHTER.93...@kepler.math.nwu.edu>
ric...@kepler.math.nwu.edu (Bill Richter) writes:
> Plus it would be producing code free of the silly GPL. As an example
> of how the GPL would have hurt the development of a product, consider
> TeX ... I can run a full implementation of TeX on an 8088 PC
> running MS-DOS. I cannot do the same with GNU EMACS.
>Couldn't resist this one:-) There would have been be one big advantage
>to TeX being part of the GNU project---better coding!
Really? I can follow TeX a hell of a lot better than any of the GNU
stuff I've looked at.
>Emacs is really a beautiful piece of code compared to TeX. I'm a big
>fan of TeX; the main thing I use the computer for is producing my
>shiny dean fodder. But's it's not structured programming.
It never claimed to be. Knuth's style is admittedly idiosyncratic, but
it's by no means unreadable. Or perhaps you're looking at web2c output
or worse yet the Pascal brick?
>TeX would be very much improved if it directly created a postscript
>file, for instance, rather than these dvi (device dependent) files.
No. DVI means device independent and that is really true. I can
preview a DVI file far quicker on any platform than I can go
DVI->PS->PS-Preview-Of-Any-Sort. And I can easily generate a
driver for any realistic output device from scratch in 8 weeks
or less (which includes time to learn the device characteristics
and testing).
>As a result, it's really a headache to try to include a picture. You
>can special in postscript of course, but it won't have nice TeX fonts.
It's actually very easy. As for TeX fonts in the pictures, that has
nothing to do with how the graphics are included. It has to do with
font capabilities of the graphics program (if it can only do PS fonts
then it's a matter of getting your fonts in PS form, not that big
of a problem, really) and your font choice for the base text. I
use PS fonts with the current implementation of TeX with no difficulty.
One of my clients needed TIFF inclusion so I added that into dvips
for him. Try that with a PS-generating TeX.
>Probably the reason that GNU emacs is slow to come to the PC world has
>more to do with the limitations of the PCs and the operating systems.
Partly. The point was that it was possible, but without the possibility
of getting a fiscal payback why bother?
>But an 8088? What museum did you steal it from? With Linux/X386free
>you can run emacs 19 on a a good PC, and on my 3b1 I've got an
>excellent version of GNU emacs 18 running, nice mouse support.
That's not the point. I've got GNU Emacs running on my machine too.
I personally made the leap from a Z-80 based computer in the mid-80s
to a 386/25 and skipped the steps in the middle. I don't particularly
need or want EMACS for an 8088. I'm just pointing out that there's
no real impetus to do the hard ports without a chance of reward for
the efforts. Sorry, but a warm fuzzy feeling doesn't cut it for me.
-dh
Newsgroups: gnu.misc.discuss
Path: gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!
darwin.sura.net!sgiblab!sono!imdvlf!al
From: a...@imdvlf.acuson.com (Al Petrofsky)
Subject: Re: GPL not bypassed
In-Reply-To: dhosek@jarthur.claremont.edu's message of Mon,
19 Jul 1993 19: 18:42 GMT
Message-ID: <AL.93Jul20220056@imdvlf.acuson.com>
Sender: a...@acuson.com (Al Petrofsky)
Organization: Acuson; Mountain View, California
References: <1993Jul15.202633.332@Princeton.EDU>
<CA8I4D.1Dr@news.claremont.edu>
<RICHTER.93Jul16144253@kepler.math.nwu.edu>
<CAFEzF.2y9@news.claremont.edu>
Distribution: gnu
Date: Wed, 21 Jul 1993 05:00:56 GMT
Lines: 22
From: dho...@jarthur.claremont.edu (D Hosek)
Date: Mon, 19 Jul 1993 19:18:42 GMT
That's not the point. I've got GNU Emacs running on my machine too.
I personally made the leap from a Z-80 based computer in the mid-80s
to a 386/25 and skipped the steps in the middle. I don't particularly
need or want EMACS for an 8088. I'm just pointing out that there's
no real impetus to do the hard ports without a chance of reward for
the efforts. Sorry, but a warm fuzzy feeling doesn't cut it for me.
But the 32-bit MS-DOS emacs and gcc ports were hard ports, and they were
done. And there was reward for the effort:
1. The authors can now use emacs and gcc under 32-bit MS-DOS.
2. They can use other people's improvements to the ports.
3. They have improved their reputations and increased
their money-making potential as programmers.
I would imagine the main reason nobody has done an 8088 port is that
it would have to be so crippled as to be useless.
-al