[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: clef, readline, and the GPL

From: Camm Maguire
Subject: [Gcl-devel] Re: clef, readline, and the GPL
Date: 06 Nov 2003 17:58:10 -0500
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2


[ Ah, this issue again -- I would so prefer it to just go
  away.  Yet it is important, and, like everyone else, I have 
  my own thoughts on the matter, which I feel in this case obliged to
  share.  Please note these are only my opinions, eminently fallible
  and flawed.  I've expressed them below with a bit more vehemence
  than I'm accustomed to -- I hope no one finds them offensive in any
  way, as this is the furthest from my intention.  I do want to remain
  open to differing perspectives as time goes on, yet feel the
  following needs to be said at this juncture. Please read it with
  indulgence, as it is written with too little time for review. ] 

Tim Daly  <address@hidden> writes:

> Bill,
> clef and readline tend to do the same things but at different levels.
> clef works at the axiom read level (in fact, you can "clef"-ify any
> program by typing "clef -e program").
> <flame-on>
> GPL and Modified BSD are "compatible" licenses and can freely use each
> other's code as far as I understand the issue (this is from Richard
> Stallman). So I'm proceeding on the strength of this statement. Anyone
> who disagrees with this and wants to debate it either (a) has to get
> Richard to change his position or (b) has to get an IP LAWYER'S
> statement to the contrary. I am not a lawyer so I consulted the person
> that wrote the license as to what the GPL is intended to mean and how
> it applies to Axiom. GPL requires that you distribute the changes if
> you distribute the code, which we do. So we are not in violation of
> the spirit of the GPL. Further, there is no legal reason why the code
> can't be dual-licensed under both Modifed BSD and GPL. The only time a
> legal issue would arise is if someone tried to "take Axiom private",
> AND the private copy included GPL code, AND they tried to distribute
> the private copy AND the distribution did not allow access to the
> source code. Until that happens, as far as I understand it, the
> issue is moot. 
> Again, none of us are lawyers so we're about like English majors
> debating Garbage Collection strategies. If someone can get Eric Molgen
> (the FSF lawyer) to sort thru the relationship of readline, GCL,
> Axiom, clef, etc. then I'd feel we knew where we stand.
> I'd advise caution about reading licenses. I've had a couple actual
> legal issues with licensing and lost a job because I refused to sign
> one. I've had discussions with attorneys about IP issues.  I've read 6
> books on the subject at this point and I have only one comment: Any
> license does NOT mean what you think it means.  Legalize is NOT
> english. You CANNOT, without a law degree and reading particular cases
> actually concluded by the courts, draw valid conclusions. If you think
> otherwise then I suggest you try representing yourself in court.
> I'm just tired of talking about things I'm clearly unqualified to discuss.
> If someone wants to raise legal issues I'd appreciate knowing what degree
> they hold in law and where they passed the bar exam.
> Sorry for the flame but I've been "poked" on this point so often I'm
> overly senstive about it. I think I understand where Richard was
> coming from as I am old enough to have run into the same issues he
> encountered. I don't believe it benefits any of us to "demonize" the
> GPL. I don't know anyone who is willing to spend $100K on attorney's
> fees to force GCL to carry the GPL. What would it gain anyone (except
> the lawyer)? It would have no practical effect since the code is
> already free. So if no-one will sue and no attorney will render legal
> advice why are we wasting our lives discussing things we're
> unqualified to discuss?
> <flame off>
> That said, I do appreciate that you did not copy the developer list
> as it will only touch off another pointless flame-fest.
> Bill, I know this seems directed at you but it REALLY isn't.
> It's just my frustration being vented.
> My sincere apologies if you found it offensive.
> Tim

Tim, thanks very much for sharing your perspective here.  In short, I
agree with you completely.

The facts are as you state, to the best of my knowledge.  Given this,
one wonders what possible thought process could lead one to object
even to a full GPL'ed GCL where axiom is concerned, other than a
secret hope to somehow personally cash in on the hard work given
freely by others.  I'm sure there must be some reasonable such motive,
its just that I am at a loss to fathom it.  

I don't know if your experience is similar to mine, but as far as I
can recall, there has been exactly one GCL patch submitted by a user
who has expressed concern over the GPL.  Far more frequent are GPL
complaints from people who have contributed absolutely nothing to the
GCL project.  I've also heard of exactly one, IMHO, valid concern over
the GPL from someone actually using it in a real project, and that is
from someone wanting to write a game in lisp.  Far more frequent are
GPL complaints from people not even using GCL at all, let alone in a
serious project.  Perhaps the most serious industrial user of GCL at
present is a for-profit corporation which to my knowledge has no
problems whatsoever with GCL licensing, as they are using it for their
internal design process and not trying to sell or distribute GCL
images to others.  In short, the body of people objecting to the GPL
for GCL have thus far proven much less useful to the project than the
general body of users and contributers.  These observations lead me to
tentatively conclude at present that GCL will not gain anything in
trying to distance itself categorically from the GPL, but will rather
lose potential development opportunities in the future in so doing.

Now I am not so extreme as to want to deny someone from selling a lisp
game based on GCL -- there are too many social realities surrounding
software games to envision (for me at least) a viable open source
model for this genre as a whole.  Hence my license proposal earlier
posted, in which I outline a possible separation of the GPL'ed and
LGPL'ed parts of GCL as an option for such users.  This has not yet
been effectuated.  At present, I'd just like to note that we have
three GPL'ed components in GCL, unexec, readline, and bfd.  Readline
is totally optional and can be disabled at configure time.  BFD is
similarly optional on all platforms if one wants to live with the
frustrations of dlopen, and optional only on sparc and i386 otherwise.
Unexec is absolutely unremovable at present, but could be in
principle.  Frankly, I'm almost completely unmotivated to put in the
work to get this done in the short term, as the presence of unexec,
which GPL's the GCL image as I understand Richard, is not actually
posing a difficulty to any *real* user of GCL at the present time, not
to mention the loss of functionality such removal would entail.  I'm
even rather using it as a test case to find out whether anyone wanting
this will actually contribute the necessary patch to GCL themselves.

This of course it totally apart from the banner removal issue, which I
do intend to address for axiom shortly.  I'd like to put in some
function by which the user can remove the banner as long as their own
notice is compatible with the image they're using.  Nothing enforced of
course, just leaving the user to their honor, perhaps even in the
function name, like (remove-gcl-banner :license-compatible-with 'foo).
As far as I know, this is the only GCL licensing issue meaningfully
impacting axiom at all.  Please let me know if I am in any way
misunderstanding this situation vis a vis axiom.

Furthermore, the long term future of GCL is best served, IMHO, in a
gcc-frontend option.  We will have to integrate with GPL'ed code (in
the compiler package only) in all likelihood, but we will likely get a
solid portable lisp platform for the benefit of all lisp users,
exercising a unifying influence on the community, and lasting far
beyond my or any other "expert's" overseeing of the project.  Binary
modules which don't usurp the compiler would be allowed as described
in my proposal.

I'm in this project for reasons I perceive to be very similar to yours
-- to preserve works of great value which are unsustainable in the
commercial world, and to make people happy.  But I'm beginning to come
to the conclusion that everyone cannot be made happy simultaneously,
and that at some point we must ask who are we here to serve?  School
children?  Researchers in poor and rich countries alike? Graduate
students?  Or would-be entrepreneurs seeking a profit from these users
as long as it can be had, after which time they must of a necessity
abandon the code and all the work gone into it?  Please don't get me
wrong, there is nothing the matter with such commercial activity,
(providing one doesn't coerce or paralyze the user in the process) --
indeed it is a vital, moral and ethical way for meeting the short term
needs of society in a balanced fashion.  But does this latter group of
would-be entrepreneurs really need free contributions from volunteer
developers, or is it rather the former group that needs such

While critically important and moral when performed ethically,
commerce is often incapable of producing or sustaining works of
enduring value.  Can anyone imagine a company proving Fermat's
theorem, painting the Sistine Chapel, or discovering relativity?  This
is no slight to commerce, as these sorts of things are outside the
scope of its central purpose, which is to distribute goods and
services, and to satisfy the material needs of people, efficiently.
Axiom's own history bears this conclusion out.  I believe that there
is therefore room and need for some extra-commercial force in society
to balance and complement the effects of commerce, not to replace it.

It is very important, in my opinion, when operating in this realm not
to cede automatically to the perspectives of those with different
interests in mind.  To do so would often be to compromise the very
enduring value one tries to create.  While commercial perspectives are
often loud and insistent, they are short term, ultimately much less
effective to the task at hand, in my opinion, and pale in weight when
compared with the frequently unuttered perspectives of the much larger
mass of end users.  The world is full of the "corpses" of good ideas
ruined or nearly ruined chasing after commercial interests.  Look at
the difference in building axiom on Linux and Windows.  Look even at
lisp in its hopelessly fragmented state.  Of all communities, one
would think that lisp users would be foremost in recognizing the
deleterious effects of ceding the reigns to commercial interests, yet
the opposite is more nearly the case.  Lisp users, IMHO, are
especially beholden to commercial perspectives, just waiting for that
next company to come along and champion their cause, while unified,
user-supported, and gcc-based developments leave this well-designed
language further and further behind.  I can't help but find this
tragic.  I hope that the GCL project can help rekindle end-user
interest in, and end-user contributions to, the world of lisp.

These are of course just my thoughts on the matter, and are
doubtlessly full of mistakes.  I want to keep an open mind going
forward, and encourage people with different perspectives to enlighten
me on what they've found (though I have almost no time for discussions
of this sort).  I'd just like to say that I'll be far more impressed
with people backing up their perspectives with contributed patches --
aka the "show me the code" philosophy.  I'd like to see a sustainable
model for GCL's future, and right now I'm more convinced by GPL'ed
examples than by anything else.

Take care,

Camm Maguire                                            address@hidden
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah

reply via email to

[Prev in Thread] Current Thread [Next in Thread]