Re: copyleft (was: Re: Owning a piece of my mind)

[Eric Eldred <eldred@mediaone.net>]

Richard Stallman wrote:
....[some interesting points deleted]
>

thanks for taking the time to respond to these
frequently-misunderstood
issues.  sorry to add even more comments here....

> 
>                                            Thus it is a sort
>     of shrinkwrap license, similar to the ones that UCITA
>     intends to enforce in some states.
> 
> No, it is not a shrink wrap license, nothing like that.  You've got
> the picture entirely wrong.  Shrink wrap licenses are based on the
> claim that the user has agreed to a contract by acquiring a copy.  The
> GPL does NOT claim that the user has agreed to any sort of contract by
> acquiring a copy.

i'm not a lawyer, but i fail to understand whether the GPL is
a contract or not.  If it is a contract, doesn't it have to
be understood legally in the jurisdiction to which it is agreed and
applied?  In that case the Corel lawyers might have something
to say about how contracts are interpreted under Canadian law,
and might need to negotiate a better contract.

If it is not a contract, but simply a statement of
what terms the copyright owner will license reuse or copying
of the work under, no matter what the jurisdiction, then
why isn't it a shrinkwrap license?  A shrinkwrap license can
set terms which the licensee must follow for any legal use
of the work not to infringe on the intellectual property of
the copyright owner.  It doesn't need to be a contract, does it?

its being a "shrinkwrap" license simply means that no money
is exchanged separately, and that a purchase means agreeing to
the license. In the same sense, it is true that merely receiving
and using a GPL'd product does not mean agreeing to the license,
but there is no separate agreement to the license except when
one reads and accepts the GPL attached to the product at the
point when one intends to copy or reuse the product--not just
use it. It is not negotiable as a contract it seems, and no
money is exchanged (although there is a statement to contact 
the FSF to discuss terms).

but perhaps this is academic.  it would be easier to determine
the truth if there were some case law to fall back on, to
predict how courts will interpret the license.  that point will
be reached when we have a case of copyright infringement against
someone reusing a GPL'd product without fully honoring the terms
of the GPL, and uses as a defense his own interpretation of the
GPL.

> 
>       The GPL allows
>     redistributing the work only under certain conditions.
>     Thus it "infects" the work being distributed
> 
> This is based on a misunderstanding of the way the GPL works.
> Actually, the situation is much simpler: my choice of the GPL affects
> what you do with your program if it includes a substantial amount of
> my code.

I guess that is why some people call that "infecting" other
programs.  The new license applies not only to the new work by me,
but also has to be the exact same as that of the work I am
substantially including, with no further limitations (unless
the FSF or owners of the GPL change it in the future).

you see this as being an effect on your own software by the GPL;
i might see this as an effect on my own software.  but if it
were the LGPL then it would be different.  then there would be
no question of "infection"; i could issue my software under a
new license different from GPL.

perhaps it is true that the term "infection" is unfortunate.
but using it is a way to confront the underlying issue here:
mixing in "non-free" software with "free" software is something
that the GPL was intended to prevent.  am i wrong?

> These two groups are not the only ones: we who launched the
> development of the GNU/Linux system have a different view.  We are
> concerned with free software, free in the sense of freedom; we welcome
> corporations when they respect and support our freedom, and we oppose
> them when they try to separate people from their freedom.
> When corporations help us work for freedom, we are happy to see
> them make money, and we don't want to antagonize them.
> But when corporations add non-free software to GNU/Linux,
> spoiling its free status, they are not helping us and we do not
> hesitate to say so.  As for market momentum, that is a side issue.

i must have overlooked a proper explanation of the correct
FSF position in my earlier post; thanks for this summary.

Yes, market momentum I agree is a side issue.  of course, for
the other people who are caught up in the latest south sea
bubble that is another matter.  certainly the free software
movement
has gained some respect recently because of this very position,
in contrast to the so-called "Open Source" (tm) people.

i guess the issue boils down to deciding on the practical
application, of interpreting just "when corporations add non-free 
software to GNU/Linux, [or any GPL'd work] spoiling its free
status,"

there are two parts to this.  the first is "free" in the sense
of not charging royalties (or other restrictions) to limit 
further distribution of the work.  the second is maintaining 
"open source" and permitting further improvements and modifications 
to the code.

in the case of some "open source" advocates, it seems that they
are quite willing to mix "free," "open source," "proprietary,"
and "commercial" all together as long as it is called Linux.
they may say they are making money only from the hardware, or only
from support, but really they do also make money from other
software that is mixed in with the GPL'd software.  it seems
that there is room here for a lot of clarification about the
meanings of the terms, and the underlying principles involved.

it makes a lot of sense to point out that when the software
is released without source code, there is no way that others
can understand it or improve on it.  therefore it is specious
for those advocates to say they are part of an "open source"
movement, even if the [what you call "non-free"] software
is designed to run on Linux, and is released without cost.
even if the software is "free" it might have the effect of
discouraging others to release software that is truly free,
with source code.

but it is equally true to point out that when Sun releases
Solaris or Java for "free" (even including source code)
it still holds onto rights which restrict others from freely
modifying or redistributing it.  as I understand it, the GPL
was crafted to answer this problem as well.

and this is the point (not the other points about "freedom")
that bother commercial software people (not "proprietary").
they don't want another company to come along and make money
from their work, no more than a FSF developer wishes Sun
to make her product proprietary.  and there are various ways
for the other company to make money from it, not always
by selling it--having control over development so it can be
made to run better on some other windowing operating system
might be just as important.  this latter concern needs more
reasoned exploration instead of passionate argument. it is
too bad this point is submerged under the usual prejudices
against companies that wish to make money (not "free beer").

but my point was not in this philosophical dissection, but
rather to suggest that it is not wise for non-lawyers to
start debating these matters in forums such as Slashdot.
instead, FSF should retain lawyers to enforce its license
and copyright and negotiate with licensees such as Corel or
Sun whenever these puzzling problems of interpretation arise
--even if Corel and Sun don't come to the FSF first and ask.
otherwise, the judges will be forced to decide cases that
might not be the best ones for our interpretation of the GPL.
the arguments for "freedom" unfortunately can be interpreted
by the courts only by means of the licenses and the law.

i have gotten a lot of mail from some people who think i
am all wet.  they say the answers are obvious and that they
need to explain them to me.  i respond that this is an
evolving process, and that we can't always predict how
the law in this matter will be interpreted.  we need to
discuss the underlying issues so that we can raise them
properly.  i think rms has explained the FSF position on
freedom and the Gnu philosophy quite well.  it certainly
stands as a beacon to admire, while unfortunately i am
afraid the Internet and much of software slides downward
into some money-pits and commercial world far from these 
ideals.  i am not convinced at all that "Open Source" or
encryption are going to preserve our fundamental freedoms.

finally, i want to point out that the Public Domain and
Free Software are two different things--the GPL relies
on some strong copyright to enforce its freedom, while
the freedom ought to come automatically from the Public
Domain, without "intellectual property."  What we need
is some strengthening of the Public Domain, in my opinion.

The recent article in the new york times quoting hal
varian on the microsoft case was interesting in that
regard.  he proposed that the price for the operating
system be reduced automatically after a few years.
that would give ms an incentive to innovate and room
for some competition (along with less restrictive
licensing practices).

similarly, one could argue that copyright law be changed
for software so that the term is no longer than the
normal period for commercial exploitation, say 14 years
or even five years.  and that all digital products must
have source code deposited in the library of congress
so that when the term expires there is something to
revert to the public domain.  (the alternative of the
government's refusing to enforce copyright on digital
products at all remains an alternative that needs 
exploration in another thread.)

why wouldn't the software be "free" then, in the sense of
being open source, and also modifiable and redistributable
without charge?  it would be hard to see why this would
be considered depriving software developers of any of
their creative expressions, since there is no reason to
believe that anyone would have any reason not to hire
the programmers to do further work, or to give them
credit for the work they did, and presumably few products
would have much commercial value after that period.

(as an example, visicalc was recently released for "free"
on the internet in one version. but only the object code.
one can't adapt it to a modern computer or os, or easily
take it apart to learn Frankston's programming tricks.)

when i discussed the GPL with bruce perens, though, i got
the response that "Open Source" (tm) developers would prefer
that the copyright term be as long as possible.  why?