[Top][All Lists]

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

Re: More GPL questions

From: David Kastrup
Subject: Re: More GPL questions
Date: Wed, 18 Oct 2006 01:28:49 +0200
User-agent: Gnus/5.11 (Gnus v5.11) Emacs/22.0.50 (gnu/linux)

Stefaan A Eeckels <> writes:

> On Tue, 17 Oct 2006 23:49:05 +0200
> David Kastrup <> wrote:
> You define "based on" for source code as "not runnable in compiled
> format without the library and containing references to the
> library's API". The way I would define "based on" for source code is
> "incorporating modified or unmodified source code from the library".
> Your definitions opens, IMHO, more cans of worms you'd ever wanted
> to open in an entire lifetime.

Well, the U.S. copyright system is a larger can of worms I'd ever
wanted to have opened, but it is not my choice to make.  I am merely
trying to get an estimate about how large the can might be.  The
copyright code is more explicit for literary works, so trying to
extrapolate from there seems a reasonable approach to me.

>> Anyway, for Qt the point is quite moot, since an API-compatible
>> library (Qt commercial) under a different license is available, and
>> so source code written for Qt does not require a GPLed library
>> version to run.  The code might be useless without Qt, but not so
>> without _GPLed_ Qt.
> I believe neither the existence of another library or another
> license, nor the "usefulness" of source code without the called
> libraries or the supporting OS has anything to do with its copyright
> status.

Well, I think that this can construed differently.  A book containing
annotations, for example, is useless without the original, and it is
considered a derivative work.  If I put instructions about where to
download a copy to be used against the intent of the license, am I not
party to the process?  If I sell fully functional guns without
ammunition, but tell people where to mail-order them, can I claim that
I did not sell weapons since without ammunition, the guns are just

Now if the customer chooses to link a GPLed library (but not
redistribute the result), he is allowed to do so.  But the choice is
not really for the customer to make: he _has_ to do it, as that is
what his original purpose requires.  Is the provider of source code
involved with "distributing" it if he just gives pointers?

In a sane world, probably not.  Unfortunately, we are not living in a
sane world: Bittorrent seed sites have been shut down repeatedly in
the U.S., and the Bittorrent seeds basically just tell where to
download material, not providing it itself.

Maybe the FSF interprets the powers of the GPL in the most stringent
way in order to have it challenged: if courts finally decide that
linking never affords any power of copyright, it will be a win for
software freedom.  Every power that the GPL can be shown not to have
is a power that no other license can have.

So there is little to be gained by the FSF by not assuming the most.
If it gets thrown out by some court eventually, it sets a precedent
that also restricts the power of proprietary licenses.

> Usefulness is not a criterion for software to be a derivative work
> or not. An OS (even compiled) is useless without a computer with the
> appropriate instruction set, but that doesn't make it a derivative
> work of the processor's microcode, design, or assembly language
> mnemonics (I still have Intel's 8080 Assembly Language manual, which
> mentions at the bottom of every page of the first chapter that the
> mnemonics are copyrighted: "ALL MNEMONICS © 1974, 1975, 1976, 1977

Well, thank _God_.  That's probably the reason that Zilog invented its
own, much saner mnemonics.

David Kastrup, Kriemhildstr. 15, 44793 Bochum

reply via email to

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