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

Re: Q: Combining proprietary code and GPL for in-house use

Gregor Hoffleit <gregor@mediasupervision.de>:

> My program works well without the GPL library. Now if I sell this program,
> and add a module that the customer may link with the GPL library, would I
> violate the GPL of the library, and why ? If the customer linked in the
> module, would he violate the GPL, and why (he's in the same situation as I
> was: linking a GPL module with proprietary code is not forbidden, only the
> distribution thereof).
> Now what happened if the customer paid somebody else for linking the GPL
> library with the module of my program (the consultant would do this on the
> machine of the customer, FWTW). Would somebody violate the GPL here. Who and
> why ?
> If somebody who knows could please comment on this, or point me to a
> discussion of this. I just don't see where exactly the violation might be.

I don't know of a good discussion of this question, though it is often
asked. I think perhaps some of the people who understand the issues
best don't want to discuss them in public.

I think RMS claims that by distributing a program that uses a GPL
library you are in effect distributing a derived work that includes
the library even if you are not distributing the library yourself.

On the one hand it seems strange: how can you be violating X's
copyright if you never copied X's code?

On the other hand, if it were acceptable to use a GPL library with any
program, you could take any GPL program, modify it to make it into a
library, get an accomplice to distribute the newly created GPL library
then distribute your proprietary code that uses that library, which
would defeat the purpose of the GPL.

I don't think these things have ever been tested in a court of law, so
there's little point in talking to lawyers about it. The opinions of
today's programmers might to some extent influence future law, so you
also have to ask yourself what you would like the law to be ...


Reply to: