Re: Is there any FREE alternative for l
> From: Leandro GFC Dutra
>
> Not necessarily... for example, take the KDE case. It is GPLd, ...
<IMHO>
No, it is not, even if there is written so.
It is not GPL because GPL ensures that all GPLed program can be ported
to Windoze and sold as a statically linked binary file, without paying
*any* royalty.
You are not *allowed* to do this with KDE, and this is because of a
*decision* freely taken by the developers, which in facts contraddicts
and nihils their desire to use GPL as license.
</IMHO>
>
> The point here is that even if theoretically possible, it can
> be very
> difficult to un-GPL a piece of code if you are not the sole author of
> it. If there are authors uncooperative with your will to change the
> licensing terms, or even too many authors, your program will be
> effectively GPL for eternity.
<DISC="I'm not a lawyer">
It depends on the law of the country where the main author resides (taking
the simplest case when he lives in its own country).
For example, Italian law (which is absurdelly complex) grants exclusive
rights to the compiler of a collection of copyrighted parts.
Each part is singularly copyrighted by its own author, and the collector
needs to get right to create the collection, but then HE is the copyright
owner for the "collected opera". The law says that the act of granting right
to include a copyrighted artifact into another one, transfers the copyright
to the owner of the resulting opera, but limited to the use of the part
*inside* the whole, and not elsewhere.
Although I have not checked here in Finland, I suppose the finnish law has
something similar (and some ystäviä can check more easily than me).
If so, then the act of giving a patch to Linus for inclusion in the Kernel
would transfer the copyright (for the only purpose of inclusion in the
kernel) to Linus, who is the only owner of the copyright on the Kernel (a
product created In Finland from a Finnish citizen is subject to Finnish
copyright law even outside Finland, if that law doesn't contradict local
laws).
Something similar exists also in German Law.
Therefore Linus could be free to release a new Kernel with a different
License, unless the contributor hasn't esplicitly asked Linus to use GPL for
the resulting product, or, similaryly, unless Linus has esplicitly promised
to release the Kernel under GPL even in the future.
Both these things happened really, so if Linus would release a new Kernel
under a diferent license, this would break the "contract" under which the
contributors has contributed, and then Linus would need a *new* grant from
*each* of the contributors.
It is not difficult to release a program under restricted copyright even if
it was GPL, if you are the copyright owner, but it is also very easy for a
contributor to ensure that its work would be included *olny* in a GPLed
product.
Going back to the libdb2 code (which was the original subject) I have
noticed that Sleepycats don't let patches go into the code. As a (good)
patch arrives, they refuse it, declaring that the idea was good, but they
have *rewritten* differently the fix (and they really do it).
This behaviour circumvents the "contribution problem".
anyway, this affects only *new* release of the code, as a "legally released"
code under GPL can be always redistributed under the GPL.
fab
--
Reply to: