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

Re: GPL "+" question

On Fri, May 29, 2015 at 09:32:12AM +0200, Ole Streicher wrote:
> Hi,
> I just had a discussion with an ftp-master who rejected one of my
> packages. The package in question is "missfits". It contains a
> directory, src/wcs/ with files that were originally released by Mark
> Calabretta under LGPL-2+, but changed by the upstream author (Emmanuel
> Bertin) and released in the package under GPL-3+.

Upstream authors can't change licensing of any files, under any
conditions, ever.

If I say a file is GPLv2+, it is forever GPLv2+, even if it's combined
with a GPLv3 work, in that case the *files* are still GPLv2+, that other
file is a GPLv3 work, and the *combined work* is distributed under the
terms of the GPLv3, since it satisfies the license of every file in the
combined / derived work.

> debian/copyright currently mentions only GPL-3+ for the whole package.

Yeah, debian/copyright isn't what the binary is distributed under, it's
what the source licenses are.

If it had MIT/Expat code, you'd still need it in debian/copyright if the
other files are GPLv2+

> The ftp-master now asked me to add GPL-2+ for these files to
> debian/copyright. But I think that this would be wrong, since the files
> under src/wcs are not distributable under GPL-2+ (because they contain
> GPL-3+ code from Emmanuel Bertin).

Nah, it's wrong because you said LGPLv2+, adding it sounds right.

Just because files are being combined in such a way that they're
distributed under different terms than some of the files doesn't mean we
exclude them.

Just like Expat is contained within BSD-3. Or ISC is contained in Expat.
You still need all three, since that's the licese for the file.

Only the copyright holder can change what a *work* is licensed as.
Anyone can distribute a derived work inline witht he terms of their
license. That may also contains other terms as well.

> Do I miss an important point here?
> Best regards
> Ole



Attachment: signature.asc
Description: Digital signature

Reply to: