Re: LGPL code using GPL code

[Leonard den Ottolander <leonard den ottolander nl>]

Hello Jindrich,

On Wed, 2004-10-20 at 10:24, Jindrich Novy wrote:
> Arjan van de Ven:
> it's explicitly allowed, but it means the LGPL code is used as GPL code
> at that point, eg basically the "L" goes away... ;)
> 
> Roland McGrath:
> There is no problem unless you are producing a library containing some
> GPL source code and then distributing a library binary and claiming that
> it is under the LGPL.
> 
> So in my opinion there's not a big problem in this since we are not
> distributing any mc library claiming it is LGPL.

Just what I thought, but this makes the vfs parts undistributable as a
separate entity under the LGPL as they use GPL code.

> For the license
> conversion to GPL please note, that you need all the authors of mc to
> agree with it.

No! I said this before. The LGPL specifically allows redistribution
under the GPL. See paragraph 3 that I quoted in my mail. E.g. you can
make the license stricter, but people can still use the "original" code
as LGPL.

So there are 4 possibilities (3 1/2 actually):
1) We ask the authors of the GPLed functions if we can redistribute the
involved functions in src as LGPL. This will make the vfs
redistributable as a separate entity under the LGPL.
2) We rewrite all "offending" functions. I don't think this is a very
good idea. I think Roland is overestimating the use of inappropriate
functions in the vfs code. Sure, some can go, but the bulk can not.
3) We brand all LGPLed vfs code as GPL, because vfs isn't
redistributable as LGPL in it's current implementation anyway.
4) We leave the licenses as are, and put in the README that vfs although
LGPLed is not redistributable as a separate entity under the LGPL in
it's current implementation because it uses GPLed code.

Leonard.

-- 
mount -t life -o ro /dev/dna /genetic/research