Re: Libpropc++

["Marco Scholten" <mscholtn xs4all nl>]

From: "Roel Vanhout" <roel riks nl>
To: "Ole Laursen" <olau hardworking dk>
Cc: <gtkmm-list gnome org>
Sent: Friday, August 19, 2005 3:16 PM
Subject: Re: Libpropc++


> >  > Since template code cannot be 'separated' from the rest of a program
> >  > into a shared library, all the files that depend on a template library
> >  > (such as, in fact, libsigc++) will have to be open-sourced in order to
> >  > comply with the terms of the LGPL. This does indeed implicate that 
> > you'd
> >  > have to open-source at least a part of your program in order to be 
> > able
> >  > to use libsigc++.
> > But this is your interpretation. Clearly, the authors of gtkmm and
> > libsigc++ didn't intend this interpretation. Especially for libsigc++
> > it simply does not make any sense to license it under LGPL if the
> > template parts weren't covered by the same pattern of use - i.e. as
> > long as you are just using the library, there are no restrictions on
> > your license as long as people can get to the source of the library
> > itself (and do the relinking stuff).
>
> Well this is where the fun begins. First, "the authors of gtkmm and 
> libsigc++ didn't intend this interpretation" is completely irrelevant. Not 
> the 'intent' of the authors is what gives the license its validity, a 
> reasonable interpretation of the wording of it does. So then we get to the 
> 'reasonable' part, and yes, in this case it could be argued both ways. I 
> still maintain that the case of gtkmm/libsigc++, a strict interpretation 
> of the LGPL is in order, which would not allow people to use it in 
> closed-source software.

With your interpretation, what is the difference between LGPL and GPL
in the case of gtkmm?