Re: [DotGNU]Changing pnetlib license to LGPL

[BioChem333]

On Thu, 2002-07-11 at 20:14, Mario D. Santana wrote:
> I think it's a good idea. It's exactly the sort of thing the LGPL was 
> made for. The LGPL has fallen out of favor, I think, because there's 
> enough GPL software that its benefits are outweighed by a moral 
> preference for Free Software.
> 
> But in the context of a hot, emerging arena like webservices, there are
> enough players interested in pursuing non-free solutions that the
> compromise of the LGPL, and the temptation of using already-existing code
> for no cost, could prove to be a powerful force for the standardization
> of low-level bits on Free Software. Any worthwhile proprietary uses of
> pnetlib would shortly be implemented in Free Software, as has happened
> (and is still happening) in other arenas.

I'm not sure you understand the problem. The LGPL has a loophole which
could cause all of pnetlib to be modified and released in object code
only, with only the interface being left free and under the LGPL. This
essentially eliminates all protections provided by the LGPL for the
library. The nature of the system is very different from what the
original LGPL authors could've foreseen it being used for. This can't be
exactly what the LGPL was made for because the LGPL was based on a
totally different type of system; a system on which the LGPL has no
loopholes. As for non-free web-service solutions, that's possible with
either license. The only difference is, with the LGPL you can destroy
the freedom of the supporting library, but with the GPL+linking you
can't.

Rich

(I think this illustrates why it's important that we just fix LGPL, in
order to avoid this kind of misunderstanding.)