[cairo] Re: License for cairo changed to LGPL

Mathieu Lacage Mathieu.Lacage at sophia.inria.fr
Mon Aug 9 02:38:59 PDT 2004


* This is a personal point of view which does not reflect anything my
current and previous employers might say/think  *

I would like to add that I do agree with the facts raised by David and
Georges below. 

In my previous job, I used to work on embedded systems where:

1) users cannot reprogram the flash of the device without either
reverse-engineering the hardware (which is illegal is a lot of places)
or desoldering the flash/burning it/resoldering it
2) you cannot use dynamic linking (not available and too hard for us to
implement)
3) releasing object files and making sure no useful symbols and
information leaked from the resulting binaries was felt to be too hard
and too time-consuming. Furthermore, I felt that even though I might be
able to automate this task, our customers (who modify our code and make
the final releases) would need to be educated about this issue and there
is no chance this would have ever happened.

Finally, while a lawyer should have been in the loop, there was none
which left the task of deciding whether or not to use LGPL code to
engineers (I think such issues are too often left to engineers in the
real world): as such, we (actually, I) ended up being extremely
conservative and decided not to use any LGPL code.

regards,
Mathieu


On Mon, 2004-08-09 at 08:08, George McBay wrote:
> > David Turner Wrote,
> >
> > I'll stop the nit-picking here, my point is that:
> >
> > - if you think that non-open source products are 
> > going to use statically-linked LGPL libraries, you >
> probably need a reality check. The technical and 
> > legal hurdles are just too much.
> 
> I'll second that.  As a real world example, I just
> recently released a shareware game that uses Cairo (@
> http://www.mischief.com ).  The game makes extensive
> use of vector graphics, which Cairo renders nicely
> even at lower resolutions and that opens up a lot of
> possibilities for doing things like porting the engine
> to small systems (mobile platforms), where the whole
> dynamic linking issue gets really fuzzy.  Too fuzzy
> for me to touch it with a ten foot pole, so I've been
> specifically avoiding L/GPL licensed code even though
> I do use lots of OSS and have contributed to many OSS
> projects.  And static linking while still abiding by
> the LGPL is just too much of a pain to consider.
> 
> For me, the issue has nothing to do with trying to
> hide any changes I might make to cairo and everything
> to do with the fact that there are some areas where I
> feel that the LGPL just falls apart for real world
> usage.  
> 
> I just caught up on this whole license change debate
> today and it is pretty sad news from my perspective.
> 
> For now my plans are to keep using the older version
> of Cairo that I've already been using (which is
> pre-license-change, so still MIT licensed legally) and
> just incrementally migrate away from Cairo to
> something like Antigrain.  Which is unfortunate
> because I really like/prefer the Cairo API, but I just
> can't live with some of the baggage the LGPL piles on.
> 
> 
> _______________________________________________
> cairo mailing list
> cairo at cairographics.org
> http://cairographics.org/cgi-bin/mailman/listinfo/cairo
-- 
Mathieu Lacage <mathieu.lacage at sophia.inria.fr>



More information about the cairo mailing list