Assistance/advice in choosing a license for POV-Ray 4.0

Ian Jackson ijackson at chiark.greenend.org.uk
Tue Nov 15 22:39:30 UTC 2005


Chris Cason writes ("Re: Assistance/advice in choosing a license for POV-Ray 4.0"):
> Actually I did address this (at least partially), though not in my
> original message (which was already too long). Quoting from a later message:
> 
> > One of the issues we have with the GPL and LGPL is that it doesn't really
> > attack the patent misuse issue in the way that e.g. the CDDL does. I have
> > strong feelings about how trivial patents are being used to wreck the
> > industry for small developers and frankly I'm more in line with the MPL's
> > 8.2(b) rather than the CDDL when it comes to this subject (though apart
> > from that I prefer the CDDL over the MPL).
> 
> That alone is sufficient to disqualify the GPL from contention as far as
> we are concerned.

I'm quite in agreement with you about patents being a bigd problem -
and it is clear that the FSF (licence stewards for the GPL) are too.
The GPL does defend against patents affecting the GPL'd software (such
patents must, effectively, be licenced at least for all GPL'd
programs).

But, you are right that it doesn't have a patent death-trigger clause
along the lines of MPL 8.2(b).  AIUI the main difficulty with this has
been to come up with a clear way of distinguishing software patents
from other patents, or software patent actions from other patent
actions.

Many of the attempts at patent death-trigger clauses have proved
controversial - at least in large part because of difficulties in
drafting and of course difficulties in agreeing on the correct scope.
The FSF tell us that they hope to have an answer this problem in
GPLv3, which will of course be some time coming.  (And some will no
doubt disagree with the FSF's answer, in any case.)

The reason I'm worried about your choosing a strong-copyleft licence
that is not the GPL is that it will make your code licence-
incompatible and unmixable with the vast body of existing (and future)
GPL'd software.  This is, I'm sure you'll agree, undesirable.

Have you considered options like
  ... POV-Ray Public Licence 2.0, or any later version of the
  POV-Ray Public Licence published by Trusted Body ...
?  That would mean (for example) that if GPLv3 met your patent needs
you would be able to relicence your code in a GPL-compatible way
without having to contact long-gone contributors or rewrite parts of
the program.  (Your Trusted Body could republish the GPLv3 as the
POV-Ray Public Licence v3.0 or whatever.)

You would have to think about who would be the Trusted Body.  A
committee of your major contributors, perhaps, or some existing
entity.[1]

(As an aside, it's not clear to me that 8.2(b) helps the putative
defendant Participant all that much because the patentholder can
usually just sue the Participant's customers instead, which doesn't
appear to trigger 8.2 except for customers who are themselves
Participants.)

Ian.

[1]  I hesitate to suggest SPI but it might at least be a possibility
depending on what you'd want out of them.  The most relevant SPI
policies about this kind of thing are:

 `Associated Project Framework'
  http://www.spi-inc.org/corporate/resolutions/resolution-2004-08-10.iwj.1
 `Position and Promises about Intellectual Property'
  http://www.spi-inc.org/corporate/resolutions/resolution-1998-11-16.iwj.2

You will see that I'm not what you would describe as neutral as
regards SPI - but even so I don't yet have an opinion on whether an
association with SPI would be right for you.  You will also see that
SPI's current policy is fairly pro-GPL, although in practice this is
usually irrelevant since few people assign copyrights to SPI.



More information about the License-discuss mailing list