> 1) Do we think there are circumstances under which it is best for
> Mozilla to release software under a "no copyleft" (a.k.a. "permissive")
> license? If so, what are they?

There are two kinds of reasons motivating a copyleft requirement in a software 
release: moral ones, and pragmatic ones.

I'm not particularly motivated by moral arguments, at least when it comes to 
something like open source software.  I am driven by a pragmatic desire to pick 
the license that is most likely to achieve the goals of a particular project.  
(And realize that in an open source community, where contribution is important, 
some licenses attract contributors who _do_ have a moral stance, which is fine).

In the context of an end-user product like Firefox, I always understood the 
copyleft part of the MPL to be about

 - encouraging contributions 
 - making it slightly harder for someone to take Firefox and fork it "for evil".

In the late 90s, the world was different.  In particular, open source was new, 
and licensing as a way to "force" people to contribute seemed reasonable (at 
least to me).

In 2013, I have to say I find the copyleft clause much less interesting or 
valuable, because:

  - open source has evolved, and IMO social pressure is a much more effective 
way to build an actual community of contributors than licensing restrictions
  - we know that bad guys will fork Firefox and turn it into malware, and that 
copyright is a better tool for reigning them in than licensing law.  (I'm not 
talking about open source forks, which are fine)
  - history (e.g. DVCS'es, github) has shown (IMO) that encouraging forking 
ends up resulting in more code being published than focusing on merging.

Furthermore, for a lot of the code we write and publish, there's no "problem" 
with people taking some code and repurposing it in some random other bits of 
software.

So, from a pragmatic POV, I'm less interested in copyleft clauses than I was a 
decade ago. 

In addition, the MPL isn't a "trendy" license -- MIT, BSD, Apache are all more 
widely / better understood.  MPL2 is (IMO) much better, and if we'd had a time 
machine and released MPL2 back in 1999 or whatnot, the world might be different 
today, but pushing MPL2 externally at this point doesn't feel high-value.

Where I end up therefore is that picking a license that minimizes the barrier 
to adoption & contribution is pretty high on my list of criteria.  I have no 
strong opinion w.r.t. MIT vs BSD vs  Apache, as in my experience they're often 
interchangeable in that pursuit.  YMMV, especially if the code has to integrate 
in an existing community.

--da



_______________________________________________
governance mailing list
[email protected]
https://lists.mozilla.org/listinfo/governance

Reply via email to