Re: Taking a position on anti-patent licenses (was ' Re: Bug#289856: mdnsresponder: Wrong license')
I'm sorry that Nick feels misunderstood. The point I was trying to
make was that the proposition as written was far too broad and agreeing
with it probably means agreeing with popular bogeymen like the "pet a cat"
> So the question I was trying to ask was "do we believe that there are
> *any* restrictions which would be acceptable?" -- with the intention that
> if/when the answer turned out to be "yes", then we could discuss precisely
> which restrictions they would be.
Yes, there are at least some restrictions which are acceptable. Those
are ones which do not contradict the DFSG.
I think a key concept is that patents and copyright are orthogonal and
should not be mixed together. It's fine for the copyright permission to
require fulfilling some restrictions that fit the DFSG. It's also fine
for the patent licence to do that. I think it's bad for the copyright
permission to add conditions to the patent permissions, and it runs a big
risk of making software patents affect swpat-free places but I'm not 100%
clear how to relate that to freedom directly and in general.
Even so, I suggest that general anti-patent patent licences do not follow
the DFSG. Compare it with a copyright-based situation where I give you
many pieces of software under DFSG-compatible copyright licences; then
we get in some litigation over a particular piece of licensed software
with claims going both ways. Can the other licences be free *and*
terminate when you claim against me? (I think it's clear that the
disputed licence can terminate if you breach it (like the GNU GPL), but
I don't think even that can for just a counter-claim. Less sure, though.)