On Sun, Feb 10, 2013 at 02:43:32PM +0200, Engel Nyst wrote:
Hello copyleft-next,
I'd submit for discussion the (re)introduction of an inbound-outbound
clause for contributions, and the possibility to use it as the place where
the upstream licensor makes an explicit promise: to keep the software
under copyleft.
A clause in the style of Apache License 2.0 section 5[1] has been in
the text of copyleft-next at some point[2], however it has been removed
later. I'd suggest to re-add it (and thereby make CLAs more clearly
unnecessary), and also, to consider addition of a second part.
Part #1:
"Unless You explicitly state otherwise when making it, any
Contribution intentionally submitted for inclusion in a Covered Work
by You to Us shall be under the terms and conditions of this License,
without any additional terms or conditions."
Part #2:
"By accepting the Contribution, We promise to only distribute it
under copyleft-next, or future versions."
The second part (which is not in the Apache License provision) seems
like overkill -- isn't it already a given that if you're licensing a
contribution under license X upstream, upstream has to use that
license?
Or,
"By accepting your Contribution, We acknowledge your intention
to be distributed by Us under copyleft-next only, unless You make
a public express statement of your agreement to different terms."
This gets closer to the problem. Typical contributions to projects
have no explicit licensing information.
Stylistically, I strongly dislike 'acknowledgment' clauses as well as
references to intent.
Or,
"By accepting your Contribution, We acknowledge your intention
that We distribute it solely under copyleft-next or future versions,
and We promise to offer any work we prepare based on it under
these terms (or of a future version)."
I assume these attempts are quite unworkable (please shut them
down), but consider the point itself.
I continue to think the *idea* is a good one.
As far as I can tell, this will not, and can not, make the licensor
legally bound, because Licensor (in AL 2.0 sense) can always
receive a different license, like the CLA grant. The CLA, IMO, will
apply unrestricted anyway. IANAL though.
You can't stop projects and contributors from using CLAs if they want
to. The nice thing about this kind of provision is that it provides an
additional argument for why CLAs are unnecessary.
It is also interesting that it makes the Apache License 2.0 a copyleft
license in a very limited sense.
OTOH, this opinion is not widely shared[3]: it seems the respective
AL 2.0 paragraph can be read as precisely making room for Apache
itself to not follow the conditions of the AL 2.0 grant, but instead
license on the basis of the CLA grant. Please see the thread: the
potential that AL 2.0 submissions clause *could limit* the CLA
holding entity in some way is quite interesting.
I read the thread (I lurk on the public Apache legal list).
But is legally binding of the promise the only issue worth pursuing?
The expression of intention of contributors, or the creation of an
expectation from the community in regard to the behavior of a
CLA-holding entity, could have more positive effects. Added up
to part #1.
Some of the issues with CLAs are that they might be signed with
many unwritten expectations, and they may grant much more rights,
than they were originally intended for. A clause like this is a place to
have those expectations written.
Other issues with CLAs are simply that they're separate documents,
which give licensing rights, are deemed necessary for some reason,
and are written as an agreement (not as a license) typically in
legalese.
Personally, I see a concerning potential for proprietary relicensing
to affect community contributions which were not intended to be
proprietary by contributors. (as far as I understand, Richard Fontana
disagrees with this. I admit I am not objective on the matter.)
I'm not sure I disagree with that, but maybe I misunderstand what you
mean (at least now, if your statement is based on something I said
earlier).
I have a different idea for how to deal with this, but I'll post it in
a separate response.
- RF