www-legal-discuss mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen Connolly <stephen.alan.conno...@gmail.com>
Subject Re: Understanding non-release licensing at the ASF
Date Thu, 27 Aug 2015 16:48:10 GMT
On Thursday, August 27, 2015, William A Rowe Jr <wrowe@rowe-clan.net> wrote:

> Thanks for taking the time to shed light on this discussion on both
> posts.  Clarification requested...
> On Aug 25, 2015 9:28 PM, "Roy T. Fielding" <fielding@gbiv.com
> <javascript:_e(%7B%7D,'cvml','fielding@gbiv.com');>> wrote:
> >
> > On Aug 25, 2015, at 2:08 PM, Roman Shaposhnik <roman@shaposhnik.org
> <javascript:_e(%7B%7D,'cvml','roman@shaposhnik.org');>> wrote:
> > >
> > > I've asked this question on the thread where the discussion
> originated, but I'd
> > > like to repeat it here to add to the list of points that I would like
> > > to see explicitly
> > > validated:
> > >
> > > 4. Because of language in ICLA, the "Grant of Copyright License" is
> specifically
> > > targeting "the Foundation and [to] recipients of software distributed
> > > by the Foundation"
> > > and as such does NOT guarantee that contributions NOT included into a
> formal
> > > ASF release (since only releases can be considered distributed) are
> > > licensed under
> > > the ALv2.
> >
> > No, anything we distribute has been distributed, and this license applies
> > to all recipients of all software we distribute (including software that
> > doesn't contain that contribution).
> >
> > > IOW, if an entity outside of ASF takes non-released contributions said
> entity
> > > needs to obtain a separate license Grant of Copyright License from
> original
> > > contributors OR wait for ASF to produce a release.
> >
> > Absolutely not.
> But the operative function here is a license to our Copyright.  If the ASF
> has not published, and therefore not copyrighted the code, the license has
> simply been granted by another author.
> Going to Jim's point, when does GPLv2 code granted to the ASF to relicense
> get 'published' as an ASF work?  Not if committed verbatim with GPL
> LICENCE.  So when?  The moment an ASF committer changes the
> LICENSE/NOTICE/individual copyright headers?  When it is assembled as a
> snapshot/nightly build?  When it is advanced as a vote to release?  Or when
> the code is approved by the PMC as a release?
It may make a difference if the project is CTR or RTC

With RTC the commits have been "reviewed" by the PMC who are supposed to do
due diligence on licensing and other concerns.

With CTR the recent commits (less than 72h old) have not completed review
by the PMC.

On the other hand, officially the PMC can be a bit "loosy-goosey" until
time comes for a release as long as at release time the PMC reviews the
code changes for provenance, etc.

In practice, I have never seen a PMC perform such queries at release
time... So I can only conclude that *either* the PMCs are not doing such a
review *or* all the code is always perfect.

In fact, I do not recall ever seeing a single commit's provenance
questioned (outside of incubation)...

Oh well, I'll return to my "committers only commit stuff they are entitled
to commit and committing signifies intent to license" ivory tower of
ignorance ;-)

On a side note, if we are saying that code in source control is not a
release, then that would neatly resolve the "should svn/git have LICENSE
and NOTICE files at each release root" argument with a "no" answer that
was/is unanswered in the LEGAL JIRA for some time

Sent from my phone

View raw message