incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <>
Subject Re: Q. Forks without concensus?; A. anytime / depends / never without agreement
Date Thu, 12 Jan 2012 23:42:24 GMT
On Jan 11, 2012, at 8:33 PM, Noel J. Bergman wrote:

> Roy T. Fielding wrote:
>> Noel J. Bergman wrote:
>>> The ASF is not about code; it is about community.  If a community forks,
> or otherwise emerges around a codebase, we are not accepting the CODE: we
> are accepting the COMMUNITY.
>> One company is not a community.
> As you've otherwise acknowledged, I was talking in the general case, and
> you're addressing a specific instance.
>>> And it seems to me that if we are to say that a COMMUNITZY is not
> permitted
>>> to participate despite use of code that is perfectly proper according to
> the
>>> license, then we are beggaring out own license, the whole point of which
> is
>>> to permit forks, and to prevent a sole copyright holder from assuming
> control
>>> over the community.
>> If there is no community for the original codebase, yes.
> Agreed.
>> If there is a community and that community doesn't want Apache to fork the
> code that they created,
>> then we will not fork that code at Apache.
> Why not, *IF* there is an active second community that wants to fork?
> Again, in the hypothetical, not in the specific, case, which you say is a
> single vendor, not a community.

Because it is rude to do so.

The second community is welcome to fork their own code and contribute that here.
In some cases, an open community will have joint copyright and some of those
folks can split off and contribute the whole here even if the others don't like it.
But that is a rare case, and I'd suggest we would have to look at it carefully
to avoid being used as someone else's pawn.

>> If the original developers of the code do not want their license changed,
> then we
>> will not fork the code at Apache.
> I kind of take that as a given, since how could we fork it if we can't
> relicense it?

We could fork BSD variants without relicensing the files.  By forking them
here, we relicense the end product (our releases).  We choose to do so as
an Apache fork only when it is desired by the current maintainers of that code.
Otherwise, we make it clear that we are only distributing a copy of their code,
under their original license, and place it in a location for third-party
code (srclib) or have the user download it separately at build time.

>> We only accept voluntary contributions
> The presence of a community that wants to work here implies voluntary, and
> not everyone has to agree with the fork.  Don't you remember the origins of
> Apache Felix?

By community, I mean the people who have contributed to the work and thus
have a vested interest in its future.  Such community members are welcome to
contribute here any work products that they have personally developed or own
copyright to, and any code for which they have an expressed permission from
the copyright owner that it is okay to contribute it here.  We also accept,
at face value, files under a compatible license for which the author is no
longer responding to communication, or small subsets of code for which
copying them here has no negative impact on some other community, such as
copying routines out of FreeBSD for use in httpd.

Yes, open source licenses give permission to fork.  We try to do no harm
when we fork.  It's a philosophy thing that is fairly unique to Apache
because we are so community-centric, but it is not difficult to explain
to others and many open source developers just assume it as a given.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message