Grant Ingersoll wrote:
On Sep 15, 2008, at 3:19 PM, Bernd Fondermann wrote:
Ryan McKinley wrote:
= Droids planning to move out =
Lab Droids (est. 2007-02, PI: Thorsten Scherler) is considering moving
on to the Incubator. Currently, the lab is preparing incubation,
including drafting a proposal at the Incubator Wiki and looking for a
champion and mentors. We are very excited about that, since
providing an
ecosystem for projects on the way from the first line of code until
incubation is one of the goals of the Labs project, and Droids - if
successful - would be the first lab to follow this path ultimately.
Some
discussion came up, here and on the Incubator's general list, about the
process how to proceed with labs aiming to become a project's
subproject, instead of going TLP. According to our bylaws, going
through
the Incubator is inevitable. And the Incubator surely is the right
place
to determine how to properly deal with that.
I'm reading over the previous thread and the bylaws, and trying to
get my head around the most effective next steps for Droids.
The key messages are:
http://www.nabble.com/Re%3A-To-incubate-or-start-subproject-directly--(was-Re%3A-Looking-for-a%09Champion)-p19396492.html http://www.nabble.com/Re%3A-To-incubate-or-start-subproject-directly--(was-Re%3A-Looking-for-a%09Champion)-p19440800.html http://labs.apache.org/bylaws.html
Assuming an existing PMC agrees they want labs code to start a
subproject, that PMC is responsible to make sure the code is kosher.
The responsibility is with the receiving PMC, *not* Labs. As far as
the Labs bylaws, can't we just change the droids status to "Completed"?
You can only do that by _Labs_ PMC vote. And my guess would be that
the PMC would /promote/ (to the Incubator) rather than /complete/ the
labling.
Consider the example where solr may want to add JSON parsing. The
path to include noggit in a solr release surely does not require
going through incubation!
I don't think so. The lab was conceived with the clear intention that
it should not be a way around the Incubator.
But on the other hand, as far as Droids is concerned, from what I see
I get the impression that the labling is preparing a proper Incubation
anywa. And you never know what might come out of that - some projects
entered incubation to become a subproject and went out as TLP :-)
I'm really impressed with the community support Droids is attracting
right now.
By the same token, it's ASF licensed. Any PMC can take it and include
it and is the responsibility of the receiving PMC to do it's due
diligence on, AIUI the license. See Roy's message on the "Re: To
incubate or start subproject directly? (was Re: Looking for a Champion)"
In general, I agree. In practice, I never witnessed moving code
ownership from one PMC to another one, except from Incubator to Target
PMC, which requires a IPMC vote.
I only vaguely remember the founding, but I don't think anyone is
getting around Incubation, I just don't think it is clear that it needs
to go through incubation to begin with.
However, if the receiving PMC doesn't want to do incubation b/c it feels
there will be enough a community in a decent amount of time, then why
does it matter? Isn't it the receiving PMC's responsibility? Why
should Labs care what happens to a project that has left labs for
another place within the ASF?
I don't have a problem to let Droids go, because it is already
outgrowing labs. But Droids deserves a proper growing up phase and maybe
to become a TLP. This can all be done in a non-Incubator PMC, too. I
just doubt it will ever work that well.
I see three potentially receiving PMCs currently, HC, Lucene and
Incubator. Maybe it does not need to go through incubation, but what's
so bad about it?
I'd like to here what Thorsten thinks is his preferred direction right
now. He was starting out with HC as I-champion, then Lucene showed up,
while he is still actively preparing incubation.
I fail to see how the situation is any
different from the case where Thorsten had gone directly to a PMC and
said "I'd like to start a subproject named Droids" and that PMC either
votes for it or not.
But he didn't. IIUC, partly this didn't happen because he felt that
Droids does not fit well into existing projects. It's a general purpose
crawling framework.
Asking more provocantly, why should this be burried in some vaguely
fitting TLP?
I think Droids should at least try for TLP by going through incubation.
I think the trick becomes that we don't want there to be two projects
that are Droids (even if there named differently), one in Labs and one
in the receiving PMC, even though this seems perfectly legal under the
ASF license. Thus, I don't see why the Labs project can't just be
marked as "completed" and then refers to the new subproject.
Maybe it can, though I'd prefer 'promoted'. But let's think about what
is the best for the Lab in question, not the fastest or most convenient.
Bernd
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
