incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mattmann, Chris A (3980)" <>
Subject Re: Podling request: Gerrit
Date Thu, 16 Jul 2015 00:37:33 GMT
Hi Till,

We should probably move this discussion on to the list.

In short, we shouldn’t have situations in which there are contributors
who contributions are “shepherded in” by Apache AsterixDB Incubating
PPMC members whose contributions have an indirect middle man at
UCI. All development on ASF projects must happen at the ASF.

We went to great lengths to get the Github workflow integrated into
our mailing lists for provenance and for foundational tracking
perspective and ultimately so that we can tell people who use Apache
software that it’s from a plan and provenance they can trust. Infra
did a lot of work to make sure contributions have at least an email
address that flows through to the mailing list.

Here in the Gerrit situation, it could be similar to Github I suppose
if we make sure all communication from that Gerrit instance is
mirror’ed to the list (dev@, or some similar list, probably issues@,
or something that folks can choose to subscribe to).

Ideally we need ICLAs on file for anything bigger than smallish
contributions that have clear mailing list provenance. So, one thing
you guys are doing is potentially circumventing that review from
an ASF perspective without this mailing list mirror’ing at the very

If ASF infra is willing to throw up a Gerrit instance, that’s the
most ideal situation. If they are not there is precedence for what
you guys are doing ( e.g., with Github; and also with build farm
machines, e.g., such as those contributed by Y! initially when
Hadoop started, etc.) But this is our core product; code, and it’s
not something to be taken lightly especially in light of things
“not happening at the ASF” and for provenance purposes. It’s nice
that this is going on at UCI, and we appreciate their use of
resources, however, ASF projects develop and “occur” at the ASF.

Here are the immediate actions:

1. Mirror all UCI Gerrit to Apache AsterixDB mailing list (discussed
in dev@asterixdb.i.a.o and agreed upon by the PPMC within 24-48
hours) 2. Work with ASF infra (David Nalley is the VP of infra, so
you have his attention here) to see if they are willing to run a
Gerrit instance. It’s my understanding David, that the AsterixDB
folks have a few lingering issues here where they have not heard
back so if you could reply on those I’d appreciate it.  3. Contributions
from non AsterixDB PPMC members need to be recognized as such as
we should be looking to have a discussion about who should be added
to the PPMC based on the work that’s been going on.

OK, that sound like a plan? This discussion should move to
dev@asterixdb.i.a.o if there is nothing more here. This is a community
and teaching issue that doesn’t need to be on general@.


Chris Mattmann, Ph.D.
Chief Architect
Instrument Software and Science Data Systems Section (398)
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 168-519, Mailstop: 168-527
Adjunct Associate Professor, Computer Science Department
University of Southern California, Los Angeles, CA 90089 USA

-----Original Message-----
From: Till Westmann <>
Reply-To: "" <>
Date: Wednesday, July 15, 2015 at 4:18 PM
To: "" <>
Subject: Re: Podling request: Gerrit

>> On Jul 16, 2015, at 12:25 AM, David Nalley <> wrote:
>> On Wed, Jul 15, 2015 at 5:48 PM, Mattmann, Chris A (3980)
>> <> wrote:
>>> Hi Folks,
>>> Can someone clarify in simple terms what the issue is here?
>> There's a few issues Chris:
>Let me try to describe this in terms how most members of the AsterixDB
>community probably see it.
>> 1. Contributions are being submitted, discussed, and accepted
>> externally. No record of the submission, discussion, or acceptance is
>> currently maintained at the ASF.
>AsterixDB uses a Gerrit instance hosted at UCI as a code review tool
>before submitting to the master branch.
>Discussions on modifications are indeed happening in Gerrit and are
>currently not forwarded to the ASF mailing lists, but forwarding those
>discussions should be possible.
>After discussion, review, and acceptance in the review tool, an AsterixDB
>committer manually commits the reviewed modification to the master branch
>in the ASF repository.
>If the original author of the modification of the code was an AsterixDB
>committer, the commit should be done by the original author.
>If the original author was another contributor, the commit should be done
>by the AsterixDB committer who reviewed and validated the modification.
>> 2. As in 1) contributions are being accepted externally, and then
>> synced, to the ASF repo, essentially making it the mirror, rather than
>> the required canonical copy.
>Contributions are accepted by an AsterixDB committer on a tool that is
>not hosted by the ASF.
>It is not clear why that makes the acceptance external to the project or
>the ASF.
>After acceptance, an AsterixDB committer commits the modifications to the
>ASF repo.
>The master branch of the ASF repository is considered to be the source of
>truth and the basis for releases.
>It is not obvious, why the fact that the modifications were reviewed in
>Gerrit before being committed to the ASF repo would make the ASF repo a
>“non-canonical” copy.
>To unsubscribe, e-mail:
>For additional commands, e-mail:

To unsubscribe, e-mail:
For additional commands, e-mail:
View raw message