incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alan Gates <>
Subject Re: [DISCUSS] [VOTE] HCatalog to Graduate and become part of Apache Hive
Date Thu, 14 Feb 2013 16:31:49 GMT
I'd like second and extend Benson's point about clarifying how these things should work.  In
addition to clarifying what it means to graduate into a subproject now that that is frowned
upon, clarifying how these votes work would help.  I think Chris felt that we ignored his
vote and pushed ahead.  From my reading of the docs it was supposed to be a majority vote
and thus to view the -1s as a veto would be to improperly ignore the 5 +1s.  If the rules
were clear in advance for the next group that faces this situation it will help to avoid these
misunderstandings and frustrations.


On Feb 14, 2013, at 3:29 AM, Benson Margulies wrote:

> I'm not so much opinionated as confused here, perhaps because I have a
> very linear view of governance.
> I like to know how a vote fits into a governance structure or process,
> and I've felt for some time that this case (podling goes to existing
> TLP) is not well-explained by our structure.
> Back in the days when subprojects were normal and valid, the incubator
> had a role on behalf of' an existing TLP in supervising IP and
> community behavior. Graduation meant: "OK, umbrella, we certify that
> these people can behave like a project and have clean IP." And,
> perhaps, the board actually established subprojects? It's all before
> my time.
> Now that subprojects are no longer in the picture, I don't even know
> why the IPMC should ever incubate a podling *if the plan, from the
> start, is to be part of some existing TLP.* So I have assumed that
> HCatalog started out with the intention to grow into an entire TLP,
> and came up with the Hive plan as a fallback.
> To try to make this long story shorter, I think that we should make a
> proposal to the board with a schema for handling this case that makes
> sense in current conditions. I'm happy for it to be your schema, which
> amounts, as I see it, to the board having a supervisory moment when
> this happens, with an IPMC vote providing the same sort of strong
> recommendation one way or the other that it does for establishing a
> TLP.
> On Thu, Feb 14, 2013 at 12:49 AM, Mattmann, Chris A (388J)
> <> wrote:
>> Hi Benson,
>> I saw your later email(s) and Incubator board report. It's fine and I
>> think the message of my objection comes across.
>> So thanks for that.
>> One thing I wanted to comment on:
>> On 2/13/13 4:10 AM, "Benson Margulies" <> wrote:
>>> Chris,
>>> The obvious compromise is to ask them to report the vote result as it
>>> happened, it seems to me, -1's and all. But where do you think that
>>> they are reporting anything? There's nothing happening here at the
>>> board level. There's no board resolution needed for a Hive committer
>>> to type 'svn cp' on the hcatalog tree,
>> Not by my counts. There's a *community* resolution and a recommendation to
>> be made by the IPMC, nonetheless.
>> Otherwise, the IPMC is pretty useless IMO, and more importantly, so is the
>> Incubator.
>> Why bother even incubating HCatalog? Hive could have simply svn cp'ed
>> whatever code came in, or whatever code the podling arrived at, and
>> Incubation would have stopped then. But we both know that's not the way it
>> works. Even if a podling graduates to an existing TLP, go check out the
>> past resolutions. You'll note there's a section in there that discharges
>> the responsibility of the IPMC for the podling. So, yes, the IPMC *is*
>> involved. And yes, the IPMC vote matters.
>> Cheers,
>> Chris
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> For additional commands, e-mail:
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

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

View raw message