Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Hive >> mail # dev >> [DISCUSS] HCatalog becoming a subproject of Hive

Copy link to this message
Re: [DISCUSS] HCatalog becoming a subproject of Hive
So what is the next step here?  As far as I can tell all agree that moving HCat into Hive is the right thing to do, but there is a fairly polarized division on how to handle the commit rights.  Given that I re-propose my compromise.  Like all good compromises most everyone hates it.  But I think it gives us a path forward to what we agree is a common goal.  

Here is the slightly modified version of my proposed compromise, which makes clear that promotion from HCat committer to Hive committer is not automatic.  

All active HCat committers (those who have contributed or committed a patch in the last 6 months) will be made committers in the HCat portion only of Hive.  In addition those committers will be assigned a particular shepherd who is a current Hive committer and who will be responsible for mentoring them towards full Hive committership.  As a part of this mentorship the HCat committer will review patches of other contributors, contribute patches to Hive (both inside and outside of HCatalog), respond to user issues on the mailing lists, etc.  It is intended that as a result of this mentorship program HCat committers can become full Hive committers in 6-9 months.  Becoming a full Hive committer in this time frame is not guaranteed.  It will require the vote of the Hive PMC to move the HCat committer to a full Hive committer.  No new HCat only committers will be elected in Hive after this.  All Hive committers will automatically also have commit rights on HCatalog.

If anyone has additional proposed modifications please make them.  Otherwise I propose that the Hive PMC vote on this proposal.

On Dec 20, 2012, at 3:53 PM, Ashish Thusoo wrote:

> hmm... why is this considered "preferential treatment"?
> All the work for HCat is in the public domain so we can really evaluate whether they have been following apache practices - the fact that they are graduating from the incubator would seem to indicate that they have been doing so. If this code base is contributed back to Hive, is that not counted as a significant contribution to Hive? I am failing to understand on what count they don't qualify to be committers.
> Plus if it is too onerous to enforce committer privileges on selective parts (is there a way?) of the project, then what do terms like Hive committer, HCat committer mean? Also should Hive committers have privileges to commit into HCat part of the code once it becomes a subproject. I think we are just creating walls and the problem with walls is that they just impede cross pollination and community expansion.
> Ashish
> On Thu, Dec 20, 2012 at 1:59 PM, Carl Steinbach <[EMAIL PROTECTED]> wrote:
> I agree with Namit on this issue. I don't think it's fair to the
> existing group of Hive contributors to give preferential
> treatment to HCat committers, or to automatically promote them to
> full committer status on the Hive project.
> On Thu, Dec 20, 2012 at 1:10 PM, Bhandarkar, Milind <
> > I agree with Ashish.
> >
> > When Hcat becomes a subproject of Hive, all Hcat committers should
> > immediately become Hive committers.
> >
> > After all, that worked well for Hadoop, where all Hadoop committers can
> > commit to all Hadoop code (common/HDFS/MapReduce), but not all do, instead
> > focusing only on their area of expertise, and familiarity with portions of
> > codebase.
> >
> > - milind
> >
> > ---
> > Milind Bhandarkar
> > Chief Scientist,
> > Machine Learning Platforms,
> > Greenplum, A Division of EMC
> > +1-650-523-3858 (W)
> > +1-408-666-8483 (C)
> >
> >
> >
> >
> >
> > On 12/20/12 5:58 AM, "Ashish Thusoo" <[EMAIL PROTECTED]> wrote:
> >
> > >Actually I don't understand why getting Hcat folks as committers on Hive
> > >is
> > >a problem. Hive itself became a subproject of Hadoop when it started with
> > >all the Hive committers becoming Hadoop committers. And of course everyone
> > >maintained the discipline that they commit in parts of the code that they