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

Switch to Threaded View
MapReduce >> mail # dev >> [VOTE] Release Apache Hadoop 2.0.4.1-alpha


Copy link to this message
-
Re: [VOTE] Release Apache Hadoop 2.0.4.1-alpha
On Thu, May 30, 2013 at 04:08PM, Jean-Daniel Cryans wrote:
> FWIW, not that I have a dog in this fight, but the only release with a
> 4th number (not including .0 like the 0.20.20x releases did) we had
> was:
>
> http://hadoop.6.n7.nabble.com/VOTE-Release-0-17-2-1-rc-0-td13398.html
>
> 0.17.2 was missing some native libs so 0.17.2.1 was released to fix
> that critical issue instead of calling it .3

Exactly the point - the _bigfix_ release. Thanks for pointing out the
similarities.

Cos

>
> J-D
>
> On Thu, May 30, 2013 at 3:52 PM, Konstantin Boudnik <[EMAIL PROTECTED]> wrote:
> > On Thu, May 30, 2013 at 03:11PM, Matt Foley wrote:
> >> Hi Cos,
> >> I would also request that you renumber the release candidate to just
> >> three-numbers, hence "2.0.5-alpha".
> >>
> >> Arun, are you willing to start the 2.1.x name-space for your next release,
> >> so that 2.0.x-alpha can become an intermediate stabilization branch as Cos
> >> and Konst want?
> >
> > Let's get the facts straight, Matt, please: this "want" has been expressed in
> > the official vote here http://s.apache.org/ZMf Apparently, 2.0.5-alpha is
> > blocked now because of the another vote that hasn't been closed yet for
> > whatever reason. In order to unblock a number of releases in downstream
> > component I have moved forward with this release. Do you have any material
> > objections to the release that pursue this goal?
> >
> >> I just think that using four-number schemes was symptomatic of the
> >> near-forking we had back in the 0.20.xxx.y days, and I really don't want to
> >> go back there.  Especially since you could say that "0.20.xxx.y" is just
> >> three significant numbers, the leading zero being inconsequential.
> >
> > I dare to remind that forth part of the version is reserved - not in a
> > parallel universe, of course - for "patch level" aka bug fixes. It hardly can
> > be taken a sign of 'forking' by any definition.
> >
> > Cos
> >
> >> So, would you please consider using 2.0.5-alpha?
> >>
> >> As for the "2.0.5-SNAPSHOT" in the branch-2 versioning, that's standard
> >> usage.  Whoever makes the 2.0.5 release (or any "next" release) is expected
> >> to update the parent branch's SNAPSHOT default versioning, per
> >> HowToReleasePostMavenization#Branching<https://wiki.apache.org/hadoop/HowToReleasePostMavenization#Branching>,
> >> step 6.
> >>
> >> Thanks,
> >> --Matt
> >>
> >>
> >> On Thu, May 30, 2013 at 11:52 AM, Konstantin Boudnik <[EMAIL PROTECTED]> wrote:
> >>
> >> > On Thu, May 30, 2013 at 10:57AM, Arun C Murthy wrote:
> >> > > I see you just re-opened MAPREUDCE-5211.
> >> > >
> >> > > Why not include MAPREDUCE-5211 as well rather than create one release
> >> > per patch?
> >> >
> >> > Arun, it is unclear if MAPREDUCE-5211 has implications in 2.0.4 as per
> >> >
> >> > https://issues.apache.org/jira/browse/MAPREDUCE-5211?focusedCommentId=13670574&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13670574
> >> >
> >> > Hence, there's a good chance that it never will be backported. And I don't
> >> > have any plans to created 'a release per patch'.
> >> >
> >> > > Also, this is the first time we are seeing a four-numbered scheme in
> >> > Hadoop.
> >> > > Why not call this 2.0.5-alpha?
> >> >
> >> > There were precedents in four-numbered schemes before: 0.20.20[3-5].0
> >> > comes to
> >> > mind.
> >> >
> >> > As for 2.0.5-alpha: The release numbering games and votes that had
> >> > happened in
> >> > the last few weeks are very confusing. Some of them never been concluded,
> >> > the
> >> > branches are moved and artifact versions seem to be colliding. 2.0.4.x
> >> > seems
> >> > to work well for the stabilization purposes and it will allow to unblock
> >> > downstream and integration projects quickly.
> >> >
> >> > Cos
> >> >
> >> > > Arun
> >> > >
> >> > > On May 24, 2013, at 8:48 PM, Konstantin Boudnik wrote:
> >> > >
> >> > > > All,
> >> > > >
> >> > > > I have created a release candidate (rc0) for hadoop-2.0.4.1-alpha that