Mailing-List: contact general-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: general@hadoop.apache.org
Received-SPF: pass (nike.apache.org: domain of eli@cloudera.com designates
 209.85.215.176 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <85E3BFC7-25CE-4019-80D2-182D0E1E5813@yahoo-inc.com>
References: <85E3BFC7-25CE-4019-80D2-182D0E1E5813@yahoo-inc.com>
Date: Mon, 17 Sep 2012 12:42:27 -0700
Message-ID: 
 <CAPn_vTu6NoH=2g9EWymq2vjzh4_iB_7Fon58ezPcQpMH79eoZw@mail.gmail.com>
Subject: Re: Spinning out Hadoop Eclipse Tools
From: Eli Collins <eli@cloudera.com>
To: general@hadoop.apache.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Hey Adam,

That would be awesome.   +1

We've discussed moving the various contrib projects out of Hadoop in
the past, and think it make sense for all the reasons you mention.  A
good example of how we've done this  was done previously is MRUnit,
now hosted at http://mrunit.apache.org.

Next steps, if others agree then you can draft a proposal to vote on.
See this thread and jira for an example:
http://mail-archives.apache.org/mod_mbox/hadoop-general/201102.mbox/%3C7740=
5974-6771-4604-926B-976240743F3C@mac.com%3E
https://issues.apache.org/jira/browse/MAPREDUCE-2430

Thanks,
Eli

On Mon, Sep 17, 2012 at 12:34 PM, Adam Berry <amberry@yahoo-inc.com> wrote:
> Hello all,
>
> I'm proposing spinning out the eclipse plugin from the main hadoop source=
 into a separate incubating project. I'm willing to commit to this effort.
>
> With a separate project, that isn't tied (source wise) to the Hadoop sour=
ce we could make dev tools that could support multiple versions of Hadoop f=
rom within a single IDE, which would be very helpful when migrating jobs or=
 dealing with clusters running different versions.
>
> The current tools seem to support the 0.20 line, and help with developing=
 MR jobs, and also have some HDFS tools included. So the early work would b=
e to ensure that these features were supported across the version space of =
Hadoop.
>
> Down the line we could also add support for additional tools such as MRUn=
it and other products in the Hadoop ecosystem, to eventually yield a one st=
op project for all dev tools around this space.
>
> If there is agreement on this, how do we go about getting it done?
>
> Thanks,
>
> Adam Berry