Mailing-List: contact general-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: general@hadoop.apache.org
Received-SPF: pass (athena.apache.org: domain of esammer@cloudera.com
 designates 209.85.210.176 as permitted sender)
MIME-Version: 1.0
Date: Sun, 30 Jan 2011 18:26:10 -0500
Message-ID: <AANLkTi=fOv-5sKrP9jMDC=mJuQjZgh+Tz4LWyDjYiwic@mail.gmail.com>
Subject: MRUnit
From: Eric Sammer <esammer@cloudera.com>
To: general@hadoop.apache.org
Content-Type: multipart/alternative; boundary=20cf30549ef7e51554049b189f2a

--20cf30549ef7e51554049b189f2a
Content-Type: text/plain; charset=ISO-8859-1

All:

I've created a fork of mrunit from Hadoop's contrib directory at
https://github.com/esammer/mrunit

Why:

* MRUnit could benefit from a separate development and release cycle than
that of Hadoop itself. This would allow MRUnit to evolve independently of
Hadoop.
* Moving projects out of contrib simplify Hadoop's build and the artifacts
that need to be maintained.
* I've heard rumblings that contrib is generally frowned upon by some of the
committers. I agree.

What is already done:

* Build system ported to Maven.
* No more build warnings re: generics or deprecation.

I think it makes sense to remove mrunit from contrib (after a deprecation
period) but I'm curious to hear others' opinions. If the community at large
decides it makes sense to retain mrunit in contrib I'm happy to provide
contributions back. I'd also like to offer commit status on the fork to any
current Hadoop committers who are interested. The project is being hosted on
Github to ease collaboration. Pull requests are greatly appreciated.

Regards.
-- 
Eric Sammer
twitter: esammer
data: www.cloudera.com

--20cf30549ef7e51554049b189f2a--