Mailing-List: contact notifications-help@accumulo.apache.org; run by ezmlm
Precedence: bulk
Reply-To: jira@apache.org
Date: Wed, 15 May 2013 22:45:15 +0000 (UTC)
From: "Benson Margulies (JIRA)" <jira@apache.org>
To: notifications@accumulo.apache.org
Message-ID: <JIRA.12646846.1368128183932.327547.1368657915813@arcas>
In-Reply-To: <JIRA.12646846.1368128183932@arcas>
References: <JIRA.12646846.1368128183932@arcas>
Subject: [jira] [Commented] (ACCUMULO-1402) Hadoop2 maven artifacts
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/ACCUMULO-1402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13658946#comment-13658946 ] 

Benson Margulies commented on ACCUMULO-1402:
--------------------------------------------

I might not be able to update a tree and look at this right away.  Does the following resemble the idea?

A 'straight' build will produce a set of class files that go with 'hadoop 1', and a set of poms that have the right dependencies for hadoop 1, and that will get pushed to Maven Central under the normal G/A/V coordinates for Accumulo.

A build with -Psomething will build a set of class files that go with 'hadoop 2', and pom files that have the right dependencies for 'hadoop 2' only for folks who add the same -P to their builds, all using the same old G/A/V set, which means that they better not get pushed to Maven Central.

I don't quite see how classifiers come into it; there is no Maven option I know of to cause the main jar artifact of a build to have a non-null classifier. Have you come up with some way to grab the 'hadoop2'-compatible jars and attach them as additional classified artifacts?

                
> Hadoop2 maven artifacts
> -----------------------
>
>                 Key: ACCUMULO-1402
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-1402
>             Project: Accumulo
>          Issue Type: Bug
>            Reporter: John Vines
>            Assignee: Christopher Tubbs
>              Labels: bad
>             Fix For: 1.5.0
>
>         Attachments: ACCUMULO-1402.v1.patch
>
>
> jars need to be compiled separately for hadoop1 vs. hadoop2. Both versions are used in the wild, and will cause a lot of work helping users who do not realize this mistake when they add a maven dependency and expect the one from central to work for their hadoop 2 installation. We should package a separate hadoop2alpha set of jars to release to central for people working against hadoop2 to use.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira