[
https://issues.apache.org/jira/browse/HBASE-22076?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16798169#comment-16798169
]
stack commented on HBASE-22076:
-------------------------------
{quote}In the past I just do this:
1. Set the version to x.x.x, i.e, remove the -SNAPSHOT suffix, it will be an
issue on jira.
2. Do everything for a release.
3. Set the version to x.x.x-SNAPSHOT, it will be another issue on jira.
In step 2, we may push out several RCs, and the version is always x.x.x, so I
do not think we need to change the versions in all the pom when we make a
RC?{quote}
Yeah, I was trying to eliminate the above steps.... automate them. Remove
having to do the JIRAs and commits by hand and having to remember to do it. If
only one RC (the usual?), then the scripts are doing the right thing? I was
thinking scripting it a positive thing.
I can undo it no problem if you feel strongly about it (or anyone else does).
> Move version on from 2.2.0 to 2.2.1-SNAPSHOT
> --------------------------------------------
>
> Key: HBASE-22076
> URL: https://issues.apache.org/jira/browse/HBASE-22076
> Project: HBase
> Issue Type: Sub-task
> Reporter: stack
> Assignee: stack
> Priority: Major
>
> Moving on the pom version. Doing this because the script that makes RCs fails
> otherwise; it needs to work against a version that ends in -SNAPSHOT. The
> script takes care of setting the version appropriately as it builds. FYI
> [~zghaobac]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
