You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@infra.apache.org by "Patrick Wendell (JIRA)" <ji...@apache.org> on 2014/08/26 07:36:58 UTC
[jira] [Commented] (INFRA-8241) Create ASF account to be shared by
Spark team for project infrastructure
[ https://issues.apache.org/jira/browse/INFRA-8241?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14110305#comment-14110305 ]
Patrick Wendell commented on INFRA-8241:
----------------------------------------
Hey David - I've already explained this to infra several times on different JIRA's. But the issue is closing pull requests that have gone idle, not ones associated with a merged patch. About 10% of our pull requests are never responded to or go idle... within a few months this has accumulated to hundreds of idle pull requests. We have no way to close them other than creating "dummy" commits which actually modify no files except for having the message "closes $foo". I've been personally creating these from my account, but we'd like to do it from a group account because if this continues I'll end up with hundreds of commits to Spark under my personal account that do nothing but close patches.
https://github.com/apache/spark/pulls
It's already bad enough that we have to add dummy commits all over the place (and in the long term this means that about 10% of the total commits in Spark history will just be closing pull requests - which really looks unprofessional). But can we at least have a group account from which to do it?
I previously discussed parts of this with Gavin McDonald in INFRA-7918.
> Create ASF account to be shared by Spark team for project infrastructure
> ------------------------------------------------------------------------
>
> Key: INFRA-8241
> URL: https://issues.apache.org/jira/browse/INFRA-8241
> Project: Infrastructure
> Issue Type: Bug
> Reporter: Patrick Wendell
>
> For a few reasons, it would be really helpful if we could have an ASF account for shared use by the Spark committership.
> One is to have an account for maintenance commits for our github repo. Because we aren't able to close pull requests in our github account, we need to create "dummy" commits that exist for the purpose of closing issues. It doesn't really make sense for a specific commiter to have a bunch of commits floating around just for the purpose of closing issues.
> Metrics like commit count are something that people pay attention to, so I was hoping to have a separate git identity for this. Have any other projects done anything like this?
> It would also be nice to have an account like this for posting release candidates so that we can provide a reference location where release candidate are always posted and it isn't sensitive to the specific release manager. I.e. something like
> http://people.apache.org/~apachespark/latest-rc
> The credentials of the account can be shared internally on our private list.
--
This message was sent by Atlassian JIRA
(v6.2#6252)