On Tue, Mar 29, 2011 at 7:47 AM, Aaron Kimball <[email protected]> wrote: > related: does there still need to be a 1:1 mapping from commits to JIRAs?
I'm not sure if it's an Apache requirement (does anyone know for sure?). The committer guide doesn't talk about it afaict: http://www.apache.org/dev/new-committers-guide.html IMO it's a good best practice, I've found that having the patch in JIRA helps later on - given that JIRA has a decent interface which includes search. Chris may have more insight, I believe other project (other than hadoop related which I'm familiar with) seem to rely more on mailing list than jira. CTR sounds fine to me, but I'd encourage you to document stuff like this (wiki "how to contribute" is probably a good place to start...) http://www.apache.org/foundation/glossary.html#CommitThenReview Patrick > - Aaron > > On Tue, Mar 29, 2011 at 4:57 AM, Eric Sammer <[email protected]> wrote: > >> I don't have a strong opinion and I think for something like mrunit it >> makes sense to default to commit. I like review first for larger commits, >> but I'm happy to leave it to the individual's best judgement. >> >> >> On Mon, Mar 28, 2011 at 10:04 PM, Aaron Kimball <[email protected]>wrote: >> >>> Hey folks, >>> >>> It was my understanding that in the Hadoop community, all patches had to >>> get >>> a +1 on the review before being committed, even those from long-time >>> committers on the project. The new committer guide [1] suggests that being >>> a >>> "committer" on a project implies that you should commit first. >>> >>> How would we like to go about this here? >>> >>> >>> [1] http://www.apache.org/dev/new-committers-guide.html >>> >>> Thanks, >>> - Aaron >>> >> >> >> >> -- >> Eric Sammer >> twitter: esammer >> data: www.cloudera.com >> >
