Just would like to add to it, to make sure we maintain quality. Today, reviewers usually give a +1 pending pre commit run. I think we should not push our patch ourselves if the precommit run has failures/ or has not been run. It is better to get a second go from the reviewer if there are unrelated test failures.
On Tue, Dec 2, 2014 at 10:12 AM, Sravya Tirukkovalur <[email protected]> wrote: > +1 on committers pushing their own change after a +1. And I think we can > keep the "Reviewed by:" to avoid another hop to jira. > > > On Mon, Dec 1, 2014 at 8:26 PM, Prasad Mujumdar <[email protected]> > wrote: > >> +1 >> Sounds good to me. It would help us get the patches committed quickly. >> >> I guess the reviewed by indicate that there's a +1 from another committer, >> it's in line with the regular commit message format. >> >> @Lenni, if we have enough consensus on this, please start a vote thread on >> the private list. The project bylaw changes should be formally approved >> the >> by PPMC. >> >> thanks >> Prasad >> >> >> On Mon, Dec 1, 2014 at 5:41 PM, Gregory Chanan <[email protected]> >> wrote: >> >> > +1. >> > >> > What's the purpose of the "Reviewed by" given we could find that >> > information on the JIRA? >> > >> > Greg >> > >> > On Mon, Dec 1, 2014 at 5:34 PM, Lenni Kuff <[email protected]> wrote: >> > >> > > Hi, >> > > I would like to propose a change in the Sentry commit process >> > > < >> > > >> > >> https://cwiki.apache.org/confluence/display/SENTRY/How+to+commit#Howtocommit-Committingpatches >> > > >. >> > > Currently, the process recommends that the contributor and person who >> > > pushes the code change are different people, with the following >> > > exception: "Committers may have to bypass the long drawn process to >> > commit >> > > the change in order to fix a broken build, or work through a release >> > etc." >> > > >> > > I suggest that we relax this restriction to make it acceptable for a >> > > comitter to push their own change, as long as it has a +1 from another >> > > committer on the project. The commit messages for these changes should >> > > always include a new "(Reviewed by: <Name of Reviewer>)". >> > > >> > > The motivation is to improve the velocity with which changes can be >> > > submitted, while retaining the same level of quality for the project. >> > > >> > > Please let me know if you object to this proposed change. >> > > >> > > Thanks, >> > > Lenni >> > > >> > >> > > > > -- > Sravya Tirukkovalur > -- Sravya Tirukkovalur
