Per my reading http://www.apache.org/legal/release-policy.html#release-approval <http://www.apache.org/legal/release-policy.html#release-approval> it’s not OK to stop when you reach 3 +1 votes. Because it’s not 3 +1s absolute, it’s 3 more +1s than -1s. So, you have to have a fixed, reasonable timeframe for the vote to give those -1s time to accrue.
Release policy says the timeframe SHOULD be 72 hours. I can think of scenarios where it could be less, but you still have to give PMC members adequate notice. When Ross wrote “72hrs or 3 +1” I think he was using shorthand, not intending to rewrite foundation policy. Also note in that same section: > Before casting +1 binding votes, individuals are REQUIRED to download > all signed source code packages onto their own hardware, verify that they > meet all requirements of ASF policy on releases as described below, > validate all cryptographic signatures, compile as provided, and test the > result on their own platform. This clearly establishes the minimum required of mentors voting for a podling release. Lastly, I absolutely agree with the question in the subject line. Yes, Mentors SHOULD vote on podling releases prior to asking IPMC to vote. As mentors are busy, and are not 100% focused on the podling, the podling should give them a little nudge and a little leeway. Case in point: Druid 0.14 started a vote at 4.08pm on Friday. It is now 3.08pm on Monday, i.e. we’re at 71 hours. I voted but the other 2 mentors (whom I would describe as “engaged”) have not yet. I just emailed them to ask them to vote. If they do not vote in the next hour, the release manager should keep the vote open for a while to give them chance to vote. Julian > On Apr 1, 2019, at 2:59 PM, Justin Mclean <jus...@classsoftware.com> wrote: > > Hi, > >> 72 hours is a guide. As long as your project community is agreeable you can >> say "72hrs or 3 +1”. > > 72 hours is a guide, but recommended, as it gives a chance for all PPMC > members to have a look at the release. They may be in different timezone or > have day jobs or other things going on, we are all volunteers here and not > all of us work full time on a project. If that 72 hours was constantly > reduced it would exclude certain groups of people looking at the release and > I don’t think we want to do that. For instance I don’t think it would be OK > if those 3 +1’s came in an hour of announcing the release candidate. > > Thanks, > Justin
