Thanks Roman and Greg for the opportunity/ challenge! Happy to help with the initial release if the group agrees to it, and with your guidance/ support, of course. When do we start!? Best, Nitin _____________________________ From: Roman Shaposhnik <[email protected]<mailto:[email protected]>> Sent: Thursday, September 10, 2015 3:42 PM Subject: Re: Releasing Geode (was Re: September 2015 Report) To: <[email protected]<mailto:[email protected]>>, <[email protected]<mailto:[email protected]>>
On Thu, Sep 10, 2015 at 10:01 AM, Ashvin A <[email protected]<mailto:[email protected]>> wrote: > http://www.apache.org/dev/release-publishing.html#release_manager > > *"The common practice at Apache is for a single individual to take > responsibility for the mechanics of a release. That individual is called > the 'release manager.' Release managers take care of shepherding a release > from an initial community consensus to make it to final > distribution.Release managers do the work of pushing out releases. However, > release managers are not ultimately responsible. The PMC in general, and > the PMC chair in particular (as an officer of the Foundation) is > responsible for compliance with requirements.Any committer may serve as the > manager of a release.A release starts when the project community agrees to > make a release. However, no release manager can make a valid release unless > the community has taken the necessary steps to prepare in advance. The > source code and build process must comply with the legal and intellectual > property requirements for a valid release, and the project must have the > infrastructure in place to correctly sign the release artifacts."* > > > Is it important to be a committer to be a release manager? Somewhat, but it is much more important for a TLP. But even there I was an RM for a Hadoop release before I was a committer on the project. Literally the only thing you can't do as an RM if you are NOT a committer on the project is cherry-picking commits into a release branch. But you know what? This is a bad practice anyway. This is something that appropriate original commiter should be doing instead (which makes RM a bit more of a cat herder, but that's ok ;-)). Thanks, Roman.
