On Mar 14, 2010, at 03:52 AM, Jesus Cea wrote:
The release manager could clone the repository to getting it ready for release, while the main repository accepts new changesets. The point would be not stopping commits, ever.
When done, the release manager could tag the release in his/her clone and push the changes to the main branch, like any other commit.
If the release manager needs some patches for the main repository, he/she can cherrypick them manually or using a tool like http://mercurial.selenic.com/wiki/TransplantExtension .
The devil is in the details, like how the release manager repository clone accesses buildbots, for instance.
This is possible workflow. I don't know what is the actual plan, yet.
There's no doubt that having a dvcs will change our workflow. I'm not sure that the above is the best way to do it, but the only way we'll find out is through experimentation.
Another way could be: the freeze still happens on the official branch, from where the release manager does the release. Individual feature and bug branches aren't affected at all, and once the release is done, those branches can be merged to the mainline by their authors. During branch freezes, authors can nominate their branches to fix release blocker issues, managed and reviewed through the tracker.
It's definitely true that with a dvcs, work does not have to "stop" during rc or freeze periods. The question is just, where does that work continue?
-Barry