Here is how we might converge on a 1.1 release.
- We occasionally report a git build that we have been using and having success with. We list the mods that we are using.
- This is like a high-water mark that gradually moves forward. It is really about putting "play hours" on a build (as in a mean-time-between-failure metric). I am currently using a Jan 2 build with lots of mods but if, say, Falko is using a Jan 10th build successfully, I would move my install forward.
- Each time the high-water mark moves forward, we have a new candidate that could be a 1.1 build.
- At some point, a dev lead announces a timetable for a 1.1 release, and we all start using the same candidate.
- If the candidate has enough quality, it becomes 1.1 with all the required ceremony (e.g. an installer, release notes, download link)
I am a huge fan of getting stuff out there for people to try. But many users prefer named builds with an installer. The nightlies are generally high-quality and almost any of them could be announced as a 1.1 candidate. At that point we would all move to the candidate and play for a month, then we go gold.