To ensure that there is a long enough amount of "bake time" for blockers, I have decided that we will be delaying the upcoming code freeze for Firefox 4 Beta 6, originally scheduled to be tomorrow.
There are still a large number of blockers remaining:
https://bugzilla.mozilla.org/buglist.cgi?quicksearch=blocking2.0:beta6
and the JaegerMonkey code has yet to be merged over to mozilla-central. Many of these changes - especially those to the JavaScript engine - will require at least two days of nightly tester coverage before we're comfortable starting builds for the next beta.
Tomorrow I will be reviewing the remaining blockers, working with component leads to re-target those which are not required for Firefox 4 to a future release. Once again, I encourage everyone to focus on blocking bugs as their primary work task.
cheers,
mike
> Tomorrow I will be reviewing the remaining blockers, working with component leads to re-target those which are not required for Firefox 4 to a future release. Once again, I encourage everyone to focus on blocking bugs as their primary work task.
An update: the JagerMonkey team informs me that they hope to merge either later today or tomorrow.
Given the number of outstanding blockers, I am going to propose the following revised schedule:
- code freeze Wednesday, Sept 15, late PT
- bake for two days
- send to build on Friday, Sept 17
This would mean holding the tree frozen (or at least to very limited checkins) for two days. I propose exceptions be made for:
- followups to beta 6 blockers
- fennec blockers
- test, bustage and NPOTB fixes
We can discuss the merits of holding the tree closed for this long next Tuesday; let's first get to code freeze on Wednesday. Do people think restricting the tree to beta6+ blockers only would help us focus?
cheers,
mike
In addition to helping focus, it will reduce risk by removing
confounding factors from the tree, reducing load on infrastructure,
and reducing contention for review resources. I think it is a fine
idea.
(Honestly, it might or might not help us focus; that's more an issue
of motivation than policy, I think. But it's unlikely to hurt.)
Mike
Axel