- Beta user stats
- Stability report
- Go/no-go decision
- Official "Go To" Beta is still May 17th.
- The code that's in Aurora on the 16th, will move to Beta and we will do a build from there.
- The merge happens and we have the choice at that time whether or not the update is put out for all the beta users or not.
- We have 6 bugs we are tracking and we will either get them resolved or decide they are not severe enough to block.
- Laura: When will people see this beta live? Probably a couple of days after we do the merge. We will want some time to test this.
- Christian will work with Laura offline to figure out time line - bugs will be logged for all the web pieces and everything that needs to happen when.
- Christian will also publish a detailed schedule of the merge timeline so that rel eng and marketing can plan accordingly.
Beta user stats
- According to crash stats, yesterday's volume on 5.0b1 was 37K.
- Over the weekend, getting about 11K hits on the first run page per day.
- Last Thurs/Fri, about 15K hits per day.
- John: We have been running with 50% throttling so we didn't overshoot. Should we up that?
- We are going to go to change to 100% unthrottled for that group of people.
- We will not stop at 100K users either; will offer to all ~417K eligible users.
- Details in bug 651982
- Bump in numbers is en-US only - this is because FF5.0b1 is en-US only, not localized.
- Top crash list similar to top crash list for Aurora.
- No new or explosive crashes have been identified.
- Still need more users on the channel to get better stability data.
- A couple of issues with the crash metrics have surfaced
- We are currently throttling 5.0 crash reports at 10%. We want to change this so that on Beta, we have 100% and throttle on release. Bug 655838 will be fixed today.
- There is a second issue where the crash rate appears to be miscalculated. For 37K ADUs on 5.0 we have 1600 crashes which seems high if you compare to a similar number of users on the trunk. This is still under investigation. Bug 655846
- There was a thread on the list about this.
- What data would QA present to help us make the Go/no-go decisions?
- What has QA done up to that point.
- Stability report.
- Performance? We don't have anybody reporting on this. Most of these should be handled in the Nightly to Aurora transition.
- Bugs we are tracking that need to be resolved before going to Beta - 6 bugs.
- Next Monday we will assemble this information to make a go/no-go decision for doing the merge.
- Matt owns packaging this up for Monday's meeting.
- Extension compatibility
- Do we care about this? We probably can't make it part of the calculation for this time but in the future, this will be a bigger part of the criteria.
- Would we delay going to Beta if an extension is not compatible? Probably something we need to take to the list.
- There are 2 explicit decisions here
- Do a beta?
- Did we do it correctly?
- All ready for 17may.
- Still an issue with l10n but it does not affect the merge on May 17th. This *might* relate to the next merge from Aurora-Beta.