Closed Bug 1162032 Opened 11 years ago Closed 11 years ago

Turn off the mozilla-b2g30_v1_4 branch and JB Flame device image builds in automation

Categories

(Release Engineering :: General, defect)

defect
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: RyanVM, Assigned: RyanVM)

References

Details

Attachments

(9 files)

Per email w/ Kai-Zhen: > I got the answer from PM, there are still devices with 1.4 are on market, > it would be better we keep maintain this branch. And we can stop the > build on this branch. IMO, I think we should disable all nightly builds > and tests. I am thinking if we need to keep the build for push? > How do you think? While I'd love to kill this branch entirely, nightlies would be a start anyway. I'm continuing the email thread because the rationale given there for keeping other builds alive doesn't seem to fit with the reasons we've given for killing other branches (i.e. 1.3t was turned off while devices were still on the market AFAIK). It seems to me that the more-relevant questions are: 1) Are we expecting partners with 1.4-based phones on the market to ship 1.4-based updates to those phones to pick up fixes that have landed since? 2) Are there new 1.4-based phones still coming to market or have they moved on to newer releases? Unless the answer is yes to either of those, I don't see what we're gaining by keeping support for b2g30 around. It's our oldest-supported Gecko version right now and backporting security fixes is becoming increasingly difficult. It's a waste of engineering time spending non-trivial effort backporting and testing fixes that nobody's ever actually going to see on their phone.
Ryan, I agree with you. Steven, Vance, How do you think?
Flags: needinfo?(vchen)
Flags: needinfo?(styang)
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #0) > Per email w/ Kai-Zhen: > > I got the answer from PM, there are still devices with 1.4 are on market, > > it would be better we keep maintain this branch. And we can stop the > > build on this branch. IMO, I think we should disable all nightly builds > > and tests. I am thinking if we need to keep the build for push? > > How do you think? > > While I'd love to kill this branch entirely, nightlies would be a start > anyway. > > I'm continuing the email thread because the rationale given there for > keeping other builds alive doesn't seem to fit with the reasons we've given > for killing other branches (i.e. 1.3t was turned off while devices were > still on the market AFAIK). It seems to me that the more-relevant questions > are: > > 1) Are we expecting partners with 1.4-based phones on the market to ship > 1.4-based updates to those phones to pick up fixes that have landed since? No > 2) Are there new 1.4-based phones still coming to market or have they moved > on to newer releases? AFAIK, move on to newer release such as 2.0, 2.1. or even 2.2... > > Unless the answer is yes to either of those, I don't see what we're gaining > by keeping support for b2g30 around. It's our oldest-supported Gecko version > right now and backporting security fixes is becoming increasingly difficult. > It's a waste of engineering time spending non-trivial effort backporting and > testing fixes that nobody's ever actually going to see on their phone. Agree. I have no question to stop the nightly builds for 1.4 (aka. b2g30)
Flags: needinfo?(styang)
(In reply to Steven Yang [:styang] from comment #2) > Agree. I have no question to stop the nightly builds for 1.4 (aka. b2g30) I think there was already agreement about turning off nightlies. The question is whether we can turn off *all* builds/tests on b2g30 rather than just a very small subset. From the answers you gave, it sounds like we can.
Flags: needinfo?(styang)
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #3) > (In reply to Steven Yang [:styang] from comment #2) > > Agree. I have no question to stop the nightly builds for 1.4 (aka. b2g30) > > I think there was already agreement about turning off nightlies. The > question is whether we can turn off *all* builds/tests on b2g30 rather than > just a very small subset. From the answers you gave, it sounds like we can. Yes, that's what I mean. We can turn off *all* builds/tests. Thanks.
Flags: needinfo?(styang)
Thanks for confirming!
Assignee: nobody → ryanvm
Flags: needinfo?(vchen)
Summary: Turn off nightly builds on b2g30 → Turn off the mozilla-b2g30_v1_4 branch in automation
Blocks: 1164021
Attachment #8604655 - Flags: review?(rail)
Blocks: 1164024
Attachment #8604662 - Flags: review?(rail)
Bonus, this means we can kill the Flame JB builds now too.
Attachment #8604718 - Flags: review?(rail)
Going to finish the b2g30 patches, then will post JB Flame removal patches as well.
Summary: Turn off the mozilla-b2g30_v1_4 branch in automation → Turn off the mozilla-b2g30_v1_4 branch and JB Flame device image builds in automation
AFAICT, this file doesn't have much use anymore, so I went ahead and removed it. I'm open to keeping it and making more useful changes to it otherwise.
Attachment #8604751 - Flags: review?(rail)
The versioned B2G release branches are on their own mozharness branches. Am I going to need to cherry-pick this patch and the forthcoming Flame patch to them as well?
Attachment #8604758 - Flags: review?(rail)
Not quite sure about the change to the flame-kk line in platforms.py
Attachment #8604763 - Flags: review?(rail)
Attachment #8604764 - Flags: review?(rail)
buildfaster_report.py is a mess in general with respect to the list of builds it contains, but oh well.
Attachment #8604765 - Flags: review?(rail)
Attachment #8604766 - Flags: review?(rail)
Note to self, we should also close the mozilla-b2g30 branch on mozharness. And will need to file a BMO bug for disabling the various 1.4 approval & tracking flags.
I've closed b2g30 in Treestatus with the intent to remove is fully once these patches are in production. I've also updated the B2G Landing wiki to reflect the EOL status.
Attachment #8604655 - Flags: review?(rail) → review+
Attachment #8604662 - Flags: review?(rail) → review+
Attachment #8604718 - Flags: review?(rail) → review+
Attachment #8604751 - Flags: review?(rail) → review+
Attachment #8604758 - Flags: review?(rail) → review+
Attachment #8604763 - Flags: review?(rail) → review+
Attachment #8604764 - Flags: review?(rail) → review+
Attachment #8604765 - Flags: review?(rail) → review+
Attachment #8604766 - Flags: review?(rail) → review+
And closing the mozilla-b2g30_v1_4 mozharness branch: https://hg.mozilla.org/build/mozharness/rev/85d46afaed11
Blocks: 1165357
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #17) > And will need to file a BMO bug for disabling the various 1.4 approval & > tracking flags. Bug 1165357. (In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #18) > I've closed b2g30 in Treestatus with the intent to remove is fully once > these patches are in production. Removed. Remaining cleanups can be tracked in their own bugs. Closing this out.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Component: General Automation → General
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: