1015942 - tracking bug for build and release of thunderbird 31.0

Status: RESOLVED FIXED

(Release Engineering :: Release Requests, defect)

Description

[Rail Aliiev [:rail]]

+++ This bug was initially created as a clone of Bug #905138 +++

Comment 1

[Rail Aliiev [:rail]]

FTR, mgerva was kind to setup TBPL, graph server, the hooks and treestatus as a part of Firefox esr31.

The patches above are supposed to land whenever we are ready for Thunderbird 31 release (on the merge day).

Comment 2

[Rail Aliiev [:rail]]

Attachment: tb31-tools.diff

Comment 3

[Rail Aliiev [:rail]]

Attachment: tb31-buildbot-configs.diff

* s/comm-esr24/comm-esr31/
* rename release configs and templates
* rename l10n file for nightlies
* rename a directory under mozilla2 (we still use that mozconfig for source tarballs)

Comment 4

[Nick Thomas [:nthomas] (UTC+12)]

Comment on attachment 8452312 [details] [diff] [review]
tb31-tools.diff

r+ if there is going to be no overlap at all between comm-esr24 and 31.

Comment 5

[Nick Thomas [:nthomas] (UTC+12)]

Comment on attachment 8452313 [details] [diff] [review]
tb31-buildbot-configs.diff

r+ with changes on landing.

>diff --git a/mozilla/release-thunderbird-comm-esr24.py.template b/mozilla/release-thunderbird-comm-esr31.py.template
>     'mozilla': {
>         # XXX remove js/src/config/milestone.txt when setting up ESR31

Please do this so tagging doesn't fail.

>diff --git a/mozilla/staging_release-thunderbird-comm-esr24.py.template b/mozilla/staging_release-thunderbird-comm-esr31.py.template
>     'mozilla': {
>         # XXX remove js/src/config/milestone.txt when setting up ESR31

Comment can go too.

Comment 6

[Rail Aliiev [:rail]]

Attachment: tb31-buildbot-configs-1.diff

Ooops. Updated patch with comments and js/src/config/milestone.txt removed.

Comment 7

[Rail Aliiev [:rail]]

Comment on attachment 8452312 [details] [diff] [review]
tb31-tools.diff

https://hg.mozilla.org/build/tools/rev/cb01da93f219

Comment 8

[Rail Aliiev [:rail]]

Comment on attachment 8452829 [details] [diff] [review]
tb31-buildbot-configs-1.diff

https://hg.mozilla.org/build/buildbot-configs/rev/9f0b1c086769

Comment 9

[Rail Aliiev [:rail]]

https://hg.mozilla.org/build/buildbot-configs/rev/561f13274c6e to make tests happy.

Comment 10

[Rail Aliiev [:rail]]

I synced the repo from comm-beta https://tbpl.mozilla.org/?tree=Thunderbird-Esr31&rev=889e445119fa to trigger the automation. Some of the in-tree configs/scripts (client.py at least) should be updated as a part of the merge push (ATM the builds use releases/mozilla-beta instead of releases/mozilla-esr31).

Comment 11

[Rail Aliiev [:rail]]

Attachment: l10nUsePymake

l10nUsePymake should be set to True (as in beta)

Comment 12

[[e-mail] (:bhearsum)]

Comment on attachment 8456137 [details] [diff] [review]
l10nUsePymake

Review of attachment 8456137 [details] [diff] [review]:
-----------------------------------------------------------------

Probably don't need to touch the non-templates, but it doesn't hurt.

Comment 13

[Rail Aliiev [:rail]]

Comment on attachment 8456137 [details] [diff] [review]
l10nUsePymake

(In reply to Ben Hearsum [:bhearsum] from comment #12)
> Probably don't need to touch the non-templates, but it doesn't hurt.

Yeah. I just wanted to see release runner commit diffs clearer.

https://hg.mozilla.org/build/buildbot-configs/rev/e52ce5cfec61
and merged to production (no reconfig required ATM): https://hg.mozilla.org/build/buildbot-configs/rev/c20448c69517

Comment 14

[Rail Aliiev [:rail]]

Attachment: update_configs.diff

Just to update the patch list from comment 9:
https://hg.mozilla.org/build/buildbot-configs/raw-rev/561f13274c6e

Comment 15

[Rail Aliiev [:rail]]

Attachment: comm-esr24-tools.diff

In case we decide to build 24.7.0 without updates.

* Adds back comm-esr24
* scripts/updates/balrog-release-pusher.py conditially appends release_config['releaseChannel'] to the list of updateChannels to make it possible to not barf if the config has those set to [] and None (see next patch).

Comment 16

[Rail Aliiev [:rail]]

Attachment: esr24-dead-buildbot-configs.diff

Comments will be posted in the next comment.

Comment 17

[Rail Aliiev [:rail]]

11:01:14       rail │ Standard8: are we ok with havin 24.7.0 as a "fork" without updates?
11:09:23 @Standard8 │ rail: yes, I think so - would we be able to get users to update from it in future once we dump 24.x?
11:10:53       rail │ Standard8: we will be generating updates for everything up to 24.6.0 right away; later we can manually generate updates for those who manually downloaded in installed 24.7.0+, sure 

I'm going to proceed with this plan and land the patches.

Comment 18

[Rail Aliiev [:rail]]

Comment on attachment 8456384 [details] [diff] [review]
comm-esr24-tools.diff

https://hg.mozilla.org/build/tools/rev/8fb80a7b92d7

Comment 19

[Rail Aliiev [:rail]]

Comment on attachment 8456386 [details] [diff] [review]
esr24-dead-buildbot-configs.diff

https://hg.mozilla.org/build/buildbot-configs/rev/dbf256bfbb8a

Comment 20

[Rail Aliiev [:rail]]

in production

Comment 21

[Rail Aliiev [:rail]]

Replaced None with {} to make release.py happy:

remote:   https://hg.mozilla.org/build/buildbot-configs/rev/37abdcfddcf8
remote:   https://hg.mozilla.org/build/buildbot-configs/rev/75635858152f

Comment 22

[Rail Aliiev [:rail]]

Attachment: throttle

According to IRC we are going completely throttled

Comment 23

[Rail Aliiev [:rail]]

Attachment: releaseChannel.diff

It turns out that the build system (ab)uses MOZ_UPDATE_CHANNEL in many places. It's better to revert this change and probably go with a build2. :(

Comment 24

[[e-mail] (:bhearsum)]

Comment on attachment 8458095 [details] [diff] [review]
releaseChannel.diff

Review of attachment 8458095 [details] [diff] [review]:
-----------------------------------------------------------------

Ugh. Is there a bug on fixing the build system to make this decision in a better way? Why isn't this a mozconfig option?

Comment 25

[Rail Aliiev [:rail]]

(In reply to Ben Hearsum [:bhearsum] from comment #24) 
> Ugh. Is there a bug on fixing the build system to make this decision in a
> better way? Why isn't this a mozconfig option?

No idea. I'll check when I'm done with this.

Comment 26

[Rail Aliiev [:rail]]

Comment on attachment 8458095 [details] [diff] [review]
releaseChannel.diff

remote:   https://hg.mozilla.org/build/buildbot-configs/rev/2d5b589ed288
remote:   https://hg.mozilla.org/build/buildbot-configs/rev/79c1fb2b3b3d

Comment 27

[Aki Sasaki (not active)]

In production.

Comment 28

[Rail Aliiev [:rail]]

Comment on attachment 8457968 [details] [diff] [review]
throttle

cvs commit -m "Bug 1015942 - tracking bug for build and release of thunderbird 31.0. r=bhearsum"
Checking in config-dist.php;
/cvsroot/mozilla/webtools/aus/xml/inc/config-dist.php,v  <--  config-dist.php
new revision: 1.304; previous revision: 1.303
done

Comment 29

[Rail Aliiev [:rail]]

Done!

Comment 30

[Pete Moore [:pmoore][:pete]]

Attachment: bug1015942_buildbot-configs_setup-master.py_v1.patch

Required when setting up a new staging master...

Comment 31

[Pete Moore [:pmoore][:pete]]

Awesome, thanks Kim! :)

Landed on default: https://hg.mozilla.org/build/buildbot-configs/rev/8c5ae0f39459

Comment 32

[Rail Aliiev [:rail]]

Thanks for the fix!

Comment 33

[Simone Bruno [:simone]]

In production with reconfig on 2014-08-25 08:49 PT