1312000 - Sign Linux64, Linux32, and Android l10n desktop nightly repacks to TaskCluster (on date)

Status: RESOLVED FIXED

(Release Engineering :: General, defect)

Description

[Justin Wood (:Callek)]

+++ This bug was initially created as a clone of Bug #1171738 +++

This is a bug to sign the nightly repacks (again, on date) so date-appropriate feedback-only requested.

Annoyances here, which I plan to address:

  * No Simple Names, (this is really a problem with the real repacks, but this is the first bit of code that needs the non-simple-names)
  * Hardcoded locale values (Will be instead parsing the locales file directly in the future)

Comment 1

[Justin Wood (:Callek)]

Attachment: Bug 1312000 - Sign Linux64, Linux32, and Android l10n desktop nightly repacks to TaskCluster (on date).

Review commit: https://reviewboard.mozilla.org/r/87674/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/87674/

Comment 2

[Kim Moir [:kmoir] ET]

Comment on attachment 8803364 [details]
Bug 1312000 - Sign Linux64, Linux32, and Android l10n desktop nightly repacks to TaskCluster (on date).

https://reviewboard.mozilla.org/r/87674/#review86646

I would echo what Dustin has said here https://bugzilla.mozilla.org/show_bug.cgi?id=1171738#c7 about the duplication and not creating a new signing transforms.  If it lands on date to allow you iterate on it, that would be fine but we shouldn't be creating more signing transforms on m-c as this will lead to more maintenance as we need to make changes.

Comment 3

[Justin Wood (:Callek)]

This patch is being run on try right now with:

https://tools.taskcluster.net/push-inspector/#/Qv8Ei_RdS_mPYRwFZeNXEg?_k=13je3t
and
https://tools.taskcluster.net/push-inspector/#/afe0TwlGTRicepWm2EhngA?_k=hchj87

Comment 4

[Justin Wood (:Callek)]

(In reply to Kim Moir [:kmoir] from comment #2)
> Comment on attachment 8803364 [details]
> Bug 1312000 - Sign Linux64, Linux32, and Android l10n desktop nightly
> repacks to TaskCluster (on date).
> 
> https://reviewboard.mozilla.org/r/87674/#review86646
> 
> I would echo what Dustin has said here
> https://bugzilla.mozilla.org/show_bug.cgi?id=1171738#c7 about the
> duplication and not creating a new signing transforms.  If it lands on date
> to allow you iterate on it, that would be fine but we shouldn't be creating
> more signing transforms on m-c as this will lead to more maintenance as we
> need to make changes.

The new transform was necessary, at least for now, to account for the divergence of unsignedartifacts, but I do have ideas on how to bring this all in to have less divergence and more DRY-ness on the definitions and such.

Comment 5

[Justin Wood (:Callek)]

This landed in https://hg.mozilla.org/projects/date/rev/79a2f66ff5c00661f943ff2c5107370008ba7ed8