835467 - Create v1.0.0 and v1-train branches in b2g-manifest

Status: RESOLVED FIXED

(Firefox OS Graveyard :: GonkIntegration, defect)

Description

[Chris Jones [:cjones] inactive; ni?/f?/r? if you need me]

The "nightly" branch has now been made useless, but unfortunately that's where most developers (should) have been on.  Let's create two new branches for dev

 - "v1.0.0" tracking gecko/v1.0.0 and gaia/v1.0.0

 - "v1-train" (or "v1-next"), tracking gecko/gecko-18 and gaia/v1-train

We should then deprecate the "nightly" branch, sending emails to newsgroups, and probably temporarily make it a synonym for "v1-train".

Developers should be encouraged to switch to "v1-train", unless they a priori they should be on another branch.

jhford, want to take this?

Comment 1

[Chris Jones [:cjones] inactive; ni?/f?/r? if you need me]

Setting bb? to get this on various radars.  At the moment, it takes rocket science for anyone at Moz to produce a build anywhere near what we'll be shipping.

Comment 2

[John Ford [:jhford] CET/CEST Berlin Time]

Yup, I'll take this.

Comment 3

[Michael Wu [:mwu]]

I would also suggest that v1.0.0 manifests have gonk repo revisions locked in places that aren't tracking a specific tag. Alternately we can tag those repos if we control them.

Comment 4

[Chris Jones [:cjones] inactive; ni?/f?/r? if you need me]

Everything we rely on should be on g.m.o so we can tag.  Agreed on doing that, but can be a followup.

Comment 5

[John Ford [:jhford] CET/CEST Berlin Time]

Attachment: concatenated diffs for the v1.0.0 and v1-train branch creation

$ git push -n origin v1.0.0 v1-train
To github.com:mozilla-b2g/b2g-manifest
 * [new branch]      v1.0.0 -> v1.0.0
 * [new branch]      v1-train -> v1-train

Comment 6

[John Ford [:jhford] CET/CEST Berlin Time]

Comment on attachment 707241 [details]
concatenated diffs for the v1.0.0 and v1-train branch creation

Does this look sane?

Comment 7

[Chris Jones [:cjones] inactive; ni?/f?/r? if you need me]

Comment on attachment 707241 [details]
concatenated diffs for the v1.0.0 and v1-train branch creation

lgtm

Let's file a followup on getting a v1.0.0 tagd for our sundry repos, if that looks like it'll take a while let's take the intermediate step of adding explicit SHA1s to the v1.0.0 manifests.

Comment 8

[John Ford [:jhford] CET/CEST Berlin Time]

(In reply to Chris Jones [:cjones] [:warhammer] from comment #7)
> Comment on attachment 707241 [details]
> concatenated diffs for the v1.0.0 and v1-train branch creation
> 
> lgtm
> 
> Let's file a followup on getting a v1.0.0 tagd for our sundry repos, if that
> looks like it'll take a while let's take the intermediate step of adding
> explicit SHA1s to the v1.0.0 manifests.

I can add the revisions to the manifests easily, so long as I'm allowed to reserialize the manifests.  Using libxml2, the only thing that changes when I do that is that some unneeded line breaks are removed from the output xml.

Comment 9

[John Ford [:jhford] CET/CEST Berlin Time]

Switched the default branch that config.sh uses to 'v1-train' in 

https://github.com/mozilla-b2g/B2G/commit/fd850a45d90fbdbd92d27a0b98c74d0fd222fc42

Comment 10

[John Ford [:jhford] CET/CEST Berlin Time]

Created and populated:

https://github.com/mozilla-b2g/b2g-manifest/tree/v1.0.0
https://github.com/mozilla-b2g/b2g-manifest/tree/v1-train

Comment 11

[John Ford [:jhford] CET/CEST Berlin Time]

clarifying summary so it matches the what was requested and performed.

Comment 12

[amiller]

No test case needed in Moztrap for this issue.

Comment 13

[Darren]

Cannot verify, need steps to blackbox test this issue.

Comment 14

[John Ford [:jhford] CET/CEST Berlin Time]

(In reply to Darren from comment #13)
> Cannot verify, need steps to blackbox test this issue.

It is not possible to blackbox test this issue because it was creating branches in a repository.