867847 - navigator.id.get() is busted on b2g

Status: RESOLVED FIXED

(Core Graveyard :: Identity, defect)

Description

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

navigator.id.get(), which was landed for native persona on b2g in bug 790141, no longer works.

It needs a one-line patch for b2g

My heart is in pieces on the floor.

Comment 1

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

Attachment: Fix busted navigator.id.get() on b2g

yep, one line.

Comment 2

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

I'm also investigating alternative ways of fixing this on the server-side.  It may be possible to re-implement get in the included js using the observer api, which is how it's done here.

Comment 3

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

Re-built and tested both on desktop and device

Comment 4

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/projects/birch/rev/c7b6f809741c

Should this have a test?

Comment 5

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

(In reply to Ryan VanderMeulen [:RyanVM] from comment #4)
> https://hg.mozilla.org/projects/birch/rev/c7b6f809741c
> 
> Should this have a test?

Yes, both client UI tests on b2g and mochitests to prove the dom/identity functions

I will file a separate bug to get this http://people.mozilla.org/~jparsons/test_b2g_get.html into the b2g UI Tests (like the current navigator.mozId tests there)

Ryan, to keep this patch as minimal as possible, I think it makes sense for me to land the dom tests as a separate patch.  What do you think?

Comment 6

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/c7b6f809741c

Comment 7

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

Nomming for tef.  I'm afraid this is a bad bug.

Comment 8

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

More info:

Without the fix, persona will not work on b2g for many web sites (specifically the ones that use the older but still supported navigator.id.get api)

The fix is about as low-risk as I think a patch can be.  It adds a boolean flag to an object.  There are no new code paths, and no new features.

Comment 9

[Jason Smith [:jsmith]]

This has already been known for quite some time in bug 823761 where an argument was already made that this was a not a blocker. Why has this story suddenly changed?

Comment 11

[Jason Smith [:jsmith]]

(In reply to Jason Smith [:jsmith] from comment #9)
> This has already been known for quite some time in bug 823761 where an
> argument was already made that this was a not a blocker. Why has this story
> suddenly changed?

Clarified in IRC. There was a miscommunication that happened in analyzing bug 823761's priority, which happens to be a dupe of this bug. bug 823761 should have been a blocker from the very beginning.

Comment 12

[Jed Parsons (use needinfo, please) [:jedp, :jparsons]]

(In reply to Jason Smith [:jsmith] from comment #11)
> Clarified in IRC.

Thanks very much for checking in, Jason, and for making sure we sort this out.

Comment 13

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g18/rev/6b699b38e41d
https://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/ce67220b877d