823173 - nsIDOMNavigatorTime exposed on navigator objects on desktop but xpt is not packaged

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Boris Zbarsky [:bzbarsky]]

This is at least a problem on beta 18.  It causes "navigator instanceof Window" to throw instead of returning false.

This was not a problem in 17.

Comment 1

[Boris Zbarsky [:bzbarsky]]

m-c and aurora seem to be unaffected.

Looks like this was fixed on those in bug 807222 part 2.

Justin what should we do for 18 here?  Just backport that patch?  Do you think it's safe enough?

Triage info: this can cause web compat issues due to scripts throwing exceptions when they shouldn't.

Comment 2

[Justin Lebar (not reading bugmail)]

> Do you think it's safe enough?

That patch would definitely not be among the scariest I've landed on beta in recent memory.

Comment 3

[Alex Keybl [:akeybl]]

We're in our final beta though, and we haven't heard of any major web compat issues at this point. That's not to say we won't on release, but we need to be very very cautious here. What's the worst thing that can happen if an uplift of bug 807222 part 2 regresses?

Comment 4

[Boris Zbarsky [:bzbarsky]]

I don't think anything bad can happen on non-b2g, for what it's worth.  If it compiles to start with, it'll be strictly better than what we have now, as far as I can tell....

Comment 5

[Boris Zbarsky [:bzbarsky]]

Attachment: Beta backport of fix

Comment 6

[Boris Zbarsky [:bzbarsky]]

Attachment: test. Make sure that navigator doesn't have any broken interfaces on it.

Comment 7

[Justin Lebar (not reading bugmail)]

Comment on attachment 694138 [details] [diff] [review]
Beta backport of fix

Ship it.

Comment 8

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 694138 [details] [diff] [review]
Beta backport of fix

[Approval Request Comment]
Bug caused by (feature/regressing bug #): Some b2g landing sometime
User impact if declined: Some websites will break in Firefox 18 that worked in 
   Firefox 17.  How many, unclear.
Testing completed (on m-c, etc.): This patch has been on 19 and 20 for a while
   now.
Risk to taking this patch (and alternatives if risky): Risk is low.  The other
   option is to ship the known-broken behavior.
String or UUID changes made by this patch: None

Comment 9

[Boris Zbarsky [:bzbarsky]]

Checked in test on inbound and aurora:

https://hg.mozilla.org/integration/mozilla-inbound/rev/85055fcfb211
https://hg.mozilla.org/releases/mozilla-aurora/rev/eaccb5edc854

Comment 10

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/85055fcfb211

Comment 11

[Alex Keybl [:akeybl]]

Comment on attachment 694138 [details] [diff] [review]
Beta backport of fix

We've weighed your risk/reward analysis and agree that this is a good fix for FF18, especially given the extra flex week in this cycle. Please land no later than 12/26.

Comment 12

[Boris Zbarsky [:bzbarsky]]

Pushed patch and test to beta:

   https://hg.mozilla.org/releases/mozilla-beta/rev/5b900eea305f
   https://hg.mozilla.org/releases/mozilla-beta/rev/3fa6dcc664fc

Comment 13

[Boris Zbarsky [:bzbarsky]]

And https://hg.mozilla.org/releases/mozilla-beta/rev/3094a79aa73b to disable the new test on Android, because on beta their xpt situation is all broken.  See bug 807222 part 5.

Comment 14

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

https://hg.mozilla.org/releases/mozilla-beta/rev/48da39be33e1

Comment 15

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g18/rev/5b900eea305f
https://hg.mozilla.org/releases/mozilla-b2g18/rev/3fa6dcc664fc
https://hg.mozilla.org/releases/mozilla-b2g18/rev/3094a79aa73b
https://hg.mozilla.org/releases/mozilla-b2g18/rev/48da39be33e1

Comment 16

[Paul Silaghi, QA [:pauly]]

Verified fixed FF 18 release Win 7 x64.

Comment 17

[Paul Silaghi, QA [:pauly]]

STR in https://bugzilla.mozilla.org/show_bug.cgi?id=782524#c0