Closed Bug 895104 Opened 6 years ago Closed 6 years ago

Call to non-existent DOMRequestIpcHelper.observe

Categories

(Firefox OS Graveyard :: General, defect)

defect
Not set

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: nsm, Assigned: justin.lebar+bug)

References

Details

Attachments

(1 file)

Bug 889984 moved some DOMRequestIpcHelper code into DOMRequestIpcMessageListener, including observe().

SystemMessageManager.observe() still has a call

    this.__proto__.__proto__.observe.call(...)

which errors. cpeterson managed to hit this on Android. I'm surprised B2G hasn't been hit yet.

It seems to me this line should be removed. If I understand 889984 correctly, then it fixes the reason the line was added in bug 802564.
OS: Linux → All
Hardware: x86_64 → All
I hit an undefined `this.__proto__.__proto__.observe` when porting desktop system messages (bug 868322) to Android (bug 895689).
Blocks: 895689
I agree this line should go, but I'm confused as to why it's causing problems.  The observer service should swallow errors thrown when it calls observers, so if the observer service is the only one calling observe(), this exception shouldn't matter.

Is this line called from elsewhere?

If this causes problems, we should backport to b2g18, so please advise.
Attached patch Patch, v1Splinter Review
Attachment #779259 - Flags: review?(fabrice)
Attachment #779259 - Flags: review?(fabrice) → review+
Flags: needinfo?(nsm.nikhil)
Assignee: nobody → justin.lebar+bug
https://hg.mozilla.org/mozilla-central/rev/7bb5b6909d3e
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Justin,
I'm not sure what the ni? here is for. observe() is being called manually in SystemMessageManager, and not through observer service, but I see your patch already understood and fixed that.
Sorry to be unclear.

I'd like to know whether this line causes problems, or if it's just an innocuous error.  If it causes problems, we need to backport to b2g18.

AFAICT the error is innocuous, but I wanted to check.
That line is the last execution of SystemMessageManager.observe() so it won't stop any codes after that. I think it's innocuous. Keeping that only in central should be fine.
I'm going to go with Gene's judgement here, since I don't know the code that well.
Flags: needinfo?(nsm.nikhil)
You need to log in before you can comment on or make changes to this bug.