Closed Bug 1172316 Opened 10 years ago Closed 10 years ago

Notifications not re-displayed after reboot

Categories

(Firefox OS Graveyard :: Gaia::System::Status bar, Utility tray, Notification, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(blocking-b2g:2.5+, b2g-master verified)

VERIFIED FIXED
2.2 S14 (12june)
blocking-b2g 2.5+
Tracking Status
b2g-master --- verified

People

(Reporter: gerard-majax, Assigned: gerard-majax)

References

Details

(Keywords: regression, Whiteboard: [systemsfe])

Attachments

(1 file)

46 bytes, text/x-github-pull-request
alive
: review+
alive
: feedback+
Details | Review
I have just spotted on current master that notifications are not retriggered after reboot :(
Checked on my device, notifications are correctly stored.
Calling |navigator.mozChromeNotifications.mozResendAllNotifications(function(e) { console.debug(e); });| by hand from WebIDE in the scope of system app makes it working. So it means this code is probably not called during the boot time anymore? Maybe a fallout from bug 1094759 ?
Flags: needinfo?(alive)
Ok, so the code from notification_screen.js used to wait for the "load" event to trigger the resend mechanisms. This seems not to work anymore now.
Summary: Notifications lost after reboot → Notifications not re-displayed after reboot
Removing the event listener on the load event fixes ...
Cool, thanks for filing bug. So does it mean window.onload is not triggered anymore?
Flags: needinfo?(alive)
I will check what happened tomorrow morning; but you are welcome to submit patch.
Alive, I suspect the logic was good when it used the 'load' event because the js was loaded from a <script> statement. Now, it's being loaded by the module logic.
Attached file pr
Alive, this approach should be enough, at least it works properly locally. Do you have any opinion, or suggestion to share?
Attachment #8616483 - Flags: feedback?(alive)
Comment on attachment 8616483 [details] [review] pr No, it's the same and as good as I think. Thank you!
Attachment #8616483 - Flags: feedback?(alive) → feedback+
blocking-b2g: spark? → spark+
Depends on: 1172426
Comment on attachment 8616483 [details] [review] pr Here we go, fix with tests. I'm also fixing the tests because: - we were lacking some this.sinon.tick() to properly deal with mozSettings - the test 'desktop-notification-resend not sent' was just wrong since we early return when the setting is set to false
Attachment #8616483 - Flags: review?(alive)
Whiteboard: [systemsfe]
Target Milestone: --- → 2.2 S14 (12june)
Depends on: 1172548
So it looks like Gij12 is failing, but I'm not able to investigate for now because running those tests locally do not work anymore ...
Finally reproduced locally, then rebasing on master and it's not reproducing locally anymore?
Ok, it's reproducing and only with my patch. Alberto, since you wrote the test, do you have any idea what might be going wrong here ? So far I have been checking and I don't see anything obvious :(
Flags: needinfo?(apastor)
May be we need to wait for 'system.waitForFullyLoaded()' in the setup, now that the new bootstrap has landed?
Flags: needinfo?(apastor)
(In reply to Alberto Pastor [:albertopq] from comment #14) > May be we need to wait for 'system.waitForFullyLoaded()' in the setup, now > that the new bootstrap has landed? Nope that's already present. Forcing the setting |notifications.resend| to false gets me green. Given my observation, what happens is that with this patch the call to resend all stored notifications will be delayed compared to before the landing of new bootstrap. In this test, this makes it triggered after the call to |new Notification|, thus messing up.
Comment on attachment 8616483 [details] [review] pr Thank you Alex.
Attachment #8616483 - Flags: review?(alive) → review+
And it's all green.
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
blocking-b2g: spark+ → 2.5+
This issue is verified fixed on Aries and Flame. However I noticed that notifications on lockscreen upon reboot are NOT immediately shown. It takes several seconds to show which is a regression from Flame 2.2. NI myself to bug this tomorrow. Device: Aries (dogfood debug build) BuildID: 20150721153949 Gaia: 805cf546729ba742bf23febda52970fcb35c0e8f Gecko: 512c7e8f0030 Gonk: 2916e2368074b5383c80bf5a0fba3fc83ba310bd Version: 42.0a1 (2.5 Master) Firmware Version: D5803_23.1.A.1.28_NCB.ftf User Agent: Mozilla/5.0 (Mobile; rv:42.0) Gecko/42.0 Firefox/42.0 Device: Flame (KK 319MB full flashed) BuildID: 20150721010202 Gaia: 4fe0507781f3ed56c8ae5e66dd9489165d1ff68e Gecko: 3a4bfa5d2d02 Gonk: 41d3e221039d1c4486fc13ff26793a7a39226423 Version: 42.0a1 (2.5 Master) Firmware Version: v18D nightly v4 User Agent: Mozilla/5.0 (Mobile; rv:42.0) Gecko/42.0 Firefox/42.0
Status: RESOLVED → VERIFIED
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(pcheng)
Flags: needinfo?(ktucker)
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(ktucker)
Filed bug 1186568 for issue observed on comment 19
Flags: needinfo?(pcheng)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: