1478053 - Download progress notification can't be dismissed after closing app

Status: VERIFIED FIXED

(Firefox for Android Graveyard :: General, defect, P1)

Description

[Andrei Lazar]

Steps to reproduce:
1) Go to https://www.thinkbroadband.com/download
2) Start downloading something
3) Close the application before the download is finished
4) Try to interrupt the download by using the Cancel action within the notification

Expected result:
The download will stop and the notification will disappear

Actual result:
Nothing happens

Comment 1

[Sorina Florean [:sflorean]]

Regression window:
Last good build: 6272dd5e
First bad build: 3931f461

Pushlog: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=6272dd5e7417adf68255abfb1d691a36e5a2d6e8&tochange=3931f461c8e8668a264d52b51a4524aac39a7a16

Comment 2

[Pascal Chevrel:pascalc (PTO until April 26)]

Most of the patches in the regression range are DOM/CSS/l10n/JS engine patches, I see a necko patch in that range that might look more likely to produce such a regression. Jan, do you think that your patch in bug 1442178 could have caused this regression?

Comment 3

[Jan Honza Odvarko [:Honza] (always need-info? me)]

(In reply to Pascal Chevrel:pascalc from comment #2)
> Most of the patches in the regression range are DOM/CSS/l10n/JS engine
> patches, I see a necko patch in that range that might look more likely to
> produce such a regression. Jan, do you think that your patch in bug 1442178
> could have caused this regression?

Bug 1442178 has been fixed by Jan Bambas (:mayhemer).
I can't forward NI to him atm (he's on PTO), so I pinged him through email.

Honza

Comment 4

[Jan Honza Odvarko [:Honza] (always need-info? me)]

Jason, do you have more info on this?

Honza

Comment 5

[Honza Bambas (:mayhemer)]

(In reply to Jan Honza Odvarko [:Honza] from comment #3)
> (In reply to Pascal Chevrel:pascalc from comment #2)
> > Most of the patches in the regression range are DOM/CSS/l10n/JS engine
> > patches, I see a necko patch in that range that might look more likely to
> > produce such a regression. Jan, do you think that your patch in bug 1442178
> > could have caused this regression?
> 
> Bug 1442178 has been fixed by Jan Bambas (:mayhemer).
> I can't forward NI to him atm (he's on PTO), so I pinged him through email.
> 
> Honza

Hard to say if this is related.  Logs [1] might help answer this.

[1] https://developer.mozilla.org/en-US/docs/Mozilla/Debugging/HTTP_logging

Comment 6

[Marcia Knous [:marcia]]

Andrei: What device and API are you using? Would a logcat help?

Comment 7

[Andrei Lazar]

Hei Marcia, this is reproducible on any device that is at least API 26, though, I am not pretty sure what is the expected result. Should the download stop when closing the app from background?

Comment 8

[Marcia Knous [:marcia]]

adding Paolo in case he knows the answer to Comment 7. I am not sure what the expected behavior is for either desktop or android for download manager.

Comment 9

[:Paolo Amadini]

For Desktop, downloads are cancelled when closing the application, but we show a quit confirmation dialog box.

I don't know what is the expected behavior on Android. When we are running in multi-process mode, our internal download state and progress notifications are handled in the parent process, so if this process terminates then downloads have to stop as well.

Comment 10

[Mike Taylor [:miketaylr]]

Andreas, does product have any opinions on expected behavior here?

Comment 11

[Andreas Bovens [:abovens]]

When trying this in Chrome, the download continues in the background (and can be tracked via the built-in Download manager). We should probably do the same.

Comment 12

[Pascal Chevrel:pascalc (PTO until April 26)]

Susheel, could you prioritize this regression? Should it be fixed and uplifted for 63? Thanks

Comment 13

[:sdaswani]

Vlad can you work on this ASAP?

Comment 14

[Chris Peterson [:cpeterson]]

Setting priority P1 because Susheel asked Vlad to work on this bug ASAP (in comment 13).

Could this bug affect GeckoView in Focus?

Comment 15

[Vlad Baicu]

Attachment: Bug 1478053 - Remove persistent notifs from the clearable list. r=sdaswani

Persistent notifications were not properly distinguished from clearable notifications
because they would first start as not persistent and got updated afterwards. Also added
a handle action in BrowserApp's onCreate for whenever the process gets removed.

Comment 16

[(inactive) Jim Chen [:jchen] [:darchons]]

Comment on attachment 9011878 [details]
Bug 1478053 - Remove persistent notifs from the clearable list. r=sdaswani

Jim Chen [:jchen] [:darchons] has approved the revision.

Comment 17

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/55660b661412
Remove persistent notifs from the clearable list. r=jchen

Comment 18

[Cristina Coroiu [:ccoroiu]]

https://hg.mozilla.org/mozilla-central/rev/55660b661412

Comment 19

[Sorina Florean [:sflorean]]

Tested with 
- Xiaomi Mi4i(Android 5.0.2), 
- Nokia 6(Android 7.1.1) and after the browser is closed, the notification is not displayed. Is displayed again and the files are downloaded if I re-open the browser. 
Nightly build 64.0a1 (2018/01/10) downloaded from Google Play. 

For some devices: Nexus 5(Android 6.0.1), Motorola Nexus 6(Android 7.1.1), HTC Desire 820 (Android 6.0.1), Google Pixel (Android 9) the issue is not fixed and filed bug 1495443.

Comment 20

[Pascal Chevrel:pascalc (PTO until April 26)]

Vlad, could you request beta uplift please? thanks

Comment 21

[Vlad Baicu]

Not sure if this issue is ready for a beta uplift because it appears there are still some isssues left to sort out regarding this functionality, signaled in bug 1495443 and bug 1495743.

Comment 22

[Pascal Chevrel:pascalc (PTO until April 26)]

Since bug 1495443 is stalled for now and we are in our last week of betas, this is wontfix for 63.

Comment 23

[miralobontiu]

Verified as fixed on Firefox 64.0.2.