1208807 - Selecting timezone and reselecting set automatically in settings shows "UTC±offset" instead of the timezone name

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: General, defect, P3)

Description

[Zibi Braniecki [:zbraniecki][:gandalf]]

When I select timezone America/New_York and then switch to "Set Automatically" the Settings UI shows "UTC-07:00" instead of the name.

When I deselect "Set Automatically", then select America/Los_Angeles and then "Set Automatically" again the Time Zone box is empty.

Comment 1

[Zibi Braniecki [:zbraniecki][:gandalf]]

The device is Z3-kk, master from yesterday

Comment 2

[Zibi Braniecki [:zbraniecki][:gandalf]]

Attachment: America/Los_Angeles behavior

Comment 3

[Zibi Braniecki [:zbraniecki][:gandalf]]

Attachment: America/New_York behavior

Comment 4

[Zibi Braniecki [:zbraniecki][:gandalf]]

I believe it is a regression

Comment 5

[Dylan Oliver [:doliver]]

[Blocking Requested - why for this release]:

Comment 6

[Mahendra Potharaju [:mahe]]

Blocking for 2.5 with a P2 priority.

Comment 7

[[e-mail]]

ask for regression-window

Comment 8

[Jayme Mercado [:JMercado] (Inactive 3/4/2016)]

Let's get branch checks first to verify this actually is a regression.

Comment 9

[Max Murrell [MMurrell]]

Able to repro on Aries 2.5, Flame 3.0, and Flame 2.5.  Flame 2.0 repros halfway. Deselecting "Set Automatically" after selecting America/New_York causes the timezone to change to "UTC-07:00". Deselect "Set Automatically", after selecting America/Los_Angeles does not cause the timezone to change.

Repros on the following builds
Device: Aries 2.5
BuildID: 20151009104130
Gaia: 1609aecaba381c14eff95d5084e59280f53b7d8b
Gecko: d01dd42e654b8735d86f9e7c723cc869a3b56798
Gonk: 2916e2368074b5383c80bf5a0fba3fc83ba310bd
Version: 44.0a1 (2.5) 
Firmware Version: D5803_23.1.A.1.28_NCB.ftf


Environmental Variables:
Device: Flame 2.5
BuildID: 20151009030234
Gaia: 1609aecaba381c14eff95d5084e59280f53b7d8b
Gecko: d01dd42e654b8735d86f9e7c723cc869a3b56798
Gonk: a4f6f31d1fe213ac935ca8ede7d05e47324101a4
Version: 44.0a1 (2.5) 
Firmware Version: v18D


Environmental Variables:
Device: Flame 2.2
BuildID: 20151006032504
Gaia: 5dd95cfb9f1d6501ce0e34414596ef3dd9c2f583
Gecko: fc588eb28eab
Gonk: bd9cb3af2a0354577a6903917bc826489050b40d
Version: 37.0 (2.2) 
Firmware Version: v18D


Partially repros on the following build

Environmental Variables:
Device: Flame 2.0
BuildID: 20150723000207
Gaia: b16ba05481e577bc644ed8966f587a70fe2148e6
Gecko: 2e6f1d4deff9
Gonk: bd9cb3af2a0354577a6903917bc826489050b40d
Version: 32.0 (2.0) 
Firmware Version: v18D

Comment 10

[Jayme Mercado [:JMercado] (Inactive 3/4/2016)]

Zibi this issue seems like two separate issues.  One is a regression (blank after selecting Los Angeles and resetting Set Automatically) the other is not.  I'll leave regressionwindow-wanted for now but should this bug be split into two separate issues?  If so please set regressionwindow-wanted on whichever one is for the blank portion.

Comment 11

[[e-mail]]

I think 1. select timezone America/New_York and then switch to "Set Automatically" the Settings UI shows "UTC-07:00" instead of the name is a limitation, because of we cant keep the mapping between geolocation and timezone (for all UTC-07 timezone, it might be at New York or other place).

Comment 12

[Johan Lorenzo [:jlorenzo]]

FWIW, I managed to reproduce just by changing of timezone. Both of my phone are always on "Set automatically", so I didn't tap anything on the UI. Hence, I wonder if it's not an underlying gecko issue. 

I tried to get a logcat, but I didn't get anything interesting, except that the timezone is correct in the logs. I manage to switch from UTC+2 to UTC-7 correctly:
> 10-13 17:37:14.063   321   321 I Gecko   : AlarmService: _onSystemClockChanged
> 10-13 17:37:14.063   321   321 I Gecko   : AlarmService: _restoreAlarmsFromDb()
> 10-13 17:37:14.063   321   321 I Gecko   : -*- TimeKeepService: Received time update: 1444750634064
> 10-13 17:37:14.063   321   321 I Gecko   : -*- TimeKeepService: Reading epoch from /sys/class/rtc/rtc0/since_epoch
> 10-13 17:37:14.063   321   321 I Gecko   : -*- TimeKeepService: Opening /sys/class/rtc/rtc0/since_epoch
> 10-13 17:37:14.063   321   321 I GeckoDump: [system] [TimeCore][41.909] handling moztimechange
> 10-13 17:37:14.163   321   321 I Gecko   : AlarmService: Callback after getting alarms from database: [{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":6},{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":7},{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":8}]
> 10-13 17:37:14.163   321   321 I Gecko   : AlarmService: Current alarm: {"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":6}
> 10-13 17:37:14.163   321   321 I Gecko   : AlarmService: Alarm queue: [{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":7},{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":8}]
> 10-13 08:37:14.213   321   321 I Gecko   : AlarmService: _onTimezoneChanged()
> 10-13 08:37:14.213   321   321 I Gecko   : AlarmService: _restoreAlarmsFromDb()
> 10-13 08:37:14.223   321   321 I GeckoDump: [system] [TimeCore][42.068] handling moztimechange
> 10-13 08:37:14.313   321   321 I Gecko   : AlarmService: Callback after getting alarms from database: [{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":6},{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":7},{"date":"2015-10-31T23:00:00.000Z","ignoreTimezone":true,"data":{"type":"nextReset"},"pageURL":"app://costcontrol.gaiamobile.org/message_handler.html","manifestURL":"app://costcontrol.gaiamobile.org/manifest.webapp","timezoneOffset":-120,"id":8}]


Mahe, can we get this bug raised as P1? This timezone problem prevents me from checking bug 1180257, which is a P1.

Comment 13

[Sxean Lee-David [:sxean]]

B2G inbound: 

Last Working: 
BuildID: 20150511192241
Environmental Variables:
Device: Flame 2.5
BuildID: 20150511192241
Gaia: e6d21b2e3d9d366dddb3c0f7912639cd41665779
Gecko: fa102b192965
Gonk: c4779d6da0f85894b1f78f0351b43f2949e8decd
Version: 40.0a1 (2.5) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0

First Broken: 
BuildID: 20150511211342
Environmental Variables:
Device: Flame 2.5
BuildID: 20150511211342
Gaia: ad61a0e4db38fdb42feca054725fc1f4f33e743a
Gecko: 30ca91012bf2
Gonk: c4779d6da0f85894b1f78f0351b43f2949e8decd
Version: 40.0a1 (2.5) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0


FBGaia.LWGecko
Device: Flame 2.5
BuildID: 20150511211342
Gaia: ad61a0e4db38fdb42feca054725fc1f4f33e743a
Gecko: fa102b192965

LWGaia.FBGecko
Device: Flame 2.5
BuildID: 20150511192241
Gaia: e6d21b2e3d9d366dddb3c0f7912639cd41665779
Gecko: 30ca91012bf2

Gaia Pushlog: 
https://github.com/mozilla-b2g/gaia/compare/e6d21b2e3d9d366dddb3c0f7912639cd41665779...ad61a0e4db38fdb42feca054725fc1f4f33e743a

This issue may have been caused by changes made by Bug 1162276

Comment 14

[Jayme Mercado [:JMercado] (Inactive 3/4/2016)]

Zibi or Johan, Arthur is marked as inactive.  Do you know who should look into the pushlog from comment 13?  Seems this is a fallout from bug 1162276

Comment 15

[Johan Lorenzo [:jlorenzo]]

Bug 1162276 introduced the box where UTC-7:00 is displayed. The blank area that happens after the second switch is already covered by bug 1169154.

We don't have a bug that tracks the use of "UTC-07:00", let's use the current one to track it. Removing the regression keyword as this bug wasn't shown until the box appeared in 2.2 and master.

Side note: There is also a bug showing the discrepancy of the time, when you change the timezone to another and then switch to "set automatically" again: bug 1178259.

Clearing Mahe's NI as I've been able to set my phone stuck on UTC back on PDT. Steps:
1. Turn off "set automatically"
2. Manually enter the timezone you're in
3. Turn on "set automatically".

Comment 16

[Ken Chang[:kenkai]]

Hi Fred,  According to comment 14, please see help to check.

Comment 17

[[e-mail]]

I can reproduce bug 1169154 (blank timezone) with STR described in comment 15, so I'll work on that bug instead. 

For the first issue, if the UTC±offset value is correct, I think its a limitation as described in comment 11.


johan could you point out what issue really blocks you? Is covered by bug 1169154 or bug 1178259?

Comment 18

[Johan Lorenzo [:jlorenzo]]

This sounds like a new feature, but Android manages to determine a cities thanks to the offset given in the NITZ packet and MCC and some other heuristics[1]. After looking more closely at our RIL[2], this is something that we can handle there. Moving the bug there.

[1] http://osxr.org/android/source/frameworks/base/telephony/java/com/android/internal/telephony/gsm/GsmServiceStateTracker.java#0897
[2] https://dxr.mozilla.org/mozilla-central/source/dom/system/gonk/RadioInterfaceLayer.js#929

Comment 19

[Johan Lorenzo [:jlorenzo]]

[Blocking Requested - why for this release]: Renom'ing this issue as this was initially filed as a regression. In comment 15, the issue has been split down into many smaller ones. In comment 18, the RIL might have to be modified in order to do the mapping between offset/mcc and the continent/city.

Comment 20

[[e-mail]]

Mapping between offset/mcc looks like a clever way for phone user. 
But for device without SIMCard, we will still need fallback with current "UTC±offset" string.

Comment 21

[Johan Lorenzo [:jlorenzo]]

I just checked the case where you don't have a SIM card (and you're connected to Wi-Fi). The UI doesn't show "UTC±offset" and I'm able to manually change the continent and the city. My guess is that we don't retrieve the timezone information if you don't have a SIM card. Is that right Fred?

Comment 22

[[e-mail]]

Yes, as I know timezone info is retrieving with auto time zone update protocol(NITZ) by operator, some of them doesn't support that protocol and for that case we'll let user to manually change the continent and the city.

For no SIM card case it for sure doesn't support NITZ so we also let user to manually change the continent and the city.


ni Hsinyi who might have more insight for the under hood behavior.

Comment 23

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Please see https://bugzilla.mozilla.org/show_bug.cgi?id=1209892#c19 for Samael's explanation.

I'd say this bug is more like a new feature that requires redesign on gaia, gecko, and gonk. We shouldn't block on it.

Please re-nominate if you don't agree.

Comment 24

[Hsin-Yi Tsai (she/her) [:hsinyi]]

(In reply to Fred Lin [:gasolin] (PTO) from comment #22)
> Yes, as I know timezone info is retrieving with auto time zone update
> protocol(NITZ) by operator, some of them doesn't support that protocol and
> for that case we'll let user to manually change the continent and the city.
> 
> For no SIM card case it for sure doesn't support NITZ so we also let user to
> manually change the continent and the city.
> 
>
>
> ni Hsinyi who might have more insight for the under hood behavior.

Fred is right!
If  there's no sim card or user doesn't register on a network, then there's no NITZ coming from carrier.

Comment 25

[Ken Chang[:kenkai]]

According to comment 23, we need to have this feature eventually, move it to backlog.

Comment 26

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Moving  to General component as this is a feature touching UX, gaia, gecko and gonk. Or, which component is the best fit?

Comment 27

[Zibi Braniecki [:zbraniecki][:gandalf]]

Removing an old NI.

Comment 28

[BMO Automation]

Firefox OS is not being worked on