959414 - [B2G][Contacts] Cancelling out of editing or Updating a contact after importing displays broken UI

Status: VERIFIED FIXED

(Firefox OS Graveyard :: Gaia::Contacts, defect)

Description

[Tony Nguyen]

Attachment: screenshot

Description:
Cancelling out of editing or Updating a contact information promptly after importing a contact will have device display broken UI. 

Repro Steps:
1) Updated Buri to BuildID: 20140113004002
2) Navigate to Contacts App
3) Tap on the Gear icon to open Contact Settings
4) Import a contacts from Gmail or Outlook
5) Return to the contacts list 
6) Tap on an contact just added
7) Edit contact details by tapping on top right icon
8) Tap on any text field area, then tap on a few random text keys
9) Tap the "X" to cancel or "Update"

Actual:
Contact App has broken UI when the user tries to edit a contact promptly after importing a contact

Expected:
There are no UI issues in the contacts app

Environmental Variables:
Device: Buri v1.3 mozRIL
BuildID: 20140113004002
Gaia: b3fc4f712562ee92b0ed0bd17abc61be9a36a8da
Gecko: 5bb1837de7c0
Version: 28.0a2
Firmware Version: v1.2-device.cfg

Notes:

Repro frequency: 5/5 100%
See attached: screenshot, logcat

Comment 1

[Tony Nguyen]

Attachment: logcat

Comment 2

[Tony Nguyen]

This issue does not reproduce on the latest Buri 1.2 build.

Environmental Variables:
Device: Buri v1.2 comRIL
BuildID: 20140113004002
Gaia: 539a25e1887b902b8b25038c547048e691bd97f6
Gecko: e672faf1e6a1
Version: 26.0
RIL Version: 01.02.00.019.102
Firmware Version: v1.2-device.cfg

This issue does occur on the latest Buri Master M-C build but seems to be a different issue. Typing just a single character into any of the text fields results in the broken UI displaying. I think the two issues may be related and can write a separate issue if requested.  

Environmental Variables:
Device: Buri Master M-C
BuildID: 20140113040202
Gaia: fd3b9a97cb3c41cfa56be387b46a51db136b4422
Gecko: 12d3ba62a599
Version: 29.0a1
Firmware Version: v1.2-device.cfg

Comment 3

[J Zimbrick]

Regression Window:

Last Working Environmental Variables:
Device: Buri v1.3 Mozilla RIL
BuildID: 20140111004005
Gaia: 1f029aa4c924d6c59fd7e8c6d9786a7370755d49
Gecko: 89c68fcb1448
Version: 28.0a2
Base Image: V1.2-device.cfg

First Broken Environmental Variables:
Device: Buri v1.3 Mozilla RIL
BuildID: 20140112004002
Gaia: 1f029aa4c924d6c59fd7e8c6d9786a7370755d49
Gecko: c42fd6b6d1fc
Version: 28.0a2
Base Image: V1.2-device.cfg

Comment 5

[Francisco Jordano [:arcturus] [:francisco]]

I'm seeing this javascript error when trying it against current master:

E/GeckoConsole(  787): [JavaScript Error: "TypeError: keyboard.sendKey(...) is undefined" {file: "app://keyboard.gaiamobile.org/js/imes/latin/latin.js" line: 406}]

Jan is this familiar to you?

Comment 6

[Jan Jongboom [:janjongboom] (inactive)]

Happens when APZC is enabled. See bug 959198...

Comment 7

[Francisco Jordano [:arcturus] [:francisco]]

Per comment 6, I've disabled APZC in the settings menu and I'm not able to reproduce that error.

Asking for QA to verify if it's not happening with APZ.

Comment 8

[Sarah Parsons]

When APZC is enabled, the issue reproduced. When APZC is disabled, the issue does not reproduce. 

Gaia   a81ccdc53e45a6adeaae423e104e91bcc1e12b0e
SourceStamp 2c033140eff4
BuildID 20140117004005
Version 28.0a2

Comment 9

[Preeti Raghunath(:Preeti)]

Kats,

Can you please investigate?

Comment 10

[Kartikaya Gupta (email:[e-mail])]

Waiting for bug 959198 to be resolved first.

Comment 11

[David Scravaglieri [:scravag]]

Milan, Vivien, this blocks 1.3 CS.

Comment 12

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

(In reply to David Scravaglieri [:scravag] from comment #11)
> Milan, Vivien, this blocks 1.3 CS.

I'm going to land the patch for APZ to be enable everywhere today.

Comment 13

[Milan Sreckovic [:milan] (needinfo for best results)]

(In reply to David Scravaglieri [:scravag] from comment #11)
> Milan, Vivien, this blocks 1.3 CS.

Yes, see Comment 10.

Comment 15

[Tony Nguyen]

After bug 959198 has been resolved as fixed, this issue still reproduces so I am reopening this bug. 

Environmental Variables
Device: Buri v 1.3.0 Mozilla RIL
Build ID: 20140210004002
Gecko: https://hg.mozilla.org/releases/mozilla-b2g28_v1_3/rev/9c9382f433c0
Gaia: 5c8416fb1ea4a27f172ee6386ab3c19135448506
Platform Version: 28.0
Firmware Version: v1.2-device.cfg

Comment 16

[Milan Sreckovic [:milan] (needinfo for best results)]

Kats will check.

Comment 17

[Kartikaya Gupta (email:[e-mail])]

So I can reproduce an issue where I see the thing in the screenshot in comment 0, but it's only transient (stays for less than 1 second) and then repaints properly. Is that the entirety of the issue? Or are you seeing the "broken UI" persist?

Comment 18

[Kartikaya Gupta (email:[e-mail])]

FWIW it looks like when the keyboard is dismissed it shows some stuff that is sitting underneath. Sometimes it is the error page you see in the screenshot and sometimes it is just blank white. I see this both with and without APZ so I'm thinking this might be a contacts app bug. But I can look into it more and see what I find.

Comment 19

[Kartikaya Gupta (email:[e-mail])]

The error page from the screenshot comes from the <iframe id="fb-extensions"> and temporarily becomes visible when the keyboard is hidden. I presume this is because the content that's supposed to be showing doesn't resize fast enough. If I set this iframe to display:none in the app manager then I see something else instead of that error page (I guess whatever is underneath the iframe).

Comment 20

[Tony Nguyen]

Hi Kats, the issue is that the error as seen in the screenshot only appears briefly and then repaints properly.

Comment 21

[Kartikaya Gupta (email:[e-mail])]

Ok, thanks. And good job on grabbing a screenshot of that then! :) I would have found it pretty hard to time that just right.

For the record I don't think is a severe enough glitch to warrant status of "CS blocker" but I don't really know what criteria Mandyam is using for that so you guys can figure that out.

I'm going to try to figure out why the stuff on top (the contact editor and display sections) seem to be resizing slower than the iframes under it, which seems to be the underlying problem that causes the glitch.

Comment 22

[Kartikaya Gupta (email:[e-mail])]

Ok, so I think what's happening is this:

When you're in the "edit contacts" screen with the keyboard displayed, the window size doesn't include the keyboard area. So all the content is sized to fit in the top half of the screen. The <iframe id="fb-extensions"> has style properties "position:fixed; top: 100%" which means that it is located just below the content area (i.e. behind the keyboard). When the keyboard is dismissed but before the reflow happens, we paint the same content (with the smaller window size) but without the keyboard. Therefore the iframe hiding under the keyboard becomes visible. As soon as finish processing the resize and reflow, things go back to normal.

This seems like a problem with how the contacts app chooses to "hide" things offscreen (i.e. using "position:fixed;top:100%" instead of say "display:none" or "visibility:hidden" or even "position:fixed;top:-100%"). Since the underlying problem exists even without APZ I'm moving this bug to the contacts app. Somebody who knows the contacts code should look into this and find some other way to hide elements that they don't want to be visible.

Comment 23

[Kartikaya Gupta (email:[e-mail])]

Also note that there's a bunch of stuff stacked in the same spot as that iframe, so just setting display:none on the iframe alone doesn't fix the problem - I end up seeing whatever else is stacked under there.

Comment 24

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

Jose or Francisco, any chance that you can tweak how the iframe goes offscreen here in order to solve the issue ?

For example once the animation of the iframe is done, you can likely set the iframe visibility to hidden.

Comment 25

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

(In reply to Vivien Nicolas (:vingtetun) (:21) from comment #24)
> Jose or Francisco, any chance that you can tweak how the iframe goes
> offscreen here in order to solve the issue ?
> 
> For example once the animation of the iframe is done, you can likely set the
> iframe visibility to hidden.

Same for whatever frame is at this place.

Comment 26

[Jose Manuel Cantera]

will take care

Comment 27

[Cristian Rodriguez (:crdlc)]

Attachment: Patch v1 for 1.3

Comment 28

[Cristian Rodriguez (:crdlc)]

Comment on attachment 8376133 [details]
Patch v1 for 1.3

Jose is peer, Jordano's review is not needed, just one is enough

Comment 29

[Jose Manuel Cantera]

Comment on attachment 8376133 [details]
Patch v1 for 1.3

r+ for the patch that hides all the iframes and status content in a way that it will not be visible if there is a window resize.

Please note that this patch only solves the problem at contacts level but the issue (without seeing contacts hidden content) is still reproducible (see attached screenshot).

Comment 30

[Jose Manuel Cantera]

Comment on attachment 8376133 [details]
Patch v1 for 1.3

NOTE: Please see https://wiki.mozilla.org/Release_Management/B2G_Landing to better understand the B2G approval process and landings.

[Approval Request Comment]
[Bug caused by] (feature/regressing bug #):
[User impact] if declined: high, spureous content will be seen. low quality perceived
[Testing completed]: smoke test done
[Risk to taking this patch] (and alternatives if risky): low risk
[String changes made]:

Comment 31

[Jose Manuel Cantera]

Attachment: issue after applying contacts patch

while closing the edit screen if the keyboard was open the transition plays as it  the window were smaller.

Comment 32

[Jose Manuel Cantera]

Vivien,

The contacts issues regarging hiding auxiliary content have been solved. However, there is still the issue  described in comment #31 and its corresponding attachment.

thanks!

Comment 33

[Cristian Rodriguez (:crdlc)]

Attachment: Patch v1 master

It the same code but for master

Comment 34

[Jose Manuel Cantera]

Comment on attachment 8376208 [details]
Patch v1 master

thanks, Cristian. Please land this on master once you get a green Travis

Comment 35

[Cristian Rodriguez (:crdlc)]

ok, will do

Comment 36

[Vivien Nicolas (:vingtetun) (:21) - (NOT reading bugmails, needinfo? please)]

(In reply to Jose M. Cantera from comment #31)
> Created attachment 8376203 [details]
> issue after applying contacts patch
> 
> while closing the edit screen if the keyboard was open the transition plays
> as it  the window were smaller.

The screen is actually smaller. I have a plan to help here in bug 970093 but it will require some time.

I think we need to ask QA to reevaluate if the bug is still a blocker with the patch from Christian.

Comment 37

[Cristian Rodriguez (:crdlc)]

Merged in master:

https://github.com/crdlc/gaia/commit/7e9b4654b888bfdb93c6d4ac0829fcfdd65ccc59

Comment 38

[[:fabrice] Fabrice Desré]

Can someone from QA verify this before we decide on approval? thanks!

Comment 39

[Jason Smith [:jsmith]]

(In reply to Fabrice Desré [:fabrice] from comment #38)
> Can someone from QA verify this before we decide on approval? thanks!

Sure - I'll look at this on Tuesday.

Comment 40

[Jason Smith [:jsmith]]

(In reply to Jason Smith [:jsmith] from comment #39)
> (In reply to Fabrice Desré [:fabrice] from comment #38)
> > Can someone from QA verify this before we decide on approval? thanks!
> 
> Sure - I'll look at this on Tuesday.

Going to put needinfo on the reporter to retest this on master with the patch.

Comment 41

[Brogan Zumwalt [Inactive]]

This issue still occurs on latest 1.3

Reproduces with the original STR in comment 0, as well as when tapping on textbox while editing a contact, then deselecting same textbox. Both methods reveal same broken UI with shown in Attachment 8359514 [details]. Clearing needinfo after discussing with tnguyen. This appears to be same issue, should this bug be reopened, or should a new bug be filed?

Environmental Variables:
Device: Buri v1.3 Mozilla RIL
BuildID: 20140218004003
Gaia: df60e9b49207d64da5647ab15760c122adabfba4
Gecko: b5afe0b10e93
Version: 28.0
Firmware Version: v1.2-device.cfg

Comment 42

[Jason Smith [:jsmith]]

(In reply to Brogan Zumwalt from comment #41)
> This issue still occurs on latest 1.3
> 
> Reproduces with the original STR in comment 0, as well as when tapping on
> textbox while editing a contact, then deselecting same textbox. Both methods
> reveal same broken UI with shown in Attachment 8359514 [details]. Clearing
> needinfo after discussing with tnguyen. This appears to be same issue,
> should this bug be reopened, or should a new bug be filed?
> 
> Environmental Variables:
> Device: Buri v1.3 Mozilla RIL
> BuildID: 20140218004003
> Gaia: df60e9b49207d64da5647ab15760c122adabfba4
> Gecko: b5afe0b10e93
> Version: 28.0
> Firmware Version: v1.2-device.cfg

We should backout & reopen the bug.

Cristian - Can you backout & reopen?

Comment 43

[Jason Smith [:jsmith]]

Comment on attachment 8376133 [details]
Patch v1 for 1.3

Marking a- as the patch here fails to solve the problem in the bug.

Comment 44

[Cristian Rodriguez (:crdlc)]

But if the code is not landed on 1.3 yet. According to my managers we need a approval‑gaia‑v1.3+ to land something in v1.3. For this reason I didn't push the fix :)

Comment 45

[Jason Smith [:jsmith]]

Oh wait. I understand now. The testing done above was wrong - it was done on 1.3. Sorry about that. I'll reflag again for this to be tested on master.

Comment 46

[Jason Smith [:jsmith]]

Comment on attachment 8376133 [details]
Patch v1 for 1.3

Resetting approval flag since the testing done here was done incorrectly.

Comment 47

[Brogan Zumwalt [Inactive]]

Issue described in comment 31 still occurs in latest 1.4 master following STR from comment 0

Result: Animated transition when closing edit contact screen behaves as though window is smaller than it actually is.

Video showing issue: http://www.youtube.com/watch?v=-RJkwq-RcAY

Environmental Variables:
Device: Buri v1.4 Master Mozilla RIL
BuildID: 20140219040204
Gaia: ac06cfbd2baf6494ffbb668cc599e3892cd5e17b
Gecko: bf0e76f2a7d4
Version: 30.0a1
Firmware Version: v1.2-device.cfg

Comment 49

[[:fabrice] Fabrice Desré]

Should this be re-opened?

Comment 50

[Kartikaya Gupta (email:[e-mail])]

The issue in comment 31 is expected, as Vivien said in comment 36. There's no easy way to fix that without implementing a floating keyboard.

If that's the only issue left then I think the fix can be considered verified and uplifted.

Comment 51

[Jason Smith [:jsmith]]

(In reply to Kartikaya Gupta (email:[e-mail]) from comment #50)
> The issue in comment 31 is expected, as Vivien said in comment 36. There's
> no easy way to fix that without implementing a floating keyboard.
> 
> If that's the only issue left then I think the fix can be considered
> verified and uplifted.

Okay - sounds okay to me then.

Comment 52

[Cristian Rodriguez (:crdlc)]

Sorry guys, it is not been uplifted to v1.3 because we don't have the approval

Comment 53

[Noemí Freire (:noemi)]

Fabrice, could please help us with the approval here?. Thanks!

Comment 54

[Ryan VanderMeulen [:RyanVM]]

v1.3: 34c45c7405ffd4846a49af4ba4927042008ef5e1

Comment 55

[Loli (:lolimartinezcr)]

Tested 02/24/2014 and working
1.3
Gecko cf974e3
Gaia 5084b83