Closed Bug 1016698 Opened 10 years ago Closed 6 years ago

[Settings] When changing APN settings, data roaming is not enabled automatically

Categories

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

ARM
Gonk (Firefox OS)
defect
Not set
major

Tracking

(Not tracked)

RESOLVED WONTFIX

People

(Reporter: promise09th, Unassigned)

References

Details

1. Title : When changing APN settings, data roaming is not enabled automatically
2. Precondition : Data connection and data roaming are enabled
3. Tester's Action : Settings >> Celluar & Data >> Data settings >> Change APN settings
4. Detailed Symptom : Data connection and data roaming are disabled. After 2.5s, data connection is enabled automatically, but data roaming is not enabled automatically
5. Expected : Data roaming is enabled also
6. Reproducibility: Y
Wayne! Please check this issue.
blocking-b2g: --- → 1.4?
Flags: needinfo?(wchang)
I don't see this as a problem - changing APN may cause unexpected roaming charges so having the user to explicitly enable roaming again may seem reasonable.

Arthur, can you check if this is a bug?


Mike can you comment from UX perspective?

Anyway I don't see this as a severe problem to block 1.4. We'll see UX and dev feedback to nominate appropriately for 2.0/2.1 if required.
blocking-b2g: 1.4? → ---
Flags: needinfo?(wchang)
Flags: needinfo?(mtsai)
Flags: needinfo?(arthur.chen)
Flags: needinfo?(ofeng)
Flags: needinfo?(mtsai)
Flags: needinfo?(jelee)
From UX perspective, changing APN settings shouldn't affect the original status of data connection/roaming, as the user might not notice the action (change APN setting) could cause other settings to change too (data roaming get turned off). If data connection/roaming are enabled before the APN settings change, they should remain enabled after the change; if data connection/roaming are disabled before the APN settings change, they should remain disabled after the change.
Flags: needinfo?(ofeng)
Flags: needinfo?(jelee)
Jose, when changing APN settings, is the change of ril.data.roaming_enabled expected?
Flags: needinfo?(arthur.chen) → needinfo?(josea.olivera)
(In reply to Arthur Chen [:arthurcc] from comment #4)
> Jose, when changing APN settings, is the change of ril.data.roaming_enabled
> expected?

Yes, it is and we force that. Let me explain myself, we reconnect the data call every time the user clicks on the OK button in the APN list panel and every time we reconnect the data call we disable data roaming. This is just for safety. I see the UX feedback for it to keep the state of the roaming setting as It was before but on the other hand Wayne's feedback is the opposite, so what's the decision here?
Flags: needinfo?(josea.olivera)
(In reply to José Antonio Olivera Ortega [:jaoo] from comment #5)
> (In reply to Arthur Chen [:arthurcc] from comment #4)
> > Jose, when changing APN settings, is the change of ril.data.roaming_enabled
> > expected?
> 
> Yes, it is and we force that. Let me explain myself, we reconnect the data
> call every time the user clicks on the OK button in the APN list panel and
> every time we reconnect the data call we disable data roaming. This is just
> for safety. I see the UX feedback for it to keep the state of the roaming
> setting as It was before but on the other hand Wayne's feedback is the
> opposite, so what's the decision here?

Jose, by safety did you mean the same thing as Wayne suggested that there might be extra charges?
Flags: needinfo?(josea.olivera)
(In reply to Arthur Chen [:arthurcc] from comment #6)

> Jose, by safety did you mean the same thing as Wayne suggested that there
> might be extra charges?

Yes, I did. Safety meant avoiding extra charges.
Flags: needinfo?(josea.olivera)
Jenny, does the decision you made in comment 3 still holds considering the rationale of the original design suggested by Wayne and Jose? And maybe we could use a toast message or notification to inform users about the change.
Flags: needinfo?(jelee)
(In reply to Arthur Chen [:arthurcc] from comment #8)
> Jenny, does the decision you made in comment 3 still holds considering the
> rationale of the original design suggested by Wayne and Jose? And maybe we
> could use a toast message or notification to inform users about the change.

We still think it makes more sense that changing the APN setting shouldn't affect other settings, but as you pointed out, avoiding unintentional fee is important too. After user made changes to APN setting and tapped ok, show a dialogue box:
-------------
APN Settings 
-------------
Depending on your service agreement, changing APN setting may cost extra charges when you are in a roaming area. (this is draft message)
-----------
cancel/ok 

Tapping on cancel goes back to the previous screen, tapping ok will confirm the action.
Flags: needinfo?(jelee)
Hi Matej, I was wondering could you help review a new string introduced in comment 9? 

"Depending on your service agreement, changing APN setting may cost extra charges when you are in a roaming area."

The string exists in a dialog being prompted after users change the APN setting. Thanks a lot!
Flags: needinfo?(Mnovak)
I would simplify it a bit:

"Depending on your service agreement, changing your APN setting may result in extra charges when roaming."

Does that work?
Flags: needinfo?(Mnovak)
Assignee: nobody → thills
Please refer to bug 969298 attachment 8440607 [details] for latest wireframe update. Tks!

Hello Matej, I added a confirmation question to make it more logical, please review, thanks ;)! 

-----------------
Depending on your service
agreement, changing your APN
setting may result in extra
charges when roaming.
Do you want to proceed?

[cancel] [yes]
-----------------
Flags: needinfo?(Mnovak)
Hi Jenny,

I had a question about the latest wireframe.  In the initial description, the precondition was that roaming is enabled.  In the wirefame, on page 4 (1st Screen labeled "Cellular & Data Settings" it shows that roaming is disabled.  Do you still want the warning screen to prompt them about additional charges in the case where roaming is *disabled* to start off?

thanks,

-tamara
Flags: needinfo?(jelee)
(In reply to Jenny Lee from comment #13)
> Please refer to bug 969298 attachment 8440607 [details] for latest wireframe
> update. Tks!
> 
> Hello Matej, I added a confirmation question to make it more logical, please
> review, thanks ;)! 
> 
> -----------------
> Depending on your service
> agreement, changing your APN
> setting may result in extra
> charges when roaming.
> Do you want to proceed?
> 
> [cancel] [yes]
> -----------------

Looks great. Thanks.
Flags: needinfo?(Mnovak)
(In reply to Tamara Hills from comment #14)

Hello Tamara, if you look closely to the condition description for when should the warning message shows, you will find that only when "Data roaming is turned on and it is the current APN in use that’s being edited?" will the message appear. I realize it's a little confusing where the first screen is showing otherwise, I'll update this (but will not release until next time if we are making major changes). Tks!
Flags: needinfo?(jelee)
Hi Arthur,

Jenny's spec seems to have quite a few changes from the current flow in it.  Let me know if this should be tackled as part of this bug or if this falls into a feature flow.  I'm fine either way.

Thanks,

-tamara
Flags: needinfo?(arthur.chen)
Hi Tamara,

Let's simply add the dialog in this bug as you did in the patch because the other parts are subject to change. Note that if users click on "cancel", they should be navigated back to the apn settings page without the apn settings changed

Arthur
Flags: needinfo?(arthur.chen)
Hi Arthur,
Navigating back to the apn settings page is a bit tricky after they have hit submit.  The normal flow is for it go back to the previous panel (in this case it's #carrier).  When I popup the warning dialog, the default behavior flips it in the background back to the carrier panel... this works fine if they hit "yes" on the warning (I just save their changes and restart the data connection and make sure roaming is enabled (if it was before)).  However, if they hit cancel, by that time, it's already displaying the carrier panel when we really want it to display the carrier-dataSettings panel when they cancel.  Looking for any hints on how to override the default behavior for onSubmit so that it doesn't go back to the previous panel before we have a chance to popup the dialog.  I don't see anything similar in the settings code.  Just wanted to check if there's something I'm missing here.
Flags: needinfo?(arthur.chen)
Indeed, there is no formal way to block navigating back to the previous panel. What I can think of now is to add an 'click' event listener on the parent of the submit button using capture mode (by passing true to the third parameter), then we have a chance to block the click event propagating to the button.

Let's use this workaround in this patch. Meanwhile, I'll also investigate how to improve the navigation system of settings app to have more flexibility on this kind of use cases.
Flags: needinfo?(arthur.chen)
Unassigning as I'm busy with dialer.
Assignee: thills → nobody
Firefox OS is not being worked on
Status: UNCONFIRMED → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.