813854 - Add APN settings for MMS and SUPL

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::Settings, defect, P1)

Description

[Michael Vines [:m1] [:evilmachines]]

Modifications to the manual APN settings are completely ignored.  The APN from the built-in APN database for the given mcc/mnc always overrides.

It is a testing requirement that we be able to provide a manual APN, and I don't think we want to start adding APNs used by random test equipment to the APN database.

Comment 1

[Fabien Cazenave [:kaze]]

Works For Me (latest Gecko 18, Gaia, Otoro).

Michael, did you click on the "OK" button after modifying the APN settings? This is a very common mistake with these panels…

Comment 2

[Michael Vines [:m1] [:evilmachines]]

Hey kaze,

Just checked at the tip.  ril.data.apn is now getting set correctly based on the manual values but ril.supl.apn is not (neither is ril.mms.apn but I don't care much about that right now).  Same for ril.supl.{user,passwd}

Thanks,
Michael

Comment 3

[Fabien Cazenave [:kaze]]

Hi Michael,

I have probably misunderstood the UX requirements here. We currently have dialog box for the ril.data.* settings, I guess we need two other dialogs for ril.supl.* and ril.mms.* as well?

Larissa, your UX input would be appreciated here.

Comment 4

[Michael Vines [:m1] [:evilmachines]]

Yes, two other dialogs would probably be nice.   It's definitely possible that a different APN is needed for SUPL or MMS, not sure if this is a v1 requirement though (from the network POV).  If we decide not to add the other two dialogs then we should just make the SUPL and MMS APNs match whatever is manually set for "the" APN.

Comment 5

[Fabien Cazenave [:kaze]]

Just to clarify, here’s the current situation (or what it should be):

 * if the APN settings are unset, the device auto-selects the first APN settings that are compatible with the iccInfo (mcc, mnc) tuple — and that applies to ril.data.*, ril.supl.*, ril.mms.*;

 * we have a dialog box that allows to:
    • override the APN settings that have been automatically applied to ril.data.*;
    • select another APN when several entries match the same (mcc, mnc) tuple (e.g.: 208, 10).

(In reply to Michael Vines [:m1] from comment #4)
> Yes, two other dialogs would probably be nice.

OK, waiting for UX input on this.

> If we decide not to add the
> other two dialogs then we should just make the SUPL and MMS APNs match
> whatever is manually set for "the" APN.

This is the part I don’t understand.

A user would probably expect to select his operator in a single list matching the mcc/mnc tuple and getting these settings applied to data+supl+mms, but that’s not always possible — the “default” and “mms” carrier identifiers are often different.

If a user wants to manually override the data settings that have been taken from the APN database (i.e. not select another APN set but modify each field manually), these changes can’t be propagated to ril.supl.* or ril.mms.*.

So I guess I can’t do much better than proposing to manually select an APN set from a matching list for each type (data, supl, mms), but maybe I’m missing a point?

Comment 6

[Shian-Yow Wu [:swu]]

Agreed with comment 5.  We need two other dialogs for ril.supl.* and ril.mms.*, because they could be different from data APN.

For auto-select case, when user makes a choice, he should be able to see and modify the settings applied to data/mms/supl.

It means no matter auto or manual case, there should be three dialogs.

Comment 7

[Beatriz Rodríguez [:brg]]

(In reply to Shian-Yow Wu from comment #6)
> Agreed with comment 5.  We need two other dialogs for ril.supl.* and
> ril.mms.*, because they could be different from data APN.
> 
> For auto-select case, when user makes a choice, he should be able to see and
> modify the settings applied to data/mms/supl.
> 
Sorry but I do not understand it, do you mean that the user will be able to edit the settings that come from the automatic configuration? If this is the case, I disagree here. The settings loaded automatically should not be editable by final user. Did you mean it or I am wrong and I should file a new bug?

Comment 8

[Shian-Yow Wu [:swu]]

(In reply to Beatriz Rodríguez [:brg] from comment #7)
> Sorry but I do not understand it, do you mean that the user will be able to
> edit the settings that come from the automatic configuration? If this is the
> case, I disagree here. The settings loaded automatically should not be
> editable by final user. Did you mean it or I am wrong and I should file a
> new bug?

When user choose automatic, the data/mms/supl settings will be coming from APN database, but it's read only.  If user switch to manual, those settings still there and user can modify them.  If user switch back to automatic, the settings will be copied from APN database again.  This is what I mean, is there any concern on this way?

Maintaining manual settings separately (not been overwritten by automatic settings) is fine too.

Comment 9

[Beatriz Rodríguez [:brg]]

(In reply to Shian-Yow Wu from comment #8)
>
Thanks a lot for the explanation!

Comment 10

[Fabien Cazenave [:kaze]]

Rafa, Marco: could we talk about this please?

I need some UX input here, Larissa is overwhelmed and I think it shouldn’t take more than a few minutes to take a decision on this.

Comment 11

[Rafael Rebolleda]

Hey, 

I agree: we should have the option to manually input different APNs for different services. In my experience with a virtual carrier, I've had to manually entry the settings on most devices (iOS, WP, Android 2.X)

So:

* Automatic configuration should show non-editable values for each entry
* Manual configuration should provide blank inputs for each.

Comment 12

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

(In reply to Beatriz Rodríguez [:brg] from comment #7)
> (In reply to Shian-Yow Wu from comment #6)
> > Agreed with comment 5.  We need two other dialogs for ril.supl.* and
> > ril.mms.*, because they could be different from data APN.
> > 
> > For auto-select case, when user makes a choice, he should be able to see and
> > modify the settings applied to data/mms/supl.
> > 
> Sorry but I do not understand it, do you mean that the user will be able to
> edit the settings that come from the automatic configuration? If this is the
> case, I disagree here. The settings loaded automatically should not be
> editable by final user. Did you mean it or I am wrong and I should file a
> new bug?

Having them not being editable sounds like a new bug for me. This one is not about changing the current behavior.

Comment 13

[Beatriz Rodríguez [:brg]]

(In reply to Vivien Nicolas (:vingtetun) from comment #12)
> (In reply to Beatriz Rodríguez [:brg] from comment #7)
> > (In reply to Shian-Yow Wu from comment #6)
> > > Agreed with comment 5.  We need two other dialogs for ril.supl.* and
> > > ril.mms.*, because they could be different from data APN.
> > > 
> > > For auto-select case, when user makes a choice, he should be able to see and
> > > modify the settings applied to data/mms/supl.
> > > 
> > Sorry but I do not understand it, do you mean that the user will be able to
> > edit the settings that come from the automatic configuration? If this is the
> > case, I disagree here. The settings loaded automatically should not be
> > editable by final user. Did you mean it or I am wrong and I should file a
> > new bug?
> 
> Having them not being editable sounds like a new bug for me. This one is not
> about changing the current behavior.

Yesterday, I saw a nice proposal from Kaze. We liked it, for V1 is great. If we keep that behaviour, I will not modify it.

Comment 14

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

(In reply to Beatriz Rodríguez [:brg] from comment #13)
> Yesterday, I saw a nice proposal from Kaze. We liked it, for V1 is great. If
> we keep that behaviour, I will not modify it.

Sounds good to keep it!

Comment 15

[Fabien Cazenave [:kaze]]

Attachment: patch proposal

This patch adds an “A-GPS Settings” panel to support the SUPL settings. It also fixes the “MMS Settings” panel, though it’s still hidden in the UI for v1.
Evelyn, would you review it please?

Staś, I’ve removed the `apnSettings' l10n entity and added two new ones:
 • dataSettings = Data Settings
 • suplSettings = A-GPS Settings

Comment 16

[Evelyn Hung]

Comment on attachment 689010 [details]
patch proposal

r=me. the patch is good. 
Add Staś for late-l10n review.

Comment 17

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

Comment on attachment 689010 [details]
patch proposal

Sounds good for me. Strings removal / additions is not perfect but needed at this point and as long as the key is changed it should be fine.

Comment 18

[Fabien Cazenave [:kaze]]

https://github.com/mozilla-b2g/gaia/commit/2b436ac85a52ea98976a0ae7ffb2126a3cd05019

Comment 19

[Alex Keybl [:akeybl]]

Resolved on v1.0.0 and v1.0.1 already.

Comment 20

[Alex Keybl [:akeybl]]

(given 12/6 commit)

Comment 21

[Darren]

Cannot verify, need information on what to add to create a custom APN.