Connection data change into sim 2 when introducing a new one

Bug #1447114 reported by Noemí
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Canonical System Image
Confirmed
High
Bill Filler

Bug Description

Enviroment

Product: Krillin bq Aquaris E4.5 Ubuntu Edition
FW version: R21
HW version: MP
Material:

Description

Steps to Reproduce:
- Introduce a second sim on slot 2.
- Active airplane mode.
- Disable the airplane mode and enter the pin for both sims.

Actual Result:
Data connection is automatically established on sim 2 although it was set on sim 1.

Expected Result:
Keep the previous configuration.

Reproducibility: 100%

Tags: bq
Revision history for this message
John McAleely (john.mcaleely) wrote :

@jocave, please repro, and confirm is still a bug on vivid

tags: added: bq
Changed in barajas:
assignee: nobody → Jonathan Cave (jocave)
status: New → Incomplete
milestone: none → ww21-2015
Revision history for this message
Jonathan Cave (jocave) wrote :

Version: ubuntu-touch/stable/bq-aquaris.es r22 krillin

Steps to reproduce:
 * both SIMs pin locked - this seemed a requirement
 * flashed phone
 * booted with SIM in slot 1
 * allowed data connection to establish on first sim
 * enabled flight mode
 * inserted SIM in slot 2
 * disabled flight mode
 * unlocked SIM 1
 * unlocked SIM2

Data connection established on slot2

Testing with SIM present in slot 2 before SIM introduced to slot 1 left data connection on SIM 2.

Testing with vivid next...

Revision history for this message
Jonathan Cave (jocave) wrote :

Version: ubuntu-touch/rc-proposed/bq-aquaris.en r14 krillin
 + network-manager-0.9.10.0-4ubuntu15.1.1 (silo 028)

Repeating the above steps produces the same result, despite a slightly different set of actions as the SIM1 PIN does not need to be re-entered on turning off flight mode.

Changed in barajas:
status: Incomplete → Confirmed
assignee: Jonathan Cave (jocave) → John McAleely (john.mcaleely)
Changed in barajas:
assignee: John McAleely (john.mcaleely) → Alfonso Sanchez-Beato (alfonsosanchezbeato)
Revision history for this message
Alfonso Sanchez-Beato (alfonsosanchezbeato) wrote :

The current design of the system and ofono makes this very difficult to solve. In my opinion, to solve this properly we should make these changes:

1. ofono should stop to enable cellular data by default when a new SIM is detected

2. The set-up wizard which launches on first boot should let the user choose if she wants cellular data and which particular SIM to use for that. This has the advantage that we can switch that slot to 3G if needed at that moment.

3. ofono should send a (new) DBus signal when it is the first time a SIM has been used on a phone. A dialogue to ask the user whether he wants to use that SIM for data or not should open when this is detected.

This needs input from design.

Changed in barajas:
milestone: ww22-2015 → ww24-2015
Revision history for this message
John McAleely (john.mcaleely) wrote :

I'm not sure the changes need to be as wide ranging as you suggest. The bug is about a surprising change caused by the introduction of a *second* SIM.

I think the system should function as-is with one SIM (ie automatically select that SIM for data). I think the behaviour when you add a second sim should be refined.

In other words the rules of automatic selection/prompting the user are different when 1 or 2 sims are present.

What is the underlying reason that SIM2 is selected for data in this case? is it code that sees a new SIM and therefore selects it for data?

Revision history for this message
Noemí (noemi-gallego) wrote :

It also establishes data connection under SIM 2 when booting the device. Couldn't it be posible to have disabled de data connection until the user chooses which SIM he/she wants for it and then keep that option?

Revision history for this message
Alfonso Sanchez-Beato (alfonsosanchezbeato) wrote :

@John,

The system does not distinguish between slots: when a new SIM is detected, cellular data is automatically enabled regardless of the SIM slot. As we can have data enabled only for one slot in krillin, there is a race condition in case the previously inserted SIM had data enabled, and any of the two can end up with data enabled. Note, however, that settings are stored at that moment so provided that you do not change any of the SIMs the situation is stable.

There are technical issues with handling the situation differently depending on whether you have detected a SIM for the first time or not. The main one is when PIN is enabled. We store our settings per IMSI, which uniquely identifies a SIM. We cannot know the IMSI until the user enters the PIN, therefore we cannot know if the SIM has been inserted for the first time until that happens. Let's say that we had SIM1 in the phone, with cellular data enabled, and with PIN enabled. We insert SIM2, which does not have PIN enabled, and reboot. Should we enable data for SIM2 or not? We have no idea on whether SIM1 was there before or has been introduced at the same time, and we cannot wait for the user to enter the PIN to take a decision, as that might not even happen.

But it is not only the technical side. If we already decided that enabling SIM1 data was what the user wanted, why are we supposing things are different with SIM2? Maybe the user prefers SIM2 for SIM1 for data, for myriad of reasons. Why are we more wrong if we enable data for a new SIM when we have 2 in the phone than if we enable data for one SIM in case there is only one present?

We simply do not know what the user wants.

That's the reason for which I think we should not enable data by default and we must let the user choose. I do not see this different to when the user chooses a WiFi AP in the start wizard: we do not select a random open access point at that moment. We should do the same when there are 2 SIMs in the system, on first boot or when a new one appears. And it is also in the line of what Noemí has suggested in comment #6.

Changed in barajas:
assignee: Alfonso Sanchez-Beato (alfonsosanchezbeato) → John McAleely (john.mcaleely)
Revision history for this message
Tony Espy (awe) wrote :

@John

Note Alfonso's proposed solution ( introducing a new question in the wizard ) was NAK'd by design. If we intend to fix this, we will need to raise this with design again. Please see bug #1379412 for the details.

Changed in barajas:
milestone: ww28-2015 → ww34-2015
Revision history for this message
John McAleely (john.mcaleely) wrote :

I am working with JoeO to create a design brief for this area

Changed in barajas:
milestone: ww34-2015 → ww40-2015
Changed in barajas:
milestone: ww46-2015 → none
Revision history for this message
Julia Palandri (julia-palandri) wrote :

I tried to reproduce this with r26 and couldn't - can someone else check?

information type: Embargoed → Proprietary
affects: barajas → avila
information type: Proprietary → Public
Changed in canonical-devices-system-image:
assignee: nobody → Bill Filler (bfiller)
milestone: none → backlog
no longer affects: avila
Changed in canonical-devices-system-image:
status: New → Confirmed
importance: Undecided → High
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.