1148314 - [Bluetooth] Discoverable property will be true to false immediately while turn Bluetooth state from off to on.

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: Bluetooth, defect)

Description

[[e-mail], [e-mail]]

== Description ==
Discoverable property will be true to false immediately while turn Bluetooth state from off to on. Since I'm working on Bug 1144532, I find out the problem that we have to fix.

== STR ==
1. Turn Bluetooth on and set visibility to be enabled.
2. Turn Bluetooth off while visibility is enabled.
3. Turn Bluetooth off --> on.

== Expected ==
The visibility should be false.

== Actual ==
The visibility is changed from enabled to disabled immediately.

Comment 1

[[e-mail], [e-mail]]

Jocelyn, 

I have checked that the 'discoverable' property is changed to be false while I just turn Bluetooth state off. Under the condition and general use case, I think Gaia would not need to do adapter.setDiscoverable(false) before we do adapter.disable(). If there is any limitation, please feel free to let me know. Thanks.

Comment 2

[[e-mail], [e-mail]]

I put the bug to meta for note.

Comment 3

[Jocelyn Liu [:jocelyn] [:joliu]]

Hi Bruce,

Could you help to check on this?

Thanks,
Jocelyn

Comment 4

[Bruce Sun [:brsun]]

This is caused by the behavior of underlying bluedroid stack on flame. |bt_callbacks_t::adapter_properties_cb| is called with |BT_SCAN_MODE_CONNECTABLE_DISCOVERABLE| first and then with |BT_SCAN_MODE_CONNECTABLE| afterwards. I'm going to try other devices for references.

Comment 5

[Bruce Sun [:brsun]]

On nexus-4 (JB), |bt_callbacks_t::adapter_properties_cb| is called with |BT_SCAN_MODE_NONE| first and then with |BT_SCAN_MODE_CONNECTABLE| afterwards.

Comment 6

[Bruce Sun [:brsun]]

On nexus-5 (KK), |bt_callbacks_t::adapter_properties_cb| is called with |BT_SCAN_MODE_NONE| first and then with |BT_SCAN_MODE_CONNECTABLE| afterwards, which is the same as nexus-4 (JB).

Comment 7

[Bruce Sun [:brsun]]

The behavior of Bluetooth API v2 on flame is the same as that of Bluetooth API v1.

Comment 8

[Bruce Sun [:brsun]]

This issue is related to the underlying behavior of different Bluetooth backend. Gecko handles the status correctly. Un-assign myself from this bug.

Comment 9

[Jocelyn Liu [:jocelyn] [:joliu]]

No longer blocks APIv2 testing since it's POVB and the behavior is also the same as v1.

Comment 10

[BMO Automation]

Firefox OS is not being worked on