[pulseaudio-discuss] [PATCH next v1 2/7] bluetooth: Add state to transport objects

[Mikel Astiz]

Hi Tanu,

On Tue, Dec 11, 2012 at 5:40 AM, Tanu Kaskinen <tanuk at iki.fi> wrote:
> On Mon, 2012-12-10 at 08:30 +0100, Mikel Astiz wrote:
>> +typedef enum pa_bluetooth_transport_state {
>> +    PA_BLUETOOTH_TRANSPORT_STATE_DISCONNECTED,
>> +    PA_BLUETOOTH_TRANSPORT_STATE_IDLE, /* Connected but not playing */
>
> Why "idle" instead of "connected"? "Connected" would be more consistent
> in my opinion. Is the BlueZ 5 interface going to call this state "idle"?

Yes, the reason was to be consistent with the terminology in BlueZ 5.

Having said that, I'm not 100% convinced since I didn't actually
follow BlueZ 5 by adding PA_BLUETOOTH_TRANSPORT_STATE_PLAYING, as
opposed to two different states exposed in BlueZ 5: "pending" (audio
playing but transport not acquired) and "active" (audio playing and
transport acquired). The reason to merge these two inside PA was that
you can't actually have such a separation in BlueZ 4.

Therefore, I would also be fine with
PA_BLUETOOTH_TRANSPORT_STATE_CONNECTED instead of _IDLE. I'm slightly
in favor of the latter because it would eventually make it easier to
support _PENDING, but this sounds nonsense while we have the support
for BlueZ 4.

Cheers,
Mikel