Re: [RFC PATCHv2] autopair: add autopair plugin

Em Mon, 2012-01-30 =C3=A0s 14:35 -0800, Scott James Remnant escreveu:

On Mon, Jan 30, 2012 at 2:33 PM, Bastien Nocera [off-list ref] wro=

te:

quoted

Em Mon, 2012-01-30 =C3=A0s 13:42 -0800, Scott James Remnant escreveu:

quoted

quoted

The PIN length hint is also ignored in the current code:
+       <device type=3D"audio" oui=3D"00:1A:80:" name=3D"CMT-DH5BT=

"

quoted

quoted

pin=3D"max:4"/>

quoted

Because that's a UI restriction, no?

I didn't put any PIN-generation code here, because I'm not entirely
sure what the UI for "generate random PIN' should look like or which
spec you're following.

It means that a generated PIN shouldn't have more than 4 digits for i=

t

quoted

to work with this device.

What generated PIN? Which spec/profile says you have to generate a PIN
for this device?

It's recommended that the UIs try to generate PINs for the users, rather
than relying on them entering random numbers.

How does the user know what PIN has been generated?

You will display it to them.

Do you display the PIN in the UI in any way?

Yep, in the UI, because that's the PIN you will have to enter on the devi=
ce.

Are they expected to confirm the PIN with a remote display?

Not confirm, but enter.

Are they expected to type the PIN into the remote device? Or even the
host, perhaps it's displayed there?

They're supposed to enter it. The device is a Hi-Fi with Bluetooth
functionality:
http://www.sony.co.uk/support/en/product/CMT-DH5BT

You have a UI to enter the PIN but it will only allow you to enter 4
digits. So any number longer than 4 digits will fail to pair.

Or do these devices simply just accept any PIN without question? In
which case, why does it matter whether you send a random PIN or 0000?

It doesn't.