W3C home > Mailing lists > Public > public-web-bluetooth@w3.org > January 2016

Re: Questions for user gestures of |Bluetooth.requestDevice(...)| provided by UA

From: Jeffrey Yasskin <jyasskin@google.com>
Date: Wed, 13 Jan 2016 08:30:20 -0800
Message-ID: <CANh-dXmyT3RW06B7r_aYgBBJTb6wepWFTH=WcEdW9_7dkQvzZQ@mail.gmail.com>
To: Bruce Sun <brsun@mozilla.com>
Cc: public-web-bluetooth <public-web-bluetooth@w3.org>
The exact arrangement of prompts is up to the UA, but in Chrome we're
implementing a single bubble opened by requestDevice(). If the
Bluetooth adapter is enabled, the browser immediately starts scanning
for nearby devices and starts listing found ones in the bubble. If the
adapter is disabled, or something else prevents a scan (e.g. missing
Location permission on Android M+), the bubble explains the problem,
and gives the user a link to fix it. Once the user fixes the problem,
if the bubble is still open, the scan starts automatically.

We don't yet have a way to explain which services are being requested
(we'll need a registry of human-readable names for each service UUID),
so we just say "pair with" and link to an explanation that pairing can
give the site complete control of the device (which is unsatisfying,
but matches existing uses of "pair" in the BLE ecosystem).

If Mozilla wants to drive the creation of that registry, we'd almost
certainly adopt it. If you want to give the user the ability to strip
services from the filters+optionalServices before the scan starts, we
probably wouldn't follow you, but it's a reasonable UA decision within
the bounds of the spec.

Does that make sense?
Jeffrey

On Wed, Jan 13, 2016 at 2:40 AM, Bruce Sun <brsun@mozilla.com> wrote:
> Hi,
>
> Well...not exactly questions, more like double confirming the behavior of
> how UA interacts with users after |Bluetooth.requestDevice(...)| has been
> called.
>
> There are some wordings on the spec mentioning that a user gesture is
> required to reduce the frequency of scans some[1][2]; and UA must inform the
> user what capabilities the requested services can provide[3]; and a prompt
> is required to show the user the human-readable name of each device[3].
>
> So if I understand it correctly, UA ought to be able to provide at least two
> kinds of prompts:
>  - One is used to declare the usage of each service in the filter, and to
> provide an option at the same time for users to switch Bluetooth on to
> perform scan. The timing to show this prompt is right after
> |Bluetooth.requestDevice(...)| has been called but before the real scanning
> process begins.
>  - Another one is used to display all scanned devices after adapting the
> filter, and to let users choose one single device for further interaction.
> The timing to show this prompt is right after the real scanning process
> begins.
>
> Are the above two kinds of prompts exactly the same as what spec expects?
> Kindly help to share information if I misunderstand something or miss
> something. Any comments are appreciated.
>
> Best regards,
> Bruce Sun.
>
> [1] https://webbluetoothcg.github.io/web-bluetooth/#ua-bluetooth-address
> [2]
> https://github.com/WebBluetoothCG/web-bluetooth/issues/127#issuecomment-119698067
> [3]
> https://webbluetoothcg.github.io/web-bluetooth/#device-access-is-powerful
> [4]
> https://webbluetoothcg.github.io/web-bluetooth/#dom-bluetooth-requestdevice
Received on Wednesday, 13 January 2016 16:31:09 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 13 January 2016 16:31:09 UTC