W3C home > Mailing lists > Public > public-device-apis@w3.org > February 2010

Re: Sensors simplified (or not)

From: Max Froumentin <maxfro@opera.com>
Date: Thu, 11 Feb 2010 17:00:27 +0100
Message-ID: <4B74299B.5090401@opera.com>
To: "Nilsson, Claes1" <Claes1.Nilsson@sonyericsson.com>
CC: "public-device-apis@w3.org" <public-device-apis@w3.org>
Hi Claes,

On 11/02/2010 16:01, Nilsson, Claes1 wrote:
> Hi Max,
>
> I like this idea. It is more generic and provides possibilities for
> extensions. Assume we need some sensor discovery method?

To me, the best way would be to use the extensibility mechanism built 
into SysInfo. The specification defines a finite number of properties 
("CPU", short for "http://www.w3.org/2009/dap/sysinfo/CPU", "Battery", 
"AmbientNoise, etc.), among which the 5 sensor types currently in the 
specification.

If someone wishes to add an extra property not defined in the 
specification, then they use their own URIs, e.g.
http://bondi.omtp.org/2009/05/vocabulary/bluetoothVersion,
but without short names.

Therefore, there's no discovery. DAP currently defines 5 sensor APIs, 
which a conforming implementation must support if it has the 
corresponding sensors. Outside of that, it's not DAP's problem.

Adding a discovery mechanism on top wouldn't provide the semantics of 
discovered sensors. If my webapp discovers a list of sensors, how will 
it know which one is a sphygmomanometer?

Max.
Received on Thursday, 11 February 2010 16:01:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:14:05 GMT