DAP-ISSUE-121 (Claes): Scope of sensor API and sensor discovery [Sensor API]

DAP-ISSUE-121 (Claes): Scope of sensor API and sensor discovery [Sensor API]

http://www.w3.org/2009/dap/track/issues/121

Raised by: Claes Nilsson
On product: Sensor API

We must agree on the scope of sensor discovery. There are several options here:
1. Enable in-device sensors only.
2. Enable in-device sensors as well as locally connected, USB, BT, ANT+ etc, sensors. This would for example make it possible for web applications to access sensors in accessories and support sports and health related use cases.
3. Enable in-device sensors, locally connected sensors and sensors in local or global networks.

We also have to consider which level of sensor discovery that has to be built into the sensor API itself and what level of discovery that has to be relied 
upon an overlaying service discovery framework. 

The Webinos sensor API, http://dev.webinos.org/specifications/draft/sensors.html, relies on the Webinos service discovery API, http://dev.webinos.org/specifications/draft/servicediscovery.html, for finding sensors that could be in-device, locally connected or within a network. However, we also discussed that a separate sensor discovery method might be needed to make it possible for the sensor API to stand for itself. I am not sure on the approach here.

In Intel's/WAC's proposal as well as in the Webinos service discovery API the discovery function issues a callback for each sensor/service found. This is a flexible approach that allows the web application to build a selection list for the user or to select a found sensor/service according to the application needs. However, this has been discussed for the service discovery API proposals and the issue of privacy has been raised. The ability for web page scripts to scan the environment is a risk to privacy, and as such this must be subject to user control. An alternative to the above would be to hand the task of presenting the list of sensors and binding to a sensor selected by the user over to the browser.  The web page script would in essence be limited to defining the search parameters and requesting the browser to ask the user to make a choice and then the user agent would issue a success callback when the user has selected the sensor.

Received on Friday, 14 October 2011 13:40:26 UTC