Concluding on Availability, Enumeration, Multiple Displays

Hi all,

thanks for your active contributions to the discussion on the new proposal we brought in.

I would like to summarise and ideally conclude the discussion on the requirements with regards to display availability, enumerability, multiple display support and naming.

>From Google’s, Fraunhofer FOKUS’ and Intel’s feedback:

* In terms of availability querying: There is no need for enumerating multiple displays on the script side, i.e. the API does not need to have a concept of a display list. A binary yes/no notion of “one or more displays are available” is sufficient.

* There is however a requirement for identifying/selecting one of multiple screens on the UA side. That means, when a page makes a request to present, the UA should present a selection of possible destinations screens.

* Receiving human readable names on the script side, for example after the user has selected display from the UA selection dialog, is optional, but not strictly required for any of the use cases. 

* Google suggests not to associate and store usage permissions per particular device, but instead perform device selection for each case, since the list of available devices is dynamic.

* We might want to add a way of showing “synthetic” or “out-band” devices in such a selection dialog (cmp. Youtube.com/tv). There are concerns though with regards to spoofing attacks when providing such information from content to a native UA selection dialog.  

Next steps:
* We will update the proposal on the Wiki to model only binary availability, not a list of displays.
* Adapt the rest of the API accordingly.
* Describe ideas for permission prompting in more detail.

Are there any concerns with moving forward as outlined above?

Dominik

Received on Friday, 7 February 2014 13:06:33 UTC