Re: <device> advice

On Wed, 16 Dec 2009, Matt May wrote:
> On Dec 15, 2009, at 3:47 PM, Ian Hickson wrote:
> > Obviously I have considered simple things like making sure it can be 
> > keyboard accessible and ensuring it is media-independant, but if there are 
> > specific aspects of the design that you think should be improved, I would 
> > be very grateful if you could let me know.
> 
> I don't see anything about prompting the user to allow access to the 
> device.

The UI would be up to the UA, but the expected behaviour is described 
here:

   http://lists.w3.org/Archives/Public/public-device-apis/2009Dec/0194.html

(See also the rest of that thread:

   http://lists.w3.org/Archives/Public/public-device-apis/2009Dec/thread.html#msg149

...)


> Considering the consequences of allowing a document unfettered access to 
> a user's own devices (particularly storage and capture devices), I have 
> to assume that some form of user access control will also be specified. 
> In such a scenario, as has been discussed in similar geolocation 
> discussions, those controls MUST be directly accessible. Additionally, 
> notification of those devices being accessed (e.g., a webcam being 
> turned on) MUST be relayed, and the device's current status SHOULD be 
> available at all times (if that info is available to other users, then 
> this would be a MUST as well). Otherwise, someone who can't see the 
> screen would never know whether a given device is in use.

Thanks, I'll make sure to include this in the proposal if it goes any 
further.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Wednesday, 16 December 2009 19:38:06 UTC