W3C home > Mailing lists > Public > public-media-capture@w3.org > November 2012

Re: How to check if permission denied?

From: Martin Thomson <martin.thomson@gmail.com>
Date: Thu, 29 Nov 2012 09:50:03 -0800
Message-ID: <CABkgnnUHsJHpLw+utqXKoYZ=5tjoeouFpB73GHduyC6S_Fs92A@mail.gmail.com>
To: Stefan HÃ¥kansson LK <stefan.lk.hakansson@ericsson.com>
Cc: jonathan chetwynd <jay@peepo.com>, "public-media-capture@w3.org" <public-media-capture@w3.org>
On 29 November 2012 05:21, Stefan HÃ¥kansson LK
<stefan.lk.hakansson@ericsson.com> wrote:
> I am a bit frustrated about this situation. We could hide almost everything
> from the application to leak very little (but not get the best user
> experience), or we could expose a lot of data to the app and leak at lot
> (but get a better user experience). Then you hear two extremes to guide the
> decisions: leak as little as ever possible; or this is a lost battle anyway
> so don't bother. Sigh.

Ultimately, it's just another trade-off we have to make.  I still hold
out hope for a compromise.

My current hope is that gUM provides one of the following two
extensions to the fingerprinting surface:

 (a) boolean flags for audio and video device availability
 (b) counts for audio and video devices

I tend to think that (a) is sufficient.  I harbour a small concern
that the distinction between user-facing and world-facing is something
that an application might want to learn, but I'd be willing to lose
that as long as that was something I could set in constraints.

Providing complex capabilities once consent is granted is (in my
opinion) sufficient to meet the needs of applications without adding
more fingerprinting bits.

--Martin
Received on Thursday, 29 November 2012 17:50:31 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:26:12 UTC