- From: Harald Alvestrand <harald@alvestrand.no>
- Date: Fri, 06 Sep 2013 15:07:29 +0200
- To: robert@ocallahan.org
- CC: "public-media-capture@w3.org" <public-media-capture@w3.org>
- Message-ID: <5229D391.7090007@alvestrand.no>
On 09/06/2013 02:07 PM, Robert O'Callahan wrote: > On Fri, Sep 6, 2013 at 11:25 PM, Harald Alvestrand > <harald@alvestrand.no <mailto:harald@alvestrand.no>> wrote: > > Anne, your first link doesn't say that the approach was abandoned; > it shows that in the notifications case, you have also given up on > finding a better alternative. > > > With notifications it's especially hard to associate a permission > grant with a user action, because the whole point of notifications is > that they occur in the absence of a user action or even user > attention. I assume that's why they gave up trying to find a better > solution for notifications. > > It's unclear why the same would apply to "device access permissions". > > The second link is Roc's opinion; I respect that opinion, but > disagree with the conclusion. > > > :-) > > Besides, our feet have been solidly planted on the "ask for > permissions" path; if anything, this change helps alleviate the > popup overload pain. > > As an IETF friend said in another context: > > "This is not about starting down the slippery slope. It's how far > we slide into the muck at the bottom." > > I don't follow this logic. > To be a little more explicit: If prompting for permission to access the camera is bad, our current approach where we prompt for permission to access the camera is bad. If it's not too bad (or if we have nothing better to offer), adding the rest of the permissions that we need to have and leaving to the UA the task of explaining it to the user is not making life much worse for either the user or the UA implementor.
Received on Friday, 6 September 2013 13:07:59 UTC