I thought that we were clear. We agree with the proposal to remove the fast fail on constraint failure fir gUM, and disagree with Jim. I simply do not find the usability arguments convincing. I consider the mitigation of fingerprinting risk to be secondary to usability, but here there is a clear win for fingerprinting resilience and no substantial loss of function for applications. Legitimate ones anyhow. You can stipulate your requirements, the browser asks the user, you get access to media...or not. Simple. The only loss is in some of the corner case expressiveness in selection. Which is not substantially different to where we have been for around 2 years now. This is only a factor if you consider application-driven selection to be somehow more important than having a user choose. ...when there are multiple sources in play. I just don't see this as worth getting antsy over. On Nov 28, 2013 1:19 PM, "Silvia Pfeiffer" <silviapfeiffer1@gmail.com> wrote: > I am confused about these replies. Are you in favour of the bug proposal > or in favour of the quoted email which is in opposition to the proposal? > > Silvia. > On 28 Nov 2013 17:38, "Adam Bergkvist" <adam.bergkvist@ericsson.com> > wrote: > >> On 2013-11-27 18:47, Stefan Håkansson LK wrote: >> >>> On 27/11/13 18:30, Martin Thomson wrote: >>> >>>> On 27 November 2013 06:40, Jim Barnett <Jim.Barnett@genesyslab.com> >>>> wrote: >>>> >>>>> I am opposed to this. I think that it would complicate app >>>>> development and >>>>> make the user experience worse. >>>>> >>>> >>>> I am in favour of this (modulo a persistent grant of permissions). >>>> >>> >>> I'm in favor too. This is about privacy, and I want to be careful. >>> >> >> I'm also in favor of this (for reasons mentioned in a bunch of other >> mails). >> >> /Adam >> >>
Received on Friday, 29 November 2013 02:29:48 UTC