Re: gUM and persistent permissions

Le mardi 29 avril 2014 à 09:29 +0200, Harald Alvestrand a écrit :
> Should we repeat the requirement here in the W3C document?

My take is that the W3C document should have something on the topic;
based on my experience with other specs, I would not make it a normative
requirement — the way user agents manage long term permissions has been
traditionally left to implementors to determine and innovate on. More
specifically, I don't think there are expectations of interoperability
on how persistent permissions are managed, so we should not try to spec
it up.

We should definitely call out the risks associated with persistent
permissions on HTTP content, but I would not make it a "MUST", simply a
statement of fact in the privacy & security considerations.

Dom

Received on Tuesday, 29 April 2014 15:38:43 UTC