Re: [w3c/push-api] Allow firing pushsubscriptionchange on all permission changes (PR #400)

antosart left a comment (w3c/push-api#400)

> Why not mandate though if all browsers are going to do it?

It's not clear to me that we want to always mandate firing the event. There might be a lot of cases in which the permission state changes (part of this is vendor specific) and I'm not sure that the user agents always want to start a service worker. For example, if the permission is ungranted because of deleting site data (and yes that would also delete the service worker registration, but maybe afterwards) maybe the user agent wouldn't want to fire the event.

> Both Gecko and Blink support test_driver.set_permission() so this is not really true. (WebKit indeed doesn't support this, but having some test is better than having nothing)

Right, but in Blink this only sets overridden permissions for testing and does not integrate properly with the permissions implementation. It would be quite some work to fix that, but I agree this is something we should eventually work on.

(I realized this while trying to write WPTs. I can try to finish them up, and test they work at least in Gecko.)



-- 
Reply to this email directly or view it on GitHub:
https://github.com/w3c/push-api/pull/400#issuecomment-2820975703
You are receiving this because you are subscribed to this thread.

Message ID: <w3c/push-api/pull/400/c2820975703@github.com>