Re: [w3c/ServiceWorker] Add an option to include frozen documents. (#1442)

I spoke with Dave and I think I've been conflating things a bit.

Its true that we mark documents going into bfcache as frozen, but that is not adequate to describe bfcache behavior as a whole.  There is more work needed to fully describe bfcache.

In this case there is a semantic difference between a frozen client in a background tab the user hasn't visited in a week and a frozen client in bfcache.  The user can see the background tab exists and focus it.  Conceptually the client exists from the user's perspective.  In bfcache, though, they have navigated away from the frozen client, can't see it exists any more, and can't focus it without a history operation.

I think this means our previous discussions about bfcache still hold true, but that they should not apply to all frozen clients.  We need to spec the additional constraints on bfcache above and beyond the frozen state.

For the purposes of this PR I recommend adding a non-normative not indicating that, while clients in bfcache are frozen, they should not be exposed to clients.matchAll(), etc.  This can then be improved when bfcache is more fully spec'd.

WDYT?

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/w3c/ServiceWorker/pull/1442#issuecomment-506788719