Hi @igrigorik, hi @jaffathecake, we encountered a big problem with HTTP/2 priorities and Service Workers: https://medium.baqend.com/chromes-service-workers-break-http-2-priorities-649c4e0fa930 … Since the option to pass priorities for fetch() are nothing for the near future, are there plans to improve Service Worker prio logic in Chrome?
-
-
Thanks for linking our paper in the bug. We indeed noticed this, but assumed this was known/because service workers weren't finished yet.
@slightlylate's rant on this is exactly what we found as well. Interested in the solution :)2 replies 0 retweets 2 likes -
Replying to @programmingart @EWErikWitt and
Two changes should happen (in decreasing order of likeliness): 1.) untouched `Request` objects should plumb their original priorities through when constructor-copied (maybe) 2.) real priority information should be exposed (and be settable) on `Request`
1 reply 0 retweets 1 like -
Replying to @slightlylate @programmingart and
@yoavweiss has been working on variants of the second for the past year or so. See:https://github.com/WICG/priority-hints/blob/master/EXPLAINER.md …2 replies 1 retweet 1 like -
Replying to @slightlylate @programmingart and
FWIW, in one of the last webperfwg meetings,
@toddreifsteck showed interest in explicitly defining priorities and their mapping (rather than, or on top of, "upgrade"/"downgrade" hints). So there may be multi-vendor interest in explicit definition(which we can then map to H2 prio)2 replies 0 retweets 5 likes -
Replying to @yoavweiss @programmingart and
I actually don't care about the mapping. Browsers are going to do what browsers are going to do on both sides of the fetch() API. This seems obvious?
1 reply 0 retweets 0 likes -
Replying to @slightlylate @yoavweiss and
Anyway, what's important is that within whatever limited window of expression browsers actually give a tab/SW/API, we need to be able to express the same range of options.
1 reply 0 retweets 2 likes -
Replying to @slightlylate @yoavweiss and
And the incoming priorities might not match across browsers...so what? And outgoing fetches might get rescheduled or bumped in some way because browser knows best...also, so what?
1 reply 0 retweets 1 like -
Replying to @slightlylate @programmingart and
The reason Todd wants mapping is to avoid devs aiming their priority instructions to work well with Chrome, resulting in faster loading experience there and slower experience elsewhere. The obvious problem there is that right now different browsers do very different things...
1 reply 0 retweets 1 like -
Replying to @yoavweiss @programmingart and
Sure, that's perhaps worth doing, but it doesn't need doing *first*.
1 reply 0 retweets 1 like
It's a pretty bad smell when the proposal to create a syntax for existing behavior to enable developers to control it is blocked on aligning all of that behavior first which, thanks to legacy browsers, won't actually happen in practice for 4+ years.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
& Web Standards TL; Blink API OWNER
Named PWAs w/
DMs open. Tweets my own; press@google.com for official comms.