W3C WebPerf Call - March 31 2022




NotRestoredReason API for bfcache (presentation)


Unload beacon proposal - Fergal (presentation)


Chat Log

Michal Mocny11:02 AM


Yuzu Saijo11:11 AM


Benjamin De Kosnik11:13 AM

thanks Yuzu

Yuzu Saijo11:13 AM

slides: https://docs.google.com/presentation/d/1VWl1YEIzEkB4vQq3NgtqFc1_zez7e5dMo5bDT7vLGa0/edit#slide=id.g1209aebb9ab_0_33

Michal Mocny11:16 AM

"options" bucket like measures v3?

Timo Tijhof11:22 AM

privacy/security; not unlike what we have with event timing and input delay attribution etc.

element timing*

Katie Sylor-Miller11:24 AM

+1 to Fergal's point, it's important to note that a lot of times iframes are cross-origin and inserted via things like GTM and not under direct control

Katie Sylor-Miller11:27 AM


Michal Mocny11:45 AM

setTimeout(beacon.sendNow, 2000)

Michal Mocny11:48 AM

I like set, same as Map

Timo Tijhof11:48 AM

yeah, setData would be clearer than addData.

Michal Mocny11:49 AM

beacon.data = ... vs .setData(...) maybe

Noam Helfman11:49 AM

how about setDate(key, value)? This will allow simpler control of the state instead of managing it outside the beacon object.

Michal Mocny11:58 AM

(Or Noams idea, to just add one key.. though I'm not sure if beacons are a map like that)

Kyle Sharp11:59 AM

Have to drop. Super interesting discussions, thanks everybody!

Katie Sylor-Miller11:59 AM

we do usually send beacons as a key value store yes


Michal Mocny12:01 PM

expiry will be ergonomically useful even for the beacon receiver