Closed sefeng211 closed 1 year ago
@noamr - can you take a look? Did that get lost in the refactor?
Ah okay, I am wrong.
I think I got lost while comparing the Firefox implementation and the spec. Apparently Firefox queues the entry when we are adding it to the buffer, which doesn't align the current specification. So I thought the spec would be the same.
Sorry about this!
I am just glancing through the spec. One thing that I am confused is that I don't see the spec says to
Queue
the entry when the entry is added to the resource timing buffer. Or rather, the spec doesn't specify when an entry should be queued to the observers.For instance, in the add a PerformanceResourceTiming entry step, shouldn't we say
queue the entry
after theAdd new entry to the performance entry buffer
sub-step?Another place is in the copy secondary buffer step, do we need to say
Queue the entry
after theAdd entry to the end of performance entry buffer
step?I see the mark resource timing step mentions this, however it's not referenced anywhere in the spec.