Open Bug 1417294 Opened 7 years ago Updated 2 years ago

Make the timing channel model simpler after moving all interception logic to the parent process

Categories

(Core :: DOM: Service Workers, enhancement, P3)

enhancement

Tracking

()

People

(Reporter: tt, Unassigned)

References

(Blocks 1 open bug)

Details

After bug 1415388 and bug 1353798, if we want to collect all the timing data, we need to get both the intercepted channel (for getting service worker timing data) and the http channel (for getting the normal timing data). Besides, for ensuring to have all the service worker timing data on the intercepted channel, we have to listen to "service-worker-synthesized-response" to catch the channel and start to checking the data after getting "http-on-stop-request".

Anyway, it's too complicated. We should make it be able to collect all the timing data by catching only one channel. And, maybe it's the good time to do that after moving all interception logic to the parent process. We could do it by making sure to redirect the timing data to the redirect channel (Note 1). So that we can get all the timing data by having the redirect channel. 

Note 1: We do pour over recorded timing data to the redirect channel, but we don't redirect the new timing data to the redirect channel after the redirect happens.
Priority: -- → P3
Assignee: ttung → nobody
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.