1662252 - eightsleep.com is much slower (7-10s) to paint in Firefox than Chrome (1-3s)

Status: RESOLVED WORKSFORME

(Core :: Audio/Video: Playback, defect)

Description

[:Gijs (he/him)]

Profile: https://share.firefox.dev/3hOxq1s

From a cursory look at least some of it seems to be media isTypeSupported calls tripping driver initialization in the content process which loads DLLs which is slow. But that alone doesn't seem to be all of it.

Comment 1

[Daniel Holbert [:dholbert]]

I'm not sure this is actually a perf issue; it might be more of a webcompat/tracking-protection issue.

The site has this CSS:

body {
[...]
    opacity: 0;

...and then they dynamically add opacity:1 at some point during pageload, via a GoogleTagManager snippet I think (?) based on skimming the very top of their view-source.

I think this may be a version of bug 1516552.

It's possible that the shimming work tracked in bug 1648863 will help here.

Comment 2

[BugBot [:suhaib / :marco/ :calixte]]

Bugbug thinks this bug should belong to this component, but please revert this change in case of error.

Comment 3

[Daniel Holbert [:dholbert]]

(In reply to Daniel Holbert [:dholbert] from comment #1)

I'm not sure this is actually a perf issue; it might be more of a webcompat/tracking-protection issue.

(Actually: it's hard to say, but it looks like the opacity:1 tweak might happen pretty early on -- so that might be a bit of a red herring, and/or there might be both tracking-protection-issues as well as real perf issues going on here.)

Comment 4

[Chris Peterson [:cpeterson]]

On my Windows laptop, loading takes 4+ seconds in Firefox 80, 81, and 82 but only about 1 second in Chrome and Edge.

I don't think this related to tracking protection because the page load still takes 4+ seconds whether ETP is strict, standard, or disabled.

Comment 5

[Jim Mathies [:jimm]]

Hey Gjis, curious if you could re-generate? (based on comment 1) might be fixed.

Comment 6

[:Gijs (he/him)]

Yeah, I can't reproduce this anymore.