3 ms·
Good point about extension code running while rendering is still blocked by the network. That should prevent any user-facing delay. I just ran a few quick tes
by mostlystatic 6y ago
Good point about extension code running while rendering is still blocked by the network. That should prevent any user-facing delay.
I just ran a few quick tests with the Top 50 extensions and an NYT article. The charts are ordered by largest median metric value.
https://gist.github.com/mattzeunert/de3c8aedd2936a34eeb88b62b985ff13#gistcomment-3343533 https://gist.github.com/mattzeunert/de3c8aedd2936a34eeb88b62...
The Largest Contentful Paint (LCP) chart suggests an interesting phenomenon. Extensions like Dashlane and Evernote appear to slightly increase when that paint happens. These extensions load a large amount of code, but don't block the initial render.
Grammarly does not seem to push up LCP much. Its JS code runs before the page starts rendering, so maybe it is less likely to compete with other code later on.
Normally running code before the page renders would be bad, but if the page is network-constrained at that point it might soften the impact.
- crazygringo 6y agoThanks. Very impressed not only that you took the effort to run those, but also at how quickly you responded and were able to run, collect and publish results! Major kudos.