+++ This bug was initially created as a clone of Bug #1643796 +++

I had Browser Console and Google Docs became extremely sluggish.

https://share.firefox.dev/2XFXeoV

getRepeatId is 20% of the sample, just generating keys using JSON.

Getting rid of the repeatId property and turning it into a function for checking if 2 messages are similar, instead of stringifying an object as we do now.
From profiling, this seems to reduce time spent for this in comparison to getRepeatId, but it doesn't really shows in DAMP: https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=64beeeccfb8900057dbd2d8ba35596451340ab8b&newProject=try&newRevision=747060063e6d8c4a25dad7d28cf83ea3a0661bf0&framework=12&page=1

Adding 10 000 messages, 5 times (so a total of 50 000 messages), to try to get some median number.

with my patch: https://share.firefox.dev/34BpyOP
without: https://share.firefox.dev/3sklM4E

so that's a 25% improvement, nice but not great

(In reply to Nicolas Chevobbe [:nchevobbe] from comment #2)

Adding 10 000 messages, 5 times (so a total of 50 000 messages), to try to get some median number.

with my patch: https://share.firefox.dev/34BpyOP
without: https://share.firefox.dev/3sklM4E

so that's a 25% improvement, nice but not great

So, I think my earliest patch wasn't optimal and was calling areMessagesSimilar too much.
In a new version of the patch, I'm seeing 10x better performance. I'll push that to review soon

https://hg.mozilla.org/mozilla-central/rev/31af70aae2fa