search

martok / palefill

Inject Polyfills for various web technologies into pages requiring them
https://martok.github.io/palefill/
Mozilla Public License 2.0
79 stars 9 forks source link

performance.mark function #28

Closed biopsin closed 2 years ago

biopsin commented 2 years ago

22-06-27--1656333906_scrot

Just started to happen even after updating plug to 1.14, I must hit F5 to update the page.

Pale Moon 31.1.0 - linux5.4

Vangelis66 commented 2 years ago

I must hit F5 to update the page.

Relevant upstream issue:

https://github.com/JustOff/github-wc-polyfill/issues/68

Blocking manuelpuyol_turbo scripts is a workaround for me 😉 ... YMMV...

biopsin commented 2 years ago

Seems to be resolved, can I close this?

martok commented 2 years ago

Isn't resolved for me.

Also it should work, as far as I can see. The Performance global exists and has methods on its prototype, but they can't be called for some reason. Investigating...

Different problem. hm. Yeah, different. Closing this here and will open a new issue...

biopsin commented 2 years ago

I spoke to soon, thing is if I enable webcomponents in Pale Moon (for finn.no), I dont experience the issue on github. But as webcomponents is incomplete its also disabled by default, so Im in a limbo state at this time.

Edit: nope enabling webcomponents breaks other stuff on github then this issue, sorry but please reopen.

martok commented 2 years ago

Are you currently seeing the performance.mark errors as in this report, or "parent is null"? If the latter, then that's #30 ...

biopsin commented 2 years ago

Yes, I see it in this report. I also see the performance.mark errors while enabling webcomponents in about:config, but I dont need to refresh the page like when disabled.

martok commented 2 years ago

There is clearly something weird going on, but I don't see any errors or warnings anywhere...

Can you retest with 1.15 and let me know how it goes? There might be some A/B testing going on that has me on a different feature flag...

This might actually be related to #30 - maybe some events just get misdirected. I currently have no idea how to properly fix it because I don't understand what is broken in the first place.

biopsin commented 2 years ago

I loaded up a fresh profile and the performance.mark errors are gone, so the issue is local. However v.1.15 resolves the extra reload one have to perform so Im closing this. Thank you for you time!

martok commented 2 years ago

I loaded up a fresh profile and the performance.mark errors are gone, so the issue is local.

Sounds like some A/B testing... maybe clearing local data for github on your main profile helps?

biopsin commented 2 years ago

@martok I also suspect there are some config settings that interfere here, but Im unable to pinpoint which at present time. A fresh userprofile seems to also have mitigated some other sitebugs I noticed on the old profile.