Page MenuHomePhabricator

CheckUser: Design and implement UI toggles for ja3n/ja4h and client hints data
Open, Needs TriagePublic

Description

Summary

Feedback from users of Special:CheckUser is that the ja3n/ja4h and client hint data can take up too much space in the result rows. We should find a way to make this an option to hide, either per page, or indefinitely per user.

Design

TBD

Acceptance criteria

TBD

Event Timeline

Restricted Application added a subscriber: Aklapper. · View Herald Transcript
kostajh renamed this task from CheckUser: Implement toggles for ja3n/ja4h and client hints data to CheckUser: Design and implement UI toggles for ja3n/ja4h and client hints data.Wed, Jun 3, 6:03 AM
kostajh added a subscriber: KieranMcCann-WMF.

In individual log/edit rows, it would be helpful if the ja3n/ja4h data was on a new line. This would make it easier to compare fingerprints (because the same parts of the fingerprint would be in roughly the same horizontal location instead of being moved around based on the CH data) and make it easier to skim-read the CH line.

In the table at the top of the CU results, I'm not sure what the best solution is. Basically, the problem is: where before we only would have had a few different CH entries, now there's a new one for each individual CH+fingerprint combo. Example: https://checkuser.wikimedia.org/wiki/Zimidar. As far as I'm concerned, there's no meaningful difference between two identical CH lines with different ja3n/ja4h, so all this serves to do is blow up that column of the table. Should the fingerprints be separated from the CH line entirely, maybe? Or the fingerprints grouped by identical CH? Or some toggle that just hides the fingerprints from the page entirely?

I do find the first part of the ja4h useful, since it has the language data in it, but since the rest isn't human-readable (at least, not to me), it's mostly just getting in the way.