#13575 closed defect (fixed)
Disable randomised Firefox HTTP cache decay user test groups
Reported by: | isis | Owned by: | tbb-team |
---|---|---|---|
Priority: | Medium | Milestone: | |
Component: | Applications/Tor Browser | Version: | |
Severity: | Normal | Keywords: | tbb-pref, tbb-fingerprinting, ff52-esr, tbb-easy, TorBrowserTeam201801R |
Cc: | isis | Actual Points: | |
Parent ID: | Points: | ||
Reviewer: | Sponsor: |
Description
I need to look into it more, but I think we should be setting browser.cache.frecency_experiment
to -1
to disable HTTP_CACHE_MISS_HALFLIFE_EXPERIMENT. Since we have Telemetry disabled, of course, Mozilla's experiment shouldn't be collecting any data on our users. However, if that pref is not set in firefox.js
to -1
, it will default to 0
. And if it's 0
, then it's randomised between 1
and 4
inclusive, setting different HTTP cache decay times for the four groups, which might make for a bit of a rough-edged fingerprinting mechanism.
Child Tickets
Attachments (2)
Change History (13)
comment:1 Changed 3 years ago by
Severity: | → Normal |
---|
comment:2 Changed 2 years ago by
Keywords: | ff52-esr added |
---|
We don't need any experiments active in Tor Browser.
comment:3 Changed 17 months ago by
Status: | new → needs_review |
---|
we should be setting
browser.cache.frecency_experiment
to-1
to fix this ticket without patching.
comment:4 follow-up: 5 Changed 17 months ago by
Status: | needs_review → assigned |
---|
Thanks for contributing. We use needs_review
for tickets that have patches attached we can reason about and merge if we think they are okay. I am happy to consider a patch to get this ticket moved forward. So, if you could attach one or point to some commit in a (preferably) git repository that would be great.
comment:5 Changed 17 months ago by
Keywords: | tbb-easy added |
---|
Replying to gk:
Thanks for contributing.
Thank you too. We made a great progress for the last several years. So, glad to hear it from you.
We use
needs_review
for tickets that have patches attached we can reason about and merge if we think they are okay.
needs_review
was used to show that thorough code analysis didn't reveal anything wrong with the proposed solution, and it can be merged as is.
I am happy to consider a patch to get this ticket moved forward.
Sure, but no one added it during 3 years, for some reason.
So, if you could attach one or point to some commit in a (preferably) git repository that would be great.
It is so easy that it's better suited as an exercise for new job applicants.
comment:6 Changed 13 months ago by
Status: | assigned → needs_review |
---|
Here is a patch, I did the changes in 000-tor-browser.js
instead of firefox.js
. If it is wrong, please let me know.
Changed 13 months ago by
Attachment: | 0001-Bug-13575-Disable-randomised-Firefox-HTTP-cache-deca.patch added |
---|
Changed 13 months ago by
Attachment: | 0002-Bug-13575-Disable-randomised-Firefox-HTTP-cache-deca.patch added |
---|
comment:7 Changed 13 months ago by
Keywords: | TorBrowserTeam201801R added |
---|
comment:8 Changed 13 months ago by
Resolution: | → fixed |
---|---|
Status: | needs_review → closed |
Thanks, applied to tor-browser-52.6.0esr-8.0-2
(commit f6d6512bf85d964fa5a470f2528042c527523a34).
comment:9 Changed 10 months ago by
Keywords: | tbb-backport added |
---|
comment:11 Changed 5 months ago by
Keywords: | tbb-backport removed |
---|
Mozilla continues its experiment, now with sub-hour resolution: https://bugzilla.mozilla.org/show_bug.cgi?id=1098422