New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ToDo: diffs FF99-FF100 #1439
Comments
some bugzilla tickets
|
edited for readability pref("browser.contentblocking.features.strict",
"tp,tpPrivate,cookieBehavior5,cookieBehaviorPBM5,cm,fp,stp,lvl2,rp,rpTop,ocsp");
// prev: "tp,tpPrivate,cookieBehavior5,cookieBehaviorPBM5,cm,fp,stp,lvl2,rp,ocsp" New: |
finally crlite is a default for everyone (🥳) plus they're doing double checks to avoid false positives, maybe worth uniforming in AF as well?
more good stuff. |
uniforming? I am more than happy to stick with mode 2. Plus FF101 now sanitizes entries older than 10 days. |
why tho? it would be one less flip and it's doing double checks on revoked certs only. there's also possibly some usability to gain here but I guess false positives are kinda rare, so that doesn't count much. |
so it doesn't count much, good, we're in agreement. Is it doing "double revoked checks" - I mean if it's not revoked in crlite, in mode 2 does it fall back to OSCP to check it? You're the network nerd ... test it for us And if crlite returns revoked, does this not save contacting an OSCP (which can be a privacy gain IMO)? |
nope, mode 2 and 3 do not fall back to ocsp if crlite can verify that the website is good (meaning it covers the entry and the cert is not revoked).
yes, but since there were some issues with certs that appeared revoked in crlite when they actually weren't (1683525#c21), they decided to go for cooperation to keep privacy and speed for most checks, while keeping usability for the corner cases. |
^ "but I guess false positives are kinda rare, so that doesn't count much" make up your mind :) I am happy with mode 2. This is right up AF's alley ... improved privacy gain (by removing a third party, but unlikely that it's being used against you) vs some possible breakage (not seen any yet) and it reduces all those OSCP errors people get when the observatory is down or their ISP is an ass (or whatever it is that triggers that shit) - because we hard fail |
bugzillas for @earthlng in case you want to tweak your data grabbing script, you got the pref flip but there was no bugzilla listed - IDK what/how you parse - patch = https://phabricator.services.mozilla.com/D142988 |
FYI: marked for visibility + fingerprinting
note: there is also an old macOS pref which can be forced or respect OS settings. And of course android uses overlay scrollbars not exactly the right prefs (at least on windows), and we can await RFP to enforce overlay scrollbars on all platforms - mainly to reduce the plethora of results on linux - I think the pref to use is |
FF100 is scheduled for release May 3rd
FF100 release notes
FF100 for developers
FF100 security advisories
104 diffs ( 46 new, 36 gone, 22 different )
new in v100.0:
removed, renamed or hidden in v100.0:
7009
- 1752621 - replaced bynetwork.http2.http2*
prefs - c278dd2changed in v100.0:
7016
pref("network.http.referer.disallowCrossSiteRelaxingDefault", true); // prev: false - 17343281224
pref("security.pki.crlite_mode", 3); // prev: 1rpTop
- Restrict relaxing default referrer policy for top navigationignore
click me for details
==NEW
==REMOVED, RENAMED or HIDDEN
==CHANGED
The text was updated successfully, but these errors were encountered: