15:01:24 #startmeeting Tor Browser Weekly Meeting 2024-08-26 15:01:24 Meeting started Mon Aug 26 15:01:24 2024 UTC. The chair is morganava. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:24 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:01:56 the pad -> https://pad.riseup.net/p/tor-tbb-keep 15:02:43 14.0a2 shipped last week, and if possible i'd like to get 14.0a3 built for release tomorrow/wednesday 15:03:06 I deccided to hold off on a full deployment on the google play store given the search-bar fix which didn't get in 15:03:12 so that's sitting at 25% atm 15:03:54 looking at my emails quickly looks like I have some MR reviews todo first thing to hopefully get in for a3 🙂 15:04:05 hello! 15:04:16 morganava: I know how to fix the version bug 15:04:24 But I don't know if Claire has already started working on that 15:04:43 So, if she hasn't I can create a quick patch to get that done in 14.0a3 15:04:53 last week I went through all of the 14.0 Stable && !14.5 Stable issues and triaged 15:05:08 morganava: also, we're getting tags later today or tomorrow early morning 15:05:28 the query if you're looking for issues ;) -> https://gitlab.torproject.org/groups/tpo/applications/-/boards?not[label_name][]=14.5%20stable&label_name[]=14.0%20stable 15:05:29 Shall we have 14.0a3 built on 128.2.0 before it's released by Moz, or are we staying on 128.1.0? 15:05:52 pierov: if it was started, it's not done, i dont have a review in my queue for it 15:06:07 if it's fast maybe bash that out? 15:06:07 I'm fine with keeping 14.0a3 on the prev minor ESR in favor of shipping it and not getting in the way/or behind the 13.5 update 15:06:32 there will be a 14.0a4 next week on the usual update esr update schedule 15:06:38 Consider I do rebases continuously, but I don't build them 15:06:42 ack 15:08:35 anyway, Iv'e triaged the 14.0 + not 14.5 issues and the board should be in a 'good' state 15:08:45 so I seem low on review tickets, which, is good? If I get caught up this week what would be my next priority? should I look at some of the android patchset rebase health stuff or? 15:08:54 things in Backlog or lower priority should be de-prioritised in favor of ~Next's 15:09:16 the majority of of which are Mozilla Reviews (and the majority of those are telemetry fortunately) 15:10:07 dan_b: there are a *ton* of reviews -> https://gitlab.torproject.org/groups/tpo/applications/-/issues/?sort=updated_desc&state=opened&label_name%5B%5D=14.0%20stable&search=Review%20Mozilla&first_page_size=100 15:10:35 184 atm, so if you (and anyone else) would go thouogh those between dev/patches that'd be great 15:10:40 oh cool! 15:10:43 thanks 15:11:01 there's also a handful of issues I need to create from the code-audits (like 5) I'll get to today 15:11:36 morganava: have you gone through TZP's audits? 15:11:46 i have not 15:12:09 are those the ones linked from the various Firefox audit issues in tor-browser-spec? 15:12:10 There might be some additional Bug reviews we've lost but that have been referenced there 15:12:13 Yes 15:12:21 ok, I'll put those in my queue 15:12:38 PieroV: I have not started on the version bug 15:13:52 oh! I also created an MR in tor-browser-bundle-testsuite to add users to the nightly build emails 15:14:17 if you would like emails pease opt-in in that MR (you should all be cc'd) 15:14:41 ( https://gitlab.torproject.org/tpo/applications/tor-browser-bundle-testsuite/-/merge_requests/38) 15:14:57 what do you mean TZP audits 15:15:18 thorin: the list of bugs you have for each Firefox version 15:15:25 In TZP issues 15:15:35 oh, you mean my one issue where I dump shit all year 15:15:47 :3 15:15:52 i blame mozilla personally 15:16:30 they're not TZP issues - they're issue I come across as I audit arkenfox changes 15:17:22 ^ https://gitlab.torproject.org/tpo/applications/tor-browser/-/issues/42054 15:17:30 yep 15:18:10 you know ... audit monthly so we don't play catchup once a year 15:18:15 :P 15:18:25 That's the idea... 15:18:32 i blame mozilla personally 15:18:37 lol 15:18:39 Poor 'zilla 15:18:49 mozilla gets a bad rap 15:18:57 anywya 15:19:21 thorin: yeah hopefully we'll be able to play catch-up in the fall 2 or 3 rapid-releases at at ime until we catch up from this summer 15:19:28 and then it's just smooth sailing into the future 15:19:44 a little annoying work each month instead a whole summer of it 15:19:47 is the dream anyway 15:19:47 and then in 2027 we can have a fingerprinting week 15:20:03 a girl can dream^ 15:20:13 ok 15:20:24 i've yapped enough, are there other discussion points, announcements? 15:20:33 just one from me 15:20:52 can we uplift letterboxing/betterboxing changes 15:21:32 thorin, they're still very in flux. I don't feel they're in a upliftable state yet 15:21:49 ma1: I think we should start at least with the rounding patches 15:21:57 Perfect is the enemy of good, right? 15:22:03 we get a lot of feedback/bugz from RFP users upstream 15:22:18 but i think we're too late now for ESR128 15:22:40 we should add our patches upstream as a first step 15:22:47 ^ where possible 15:23:43 PieroV, that's true. Let's see if the more controversial stuff (like gradients and other things which might change the DOM even more) can be easily separated. 15:23:44 that's it from me 15:24:18 Yeah, I don't think we'll ever be able to upstream everything in one go 15:24:41 i know the rounding still has issues - it opens my laptop with a high dpi 10px too wide and a few pixzels too tall 15:24:52 but it's not hi priority given LBing 15:27:00 thorin, so what you suggest is giving priority to "functional" letterboxing over cosmetic betterboxing and rounding (which is "mitigated" by LB)? 15:27:08 Can someone from android land confirm that the patch for tor-browser#42622 did not break the android connect assist? 15:27:43 it is due a backport for next stable release 15:28:28 ma1: IDK what i'm advocating :) 15:29:15 henry-x: android has been working in the last 3 weeks, since this was merged 15:30:46 dan_b ok, is that with bootstrapping initially failing as well? Because the change could potentially change your error handling. From what I remember 15:31:02 ah I haven't tested 15:31:20 clairehurst? you tested that flow in the last few weeks? 15:31:23 woudl use using the connect assist debug pref exercise this functionality henry-x? 15:32:23 I think that would be one way to get an error 15:33:02 kk, i can check before backporting 15:33:09 Maybe also plane mode can help 15:33:26 There is also firefox-android!118 that was opened in relation to this 15:33:42 Other than the default "no error" flow no but I can test it. I tested it before it was merged though and it seemed to not mess with stuff 15:34:08 PieroV: oh smort 15:35:35 Yeah I didn't label that draft MR very well but iirc its more for consistancy than actually fixing anything (prevent the same thing being called multiple times, which could cause future issues) 15:40:13 alright, any other topics? 15:40:13 yeah, if you could test it that would be great 15:40:25 alright i'll hand that over to clairehurst then 15:41:42 Sounds good 15:42:32 Semi-topic: updating cloudflare early to avoid "false bugs" like the CORP one? 15:43:08 (and maybe telling them how to tell subresource loads apart from document, instead of replacing scripts and stylesheets with captchas)? 15:43:40 i've a contact email/alias for cloudflare if we need to get them to change something 15:44:00 what's going on with cloudeflare? 15:44:12 tor-browser#43079 15:44:18 morganava: still a lot of captchas 15:44:32 And still a lot of broken sites because of invisible captchas in CDNs 15:44:56 Last Saturday I had the problem with Mullvad Browser and my connection 15:45:13 hmm tor-specific stuff it seems, if it's not happening in Mullvad? 15:45:29 No no, it happens also on Mullvad Browser from time to time 15:45:45 probably they trigger on any "odd" fingerprint 15:45:45 hm ok 15:45:54 i'll send them an email and cc.. ma1? 15:46:02 point them to the issue 15:46:09 morganava, sounds good. 15:46:32 I'll add another more detailed comment about Sec-Fetch-* headers to the issue. 15:49:12 thanks thanks 15:50:33 alright 15:50:39 it's nearly the hour 15:50:42 last call for topics before we end this 15:51:54 announcement: tomorrow I will be traveling, then likely offline / weird timezone in Singapore, but *at least* doing Sec backports and managing NoScript 11.4.35 release. Back on Tue 09-03 morning. 15:52:11 oooh have fun! 15:53:25 secular humanists activism stuff, I'll likely show them the god vs javascript t-shirt design PieroV gave me this morning: https://share.riseup.net/#weWz9cc24i7ssLk8B8GQeQ 15:53:56 (it was forwarded to me, I don't know the original source) 15:54:33 (like most religious stuff) 15:55:11 looool oh dear 15:55:38 oh god 15:55:43 this is why i hate js 15:56:05 wtf why does "\t" == 0 15:56:10 MOZILLLAAAA 15:56:16 alright 15:56:22 sounds good, safe travels ma1 o/ 15:56:27 tyvm 15:56:30 have a good week everyone \o/ 15:56:34 #endmeeting