15:00:28 <richard> #startmeeting Tor Browser Weekly Meeting 2022-05-09
15:00:28 <MeetBot> Meeting started Mon May  9 15:00:28 2022 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:28 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:00:39 <richard> alright you all know what to do
15:00:45 <richard> why am I even here :p
15:00:54 <Jeremy_Rand_Talos_> Hi!
15:01:04 <sysrqb> richard: that's deep
15:01:20 <richard> lol
15:09:30 <richard> ok you've had 10 minutes
15:10:43 <richard> perhaps unsurprising, I did *not* get to creating/assigning tickets for the next block of Tor Browser refactoring/build work last week
15:11:07 <richard> but i'll be on that while signing alpha today/tomorrow
15:12:31 <richard> I suspect y'all have a pretty good idea of what needs to happen there in the meantime given the myriad planning meetings and emails we've had over the past several weeks, so don't let that stop you :p
15:13:00 <richard> other than that I have nothing to discuss, so let's go over y'alls discussion points/announcements
15:13:05 <richard> sysrqb: you seem to be #1
15:13:10 <PieroV> richard: actually I'm not sure on how you want to tackle that
15:13:35 <PieroV> but I think we can discuss very quickly outside the meeting
15:13:50 <richard> ok that wfm
15:14:02 <PieroV> great, thanks
15:14:04 <richard> or we can go through the points here and devote the rest to mullvad planning :)
15:14:27 <sysrqb> my item is only an announcement :)
15:14:30 <PieroV> yes, that also works for me
15:14:51 <richard> alright, well it looks like the rest are no-ops
15:14:52 <sysrqb> and it seems I'll be working with PieroV on merging with his improvements/rewriting
15:15:13 <PieroV> I've added bold items outside the discussion area
15:16:01 <richard> ah alright
15:16:26 <PieroV> a pair are about the Android release
15:16:40 <sysrqb> PieroV: https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/issues/25835
15:16:43 <sysrqb> (just fyi)
15:17:12 <PieroV> okay, thanks! I think we had discussed that also in the meetings, but then we've never addressed it
15:17:37 <sysrqb> yep
15:17:38 <PieroV> but let me discuss about the Android points, first :)
15:17:43 <sysrqb> k
15:18:04 <PieroV> One is fenix#40202, I've patched it in master. Any objection to backport it to stable?
15:18:31 * richard reading
15:18:32 <PieroV> I don't know how many other 11.0.x releases we'll do
15:18:54 <PieroV> richard: TL; DR: out APKs do not work in Android 11+ (or 10+)
15:19:13 <richard> this seems reasonable to me
15:19:26 <sysrqb> +1 i agree, we can backport that
15:19:27 <richard> re how many more 11.0.X releases
15:19:44 <richard> I would expect that if we do release it would only be to backport CVEs
15:20:13 <PieroV> great; sysrqb: can I set you reviewer, since it's on the zipalign thing, mainly?
15:20:18 <sysrqb> yes
15:20:22 <PieroV> ok, thanks
15:20:40 <PieroV> richard: so, do you want to pause the Android rebases?
15:21:10 <richard> yeah pretty much
15:21:49 <PieroV> only on stable, or also on alpha?
15:22:49 <richard> I think that will basically depend on how much availability I have
15:22:55 <richard> either way you don't need to worry about alpha rebases
15:24:03 <PieroV> okay, works for me
15:24:38 <PieroV> Second question about Android, is whether we want to wait for tor-browser#40913 / tor-browser!286
15:24:57 <PieroV> I'm in favour of waiting for them
15:25:20 <richard> those should be relatively easy to revert/set prefs for yes?
15:25:48 <PieroV> Yes. The pref should be enough, but I'd also add some code to partially revert it
15:25:58 <PieroV> I've added the details on the issue
15:26:17 <richard> ok that works for me
15:27:15 <sysrqb> (fine with me)
15:27:21 <PieroV> Then my points were about the manual, one has already been answered (we/I could also create an issue on tpa, since we first need them to provide the space), the other one is what do we want do to with Python?
15:28:14 <boklm> I think it's a good idea to update python anyway
15:28:22 <richard> i think we should probably update python anyway
15:28:36 <richard> as it's a relatively small jump now
15:29:07 <richard> if we have time we can do it now, otherwise perhaps we should wait until we're doing esr102 tor-browser-build related work
15:29:08 <PieroV> Okay, and to 3.8.x, in case we decide to build the manual offline, or to the latest one?
15:29:47 <richard> i have no opinion here
15:29:59 <richard> boklm?
15:30:17 <boklm> I think we can use the latest one working with firefox, and see later if we need some changes for the manual build
15:30:28 <PieroV> okay, wfm
15:30:29 <richard> +1
15:30:44 <PieroV> I lied and had another point, but it's actually related to mullvad work
15:31:25 <PieroV> that is, if we want to uplift stuff, if someone can give me a quick help on the best way to prepare such requests for Mozilla
15:31:51 <sysrqb> i believe richard ro I can help with that
15:31:55 <sysrqb> *or
15:31:58 <richard> I can point you to my previous uplift attempts on bugzilla
15:32:16 <PieroV> great, thanks!
15:32:45 <richard> but in general; outline the issue and how it affects tor browser, make sure whatever patch you propose is behind a pref and defaults to w/e the current firefox behaviour is, etc
15:32:56 <richard> you'll need to do some song and dance to get into their code review system
15:33:03 <richard> whose name i don't recall atm
15:33:12 <sysrqb> (phabricator)
15:33:18 <richard> yes that's the one
15:33:38 <richard> unfortunately/fortunately the days of uploading patch files are over :p
15:33:41 <PieroV> oh, so I'll probably need to finally get the committer permission
15:34:17 <sysrqb> and add the person who made the offending change (or recent significant changes in the file) as patch reviewer
15:34:38 <sysrqb> PieroV: you don't need Try access, but it is helpful
15:34:40 <richard> yes that^
15:35:16 <sysrqb> but you should get it, in any case :)
15:35:59 <richard> and quick announcement
15:36:02 <richard> tb-build-05 is live!
15:36:08 <PieroV> \o/
15:36:17 <boklm> \o/
15:36:23 <richard> w/ 64 logical cores
15:36:25 <richard> jfc
15:36:25 <sysrqb> nice!
15:36:32 <richard> fun fun fun
15:36:40 <richard> ok
15:36:45 <richard> ok onto mullvad planning
15:37:17 <richard> well, mullvad work now
15:38:50 <richard> what's on your mind PieroV?
15:39:21 <PieroV> Well, last Friday I wrote the document on what we could need to do: https://pad.riseup.net/p/nKyiZjNnnvqE8M-middlebrowser
15:39:37 <PieroV> Most of them are naming things, almost all of them only in the source code
15:40:26 <richard> ah and you need some review on these items
15:40:41 <PieroV> But if we wanted to just produce binaries, we could just create them, without removing these names
15:41:03 <PieroV> Whereas if we want to remove them, the smartest thing to do is preparing them also for uplifting
15:41:49 <PieroV> We have a month and a half before 102 is released, and getting whatever we can before that release is the easiest way to find them in 102, from what I understand
15:42:33 <richard> about how many upliftable things do we have?
15:43:09 <PieroV> 5 can be directly uplifted
15:43:35 <richard> well regardless, in my experience, the uplift process can involve some back and forth, so the sooner you start the better
15:43:36 <PieroV> 4 have Tor browser name in it, or something similar
15:44:22 <PieroV> And I've already forgot if the 5 direct ones need some preference to lock them
15:44:26 <richard> so i would suggest (as a rough plan of attack)
15:45:32 <richard> create a new tor-browser alpha 11.5a11-2 dev branch, with the patches roughly where they need to be
15:45:48 <richard> get the uplift process started asap for the patches which should need no modification
15:46:10 <richard> then go back and rebase the remaining patches to be 'upliftable' (so generalizing, removing tor-browser name/branding/whatever)
15:46:50 <PieroV> Do we drop Android patches from desktop?
15:47:04 <richard> and hopefully by the time 102 rolls around a significant fraction of those patches will be in and we can stop carrying them
15:47:16 <richard> god now there's a question
15:47:49 <richard> i would sday leave them in for now, sysrqb i'll ping you after this meeting about *android* and why we do some things the way we do :p
15:48:12 <sysrqb> k :)
15:49:39 <PieroV> big question: do we need to move all the proxy bypass stuff after the fork point?
15:50:20 <richard> i think proxy bypass should go before
15:50:43 <richard> i can't think of any reason why mullvad would want proxy leaks in their browser
15:50:50 <richard> (or why mozilla does for that matter)
15:51:59 <PieroV> okay; should 11.5a11-2 contain tor stuff, but we only check that it's buildable at the improve letterbox commit, or does it need to stop there?
15:52:47 <richard> ideally this would become our next 11.5a12 alpha branch, so best effort to make tor-browser work after the split
15:53:45 <richard> i mean, worst case scenario we can just apply a final patch on 11.5a12 that maps 11.5a11-2 to 11.5a11-1
15:54:09 <richard> and rebase that whole thing to the next esr91.X
15:55:29 <richard> but if making tor-browser work in the new branch gets in the way of uplift, go with uplift
15:55:49 <richard> like i said we can always patch it up later
15:56:14 <PieroV> okay; we also have 2 or more patches that we could maybe drop
15:56:49 <PieroV> But I'm not really sure
15:57:13 <PieroV> Bug 27604: Fix addon issues when moving TB directory (it's for .xpi things)
15:57:46 <PieroV> Bug 16620: Clear window.name when no referrer sent (Firefox may have a better patch we've dropped, see the linked issue)
15:58:16 <sysrqb> (time check ;))
15:58:17 <richard> so for the general problem of removing tor-browser from the commit message/subject, we can always make a new issue (or reference existing) on mozilla's bugzilla, and in that ticket reference/link to our gitlab issue for historical context
15:58:18 <PieroV> but I think they can stay there for the moment
15:58:39 <richard> alright and that's about time
15:58:44 <PieroV> Okay
15:58:57 <PieroV> If you have anything else to add on my document please do :)
15:59:05 <richard> I'll review/comment today/tomorow
15:59:08 <PieroV> Thanks
15:59:11 <richard> #endmeeting