18:02:10 <gaba> #startmeeting Tor Browser Release meeting, 31 January 2022
18:02:10 <MeetBot> Meeting started Mon Jan 31 18:02:10 2022 UTC.  The chair is gaba. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:02:10 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:02:24 <gaba> pad in https://pad.riseup.net/p/tor-browser-release-meeting-keep
18:02:43 <gaba> tbb-team around?
18:02:50 <donuts> o/
18:03:35 <gaba> o/
18:03:36 <aguestuser> o/
18:03:58 <gaba> can anybody update on what is the next release and where we are at?
18:04:02 <donuts> richard: ping
18:04:08 <gaba> this is our first tb release in many months
18:04:28 <richard> o/
18:04:53 <gaba> /s/tb release/ tb release meeting ...
18:05:34 <richard> so the most major thing iirc is the UX fixes that have plagued various users
18:05:44 <richard> (where torconnect buttons fail to load)
18:05:57 <richard> assuming you're referring to stable
18:06:01 <gaba> that is for 11.5a4 ?
18:06:22 <richard> ah no that fix is already in stable
18:07:23 <richard> for 11.5a4 (not 11.5a3?) we'll have an esr update to 96
18:07:41 <richard> along with the patch reordering (which end uses won't see)
18:08:01 <richard> the DW v2 onion rewrite patch is also to be backed out
18:08:46 <donuts> I'm a little lost wrt to releases, when are we planning to release 11.5a3 for desktop (and 11.5a4 after that)?
18:09:26 <donuts> and is there a timeline for 11.0.5?
18:09:51 <richard> yeah that's on me, so the enext alpha release will be 11.5a3 which will happen after the next stable which should happen next week (rebasing to 91.6, testing, tagging etc this week)
18:10:57 <gaba> sounds good
18:11:16 <richard> (so the next stable release will coincide more-or-less with the esr91.6 release which is scheduled for Feb 8)
18:11:32 <gaba> richard, donuts: should the 3 of us meet and go through the various TB releases milestone and clean them on what we can move or close?
18:12:19 <richard> that works for me
18:12:35 <donuts> richard: got it, so next week 11.0.5 for desktop -> then 11.5a3 for desktop -> then 11.5a4 (for android and desktop?)
18:12:44 <donuts> gaba: yep sounds good
18:13:17 <richard> yeah 11.05 next week, 11.5a3 the following and 11.5a4 sometime after
18:13:58 <donuts> great! how are things looking in android-land for a new stable release? cc aguestuser
18:14:15 <donuts> are we still considering another alpha for android first?
18:14:18 <richard> that's an aguestuser/sysrqb(?) question :3
18:14:26 <aguestuser> for android: currently working on release on top of fenix v96.2.0. the branch i'm using for that is labeled tb v 11.5.1. this discrep (11.5a4 v -1) is somewhat confusing to me but can sort?
18:15:24 <richard> what do you mean?
18:15:49 <aguestuser> which part?
18:15:52 <aguestuser> the labeling?
18:16:01 <richard> sorry what' sconfusing I mean
18:17:38 <aguestuser> the branch i am rebasing fenix changes on top of is named `tor-browser-96.2.0-11.5-1`
18:17:55 <aguestuser> the release we are discussing is called `11.5a4`
18:18:02 <aguestuser> `-1` is different than `a4`
18:18:06 <aguestuser> what does each mean?
18:18:08 <richard> ah ok
18:18:26 <richard> so .5 branches are the alpha branches
18:18:46 <richard> we don't typically need ot make a new branch between alphas unless there is an esr transition
18:18:56 <donuts> followup Q: so 11.5a3 is based on fenix 94, are we planning on a stable release that's fenix 94 based too?
18:18:58 <richard> (or as is the case this time around for desktop, some major code shuffle)
18:19:33 <richard> each of the specific versions (ie a3, a4) get tagged in the tor-browser repo
18:20:59 <aguestuser> following...
18:21:11 <richard> donuts: I don't know the answer to that
18:21:42 <richard> (aguestuser: more followup, the stable/alpha releases are referred to by tag name in tor-browser build, whereas nightlies just point to a branch)
18:23:02 <donuts> np, I guess my actual question is what's the target we're working towards before we can get a new stable release for android
18:23:23 <aguestuser> richard okay, and all alphas under discussion point to branch `11.5-1`? would an alpha ever point to `11.5-2`?
18:23:36 <donuts> (we can talk about that after though)
18:23:53 <richard> aguestuser: yes such a thing is possible :)
18:24:33 <richard> donuts: I don't know, aguestuser do you have an idea on what the android releases scheudle should look like?
18:24:34 <aguestuser> under what circumstances would that happen? (ps: happy to take this off-thread to a deep dive on further understanding the release conventions. feel like i'm thread-hijacking a bit here.)
18:25:11 <richard> aguestuser: if there is an ESR update between alpha versions forcing a rebase onto a new base branch
18:25:19 <richard> is the prime example
18:25:33 <aguestuser> dig
18:25:35 <aguestuser> richard donuts -- working on getting fenix 96.2.0 out the door. which in this schedule i think is erroneously listed as a Fenix95 update on 2022-01-11
18:25:47 <aguestuser> (but should be Fenix96)
18:26:15 <aguestuser> was complicated by some weirdness PieroV encountered in android-components -- and me getting up to speed!
18:26:48 <aguestuser> in the schedule this release is called 11.5.a5, but IIRC last android release was a3?
18:27:35 <richard> yeah the current schedule is out of date, i'll planning to fix it to better match reality tomorrow
18:27:43 <donuts> aguestuser: ah the schedule is all out of wack because we've never actually released an 11.0 stable version of android
18:28:06 <richard> mmhm
18:28:29 <aguestuser> schedule also says we have a Fenix97 release on 02-08 which is 11.0.3 which is a bit confusing because that is later than when we released 11.5a5 (but 11.0 is lower than 11.5)
18:28:33 <donuts> so the plan we discussed with sysrqb before the break was to get one or two new alphas for android out, and then release to stable before we jump onto fenix 96
18:28:55 <aguestuser> hmm.
18:29:00 <donuts> aguestuser: I think you can fully ignore the schedule tbh
18:29:03 <donuts> sorry for the confusion
18:29:04 <aguestuser> lol ok
18:29:08 <donuts> (richard feel free to correct me here)
18:30:06 <donuts> we need to sit down and fix that, but in the interim it would be good to understand when we'll be ready for an android stable release independent of whatever the schedule says
18:30:16 <aguestuser> ok so what i am working on is the fenix -> 96 bump. which is close to done. (which for me at my current rate means "on the order of day(s)" depending on how many new things i encounter)
18:31:08 <aguestuser> i don't currently have enough context to know about when that will be in alpha or stable.
18:31:27 <aguestuser> i have ridden shotgun for tagging one (fenix v94) build for 11.5a3 release (IIRC)
18:31:51 <aguestuser> i don't yet know our flows for putting something into alpha then stable (or how we think about that)
18:31:55 <aguestuser> but am eager to learn!
18:32:52 <donuts> no worries! So my concern is that it's been a while since android users have received a stable update, and it may be worth planning a stable release based on fenix 94 before 96 is ready and has been tested in alpha
18:32:54 <donuts> but
18:33:04 <aguestuser> got it!
18:33:05 <donuts> that's a decision I can't make
18:33:25 <donuts> sysrqb (and possibly richard?) may be of use here though :)
18:33:26 <aguestuser> i see.
18:33:36 <donuts> sorry, I'm not 100% clear on where we are with handovers/responsibilities yet
18:33:47 <aguestuser> me neither! we'll get there!
18:33:51 <donuts> ha +1
18:34:16 <aguestuser> the bump to 96 has thrown us for loops so most of my attention has been there.
18:34:42 <aguestuser> but i hear that we'd like to get another stable v94 release, which would require a rebase of our patches onto some upstream changes (which is what i've been learning how to do on fenix -> v96)
18:34:52 <richard> what's the major difference between fenix 94 and 96?
18:35:11 <aguestuser> android-components had a bunch of internal breaking changes
18:35:15 <aguestuser> (has been the hold-up)
18:35:29 <richard> are there any major security updates that are a must have for our users?
18:35:35 <aguestuser> have not audited yet!
18:35:43 <richard> or can they live with 94 while while make sure 96 isn't broken?
18:36:04 <aguestuser> see where you're going though
18:36:39 <aguestuser> in any case: would a flow in which we shipped 96 alpha first (which gives it time to be tested)
18:36:39 <richard> yeah my instinct is to hold out and just release a new 94 based version rather than busting our butts on something not fully tested...
18:36:56 <aguestuser> and then shipped 94 stable
18:37:04 <aguestuser> while 96 alpha being tested
18:37:06 <aguestuser> work?
18:37:35 <richard> that works for me in general, but we should check-in with sysrqb for this particular caes
18:37:46 <aguestuser> i would feel very good about doing that! :)
18:38:00 <richard> ok great :)
18:38:22 * aguestuser trying to dig up the upstream changelog for fenix to help answer the question "any significant security updates?"
18:38:24 <donuts> I would love a 94 stable if possible, yes please :)
18:39:27 <richard> sounds gooood to me
18:39:33 <donuts> okay so next week: 11.0.5 for desktop
18:39:51 <donuts> w/c feb 14th: 11.5a3 for desktop
18:40:15 <donuts> w/c feb 21st: 11.5a4 for desktop (and android?)
18:40:34 <donuts> and a 94 based android stable somewhere in amongst that?
18:41:08 <donuts> does that sound doable?
18:41:58 <richard> sounds good for me re desktop
18:42:17 <richard> 11.5a4 alpha may be later
18:42:33 <richard> as I'm traveling the week prior and who knows what nonsense will come up irl
18:42:49 <aguestuser> donuts my current idea of ordering was to do 96-based android alpha (w/c jan 31) then 94-based android stable just after (this week or w/c feb 7).
18:42:56 <donuts> richard: got it, np
18:43:24 <aguestuser> i would want to consult w/ sysrqb before any of that, but want to make sure that i'm not breaking expectations
18:43:57 <donuts> we may want to try and sync up a little too, so we're releasing mobile and desktop (mostly) together again
18:44:07 <aguestuser> +1
18:44:24 <aguestuser> btw: i have *zero* strong opinions on when to do anything
18:44:30 <donuts> okay
18:44:40 <aguestuser> just relaying what would happen if i went about working on stuff the way i had been based on work handed to me
18:44:46 <aguestuser> ;)
18:45:01 <aguestuser> happy to make adjustments and do different things!
18:45:40 <aguestuser> the primary reason i'd want to stay focused on 96 for the moment is orthogonal to org goals -- i've been learning how the flow goes and want to finish one full trip through the release cycle to help cement learning.
18:45:46 <aguestuser> purely selfish! :)
18:46:00 <aguestuser> and not strongly held!
18:46:05 <donuts> no prob, if we have something stable and secure we should plan for a stable release asap (but still while giving 11.5a3 sufficient time in the wild for testing)
18:46:43 <donuts> for reference, android hasn't received a stable update since october 25th
18:47:17 <aguestuser> oh geeze. okay. does it help that 11.5a3 against 94 has been (i think?) out for a week or so?
18:47:21 <donuts> part of that has been down to trying to catch up with fenix, and part of that has been reallocating dev ours to fixing TB11 for desktop bugs
18:48:05 <donuts> aguestuser: yep but I'd recommend checking in with sysrqb & team on whether that's sufficient
18:48:40 <aguestuser> dig. okay. so: on my todo list is check in with sysrqb about what to do w.r.t. need for 94 stable release and check back... where? #tor-dev and mention you?
18:48:56 <donuts> yes please, that would be perfect
18:48:56 <richard> ping me too plz :)
18:49:04 <aguestuser> okay! :)
18:49:06 <aguestuser> clarity!
18:49:18 <donuts> then maybe we can sync desktop and android in 11.5a4
18:49:38 <aguestuser> wahoo!
18:49:43 <donuts> failing that, 11.5a5 (whenever that is) \o/
18:50:21 <donuts> okay I'm going to write this all down in the pad :)
18:51:15 <donuts> anything else or are we all good?
18:51:23 <aguestuser> good here
18:51:23 <donuts> we may need gaba to end the meeting...
18:51:48 <gaba> #endmeeting