18:02:26 <donuts> #startmeeting Tor Browser Release Meeting 2022-03-7 18:02:26 <MeetBot> Meeting started Mon Mar 7 18:02:26 2022 UTC. The chair is donuts. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:02:26 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic. 18:02:34 <donuts> I think I did that right 18:02:56 <PieroV> you should copy the command like Richard :D 18:02:58 <donuts> gaba is afk today so we should start without her 18:03:00 <richard> you did it :3 18:03:08 <donuts> haha yeah I have to look it up every time 18:04:01 <donuts> looks like someone (aguestuser?) is adding things to the pad 18:05:23 <aguestuser> indeed! 18:05:48 <donuts> okay let's get started 18:06:30 <donuts> 11.0.7 stable is getting released tomorrow, but desktop only? 18:06:42 <PieroV> bolkm is already signing 18:06:46 <donuts> perfect 18:06:50 <PieroV> so I don't think we have time to add also Android 18:06:54 <richard> yeah 11.0.7 can ship once boklm gives the thumbs up 18:06:59 <donuts> and 11.5a6 is planned for next week 18:07:05 <richard> or more likely, he will just make it happen 18:07:07 <donuts> yeah I was going to ask what the status is with android releases 18:07:11 <richard> i'm working on blog post for 11.0.7 now 18:07:22 <PieroV> donuts: but we may release 11.5a6 earlier, because of the security patches 18:07:31 <aguestuser> donuts: did you see that 11.5a5 went out with the download bugfix? 18:07:39 <richard> yeah 11.5a6 should go out asap after 11.0.7 18:07:51 <PieroV> we've already built it (at least me, I think that boklm is going to build it soon, if he's not already doing it) 18:08:09 <donuts> aguestuser: ah no, I missed that! on Thursday? 18:08:21 <donuts> that's great, I'll bring it up at tomorrow's UX team meeting to test 18:08:33 <aguestuser> https://blog.torproject.org/new-release-tor-browser-115a5/ 18:08:38 <richard> oh one thing I wanted to call out, I've added a 'Release Prep' tag 18:08:48 <donuts> what are the security patches for 11.5a6 btw? 18:09:00 <donuts> ty aguestuser 18:09:07 <richard> to Tor Browser issues on gitlab, and you can see all of our planned/closed release tracking tickets 18:09:22 <richard> (and there's a new column at the end on the defualt Tor Browser board in gitlab) 18:09:44 <PieroV> donuts: the same that we added to 11.0.7 (exactly the same, because both are based on 91.7) 18:10:16 <PieroV> commits are tor-browser@fe4d6e1d236831635d76f062e013c6bf25d58408 and tor-browser@1794b76288c6fa7fed6f4caae88335ed57145e65 18:10:54 <aguestuser> richard, PieroV: the linked mozilla bug is private so it's a bit hard to understand what attack they are preventing 18:10:59 <aguestuser> (particularly if i'm donuts trying to explain this to users) 18:11:08 <aguestuser> do we have anything public from moz explaining the bug? 18:11:19 <donuts> richard: it looks like release prep is in applications/tor-browser only, was that intentional? 18:11:23 <PieroV> we have a CVE 18:11:25 <donuts> or should it be on all of applications? 18:11:31 <PieroV> two CVEs, actually 18:11:55 <aguestuser> PieroV: do we have links? (i looked at all our tickets associated with the bugfix and couldn't find any public links) 18:11:55 <PieroV> CVE-2022-26485 and CVE-2022-26486, let me find if I can find a statement from Mozilla 18:11:58 <richard> donuts: yeah i think that's a fine place for it? 18:11:59 <donuts> yeah anything public you could point me to would be great, although it hasn't came up yet as far as I'm aware 18:12:05 <aguestuser> and was too lazy to scrollback in #tor-internal ;) 18:12:32 <PieroV> https://www.mozilla.org/en-US/security/advisories/mfsa2022-09/ 18:12:33 <sysrqb> https://www.mozilla.org/en-US/security/advisories/mfsa2022-09/ 18:12:38 <sysrqb> heh. too slow. 18:12:44 <donuts> ah fantastic, ty both :) 18:13:03 <donuts> love these release prep tickets btw richard 18:13:29 <richard> yeah I'm going to add an issue template to help track all the work that goes into it as well 18:13:32 <aguestuser> richard: speaking of which... i am supposed to make contributions to the release prep issue template... but i can't find the template 18:13:37 <aguestuser> doh 18:13:38 <richard> for tracking+documentation purposes 18:13:42 <aguestuser> messages stepped on each other 18:13:44 <richard> aguestuer: yeah it doesn't exist yet ;) 18:13:58 <donuts> also nice touch adding it to the kanban 18:14:13 <aguestuser> richard: but i'm guessing i can look at your last one and get the gist? (for my next release prep ticket?) 18:14:24 <richard> hopefully i'll have an MR ready for it this week but you know 18:14:34 <aguestuser> nw 18:14:34 <aguestuser> :) 18:14:38 <richard> aguestuser: so currently that ticket just links the relevant issues for a given release 18:14:59 <richard> but I'm *going to* add a checklist for rebase, version checks, signing tagging, etc 18:15:21 <aguestuser> coool! :) 18:15:21 <richard> anyway i'll ping you when i have something 18:15:24 <aguestuser> ack 18:15:45 <aguestuser> okay so: what discussion point are we on now? 18:15:46 <aguestuser> ;) 18:16:07 <donuts> Android releases then the discussion items? 18:17:19 <aguestuser> i had added android releases as a discussion item 18:17:31 <aguestuser> but am happy to jump to that if folks think it's prio 18:17:54 <aguestuser> i mainly want to figure out what numbers to give things 18:18:01 <aguestuser> and when to target them for release 18:18:08 <aguestuser> so i can update calendar and issues accordingly 18:18:23 <aguestuser> decided today: will begin rebasing onto v99 *now* 18:18:31 <richard> dope 18:19:14 <aguestuser> we need another stable release this month to get the fenix v96 update into stable 18:19:20 <aguestuser> but we just shipped it to alpha last week 18:19:34 <aguestuser> so if we're trying to get on an "every 2 week" cadence 18:20:02 <richard> every two weeks for a new android stable? 18:20:03 <aguestuser> i was thinking of doing the first android release this month as an alpha release rebasing onto v99 18:20:05 <aguestuser> then a stable 18:20:09 <donuts> atm it looks like android and desktop releases are kind of leapfrogging each other on the calendar, is there a plan to get desktop and android releases in sync again? 18:20:28 <richard> so we've tentatively agreed to interleave tor-browser desktop and android releases (ignore chemspill/emergency releases) 18:20:32 <PieroV> I think it was intended 18:20:38 <aguestuser> donuts: i had made the calendar assuming we wanted to "leapfrog" 18:20:42 <donuts> oh gotcha 18:20:45 <richard> yeah 18:20:47 <aguestuser> (it is admittedly somewhat confusing to try to do that!) 18:20:55 <richard> otherwise it's too much to build/sign all at one time 18:21:05 <donuts> the version numbering seems a little weird then 18:21:16 <donuts> but idk how that works 18:22:03 <richard> yeah i'm not sure how to best resolve that 18:22:14 <aguestuser> if now's an okay time, i'd love to touch base on logistics of interleaved releases 18:22:24 <richard> shoot 18:22:26 <aguestuser> i tried to fill in release dates in the calendar for ~2 months out 18:22:48 <aguestuser> with the goal of doing android builds in the weeks in-between when desktop builds were happening 18:22:49 <aguestuser> BUT 18:23:08 <sysrqb> aguestuser: (just fyi, we want to get onto a bi-weekly release schedule, but that shouldn't delay putting out a stable version while we're still behind Mozilla) 18:23:25 <aguestuser> sysrqb: wait i really wanna hear that 18:23:32 <aguestuser> but also want to communicate another thing 18:23:42 <aguestuser> sysrqb: wait now i'm confused 18:23:52 <aguestuser> i mapped out a bi-weekly release schedule in the calendar 18:23:55 <aguestuser> is that not what you are talking about? 18:24:02 <aguestuser> err *we*? 18:24:19 <sysrqb> yes, but we're month's behind mozilla 18:24:25 <aguestuser> i was trying to alternate an alpha release with a stabel release every 2 weeks 18:24:33 <aguestuser> and put that into the calendar from today until 2 months from now 18:24:37 <aguestuser> but had a hard time doing it neatly 18:24:50 <sysrqb> so, in this case, where we have what-appears-to-be a working version based on Moz96 18:24:52 <aguestuser> b/c dekstop has a ~1 week lag between signing/tagging alpha releases 18:24:56 <aguestuser> and shipping them 18:25:00 <sysrqb> we shouldn't delay putting out a stable just because we want a biweekly cadence 18:25:02 <aguestuser> gah 18:25:06 <aguestuser> this is really hard to follow 18:25:11 <aguestuser> can we have a "talking stick" or something? 18:25:22 <richard> ok 18:25:47 <richard> sysrqb: can you outline how this bi-weekly release schedule should look for android? 18:25:49 <sysrqb> aguestuser: you can go ahead, i'll stop 18:25:58 <aguestuser> sysrqb: okay, so i *think* thing (1) is that we want to prioritize stable 18:26:10 <aguestuser> like that's the very next thing b/c playing catch up 18:26:21 <aguestuser> that seems non-controversial and not-that-hard to do 18:26:37 <aguestuser> (others agree?) 18:26:42 <sysrqb> agree. 18:26:59 <aguestuser> kk. so strike the idea of next android release being alpha on top of v99 18:27:04 <richard> 👍 18:27:14 <aguestuser> that gets the next slot (whenever that winds up being) 18:27:32 <aguestuser> but regardless of what is being released when i have a question about the general timing of "interleaving" 18:27:38 <aguestuser> on the bi-weekly cadence 18:27:54 <aguestuser> do people have the calendar in front of them by any chance? 18:28:05 <PieroV> I have 18:28:09 <aguestuser> if you noticed for example the 11.5a6 Dekstop Sign/Tag is on 3/9 18:28:17 <aguestuser> (after rebasing that starts today) 18:28:19 <sysrqb> isn't alternating an ideal scenario, but not strictly needed until Android catches up? so, break the alternating schedule for the next ~1 month and get back in sync with Mozilla? 18:28:29 <aguestuser> the release based on that sign/tag doesn't happen until 3/15 18:28:33 <aguestuser> oh no! 18:28:37 <aguestuser> what happened to richard? 18:28:44 <sysrqb> richard 18:28:49 <richard> yo 18:28:51 <sysrqb> richard's okay 18:28:55 <richard> #rurallyfe 18:29:07 <aguestuser> lolol 18:29:33 <aguestuser> did you miss some scrollback? 18:29:39 <aguestuser> (i can repaste) 18:29:40 <richard> yeah I think so 18:29:41 <aguestuser> +aguestuser | if you noticed for example the 11.5a6 Dekstop Sign/Tag is on 3/9 18:30:01 <aguestuser> [+aguestuser | (after rebasing that starts today) 18:30:10 <aguestuser> +aguestuser | the release based on that sign/tag doesn't happen until 3/15 18:30:16 <aguestuser> (and we're back) 18:30:22 <aguestuser> so my question is: 18:30:32 <aguestuser> i am also rebasing this week 18:31:01 <aguestuser> and would like to release shortly after signing/tagging 18:31:15 <aguestuser> but assuming i sign/tag at the end of the week 18:31:23 <aguestuser> i can't ship until after the desktop release goes out 18:31:38 <aguestuser> (if for no other reason because the dekstop has "taken a lock" on the release number) 18:31:56 <aguestuser> this makes it tricky to neatly interleave the releases 18:32:07 <aguestuser> in our ideal-case scenario 18:32:18 <aguestuser> (once we're caught up and the android releases don't take forever) 18:32:26 <aguestuser> so i guess i'm curious: 18:32:48 <aguestuser> (1) is there a reason we delay for a week btw/ signing/tagging on desktop and releasing? (should we do that on android also?) 18:33:22 <richard> so the delay is a builtin safeguard in case shit breaks 18:33:24 <aguestuser> (2) if we *don't* want to observe that delay on android how to handle the fact that this lag inserts an artificial delay of ~1 week in releasing android once it's essentially ready? 18:34:10 <richard> in theory there's no reason to delay if everything is fine, but i'd rather have that safe window in case we need a build2 or a build3 so we aren't making people wait based on the schedule 18:34:14 <aguestuser> oh also: i *think* waiting on that lag also means we have to delay signing/tagging android releases until after dekstop release went out 18:35:03 <richard> hmm yes 18:35:14 <sysrqb> it's not strictly needed, but it is cleaner if you wait 18:35:32 <richard> technically speaking you can sign/tag once we've verified matching desktop builds 18:35:58 <aguestuser> okay, so we could just follow the same lag? 18:36:02 <richard> at that point we shouldn't need to touch the tor-browser-build branches for that build 18:36:30 <aguestuser> thus far we've always shipped every android build the same day(ish) as signing/tagging 18:36:52 <richard> same day as binary signing you mean? 18:36:53 <aguestuser> but allowed for safety by leaving new branch in nightly for a bit before signing/tagging 18:37:00 <richard> sorry i meant signing/tagging to mean signing/tagging git commits 18:37:09 <sysrqb> 1 week is a longer delay then we've used in the past - we usually put 2-3 days, and then update the changelog with subsequent buildN's, if needed 18:38:12 <sysrqb> richard: we can release an android release in ~24 hours, from tagging to publishing 18:38:15 <aguestuser> what is the difference btw/ (1) the buffer we leave between rebasing patches onto new upstream branch and merging that (shipping it to nightly) and (2) the buffer we leave between signing/tagging commits and publishing release? 18:38:37 <aguestuser> i had thought that (1) was the thing we did to catch bugs, etc. 18:38:50 <aguestuser> is (2) just extending (1)? (leaving the same code in nightly for longer?) 18:39:16 <aguestuser> since these delays are holding something from going from nightly to alpha, right? (2) is just doing more of that? 18:40:10 <aguestuser> in any case: i don't necessarily need to know the motivation to figure out the scheduling stuff. 18:40:32 <aguestuser> the fact is just that interleaving gets a bit tricky b/c of different lenghts of delay between sign/tag and release on different platforms 18:40:49 <aguestuser> i had tried to puzzle it out on my own last week on calendar but figured it would be good to check in with others! 18:40:53 <richard> the (2) buffer in my mind was just to make the release process less crunchy from a timing perspective (and to take into account the long iteration time involved with building releases) 18:41:03 <aguestuser> akc 18:41:07 <richard> (on the desktop side) 18:41:12 <aguestuser> yup yup! 18:41:26 <richard> if android is easier/faster for whatever reason then by all means do what makes sense for you 18:41:40 <aguestuser> okay, but: what about issue numbering 18:41:46 <aguestuser> let's say we start this week 18:41:57 <aguestuser> and you are building 11.5a7 (desktop only) 18:42:05 <aguestuser> i am building 11.5a88 (android only) 18:42:08 <sysrqb> there isn't a technical reason we can't release 11.5a7 before 11.5a6 18:42:19 <aguestuser> this gets to the root of my question 18:42:34 <aguestuser> right now i am trying to take later numbers (b/c i'm assuming i will take longer while still learning) 18:42:42 <aguestuser> but if eventually android finishes faster 18:42:52 <aguestuser> i had thought it should maybe have the first release number 18:43:01 <donuts> maybe it no longer makes sense using the same version numbering system 18:43:04 * donuts quickly exits 18:43:11 <richard> donuts: my thought as well 18:43:15 <PieroV> donuts: I was wondering the same 18:43:25 <donuts> can we use some adaptation of fenix's for android? 18:43:32 <aguestuser> (if, e.g., android 11.5a8 is ready to ship by thursday, but desktop is not ready to ship until next tue) 18:43:34 <sysrqb> eventually, stable android should be build with desktop, and alpha should be at the ~2 week mark which should not align (or block) a desktop-only release 18:44:06 <aguestuser> but it seems like desktop starts to do an alpha-related rebase at the top of every month 18:44:12 <aguestuser> which android is also doing 18:44:28 <aguestuser> (since that's when the mozilla releases roll over for both of us) 18:44:31 <sysrqb> but desktop releases after ~1 week, and android builds/releases after ~2 weeks 18:44:50 <aguestuser> i see. okay. makes sense 18:44:52 <aguestuser> but if android is ready to release after ~1 week 18:44:58 <aguestuser> we still want to hold until ~2 weeks to publish? 18:45:05 <sysrqb> we should stay with the 2 week cadence 18:45:17 <aguestuser> sysrqb: in the current calendar 18:45:20 <sysrqb> you can consider expiditing, but that's on a case-by-case basis 18:45:27 <aguestuser> dekstop does *not* release at ~1 week mark 18:45:35 <aguestuser> it releases at ~2 week mark 18:45:42 <aguestuser> err 1.5 18:46:04 <sysrqb> still seems okay :) 18:46:08 <aguestuser> but regardless... i think i understand the desired cadence 18:46:18 <aguestuser> (1) moz rolls over 18:46:19 <aguestuser> (2) everyone starts rebasing 18:46:25 <aguestuser> (3) dekstop releases alpha 18:46:29 <aguestuser> (4) android releases alpha 18:46:43 <aguestuser> (5) everyone releases stable together at same time? 18:47:00 <aguestuser> (with shared version number for stable release?) 18:47:11 <sysrqb> (3') desktop/android stable 18:47:16 <sysrqb> (4) desktop alpha 18:47:22 <sysrqb> (5) android alpha 18:47:36 <sysrqb> stable release almost always comes first 18:47:53 <aguestuser> hmm... not in the calendar i worked off last week 18:48:07 <aguestuser> or rather it was step (0) 18:48:10 <sysrqb> because we want to get security updates out to our users asap after mozilla releases their version 18:48:13 <aguestuser> happened before moz rolled over? 18:48:43 <sysrqb> ah, i see 18:48:46 <sysrqb> i think. 18:48:57 <sysrqb> there are multiple versions in play 18:49:06 <sysrqb> let's look at this month 18:49:14 <aguestuser> or the stable desktop release happens more or less concurrently w/ (1) 18:49:29 <aguestuser> (the day that moz rolls over and the day after we started rebasing on the next alpha) 18:49:46 <aguestuser> yes, this month! 18:50:09 <aguestuser> (which might be tricky b/c security hotfix, but yes: a concrete example would be great) 18:50:33 <aguestuser> 11.0.7 stable desktop was calendared for release tomorrow 3/8 18:50:47 <richard> mmhm 18:50:55 <aguestuser> (on top of esr91.7) 18:51:16 <sysrqb> (let me know when/where I should jump in) 18:51:50 <aguestuser> i am actually now confused by the fact that 3/8 says it is a stable release on esr91.7 but we also start rebasing alpha onto esr91.7 today 18:52:10 <sysrqb> correct :) 18:52:39 <richard> (again the 1 week lag is just built-in buffer, really I ttry to treat these dates as due dates, not start dates, with limited success) 18:53:08 <aguestuser> "+sysrqb | correct :)" <- that is not how we do it in android? 18:53:18 <aguestuser> fenix versions land in alpha beforfe stable 18:53:27 <aguestuser> i mean jump in at any point 18:53:32 <aguestuser> i'm actually sort of hopelessly lost now 18:53:40 <sysrqb> :) 18:53:43 <sysrqb> it's all good 18:53:52 <aguestuser> but i am not sure the desktop versioning question affects my question 18:54:15 <sysrqb> let's actually look at next month, under the assumpt Android successfully jumps onto the FF99 train 18:54:51 <aguestuser> by next month we mean april? 18:54:59 <sysrqb> yes 18:55:01 <aguestuser> https://nc.torproject.net/apps/calendar/dayGridMonth/2022-04-01 18:55:21 <aguestuser> on 4/4 moz rolls over and we start rebasing 18:55:21 <sysrqb> shortly after 2022-04-04 Mozilla will tag 91.8esr and 99.0 18:55:27 <sysrqb> correct. 18:55:35 <sysrqb> we start with stable 18:55:46 <sysrqb> for the past month, we've been testing 99beta in our alpha release 18:55:53 <sysrqb> (~2 weeks, really) 18:56:08 <sysrqb> we now rebase our patches onto the 99.0 stable release branch 18:56:20 <sysrqb> and desktop patches are rebased onto 91.8esr branch 18:56:39 <sysrqb> we build these new branches and release a stable version on 2022-04-05 18:56:49 <aguestuser> okay, so without need for rebasing, we quickly publish stable targetting both 91.8esr and 99 18:56:59 <aguestuser> on 04-05 18:57:05 <aguestuser> following 18:57:06 <sysrqb> (hrm. okay, the tag date was wrong, it should be around 1 week earlier) 18:57:10 <aguestuser> (and had completely missed that in my release scheduling) 18:57:44 <aguestuser> also if looking at this calendar, note: that i've got fnx99 where i should have fnx100 18:57:58 <sysrqb> sorry, we do rebase the android patches from the 99 beta branches to the stable branches 18:58:14 <aguestuser> onto maint-11.0 18:58:18 <aguestuser> (in this case) 18:58:30 <sysrqb> yes 18:58:49 <aguestuser> and then we rebase fnx100 onto new (soon-to-be-alpha-release) branch 18:59:03 <sysrqb> correct 18:59:14 <sysrqb> and that will take around 1-2 weeks, for rebasing and toolchain updates 18:59:29 <sysrqb> after 2 weeks, we should be ready for that alpha release, on schedule. 18:59:41 <aguestuser> desktop, goes first, then android 18:59:57 <aguestuser> and if helpful for android to wait after signing/tagging to publish we can 19:00:16 <aguestuser> (i guess a helpful insight for me from this is there is no blocker to signing/tagging as soon as i want) 19:00:40 <aguestuser> and: it's still aspirational that any of the rebasing for alpha branches will happen "fast" (under any definition of "fast") 19:00:46 <sysrqb> correct, and we should probably release a combined desktop/android alpha release based on 91.8esr and Fenix98 after the 1 week delay 19:01:00 <sysrqb> because there isn't really a reason not to update android at the same time 19:01:28 <richard> mhnmm 19:01:30 <aguestuser> err... fenix100? 19:01:44 <aguestuser> (if we're talking april?) 19:02:21 <aguestuser> but where did the idea of interleaved releases go then? 19:02:23 <sysrqb> err, no, Fenix 99, sorry. we release desktop/android stable based on 91.8esr and Fenix99, and then after a week I propose we release the alpha for both desktop and android 19:02:49 <sysrqb> and then the android-only alpha based on Fenix100 comes along at the ~2 week mark 19:03:40 <sysrqb> but this doesn't need to decide right now 19:03:46 <sysrqb> we can table this until the next meeting 19:03:48 <aguestuser> "then after a week I propose we release the alpha for both desktop and android" <-- gosh. currently alpha releases are "ahead." ie: alpha is on fenix v96, while stable is on v94 19:03:53 <aguestuser> is that an anomaly? 19:04:17 <aguestuser> i was trying to map out next 2 months assuming that stable is always 1 version behind alpha in terms of fenix version 19:04:32 <aguestuser> also: we are at 5 minuts past time 19:04:38 <aguestuser> sorry for swallowing entire meeting 19:04:38 <PieroV> *35 19:04:55 <PieroV> release meeting is half an hour :) 19:05:00 <aguestuser> oh really? 19:05:00 <richard> lol 19:05:02 <aguestuser> geeze 19:05:03 <sysrqb> aguestuser: no, that is how it should work: alpha -> 99beta, stable -> 99, alpha -> 99, alpha -> 100beta, stable -> 100, etc. 19:05:11 <donuts> ha yes technically PieroV is right but that's totally fine, seems like there's nothing else following this 19:05:27 <PieroV> true 19:05:38 <richard> ok aguestuser, are we good now w/ regards to android release schudling? 19:05:43 <aguestuser> lol 19:05:47 <richard> :p 19:06:12 <aguestuser> i am reading the above tyring to parse the differentce between 99beta and 99 19:06:17 <aguestuser> etc 19:06:29 <sysrqb> aguestuser: beta release verses stable release 19:06:41 <sysrqb> mid-month verses beginnning-of-next month 19:06:55 <sysrqb> https://wiki.mozilla.org/Release_Management/Calendar 19:07:21 <aguestuser> ok, right so if we track that we presereve the invariant that alpha is always a "branch" ahead of stable 19:07:23 <aguestuser> but not a version 19:07:27 <sysrqb> 99beta is mid march, and then 99 is beginning of april 19:07:28 <aguestuser> (which is the mental model i had) 19:07:44 <aguestuser> or not a major version 19:07:49 <sysrqb> yes 19:08:16 <aguestuser> okay, i think i need to take somet time to stare at calendars again before i can actually project 2 months out 19:08:18 <sysrqb> i think i threw a wrench into you mental model when I proposed releasing android-alpha based on 99 19:08:24 <sysrqb> but that's okay 19:08:47 <aguestuser> i will start with figuring out what version numbers to give our android-only emergency releases 19:09:14 <aguestuser> then figure out what to number the next android stable (v96) and android alpha (v99) releases. 19:09:41 <aguestuser> with an honest effort to try to parse whetner i mean v99beta or just v99 19:09:48 <aguestuser> and coordinate with richard on what our release numbers should be 19:10:10 <aguestuser> (punting to next time on question of doing new version number scheme should be! yes?) 19:10:20 <aguestuser> /s/should/might 19:10:30 <sysrqb> (this is why we haven't successfully scheduled future version-number release dates, only version-less release dates) 19:12:05 <aguestuser> ok. i am officially done wasting other people's time in public. :) 19:12:13 <sysrqb> donuts: you may need to end the meeting :) 19:12:17 <PieroV> nope 19:12:19 <donuts> haha yes will do 19:12:21 <PieroV> I have still two points 19:12:29 <PieroV> onion auth is fixed, tor-browser!261 19:12:30 <richard> lol gogogo 19:12:32 <donuts> please go ahead pierov 19:12:34 <aguestuser> nooooooooooo 19:12:43 <PieroV> are we going to do a .7.1? 19:12:48 <donuts> fantastic! what was the issue? 19:12:51 <richard> dope 19:12:55 <PieroV> as Richard said, a missing await 19:13:08 <richard> lol called it 19:13:10 * richard dabs 19:13:15 <PieroV> lol 19:13:30 <PieroV> or are we waiting for next release? 19:13:43 <aguestuser> (sorry: my "noooo" was at meeting ending! not at PieroV saying stuff. really excited for someone else to say stuff! :)) 19:13:49 <donuts> hahaha 19:13:56 <PieroV> and in case I would also slip the IPv6 circuit display 19:13:57 <richard> we could do a simultaneous android/desktop stable release with the fix in 19:14:18 <PieroV> okay, then IPv6 can wait 19:14:27 <richard> yeah IPV6 can wait regardless 19:14:46 <richard> PieroV: what's the summary on the onion auth issue? 19:14:51 <PieroV> I proposed it in the backport asap small features spirit 19:15:00 <richard> is it just bad UX if the bug is hit 19:15:03 <PieroV> tor-browser#40802 19:15:04 <richard> or does it just not work at all 19:15:21 <PieroV> it doesn't work at all 19:15:30 <richard> ugh damn 19:15:36 <donuts> does that mean there should be a fix for it next week (in x.x.7.1)? 19:15:54 <richard> ok let's plan on new stable+alpha desktop releases next week then 19:16:31 <richard> i'll get to merging and tagging and whatnot this week 19:16:38 <PieroV> okay, works for me 19:16:53 <richard> aguestuser: i'll follow up wiht you later abotu the right version number to use :3 19:17:24 <PieroV> no more points from me :) 19:17:40 <aguestuser> richard: ack :) 19:18:01 <donuts> okay I'm gonna press the button 19:18:03 <richard> ok 19:18:04 <donuts> 3... 19:18:06 <donuts> 2... 19:18:07 <donuts> 1... 19:18:10 <donuts> #endmeeting