14:58:19 <richard> #startmeeting Tor Browser Weekly Meeting 2023-08-28
14:58:19 <MeetBot> Meeting started Mon Aug 28 14:58:19 2023 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:58:19 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:58:31 <richard> the pad as usual: https://pad.riseup.net/p/tor-tbb-keep
14:58:32 <Jeremy_Rand_36C3[m]> Hi!
14:58:34 <dan_b> o/
14:58:52 <richard> for the firt time in a long time, gitlab's issue priorities are accurate
14:58:56 <richard> as of last friday
14:59:40 <jagtalon> o/
15:00:36 * Jeremy_Rand_36C3[m] can't actually get to the Riseup pad at the moment, not sure if it's an issue on my end or Riseup's (the onion is hitting a TCP timeout, it seems)
15:00:47 <richard> so if you haven't already, please keep grabbing things out of ~Next && ~"13.0 Stable~ (though I'm pretty sure ther aren't any non ~"13.0 Stable" dev issues in ~Next but you never know)
15:01:15 <Jeremy_Rand_36C3[m]> (Don't really have anything noteworthy to put in the pad, so I guess the pad picked a good day to not like me)
15:01:23 <richard> if its in ~Backlog then it's in hte 'nice to have category' so unless they are a one or two-liner please stick with ~Next :)
15:01:45 <richard> this week is a release week!
15:02:06 <PieroV> 12.5.3 is being built already
15:02:18 <richard> we should be building 12.5.3 this evening once rebaes and security backports are complete/reviewe
15:02:22 <richard> and they're done
15:02:24 <richard> baller
15:02:28 <richard> thx everyone
15:02:43 <PieroV> I think my computer will need still another 3-4 hours to get to the end of the tunnel
15:02:59 <PieroV> We've also included the Snowflake changes
15:03:02 <richard> is someone already building on tb-build-05 then?
15:03:05 <richard> ma1?
15:03:07 <PieroV> Yes
15:03:16 <richard> perfect
15:03:23 <PieroV> Snowflake will need a quick QA test before signing
15:03:45 <richard> on Android or in general?
15:03:48 <PieroV> To check we haven't broken it, even though the patch for tor-browser-build#40740 seemed easy
15:03:53 <PieroV> Android only
15:04:15 <richard> ok yep make sense
15:04:31 <PieroV> TBH, I don't understand why Android has the arguments both in the exec and there
15:04:43 <richard> ~*legacy*~
15:04:45 <PieroV> But I guess it's not a talk for this meeting
15:05:01 <PieroV> Yep
15:05:01 <donuts> o/ hello, sorry i'm late
15:05:05 <donuts> had to switch locations
15:05:09 <richard> donuts: unbelievable
15:06:01 <donuts> it's 2023, you'd think mobile internet would work better in major cities :P
15:06:45 <richard> lol yeah you'd think
15:07:10 <ma1> it does outside US, or so they told me
15:07:44 <ma1> (I did security backports on a train over the weekend)
15:07:59 <richard> ok I don't think I have any oher topics to go over
15:08:06 <PieroV> Irish trains had an okay WiFi, too
15:08:09 <PieroV> I have a topic
15:08:28 <PieroV> tor-browser!721 made some problems with initialization stuff more evident
15:08:38 <richard> my plan from now until alpha stabilizes is mostly releases, code reviews and mozilla issue reviews
15:09:15 <donuts> yeah i miss scottish trains q_q
15:09:35 <PieroV> One of my last versions made it possible to await the initialization of the tor integration, and even catch initialization problems
15:10:16 <ma1> PieroV, oh, it was the one with the stored Promise?
15:10:26 <PieroV> ma1: yes, exactly
15:11:06 <PieroV> That was exactly the reason to store it. I think Firefox itself does so for some components (was it the extensions manager maybe?)
15:11:24 <richard> so i remember you chatting about this a bit last week
15:11:39 <PieroV> Anyway, I'd like to proceed with !721 anyway, because I don't think I could possibly fix everything
15:11:49 <richard> and i didn't fully understand how this race condition worked then
15:12:06 <PieroV> If it loses the 13.0a3 train, it needs to take the 13.0a4 train, which is super late
15:12:11 <PieroV> Much later than I'd like to
15:12:56 <richard> 13.0a3 is already out :3
15:13:07 <PieroV> Oh, right. Then do +1 :)
15:13:12 <PieroV> 13.0a4 and 13.0a5
15:13:13 <richard> we're looking at 13.0a4 next week
15:13:14 <richard> ack
15:13:30 <donuts> what's my deadline for icon asset swaps?
15:13:31 <PieroV> 13.0a5 is the last alpha, in our current program
15:13:53 <richard> oof
15:14:29 <richard> sept 5th seems like a good cutoff
15:14:40 <donuts> richard: ack, ty
15:15:18 <richard> PieroV: once !721 is in can you prioritize the remaining 131 must-haves?
15:15:27 <PieroV> yes
15:16:44 <richard> ok, I don't think i have any other topics
15:16:46 <richard> anyone else?
15:16:59 <PieroV> I think I wasn't done with mine ^_^
15:17:08 <dan_b> i got the big android bootstrap MR out
15:17:30 <richard> aaaaay
15:17:30 <dan_b> and it catches at least one bug that irked me for ages, sometimes tb didnt transition to home on bootstrap! now fixed!
15:17:30 <richard> nice
15:17:42 <clairehurst> Remember this is a short week! We have Thursday and Friday off
15:17:46 <dan_b> yes!
15:17:49 <richard> yes!
15:17:55 <richard> and Monday iirc
15:17:58 <dan_b> so I'd love if clairehurst can at least take a look at it
15:18:06 <dan_b> just to see if anything obviously bad pops out
15:18:21 <dan_b> and yeah, if we want for 13.5 it's much more easily refactorable
15:18:21 <richard> and then come Sept flex Fridays start
15:18:24 <clairehurst> Will do! would be a good opportunity to try to get android building on my machine too
15:18:26 <boklm> richard: are we moving next meeting to tuesday?
15:18:36 <richard> yes indeed!
15:18:38 <richard> boklm^
15:18:46 <boklm> ok
15:18:46 <dan_b> also pierov's req to shrink Tor Integration commit is wrapped up in this as the bootstrap is a new commit and onboarding was removed too
15:19:07 <jagtalon> woo 3-day work week!
15:19:10 <PieroV> dan_b: please don't close that issue yet
15:19:15 <dan_b> which?
15:19:19 <PieroV> The one about shrinking
15:19:29 <dan_b> sure! 👍
15:19:29 <PieroV> The Android patch set could use much more love than that
15:19:36 <dan_b> there's prolly more we can do
15:19:40 <richard> oh yes
15:19:41 <dan_b> this is just a nice start
15:20:09 * dan_b is looking forward to 13.5 for android already
15:20:09 <richard> i mean honestly the entire Android code-base could use that base-browser/tor-browser glow-up but that's a problem for 13.5
15:20:23 <richard> dan_b: very much same for me
15:21:20 <PieroV> I wanted to add something on the topic I brought earlier
15:21:55 <PieroV> The refactors on the whole integration made more visible that we don't manage the tor not working case very well
15:22:07 <PieroV> And we end up with an empty settings page because of an error
15:22:16 <PieroV> I think I can find a workaround for that
15:22:55 <PieroV> But as an alternative we could check about not populating the settings and display an error in some way
15:23:31 <PieroV> Because even with the workaround a user will be able to mess with options, they will be saved but won't have any effect
15:24:00 <richard> so this scneario can happen if tor is still launching/control port isn't setup yet?
15:24:08 <richard> or something more sinister?
15:24:23 <PieroV> Also when tor fails to start (e.g., port numbers clash)
15:25:08 <richard> ah right
15:25:23 <richard> well our UX there wasn't great to begin with right?
15:25:30 <PieroV> Nope
15:25:45 <richard> so we've traded one broken for another essentially?
15:26:07 <PieroV> Yes, that's the plan
15:26:15 <richard> gotcha
15:26:18 <ma1> having explict UX for failures is always better
15:26:28 <richard> yes
15:26:39 <richard> but
15:26:44 <PieroV> Yeah. So, we don't have time I think, but I wanted UX people to know anyway
15:27:01 <PieroV> In case they could come up with a quick idea
15:27:03 <richard> I suppose this will force us to actually propelry consier the general TorProvider failed to init problem which we'll have to handle regardless of backnd
15:27:18 <richard> properly consider*
15:27:43 <PieroV> In an issue I suggested displaying a warning also in about:torconnect, without adding a new state, which would require much longer
15:28:12 <PieroV> Otherwise, broken for broken
15:28:42 <ma1> do we have some logs at that point?
15:29:19 <PieroV> We do have a message box asking to restart Tor, and it can't be ignored
15:29:21 <richard> donuts: is championquizzer et al keeping an eye on the tor browser alpha testing situation or are we kind of flying blind here?
15:29:30 <PieroV> But if a user presses okay they won't be reminded anymore
15:29:42 <PieroV> And log access won't be possible, because logs need the control port connection
15:29:56 <PieroV> Any explicit error will be visible only in the browser console
15:30:48 <donuts> richard: I can ask at today's stand-up, but to my knowledge there hasn't been any 13.0 alpha specific feedback
15:31:02 <richard> donuts: well that's either great or terrible
15:31:08 <donuts> it's the latter
15:31:14 <dan_b> what's the window settings cannot be applied? they can be applied before we try and bootstrap? and after, but not during? could you just have it listen for bootstrap events and disable/enable?
15:31:14 <richard> yeah i know
15:31:22 <donuts> 🙃
15:31:25 <PieroV> We've had some issues from cypherpunks (not 1) regarding alpha
15:31:30 <donuts> we should try and do as much internal QA as possible
15:31:58 <PieroV> I've been dogfooding alpha for a while on Linux, never had a problem with it, but I use a quite standard setup
15:32:05 <richard> donut: do you have availability to draft a mail to tor-internal or shall I?
15:32:13 <richard> or we can somehow tagteam
15:32:19 <PieroV> dan_b: about:preferences#connection
15:32:30 * ma1 imagined it was the "terrible" alpha testing case because we already know about some systemic layout breakages which would be quite... visible.
15:32:39 <richard> PieroV: yeah fortunately most uers use th standard setup
15:32:51 <dan_b> aaah hm
15:33:06 <richard> Jeremy: if you haven't already can you poke you the whonix people about testing their setup with the alpha?
15:33:22 <PieroV> ma1: oh, yeah, I didn't look for "cosmetic" problems
15:33:46 <richard> i've been mailing the tails people with each release as well bu thaven' theard back from boyska or integri
15:33:47 <Jeremy_Rand_36C3[m]> richard: yeah sure -- any particular alpha release you want them to test, or just "any recent alpha"?
15:33:53 <donuts> richard: I wonder if we should hold off until after 13.0a4 at least
15:34:02 <donuts> because there are a bunch of known issues we're still working through
15:34:13 <richard> ideally the most recent alpha and then each one that follows XD
15:34:13 <PieroV> Jeremy_Rand_36C3[m]: the more recent the better, because we've been working on tor integration hardly
15:34:23 <PieroV> To prepare things for Arti transition
15:34:24 <donuts> but it would be good to dedicate some apps team time to general QA
15:34:26 <Jeremy_Rand_36C3[m]> richard: ok, I'll ping them
15:34:48 <donuts> richard: e.g. designate a specific date as QA day, and get the whole team testing across each platform
15:35:14 <richard> thx jeremy
15:35:15 <donuts> Me or jag can join from a UX pov too
15:35:36 <richard> sounds like a plan
15:35:53 <richard> and fwiw rui et al have been doing regular QA w/ each alpha release and have been filing issues
15:36:05 <PieroV> There has been a small incident
15:36:07 <richard> which i realise doesn't cover the complexity of tor browser :3
15:36:09 <PieroV> Regarding MB alphas
15:36:12 <richard> oh?
15:36:14 <jagtalon> donuts: ack
15:36:16 <richard> 'incident'?
15:36:39 <PieroV> Yes, with the new naming scheme. The last release on GitHub has the incomplete system-wide installer
15:36:52 <PieroV> We must have lost part of the information somewhere
15:37:02 <PieroV> So, new alpha users won't have working updates
15:37:11 <richard> oh instead of the portable?
15:37:14 <PieroV> Yes
15:37:16 <richard> or in-addition to
15:37:21 <PieroV> Instead
15:37:21 <richard> ahhh shit
15:38:13 <richard> well its not the end of the world
15:38:27 <richard> we can make sure to note it in the 13.0a4 change notes
15:38:28 <PieroV> Yeah, it's only new Windows alpha users
15:38:44 <PieroV> JB is already on it, maybe they already changed
15:39:56 <richard> wild the alpha stats on github are almost entirely windows
15:40:18 <richard> anywya
15:40:19 <PieroV> Oh, GitHub has stats! I didn't notice them
15:40:29 <richard> https://somsubhra.github.io/github-release-stats/?username=mullvad&repository=mullvad-browser&page=1&per_page=5
15:42:07 <richard> donuts: ok I'll plan on a tor-internal mail to go out with 13.0a5
15:42:12 <championquizzer> richard: donuts: o/ yeah, qa for tb alpha is on my radar and am keeping an eye on it, along with the docs updates. i need to devote more time though and is in my plans for this week
15:42:21 <richard> <3
15:42:39 <PieroV> Thanks championquizzer!
15:43:43 <richard> anything else for today from anyone?
15:44:10 <dan_b> ah I have to go stand in line for a while today so I'll be out from 12 my time till.. an hour or two after?
15:44:21 <richard> have fun queueing!
15:44:27 <dan_b> always 😕
15:45:49 <richard> (:
15:45:53 <richard> ok
15:46:00 <richard> have a good day everyone
15:46:06 <richard> see you on irc o/
15:46:08 <richard> #endmeeting