14:57:59 <richard> #startmeeting Tor Browser Weekly Meeting 2024-01-22
14:57:59 <MeetBot> Meeting started Mon Jan 22 14:57:59 2024 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:57:59 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:58:05 <richard> le pad: https://pad.riseup.net/p/tor-tbb-keep
14:58:17 <boklm> o/
14:58:38 <richard> This week we'll be  releasing Tor+Mullvad Browsers 13.0.9
14:58:48 <richard> (thanks for signing last week boklm)
14:59:18 <richard> I've signed TB this morning so we'll be ready to deploy tomorrow on time once again
14:59:25 <thorin> /o brb in 5
14:59:29 <dan_b> o/
14:59:37 <richard> but of course the big priority this week is S96
15:00:14 <richard> I ping'd those of you focused on s96 last week on IRC, but this is the last week we have beore we need to send code off for security review
15:00:32 <clairehurst> o/
15:00:40 <richard> my understanding is that the security people will be starting browser stuffs on the 29th
15:01:11 <Jeremy_Rand_36C3[m]> Hi!
15:01:26 <richard> typically the way this works is that we send them branches with commits/commit-ranges for them to look at
15:01:57 <richard> since our stuff is typically all squashed together in ever changing 'feature commits' this shouldn't be too taxing
15:02:55 <dan_b> cool, I had some good progress by the end of last week, I'm hoping to have my new TorController wrapped today/tomorrow and get out MRs with Settings and events working using the new geckoview stuff only
15:03:56 <richard> my current thoughts on this is to create a separate 'review' branch for each of our affected projects where it makes sense (tor-browser, firefox-android) which will basically be the latest alpha with any in-flight fixups! squashed together
15:04:55 <richard> henry-x, cohosh: how is the Lox integration going?
15:05:22 <dan_b> makes sense
15:06:31 <Jeremy_Rand_36C3[m]> (it sounds like you guys are all super busy this week, so whatever stuff I'm doing will presumably not get your attention?)
15:07:04 <richard> jeremy: basically we've been s96 non-stop since the end of october, but that's coming to an end one way or the other come the end of this month
15:07:16 <richard> (beyond resolving issues from the security review)
15:07:31 <Jeremy_Rand_36C3[m]> richard: got it, thanks. Good luck on the final push!
15:07:58 <richard> (thanks!)
15:10:36 <clairehurst> Should I work on polishing up the html UI or try more to get Native to work properly? HTML right since thats what we're shipping?
15:11:20 <richard> if there are easy wins for the HTML then I 'd say we can take those this week
15:11:38 <dan_b> clairehust: my new torcontroller should get you a big step of the way there
15:11:43 <richard> the thing that comes immediately to mind is the white bar at the bottom of the page
15:11:51 <richard> in terms of 'visual impact'
15:12:35 <richard> but otherwise i would say focus on work we're planning on shipping in 13.5 come summer
15:12:39 <richard> (so Native :) )
15:13:14 <clairehurst> Cool!
15:14:02 <richard> iirc PieroV has a MR our for fixing a crash on older Android versions for you to review as well
15:14:21 <cohosh> richard: hey sorry just saw the ping
15:14:29 <clairehurst> A decenlty big thing is if you leave the app running for *a while* you come back to a "File not found" screen (a broken about:tor page)
15:14:56 <cohosh> i'm working on henry's latest feedback and should be done with that today, and the eta for the genNextUnlock feature is tomorrow hopefully
15:15:06 * eta for the genNextUnlock feature
15:15:08 <clairehurst> I've noticed it dogfooding
15:15:44 <dan_b> and you can swipe "back" on bootstrap and get to landing page with tor unloaded
15:17:08 <richard> cohosh: ack
15:18:39 <richard> ok does anyone else have topics to discuss?
15:19:06 <ma1> What about stable releases on Tuesdays at 14 UTC?
15:19:34 <richard> ma1: that sounds like a solid plan
15:19:46 <richard> made easier by doing mullvad browser earlier
15:20:10 <ma1> i.e. perfect sync theoretically with Firefox, and fixup release in the unlikely event that draft vs public advisories don't match
15:20:58 <richard> sounds like a plan to me
15:22:37 <henry-x> richard: sorry, I missed that the meeting started. I'm waiting on input from UX for tor-browser!890. Starting to write some of the other changes ready for the Lox backend
15:23:26 <richard> henry-x: ack
15:23:59 <jagtalon> henry-x: hoping to get to that soon!
15:24:10 <jagtalon> sorry for the delay
15:24:31 <richard> jagtalon: if you could prioritise it we're on a *bit* of a time crunch
15:24:41 <jagtalon> richard: you got it!
15:24:49 <richard> ugh and speaking of crunches
15:24:55 <richard> we have a surprise android problem!
15:25:04 <richard> (our x86 apk is too big for google play)
15:25:11 <dan_b> lol
15:25:18 <donuts> I'll take a look tomorrow too
15:25:22 <donuts> thanks henry-x
15:25:24 <richard> Google Play: "Reduce your file size to 100MB or use expansion files"
15:25:48 <richard> clairehurst, dan_b: do either of you know anything about this^
15:26:07 <dan_b> we aren't even currently packaging the lox bridge binary yet are we? which i assume will be adding a few more MB?
15:26:16 <dan_b> no hadn't seen that ever before
15:26:24 <clairehurst> Off the top of my head no but I can look into it
15:27:07 <richard> apparently the solution is either use an App Bundle which has a higher limit than an apk or use an expasnion file which is a type of patch if I'm understanding correctly
15:27:15 <PieroV> Can we increase the zip compression?
15:27:27 <richard> in theory vOv
15:27:31 <PieroV> And are we sure we're stripping everything? Maybe we have some binary that isn't stripped
15:27:33 <dan_b> ooh lol we're still shipping an .apk to google not a .aab?
15:27:38 <richard> but that's either a PeiroV or a boklm problem
15:27:44 <richard> :D
15:28:02 <PieroV> dan_b: never heard of aab :)
15:28:09 <richard> me neither^
15:28:11 <dan_b> yeah its googles new prefered format
15:29:01 <dan_b> i am lucky cus for cwtch builds the difference is `flutter build apk` to `flutter build appbundle`
15:29:34 <ma1> and would it be retrocompatible (heard PieroV talking about retro-support on Android this morning)?
15:29:52 <richard> also a good point
15:29:57 <dan_b> but i don't think folks can manually install .aab? so open privacy publishes the .apk for folks and ships the .aab to google play store
15:30:30 <dan_b> which means we build both, which for TB is deeeeef gonna increase android build step time some, but maybe not too bad as it should mostly be packaging
15:30:35 <richard> boklm/PieroV: can you prioritise investigating if we have any easy wins for making our existing apk smaller?
15:30:44 <boklm> tor-browser-android-x86-13.0.7.apk was 109M so I'm wondering why we didn't get the error before
15:31:16 <richard> there seems to be some different rounding involved
15:31:19 <dan_b> .aab being newer and more modular in some way means when google p;ay ships updates it can ship partials and leave unchanged bits alone, so they save on a lot of bandwidth, and its easier on users
15:31:24 <richard> the current x86 is 110mb
15:31:37 <richard> the only one less than 100 mb is armv7
15:31:48 <richard> and google play isn't complaining about them
15:32:30 <ma1> new rule maybe, or unit rounding.
15:33:00 <Jeremy_Rand_36C3[m]> boklm: is there a MB / MiB confusion in the docs perhaps?
15:36:10 <richard> seems likely
15:36:12 <richard> anyway
15:36:27 <richard> we can move this discussion over to #tor-browser-dev and let folks go from here
15:37:37 <dan_b> sure
15:41:16 <richard> alright folks
15:41:18 <richard> have a good week
15:41:20 <richard> #endmeeting