14:59:16 <richard> #startmeeting Tor Browser Weekly Meeting 2022-09-26
14:59:16 <MeetBot> Meeting started Mon Sep 26 14:59:16 2022 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:59:16 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:59:21 <richard> hello hello
14:59:25 <henry-x> hello
14:59:27 <richard> pad per usual: https://pad.riseup.net/p/tor-tbb-keep
14:59:50 <Jeremy_Rand_36C3[m]> Hi!
14:59:56 <ma1> o/
14:59:58 <dan_b> hello
15:00:04 <boklm> hi
15:05:16 <richard> ok it looks like pad activity has settled down
15:05:58 <richard> no major announcements or discussion points from me this week
15:06:13 <PieroV> I have a bolded item
15:06:43 <richard> build tags for alpha should be happening today; did a test alpha build and all went fine so should be smooth sailing there
15:06:53 <PieroV> richard: no plz
15:06:59 <richard> and stable tags hopefully by the end of the week pending YEC stuffs
15:07:06 <PieroV> I think that the font issue is very important
15:07:20 <PieroV> I think that it should be a blocker for the alpha
15:07:27 <richard> oh really?
15:07:28 <PieroV> We have a few workarounds
15:08:02 <PieroV> Yes. As I said last week, all non-Latin script users are screwed in Ubuntu/Fedora at the moment with 12.0a2
15:08:44 <PieroV> So, even though the Bugzilla ticket is still at the moment, we should use one of the workarounds
15:09:44 <richard> alright, we'll have alpha tags after the fonts are fixed :D
15:09:59 <donuts> woops, HELLO
15:10:07 <PieroV> Currently the issue is assigned to dan_b, but we can reassign if needed
15:10:35 <richard> PieroV: can you point me to w/e relevant info/patch/pref changes needed to fix?
15:10:53 <PieroV> tor-browser#41163
15:11:01 <dan_b> i have an MR up for it at tor-browser!365 with one of the work arounds
15:11:22 <richard> ok then I'll test/review/merge it after this
15:11:40 <PieroV> Thanks!
15:11:52 <dan_b> with notes from PieroV on why the current state of the MR may not be optimal
15:12:08 <dan_b> but then I switchtedto YEC22 work friday
15:12:53 <PieroV> yes, the TL; DR is that the fix is an okay workaround, but it needs to be in ifdef XP_LINUX in a preference file, rather than StaticPrefs.yaml
15:13:16 <dan_b> ok I can move that pretty quickly and update it today
15:13:25 <richard> dan_b: wfm
15:14:00 <richard> henry-x: iirc we had to solve this same font issue in last year's YEC, I'll take a peak at it after the meeting and get back to you
15:14:42 <richard> (beyond the font embedded in the SVG last year)
15:15:24 <richard> ok PieroV: what's the good news on the tor-launcher migration?
15:15:43 <PieroV> I think that either today or tomorrow I'm moving to the last file
15:15:56 <henry-x> Yeah, you embedded it as `src: url('data:application/octet-stream;base64,` last year, but the chat on the bug between you and PieroV implies it could be done some other way to avoid the very long stream
15:16:19 <PieroV> So, I think we should move the localization soon. As a workaround we could move it to torbutton
15:17:14 <donuts> I'm still catching up, but are there two fonts in the artwork?
15:17:27 <richard> donuts: yeah
15:17:32 <PieroV> donuts: yes, or 3 if you count the bold
15:17:36 <henry-x> PieroV do you know why loading a font from `resource:` is not allowed in "about:tor", but loading images from `resource:` is ok?
15:17:52 <PieroV> henry-x: nope. But you can check it on about:manual
15:18:04 <donuts> pierov: can we do variable in TB (if that's smaller than two weights)?
15:18:29 <PieroV> donuts: I've thought about that, but it depends on the OS, I think
15:18:33 <PieroV> Rather than on Firefox
15:18:39 <donuts> space mono is obvs very important to the look and feel of the campaign
15:19:02 <donuts> the sans-serif is less so, if it's resulting in many kbs
15:19:29 <donuts> anyway I'll catch up with the ticket and repeat the same there :)
15:19:44 <PieroV> henry-x: I think that a reason is that I've set chrome:// URLs in about:manual, but maybe I should have used resource://
15:20:15 <PieroV> donuts: they aren't that big, ~100kB or ~130kB when zip compressed, iirc
15:20:32 <donuts> pierov: cool, ty
15:20:33 <PieroV> I think it could be done, for a few releases
15:20:44 <richard> i suspect either way it's smaler than last year so we should probably not agonize abou tit
15:20:52 <donuts> ha ok fair
15:21:01 <donuts> LESSONS WERE LEARNED
15:21:09 <richard> lol
15:21:54 <richard> re localization migration
15:22:17 <henry-x> richard how did you create the base64 octet stream last year?
15:22:19 * emmapeel raises ear
15:22:38 <richard> henry-x: most likely a one-off script
15:23:08 <PieroV> oh, now that I think of it
15:23:12 <PieroV> I've tested the sizes of ttf
15:23:16 <PieroV> We could use woff2
15:23:24 <PieroV> And the size would diminish
15:25:05 <richard> I'm somewhat hesitant doing that w/o understanding the differneces between woff2 and ttf encoding (eg can the conversion occur losslessly?)
15:25:26 <PieroV> richard: woff2 are optimized for the web
15:25:33 <PieroV> And I think that it can be lossless
15:25:49 <PieroV> I have both sfnt2woff and woff2sfnt
15:26:00 <PieroV> (but maybe this is for woff 1?)
15:26:19 <richard> you ocan always pop them into fontforge :p
15:26:51 <PieroV> The sans is more than half!
15:26:58 <PieroV> Err, less than half
15:27:05 <richard> if it works it works, just make sure the glyphs still look right :p
15:27:11 <richard> but anywya
15:27:28 <richard> regarding localization migration:
15:28:01 <PieroV> (yes, it's lossless, just tried)
15:28:06 <richard> it seems like a reasonable stop-gap
15:29:00 <richard> maybe it will be easiser to migrate ultimately to fluent if we only have 1 system to migrate from
15:29:24 <richard> what's your reasoning in moving the strings from tor-launcher -> torbutton rather then doing the fluent migration all at once?
15:29:37 <richard> let's us purge tor-launcher sooner?
15:30:00 <PieroV> Yes
15:30:14 <PieroV> So that we can purge one for starters
15:30:27 <PieroV> And if things with torbutton go long, we can do that for 12.5
15:31:09 <richard> yeah ok that makes sense
15:31:16 <PieroV> And to parallelize works more
15:31:35 <PieroV> My plan for the fluent migration was to start with New Identity and Security Level, which are small enough
15:31:39 <PieroV> (or even with one of them)
15:31:49 <PieroV> Since we have to test the whole new pipeline
15:31:51 <richard> can we make that migration a secondary MR from the tor-launcher functionality migration MR?
15:32:27 <PieroV> Yes. I've deleted a few strings in the current MR
15:32:36 <PieroV> I can cherry pick on another branch, in case
15:32:44 <PieroV> and remove them from the current MR
15:32:51 <richard> ok perfect
15:33:21 <PieroV> But we will need to change also the script to import translations in torbutton
15:33:51 <richard> yes
15:35:26 <richard> dan_b: can you prioritize the alpha work (go upgrade patch, font fix) over the YEC changes this week?
15:35:53 <dan_b> i'll get the font fix out today
15:36:07 <dan_b> the go thing I have no idea what that's going to entail but I can start looking
15:36:30 <dan_b> my fenix YEC22 isn't too far along, i only have hte YEC21 patch applied and showing and partially date gated
15:36:53 <richard> well small steps
15:37:02 <dan_b> but if that's cool then yeah I can start digging into whats up with go
15:37:19 <PieroV> richard: we won't be able to update obfs4proxy anyway
15:37:23 <richard> well unless the go thing is a quick fix then it can wait until next week
15:37:54 <richard> ah ok then go can wait until after YEC, ignore that bit
15:37:55 <PieroV> but updating go will be a good thing anyway :)
15:37:59 <dan_b> cool
15:38:18 <dan_b> font MR > YEC22 > go :)
15:38:26 <richard> yep exactly
15:38:39 <richard> anything from anyone else?
15:38:55 <PieroV> only organizational
15:38:59 <PieroV> I will be afk on Oct 5
15:39:03 <PieroV> then nothing else from me
15:39:37 <ma1> I'm getting serious about native UI for per-tab notifications, adding a WebExtensions API
15:40:19 <ma1> donuts just asked for a rebranding of the notification bar (indeed it looks ugly with that foxy orange gradient) and I've just verified it's doeable as well
15:40:34 <donuts> 🙌
15:40:36 <donuts> fantastic!
15:40:42 <PieroV> donuts: you should invite people to figma
15:40:49 <PieroV> So that they can see the inspector
15:40:59 <ma1> (BTW we already show that ugly bar for other stuff, so it would be a win anyway)
15:41:22 <Jeremy_Rand_36C3[m]> a bit of cool stuff on my end.
15:41:26 <donuts> pierov: oh actually I think that file's open, so anyone can inspect it with an account – but will invite everyone to the whole UX account anyway
15:41:28 <ma1> donuts, the current gradient is linear-gradient(0, #9059ff 0%, #ff4aa2 52.08%, #ffbd4f 100%); Just tell me the colors you prefer
15:41:36 <donuts> ma1: my thoughts exactly :D
15:41:36 <donuts> will do!
15:41:42 <donuts> thanks for the reference
15:41:42 <richard> I'll be afk this afternoon, and then i'm also someewhat afk next Monday and all of next Wednesday (3rd and 5th)
15:42:38 <henry-x> PieroV I created a figma account by myself without an invite and I could see the inspector for YEC2022
15:42:52 <Jeremy_Rand_36C3[m]> Electrum-NMC is now about 2-2.5x lower latency for resolving names (cut down to 1 round trip, very helpful over Tor), and it also verifies scriptSigs on the name transactions it fetches (sort of like a full node, but as scalable as a usual SPV node)
15:43:05 <donuts> henry-x: yeah it's unlocked, but if I invite you to THE TEAM i think it'll make it easier to find these files in future
15:43:19 <PieroV> Jeremy_Rand_36C3[m]: cool, congrats!
15:43:19 <donuts> or it'll make no difference at all
15:43:19 <Jeremy_Rand_36C3[m]> Electrum-NMC now also supports Tor environment variables (good for system Tor / Whonix / Tails / etc) as well as SOCKS over Unix sockets
15:43:24 <donuts> _shurgs_
15:43:58 <richard> Jeremy: neat
15:44:12 <donuts> ironically richard never accepted their ux team invite :<
15:44:30 <richard> ah well
15:44:41 <richard> i was probably distracted by something else shiny
15:45:05 <PieroV> Jeremy_Rand_36C3[m]: when I'll finish with tor-launcher I'll probably go after torbutton
15:45:19 <PieroV> And we'll need to revisit your patches
15:45:52 <Jeremy_Rand_36C3[m]> PieroV: ah cool.  Yeah we should discuss that sometime -- I think part of the patch it redundant since it duplicates a UI element that the SecureDrop stuff has
15:46:03 <PieroV> (I wonder if torbutton is also needed to merge the GeckoView-TBB desktop branches)
15:46:11 <Jeremy_Rand_36C3[m]> So yes we should consider properly refactoring that stuff
15:46:24 <donuts> invites away!
15:46:52 <dan_b> donuts: got it! thanks
15:46:56 <donuts> I also noticed henry isn't on the Applications team meeting invite
15:46:59 <donuts> will see if I can fix
15:47:01 <Jeremy_Rand_36C3[m]> PieroV: please CC Arthur and me in any tickets about that
15:47:32 <PieroV> Jeremy_Rand_36C3[m]: I hope to remember to, please remind me if I start talking about torbutton stuff without doing so
15:47:43 <Jeremy_Rand_36C3[m]> PieroV: ok, no worries
15:47:48 <PieroV> (I don't know if we have an issue for it, yet)
15:49:39 <boklm> tor-browser#41286
15:51:12 <richard> ok, anything else before I endmeeting?
15:51:21 * Jeremy_Rand_36C3[m] has nothing more
15:51:26 * PieroV neither
15:51:31 <henry-x> nope
15:51:34 * boklm neither
15:51:41 <dan_b> nope
15:51:45 <richard> ok then have a good week everyone
15:51:52 <Jeremy_Rand_36C3[m]> thanks!
15:51:55 <PieroV> Thanks!
15:51:59 <richard> #endmeeting