15:02:02 <morganava> #startmeeting Tor Browser Weekly Meeting 2024-07-29
15:02:02 <MeetBot> Meeting started Mon Jul 29 15:02:02 2024 UTC.  The chair is morganava. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:02:02 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:02:14 <dan_b> o/
15:02:29 <jwilde> o/
15:02:32 <morganava> our pad as usual -> https://pad.riseup.net/p/tor-tbb-keep
15:02:34 <boklm> o/
15:03:23 <morganava> I'll be afk this week on Wednesday and Thursday
15:03:34 <morganava> i'll be somewhat online but won't be able to reliably make meetings
15:03:55 <morganava> if you've 1:1s w/ me this week let me know if you'd like to reschedule for tomorrow
15:04:04 <morganava> and next, I ran into a few issues signing 14.0a1 and only some of them were of my own doing
15:04:05 <dan_b> i do
15:04:29 <morganava> I opened an issue for the macOS signing problem with the logs, but haven't dived any deeper than that
15:04:48 <brizental> o/
15:05:13 <morganava> and  itlooks like boklm is on it per gitlab
15:05:53 <boklm> yes
15:06:26 <morganava> thanks thanks
15:06:27 <boklm> tor-browser-build#41199
15:07:03 <morganava> i was able to unblock myself otherwise by just signing from my local machine
15:07:21 <morganava> so we're good there
15:08:05 <morganava> ok that's all from me, on to discussion points?
15:08:08 <morganava> problems, blockers?
15:08:11 <PieroV> I have a couple
15:08:15 <morganava> have at it!
15:08:20 <PieroV> 1. I think we should disable translations
15:08:38 <PieroV> It's a bad UX, and I think our settings make it worse (e.g., for web-exposed locales?)
15:09:03 <PieroV> I think we should disable it for now, and if we have time/think it's important check the incompatibilities and try to solve them with upstream
15:09:27 <PieroV> I was even asked to translate a PDF from Italian to English
15:09:49 <PieroV> (the PDF was in English, but PDF.js was in Italian because I was dogfooding and checking for localization problems while doing so)
15:10:15 <morganava> wfm, side-note is it a local translation model, or does your data get uploaded somewhere?
15:10:23 <PieroV> It should be offline
15:10:38 <morganava> small victories
15:10:43 <PieroV> And you can download languages in the settings
15:10:57 <dan_b> how big are they?
15:11:00 <PieroV> But I don't know if the first time you use a language the definition is downloaded from Moz's servers otherwise
15:11:02 <morganava> but our patches don't play nicely
15:11:23 <PieroV> dan_b: I don't know, but we don't want to grow too much also on desktop
15:11:50 <dan_b> oh i was just curious that they made a translator portable enough both in size and efficency to ship
15:11:50 <PieroV> Because smaller sizes always help with alternative download methods (and also with updates over slow circuits)
15:12:22 <PieroV> Oh, I see. Still, I don't know :)
15:12:44 <morganava> i would guess, probably not :p
15:13:34 <PieroV> Anyway, if nobody opposes to this, I'll open a MR to disable them for now (at least with a pref), then we can fix any residuals before the final release
15:13:48 <morganava> well, i think we would want to disable anyway before even without the issue
15:14:14 <morganava> can you also create an issue to investigate somehow enabling for the 14.5 series
15:14:26 <PieroV> We already have an issue about investigating the translations
15:14:30 <morganava> well perfect
15:14:31 <PieroV> I can keep it open
15:14:49 <PieroV> Maybe I can create a separate issue about disabling them for the changelogs
15:14:53 <morganava> yes please
15:15:06 <PieroV> okay. My next point is about allowed_addons.json
15:15:20 <PieroV> It's the cause for the Android build failures
15:15:35 <PieroV> We updated NoScript, but we still check for the hash in that file for Android
15:15:45 <PieroV> And since 14.0a1 was desktop-only, we didn't update that file
15:16:15 <PieroV> The quick fix would be to update the file, but it isn't used anymore
15:16:22 <morganava> aren't we inheriting some change fro 128 that makes that json file unnecessary?
15:16:27 <PieroV> No no
15:16:53 <PieroV> We reverted our patch that used that file, because the addons cache was always stale as the result of having a custom addon provider
15:17:33 <PieroV> If we are sure we won't restore that patch we can just delete that file and the tools to update it
15:17:52 <PieroV> (or we can do it anyway and then restore also the tools)
15:17:58 <PieroV> Any objections to this?
15:18:22 <dan_b> sounds fine?
15:18:42 <PieroV> I think at a certain point people were in favor of a custom addons provider
15:19:20 <PieroV> But probably a smarter one than caching AMO data (even though there's still the known bug of installed addons before bootstrapping)
15:20:56 <morganava> i'm not sure i remember the trade-offs well enough to have an opinion on this
15:22:51 <ma1> My 2 cents is the less the moving parts the better. Let's try it.
15:23:10 <dan_b> +1
15:25:02 <PieroV> ack, will do
15:26:10 <morganava> brilliant
15:26:24 <dan_b> o/
15:26:28 <dan_b> https://gitlab.torproject.org/dan/tor-browser/-/tree/tor-browser-128.0esr-14.0-2-android
15:26:31 <dan_b> exists 🙂
15:26:40 <dan_b> got the last step done late last week
15:26:48 <dan_b> just trying to see if I can get iut to build now
15:26:56 <dan_b> what's the next step for getting it merged?
15:27:52 <PieroV> dan_b: what's it current status?
15:27:56 <morganava> dan_b: does that branch just have the patches sitting atop 128.0esr-14.0-2?
15:28:00 <PieroV> I think that for MR it should:
15:28:00 <dan_b> yes
15:28:02 <PieroV> 1. build
15:28:19 <PieroV> 2. not have trivial crashes :D (e.g., crashes at start)
15:28:21 <dan_b> working on that
15:28:58 <PieroV> I think you could open a MR already if you want, but maybe mark it as draft until it's ready for review
15:29:03 <dan_b> cool
15:29:06 <PieroV> We could start reviewing it already if we have time
15:29:11 <dan_b> I can list my many intermediary branches
15:29:22 <dan_b> if that would be useful?
15:29:26 <PieroV> Yes
15:29:31 <dan_b> cool will do
15:29:46 <PieroV> Do you think you could do a commented range-diff like I did for desktop?
15:29:48 <morganava> and potentially a small walkthrough/summary of the changes in the MR if that is reasonable
15:29:56 <morganava> yes what PieroV said :D
15:30:06 <dan_b> link?
15:30:40 <PieroV> https://gitlab.torproject.org/-/project/472/uploads/120cb0906bfa88b5c0c265fb61dfd2fa/tor-browser-115-128.html
15:31:35 <dan_b> how did you make this?
15:31:57 <dan_b> you did a range diff? than html ified it? and which parts are your added comments?
15:31:57 <PieroV> I used a tool called aha and piped it to a script to split the output by commit (https://gitlab.torproject.org/pierov/lazy-scripts/-/blob/main/split-range-diff-aha.py)
15:32:12 <PieroV> My comments are the parts outside the <pre>
15:32:39 <dan_b> cool
15:32:41 <dan_b> I can give it a try!
15:32:42 <PieroV> I used range-diff with the explicit --color to make it output colors also when piped (+ --color-moved and the option to ignore whitespaces, I never remember what it is)
15:32:49 <PieroV> Then I piped it to aha and to my script
15:35:25 <morganava> maybe this process hould be a script :3
15:36:07 <dan_b> I'll see about that too 🙂
15:36:09 <PieroV> The big part is on its own script
15:36:17 <dan_b> ^1 seems like
15:36:28 <dan_b> but I can always add a wrapper to automate even more
15:36:33 <morganava> \o/
15:36:40 <PieroV> I didn't make it take the ranges because only with git I get fish's autocompletion :D
15:37:06 <PieroV> Also, I don't know how to pipe with Python's subprocess
15:38:55 <morganava> ok sounds like a plan then
15:39:00 <morganava> anything else?
15:45:20 <morganava> boklm: once the macOS issue is resolved, I'm happy to pick up the remaining pieces of the 14.0a1 release
15:45:46 <boklm> morganava: ok, I'm working on it at the moment
15:46:07 <morganava> also all i've a draft release blog post that goes a little bit into the esr process and giving a lay-explanantion for what we do each year -> https://gitlab.torproject.org/tpo/web/blog/-/merge_requests/328
15:46:54 <morganava> donuts^ fyi
15:47:00 <dan_b> ha any idea why nc calendar is not showing a browser release meeting this wed, but my google calendar subscribed to it still is?
15:47:17 <morganava> vOv
15:47:20 <morganava> NextCloud!
15:47:54 <morganava> hmm one sec
15:48:19 <morganava> ah ok i def can't make that meeting
15:49:17 <morganava> alright, if there's no more poitns fro y'all
15:49:27 <morganava> points from*
15:49:31 <morganava> o/
15:49:36 <morganava> #endmeeting