15:00:52 <richard> #startmeeting Tor Browser Weekly Meeting 2024-05-28
15:00:52 <MeetBot> Meeting started Tue May 28 15:00:52 2024 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:52 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:01:03 <richard> the pad per usual -> https://pad.riseup.net/p/tor-tbb-keep
15:01:20 <richard> some folks are afk this week on vacations/recovering from the dev meeting
15:01:41 <richard> but the Mozilla machine keeps raging, so it's a rebase/build week for 13.0.16
15:01:55 <PieroV> Is it?
15:01:59 <richard> isn't it?
15:02:03 <PieroV> No
15:02:10 <PieroV> It's June 10
15:02:10 <richard> don't we have tags?
15:02:13 <richard> or am i off by a week
15:02:14 <PieroV> No
15:02:17 <PieroV> I think so
15:02:23 <richard> ok amazing
15:02:34 <PieroV> Err, June 11
15:02:36 <richard> that aligns well with irl constraints
15:02:56 <PieroV> And we intended to go directly with 13.5
15:03:13 <ma1> it won't align with irl all-tpi holidays very well in July :P
15:03:16 <PieroV> FWIW, I haven't disabled the 13.0 branch in my rebase script
15:03:25 <PieroV> * branches
15:03:35 <PieroV> So, the rebases are still automatically done and available in my fork :)
15:03:47 <richard> \o/
15:04:12 <dan_b> o/
15:04:14 <richard> ok, we'll see how the remaining android work fits in the next two weeks
15:04:39 <richard> dan_b: can you pick up the remaining tor-browser-build work to purge our old tor backend?
15:04:57 <richard> the relevant issues are all assigned to clairehurst atm
15:05:03 <dan_b> ah yeah sure happy to
15:05:21 <dan_b> or so i was gonna start this week off with digging deeper into the firefox-android rebase
15:05:34 <dan_b> but i've got some bridge config issues, and the tbb stuff, that'll likely fill my next 3 days
15:05:40 <richard> we also have a somewhat important android bug reported last week
15:05:45 <dan_b> yep
15:05:48 <richard> ah yeah tee ridge issue^
15:05:52 <richard> the* bridge*
15:05:53 <dan_b> that's my first priority
15:06:05 <richard> yes please
15:06:30 <dan_b> really looking forward to the tbb ripping of TAS out too
15:06:33 <dan_b> should be a fun week with that
15:06:47 <dan_b> the bridge stuff we'll see but super eager to make the new GV bakcend keep ... working lol
15:06:55 <dan_b> and fix issues from the switch over
15:07:10 <richard> i'm going to go through the remaining 13.5 labeled issues tonight/tomorrow and start kicking things to 14.0
15:07:28 * PieroV is working on localizing the Windows installer right now for 13.5
15:07:38 <PieroV> I don't know of any other 13.5 items for me at the moment
15:08:01 <richard> i think desktop is in a pretty good spot atm
15:08:22 <dan_b> sweet
15:08:41 <richard> also next week bea will be starting, so I expect bella will be scheduling more on-boarding sessions like w/ jwilde
15:08:54 <dan_b> I'm glad the now default GV backend on android is hopefully getting more eyes and mostly proving good. if we just have this one bug left to squash that would be amazing
15:09:05 <dan_b> ah cool
15:09:10 <dan_b> I like onboarding stuff
15:09:17 <PieroV> Next week we might have the "review day"
15:09:23 <dan_b> review day?
15:09:25 <PieroV> I know we have a lot going on
15:09:34 <PieroV> But it'd be great not to postpone it
15:09:40 <PieroV> dan_b: yep, it's an idea Bella had
15:09:54 <dan_b> i missed that or forgot it? what is?
15:10:00 <PieroV> For the ESR rebase we could say that on a certain day, we all pause our tasks
15:10:10 <PieroV> And instead review the ESR rebase
15:10:56 <dan_b> haha damn, you in desktop land are so far ahead of us in android land almost makes me feel nervous
15:11:16 <PieroV> I hope it'll be helpful ^_^
15:11:32 <PieroV> Also, desktop has much many more commits (we're constantly over 200)
15:12:00 <dan_b> did we have a plan for who's to look at which parts?
15:12:05 <dan_b> this is starting to ring a bell now 🙂
15:12:13 <PieroV> We need to create one
15:12:16 <dan_b> i think my brain is still defragging from last week
15:12:26 <PieroV> Even though hopefully everyone should go through the whole patchset
15:12:44 <PieroV> Also, we wanted to have a BBB thing about reviewew strategies
15:12:46 <richard> yes +1 re review day
15:12:58 <PieroV> (quick refresher on diff of diffs and how range-diff works
15:13:01 <PieroV> )
15:13:15 <PieroV> (and range-diff is one of my announcements for today)
15:13:17 <dan_b> oh yeah that part would be super helpful for me
15:14:10 <richard> we also should have a meeting re the code/ticket audit process
15:14:21 <PieroV> Yes
15:14:26 <richard> though honestly that can probably just be an email
15:14:32 <richard> or a wiki doc
15:14:34 <PieroV> I don't think it's a blocker for the whole team
15:14:39 <PieroV> Better to do it sooner rather than later
15:14:45 <richard> mmhm
15:14:54 <PieroV> But I think getting the review is the first blocker
15:15:00 <PieroV> To get everybody work on the updated codebase
15:15:11 <richard> yes 100%
15:16:16 <richard> also in my backlog is to review each of our session notes from last weeks and resolve action items
15:17:38 <richard> ok any folks have discussion points for this meeting? blockers? MRs that need attention?
15:17:47 <PieroV> Yes
15:17:49 <PieroV> I have a couple
15:17:54 <PieroV> First one
15:17:55 <richard> take it away
15:18:16 <PieroV> Yesterday I've created a "commentary" of Mullvad Browser's range-diff (since it was fresh in my mind)
15:18:27 <PieroV> (but I intend to do the same for other branches)
15:18:55 <PieroV> Basically, I went through each point that needed attention (usually it's double-signed stuff, so ++, +-, -+, --) and added a comment
15:19:20 <PieroV> To try to make it more understandable, I setup a machinery to export a colored HTML for the range diff
15:19:33 <PieroV> (git range-diff --color R1 R2 | aha)
15:19:43 <PieroV> And I added comments there
15:20:04 <PieroV> I'd like to get some feedback on the format, to possibly improve when I start doing the same for BB and TBB
15:20:33 <PieroV> (a possible feedback is also "it's a waste of time, don't do it", but I hope it won't be the feedback I get :))
15:20:34 <dan_b> send an example of the outout?
15:20:50 <dan_b> i'm a fan of color outputs so i doubt i'll vote "waste of time"
15:20:52 <PieroV> Sure, was about to find it :) https://gitlab.torproject.org/tpo/applications/tor-browser/-/issues/42441#note_3033274
15:21:05 <PieroV> There's an attached HTML (cannot really inline it in GitLab comments)
15:21:36 <PieroV> An alternative I thought of is to convert the HTML to PDF, and add inline comments, but I've already been told that HTML is better
15:21:49 <PieroV> Also, ma1 might do something for me to make the creation of inline comments easier :)
15:21:53 <richard> PieroV: I think this HTMl report ought to be integrated into the tor-browser-spec repo along with the code/ticket audits going forward
15:22:09 <PieroV> (I don't want to spend more time to add them inline than to write the comments themselves)
15:22:21 <ma1> happy to help
15:22:22 <dan_b> oh this is pretty nice
15:22:39 <PieroV> ack, in case for the record this is a quick script I wrote to split the various commits and encapsulate them in <details>: https://gitlab.torproject.org/pierov/lazy-scripts/-/blob/main/split-range-diff-aha.py
15:23:09 <PieroV> (aha itself is a nice C tool to transform ansi colors to HTML, I found it packaged on Debian)
15:23:24 <dan_b> tho i think it needs a key at top... i have green+ lines with only black text, bolds i dont get where they are coming from, red(-) lines with green text...
15:23:46 <PieroV> Yeah, it's the range-diff format
15:23:49 <dan_b> but if i recall thats the range diff format i saw last year... it's a little a lot
15:24:02 <PieroV> We can probably talk about that in the BBB meeting
15:24:06 <PieroV> It's probably easier
15:24:10 <dan_b> a key/refresher blurb somewhere would def help. just a header
15:24:20 <PieroV> Also, the man page has a lot of details on how range-diff works
15:24:25 <dan_b> aaaah ok
15:24:40 <dan_b> then maybe i'll read than and shoot you an MR hopefully distilling my new understanding
15:24:41 <PieroV> https://git-scm.com/docs/git-range-diff#_examples
15:24:54 <dan_b> oh sweet thanks
15:25:37 <PieroV> What you really want is to check very carefully the "double sign" lines. Lines with background +/- are upstream changes, lines with colored +/- are our patches
15:25:49 <PieroV> So, lines that have both means that both the upstream code and our patches changed
15:26:35 <PieroV> Anyway, I won't manage to start working on this today, maybe tomorrow (my) afternoon I'll start with Tor Browser
15:27:17 <PieroV> I can start with this format and in case change it later
15:28:07 <PieroV> My second point is a reminder for uplifting patches, we have only one week
15:28:25 <PieroV> Then 128 is going beta around June 10
15:28:42 <PieroV> (maybe two weeks, I'm not sure what happens while the current beta enters RC)
15:29:34 <PieroV> Anyway, it's okay if we don't get them uplifted immediately, but the sooner we get them, the lowest is the effort to update them to RR
15:29:46 <PieroV> (and update them back for us)
15:29:50 <richard> yes
15:32:42 <PieroV> (I'm done with my points)
15:33:12 <richard> ok, anyone else?
15:33:43 <dan_b> my plates full this week but other than firing off MRs I'm good
15:34:09 <dan_b> other than pointing out obviously the firefox-android to 128 to tor-browser alpha is gonna be later than the desktop one, but I think thats the case every year?
15:34:26 <PieroV> Yes, more or less
15:34:33 <PieroV> Also, I haven't started working on toolchain yet
15:34:44 <PieroV> (I'm not sure I'll be the one working on them, but I did it for 102 and 115)
15:34:59 * ma1 assumed it was boklm's elves
15:35:10 <PieroV> So, I know we more or less build on Linux + Windows
15:35:35 <PieroV> ma1: for 102 boklm was on vacation when the rebase was ready to be built :) So, eventually I did the update
15:35:43 <richard> maybe we'll need an assist from clairehurst for macOS :D
15:35:55 <PieroV> For 115 I guess I thought I had done it for 102, and did it again
15:36:24 <PieroV> Or boklm was working on other MB stuff at that point, I didn't have anything else on my plate and did the upgrade
15:37:05 <dan_b> and jwilde who's gonna be looking at windows build tooling anyways
15:38:02 <PieroV> I built 126 for Windows. We need to update mingw, but there weren't new patches
15:38:14 <dan_b> nice
15:38:16 <PieroV> And the patches we had for 115 are still needed (and still apply cleanly)
15:38:34 <PieroV> We might need to download and inject the windows create on the build container
15:38:41 <richard> \o/
15:38:50 <richard> PieroV: did webrtc build successfully?
15:38:54 <PieroV> But I'm not sure (I had to run `cargo download`)
15:38:59 <PieroV> richard: I think so, but I don't remember
15:39:04 <PieroV> Oh well
15:39:10 <richard> :)
15:39:13 <PieroV> Now that I think of it I've rebased MB only yesterday
15:39:17 <PieroV> So I built WebRTC
15:39:26 <PieroV> * I built Tor Browser only
15:39:30 <PieroV> Which doesn't have WebRTC
15:39:39 <richard> oh right
15:39:59 <PieroV> Also, very recently Moz completely reworked the default browser agent
15:40:28 <PieroV> (maybe not very recently, again, I just rebased MB very recently)
15:40:52 <PieroV> I had to drop our patches, but they might not be needed. They dropped the dependency on Wintoast, which was the main obstacle to building with Mingw
15:41:04 <PieroV> We still might have to disable some telemetry also in the refactor
15:41:11 <richard> WinRT dependency?
15:41:22 <richard> or w/e the windows 8 thing is
15:41:29 <PieroV> Yes
15:43:19 <richard> alright
15:43:38 <richard> we can figure out who's updating what and when after 13.5 is released
15:44:01 <richard> any further topics or problems?
15:44:07 <PieroV> Nothing from me
15:45:48 <dan_b> 👍
15:46:06 <richard> ok, have a good (short) week everyone
15:46:21 <richard> #endmeeting