14:59:27 <richard> #startmeeting Tor Browser Weekly Meeting 2022-10-11
14:59:27 <MeetBot> Meeting started Tue Oct 11 14:59:27 2022 UTC.  The chair is richard. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:59:27 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:59:36 <richard> https://pad.riseup.net/p/tor-tbb-keep
14:59:44 <Jeremy_Rand_36C3[m]> Hi!
15:00:10 <boklm> good 15utc!
15:03:29 <richard> ok, last week I went through gitlab and did some issue shuffling/punting
15:03:40 <PieroV> got lots of email :D
15:03:40 <richard> exciting I know
15:03:54 <richard> currently things should be pretty much organized
15:04:30 <boklm> nice
15:05:00 <richard> the `Next` label is ideally things going in to the next alpha, while `Backlog` is stuff that should be ultimately going into the November release
15:05:05 <donuts> o/
15:05:11 <richard> yes donuts?
15:05:22 <donuts> just hello :)
15:05:25 <richard> :3
15:05:25 <donuts> happy to be here
15:05:28 <donuts> thanks
15:05:49 <richard> so some general guidance for dev work priorities over the next ~1.5 months
15:07:24 <richard> we had some fairly optimistic goals for 12.0 (update to esr102, migrate torbutton, tor-launcher, localizations, multi-locale bundles, aar64 mcOS support, tons of other stuff i'm forgetting)
15:07:49 <richard> and we've hit/will hit a lot of them
15:08:08 <PieroV> \o/
15:08:12 <dan_b> woo!
15:08:14 <richard> but one nice(?) thing is that not many of these are really user-facing (eg if we don't get to them nobody *really* notices)
15:08:39 <PieroV> well, the tor-launcher will help with Arti, I think
15:08:41 <donuts> woohoo!
15:08:43 <donuts> but yeah writing the blog post is gonna be tricky
15:08:53 <PieroV> And torbutton, too, even though if we haven't reached it
15:09:16 <PieroV> donuts: well, multi-lingual is a big thing, and tor-launcher too, for some use cases, I think
15:09:22 <richard> so ideally I would like to avoid a mad-rush at the end of November trying to squeeze in as much as possible becaus ein the end, a lot of this stuff is dev facing/dev quality-of-life stuff
15:09:28 <donuts> pierov: very true
15:09:43 <PieroV> I was wondering if we should write some special paragraphs even for the first alpha containing them (possibly the next one)
15:09:51 <PieroV> in the blog post, I mean
15:10:16 <richard> so that said, in terms of what *needs* to happen for 12.0 please prioritize things which are actually user-facing :)
15:10:51 <richard> ma1: particularly that onion-auth issue we found last week >:[
15:11:30 <richard> (and obviously if gitlab is not matching the above stated guidelines/goals than it's wrong, it's only a model after all)
15:12:06 <PieroV> richard: I've almost run out of next items
15:12:23 <PieroV> So, feel free to add anything for 12.0 please :) Or anyone, if you have too much on your plates
15:13:08 <richard> oh right, so if you're running out of Next items, please grab from Backlog :)
15:13:32 <richard> (or from other people's Next)
15:13:34 <dan_b> oh PieroV, should I close tor-browser#41163 since we merged my MR fixing it, or did you want to keep open cus I thought you were tracking an upstream better fix?
15:14:32 <PieroV> dan_b: the latter one, but you can move it to low priority
15:14:49 <PieroV> Sadly no answer from Mozilla, yet :/
15:15:02 <dan_b> cool
15:15:04 <PieroV> But it's a very niche setting, after all, that they aren't using
15:15:45 <richard> donuts: has there been any further investigation into the macOS fat-bundles shared w/ you last week (your mystery launch failure)
15:16:17 <donuts> richard: oh I was going to retest on my Intel machine to see what happens there
15:16:17 <donuts> I'll do that now
15:16:27 <donuts> I had zero luck with my apple silicon machine
15:16:56 <richard> that is somewhat worrying
15:17:19 <richard> does anyone here have access to apple arm hardware?
15:17:19 <PieroV> Is separate bundles a plan B?
15:17:26 <donuts> I can test on another apple silicon machine this evening to see if it's a me problem or a silicon problem too
15:17:29 * PieroV doesn't
15:18:02 <boklm> separate bundles is still possible (but maybe it has the same issue)
15:18:15 <donuts> Intel opened fine
15:19:02 <donuts> what ticket is it? so I can add a bug report
15:19:12 <richard> boklm: can we sooniish make aarch64-only testbuilds to test?
15:19:46 <richard> (assuming this aarch64 problem is consistent)
15:19:53 <boklm> donuts: tor-browser-build#40439
15:19:58 <donuts> ty boklm
15:20:18 <boklm> richard: yes, main branch can build aarch64-only testbuilds
15:20:35 <boklm> I can start a build on tb-build-05
15:20:41 <richard> ah is that what comes out from make torbrowser-nightly-osx-aarch64 or w/e
15:20:52 <boklm> richard: yes
15:21:09 <richard> *lovely*
15:21:11 <richard> donuts: https://tb-build-05.torproject.org/~richard/builds/nightly/tbb-nightly.2022.10.10/
15:21:28 <boklm> but it's not enable as part of default builds
15:21:33 <richard> there are some aarch builds there with some extra macOS spinlock patches applies
15:21:47 <donuts> richard: explain in designer pls
15:22:05 <richard> they're nightly wiht something extra that shouldn't affect anything
15:22:27 <richard> nightly + these patches: https://gitlab.torproject.org/tpo/applications/tor-browser/-/issues/41348
15:22:49 <donuts> so I should text aarch64_en-US.dmg?
15:22:58 <donuts> *test
15:22:59 <donuts> er
15:23:46 <boklm> yes
15:23:58 <donuts> wb richard
15:24:05 <richard> sorry about that, my inernet walked away
15:24:05 <boklm> https://tb-build-05.torproject.org/~richard/builds/nightly/tbb-nightly.2022.10.10/TorBrowser-tbb-nightly.2022.10.10-osx64-aarch64_en-US.dmg
15:25:04 <donuts> ty
15:25:12 <donuts> powering up the battlestation now
15:25:15 <richard> anyway that's all i have unless i missed something
15:25:59 <richard> does anyone else have bolded items ot discuss?
15:26:05 <boklm> maybe we can sign a testbuild (with the release key) to see if that fix the issue
15:26:26 <richard> also a good idea^
15:26:30 <donuts> I can seemingly bypass the signing thing fine
15:26:30 <Jeremy_Rand_36C3[m]> I have a small discussion item, which boklm has partially answered
15:27:02 <Jeremy_Rand_36C3[m]> Basically it'd be cool to have a stable URL that HTTP-3xx-redirects to the latest stable/alpha/nightly binaries
15:27:16 <Jeremy_Rand_36C3[m]> Would make my CI scripts less painful
15:28:00 <PieroV> +1, it could help also to ask people to test with nightly
15:28:33 <richard> truue
15:28:39 <donuts> richard/boklm: same issue with the aarch64 build
15:28:57 <Jeremy_Rand_36C3[m]> PieroV: agreed, right now I point users who want to test Nightly to the directory that has date-specific subdirs, which is not great UX
15:29:21 <donuts> I can see if I can get another apple silicon user to test though
15:31:38 <richard> well that's *interesting* and also no fun
15:32:00 <boklm> can you post the error you get in the ticket?
15:32:20 <donuts> boklm: will do
15:32:28 <boklm> (and maybe we can find a similar error in mozilla bugzilla)
15:32:39 <boklm> thanks
15:32:45 <donuts> it's not descriptive at all unfortunately
15:32:47 <donuts> it just says "no"
15:32:58 <boklm> ah
15:33:13 <donuts> https://share.riseup.net/#VyjCtjtPvKVoe2I-h__V7w
15:33:35 <richard> ahh right
15:33:38 <richard> and the ? points nowhere
15:33:44 <boklm> I'm wondering if there is some way to make that more verbose
15:33:56 <donuts> richard: it opens a help window that's 100% blank
15:34:15 <donuts> yeah
15:34:45 <Jeremy_Rand_36C3[m]> Yeah I wonder if there's e.g. some magic environment variable that enables verbose logging
15:35:52 <Jeremy_Rand_36C3[m]> (I know NSPR and NSS have such vars)
15:35:59 <boklm> some people are talking about similar errors there: https://superuser.com/questions/1345755/how-to-fix-the-application-cant-be-opened-on-mac
15:36:32 <ma1> most of them say "enable the +x permission"
15:36:46 <boklm> (in that case that seems to be not having the executable permission on some file, but maybe there can be other causes for this error)
15:37:22 <ma1> Is the Console app still a thing on MacOs?
15:38:17 <donuts> ma1: yep console still exists
15:38:31 <ma1> and nothing shown there?
15:38:37 <donuts> let me retry with the console open
15:40:28 <richard> that chmod issues looks promising, we'd probably need to do it for firefox tor and all the pts I presume
15:41:23 <donuts> I don't think there's anything related to the browser in the console
15:41:30 <donuts> however it captures *a lot*
15:42:16 <PieroV> can we chmod in the /Applications or whatever it is?
15:43:32 <donuts> aha found it, maybe
15:44:36 <richard> donuts: could you do a `ls -lahR` in /Applications/TorBrowser.app (or w/e) and we can verify perms are set right
15:44:45 <richard> but we can debug after this meeting
15:45:23 <PieroV> (probably without R, to avoid all the font lines)
15:45:54 <donuts> https://share.riseup.net/#OiHnsWp3PlkbxlgL5Ky3YQ
15:46:05 <PieroV> "lsd"
15:46:46 <richard> i am so confused about wha ti'm looking at there donuts
15:47:29 <donuts> something called launch error in the console that happened when I tried to open the app, but i've got no idea what i'm looking at either
15:47:42 <PieroV> "Error -10810 is a Launch Services result code indicating an unknown error."
15:47:52 <donuts> cool
15:48:36 <PieroV> AskDifferent says to chmod +x
15:49:23 <boklm> there is some line "Security policy would not allow process" (about the firefox file)
15:49:43 <donuts> "ASP: Security policy would not allow process: 1372, /Applications/Tor Browser.app/Contents/MacOS/firefox"
15:49:44 <ma1> "ASP: Security policy would not allow process: 137, /Applications/Tor Browser.app/Contents/MacOS/firefox
15:49:50 <ma1> :)
15:49:54 <donuts> haha
15:50:12 <boklm> can you check if the file has executable permission?
15:50:15 <donuts> I'm assuming that's just the normal security warning that always pops up
15:50:25 <PieroV> Also some lines that starts with Plug, maybe pluggable transports? But they're non-fatal errors
15:50:36 <ma1> try chmod +x  /Applications/Tor Browser.app/Contents/MacOS/firefox
15:50:58 <PieroV> (with quotes)
15:51:08 <PieroV> chmod +x  "/Applications/Tor Browser.app/Contents/MacOS/firefox"
15:51:09 <ma1> yep, space after Tor :/
15:51:27 <donuts> hang on, pls bear in mind I am a design jockey and have no idea what I'm doing
15:51:38 <donuts> I should probably prep for the ux team meeting too
15:51:48 <donuts> but I can circle back to this in #tor-browser-dev afterwards?
15:51:58 <Jeremy_Rand_36C3[m]> Might be good to do the `ls` thing before trying `chmod`, so that we can see the status before changing the state of things?
15:52:26 <ma1> OK, donuts, see you in #@tor-browser-dev after the UX meeting?
15:52:26 * Jeremy_Rand_36C3[m] tries to avoid destroying evidence until the evidence is analyzed
15:52:32 <ma1> Jeremy_Rand_36C3[m], good call
15:52:46 <donuts> ma1: yep, will catch you then :)
15:57:03 <PieroV> I think we can call the bot? :)
15:57:49 <donuts> richard ^
15:59:59 <richard> #endmeeting