16:58:30 #startmeeting network team meeting november 2 2020 16:58:30 Meeting started Mon Nov 2 16:58:30 2020 UTC. The chair is ahf. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:58:30 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:58:33 hello network-team 16:58:35 o/ 16:58:41 o/ 16:58:47 our pad is at https://pad.riseup.net/p/tor-netteam-2020.1-keep 16:58:57 we have one of the slightly longer meetings today, as it is the first monday of the month 16:59:17 and there has also been time changes in the US i think 16:59:24 o/ 16:59:32 dgoulet, asn, nickm: here? :-) 16:59:33 o/ 16:59:44 right (re time change in the US) 17:00:06 o/ 17:00:22 https://gitlab.torproject.org/groups/tpo/core/-/boards 17:00:23 hihi! 17:00:29 sorry for delay, needed tea 17:00:42 are we all doing well on the board? 17:00:52 very nice, i missed my tea slot right before the meeting :-( 17:01:41 o/ 17:01:48 board is ok w me 17:01:57 same here 17:02:02 excellent 17:02:04 yah 17:02:14 looks like stuff was delegated out. we missed torspec!16 but i can grab that one 17:03:24 are folks doing OK with 0.4.4 tickets? 17:03:36 maybe we should be looking at 045 instead now? 17:03:43 err 0.4.5 yes 17:03:57 so if we just change all the strings there to 0.4.5 we should be good? 17:03:58 honestly I haven't started mine yet. 17:04:10 ahf: let's try and see? 17:04:46 done 17:04:53 I'm concerned about tor#40172 -- I can't make progress there till there are repro instructions, but I'm hoping that the tb folks will find that it's just a missing AR= 17:04:55 * ahf was planning on doing windows stuff on thursday and friday 17:05:23 nickm: ah, that one made me think that we should probably have a cross-compiler macos builder on linux on gitlab-ci as well 17:05:32 to build like the browser folks do 17:05:42 no objection from me! 17:05:56 ok, let's talk a bit more about CI under the jenkins point 17:06:13 ok! 17:06:33 do we have anything we need to discuss for s61 today? i think we might have caught that in the s61 meeting an hour ago? 17:06:42 or is there some follow up we need to have with just the network team here? 17:06:49 do you have something to add here, mikeperry ^^ 17:07:39 i assume not :o 17:07:47 okay! 17:07:51 it's the first monday of the month 17:07:54 which means we have a few more items to do 17:08:03 Look at Core Tor Releases and see if there is anything important coming up: https://gitlab.torproject.org/tpo/core/team/-/wikis/NetworkTeam/CoreTorReleases 17:08:16 anything that needs to be revised here? 17:08:29 not afaik. 17:08:34 excellent 17:08:39 one thing: it's 2 weeks till we open the merge window for 046 17:08:46 ahf: main thing is I am going to make a proper MR for https://gitlab.torproject.org/tpo/core/tor/-/issues/40168 and would (eventually) like backport 17:08:58 if at all possible, I'd like everybody to try to get the 045 bugfixes done on the early side 17:09:08 +1 17:09:25 mikeperry: okay, awesome! once thte MR is there we can make a plan for backporting! 17:09:28 nickm: yep 17:09:50 speaking of backports. next item is: Look at tickets/MRs with Backport label and figure out what to do and who is going to do the merging. https://gitlab.torproject.org/tpo/core/tor/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=Backport 17:10:02 nickm recently did a lot here 17:10:37 hm, ok, so tor#40168 already has backport label 17:11:15 nickm: so all the tickets we have with Merge Ready is the ones we need to figure out how to handle, right? 17:11:15 I'm planning to do a round of backports this week, but probably a conservative round. 17:11:43 I think what we should look at there is if there are any in that list that we should be 100% to backport for the upcoming stable releases 17:11:44 ok. do you want help? i can do some of the merging if we plan on doing it, say, thursday post-meeting? 17:11:52 ok 17:11:53 (which I hope will be next-week) 17:12:15 I think I'll be okay; mainly I'd like opinions on "absolutely must" vs "absolutely must not" backport. 17:12:41 hm, tor#33781 i think can bake for a bit longer 17:14:20 I'm planning to backport tor#40080 unless somebody shouts that it's broken 17:15:08 good with it 17:15:13 * ahf too 17:16:17 ok, i can jump to next item? it doesn't sound like there is much more feedback here 17:16:21 sounds good 17:16:25 ok 17:16:26 nickm: you good with that? 17:16:27 ok 17:16:51 okay, let's look at our meta issues: https://gitlab.torproject.org/tpo/core/team/-/issues 17:17:10 "under control" :) 17:17:13 yeah 17:17:25 we should open more stuff there maybe if we think of it 17:17:28 for some reason, gitlab juts logged me out, and i need to go into another room and get my u2f device. can someone else close core/team#18 for me ? that was solved a while ago 17:17:38 https://gitlab.torproject.org/tpo/core/team/-/issues/8 17:17:41 err, 8, sorry 17:17:49 doing 17:17:53 thank you, nickm 17:18:00 done 17:18:08 if you ever spot these duplicated labels, please let me know. they are nightmare to deal with 17:19:18 okay, the next item is me looking at whether we have anything coming in from other teams. i haven't seen anything other than tasks related to the anti-censorship team (and nickm's macos bug mentioned earlier in here). the anti-censorship team ones are all under s30 or s28, so i think we are doing OK with that 17:19:23 i don't think we are blocking anybody there 17:19:44 any questions/additions here? 17:20:58 i take that as a no 17:21:18 okay! that was all the first meeting of the month tasks 17:21:26 nickm have one discussion item: [For Nov 2] What shall we do with Jenkins? -nick 17:21:53 yup. weasel points out our builds have been been broken for a while there. 17:22:16 i think my answer here is that i am not sure. i want us to move away from it, but from what i can understand we right now depend HEAVILY on travis and appveyor, and jenkins have a bit less priority than the two first systems? 17:22:36 IMO we should use gitlab CI for everything we can 17:22:37 travis just announced a new pricing model and i don't fully understand if it will impact us :-/ they seem to want to people ot do much less macos builds 17:22:48 yes, so our goal is to get rid of jenkins here, right? 17:23:03 jenkins is a nightmare compared to the other systems in that it is not controlled with a config file in the repo 17:23:12 but some config files and deployment elsewhere 17:23:33 yeah. Maybe let's plan to decommission tor builds on jenkins on some given date, first going over the config repo to see if we're missing anything that we want to do in gitlab 17:23:57 agree 17:24:03 i can do that. i think i have probably touched jenkins the most recent i think 17:24:20 ok 17:24:24 ok. Let's say "end of november" for our target? 17:24:45 and say that this only applies to tor builds; others can still use jenkins as much as they want 17:24:48 for requesting removal of jenkins? 17:24:49 yep 17:24:52 i think that is fine 17:25:12 we'll probably want to migrate stuff to nightly builds, not commit-triggered builds. I don't know how those coexist on gitlab. 17:25:18 i cannot imagine we'll find something we are missing. it could be that we need to have a better strategy with testing actively maintained releases for nightly builds 17:25:21 heh yes 17:25:33 good we are aligned here even with bringing it up :-P 17:25:51 okay 17:25:58 i don't see anything else on our discussion list? 17:26:16 (I think older/weirder debian/ubuntu/fedora/*bsd versions are something we'll want to do on Nightly) 17:26:21 yes 17:26:35 did someone delete asn/dgoulet's updates on the pad? :o 17:26:51 i keep being thrown off it 17:27:04 oh wow ! lolll... I think I have entirely forgotten that it is a thing I need to do! woa 17:27:16 oh, so *i* have deleted it 17:27:27 but yes, please do the updates 8) i use them to check for questions 17:27:31 mine was simply never on the pad :) 17:27:37 mikeperry: is your questions for asn still relevant? 17:27:50 wait, no, that is not mike 17:27:51 never mind 17:28:02 okay, i think we are good 17:28:03 thanks all 17:28:06 #endmeeting