17:00:05 #startmeeting network team meeting, 3rd of may 2021 17:00:05 Meeting started Mon May 3 17:00:05 2021 UTC. The chair is ahf. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:00:07 hello hello 17:00:10 woo! hi! 17:00:15 https://pad.riseup.net/p/tor-netteam-2021.1-keep 17:00:18 is our pad! 17:00:30 i missed the s61 meeting, but got a brief from mike about it. sorry about that 17:00:44 first meeting of the month today, so let's get going 17:00:45 yello 17:01:03 hihi! 17:01:26 how are folks dashboards looking? see: https://gitlab.torproject.org/groups/tpo/core/-/boards 17:01:49 beautiful 17:01:58 that is the spirit, dgoulet! 17:02:05 full of colorful labels 17:02:18 so many Next and Backlog :P 17:02:29 yes :-/ 17:02:43 i have a bit stress about that too, but will hopefully get some of that down this week 17:04:07 are we just the three of us? 17:04:18 that is a possibility... 17:04:24 asn is afk 17:04:29 ah 17:04:50 how are things looking with releases? 17:05:12 I want to do an 046 and an 045 this week 17:05:14 tor#40375 is new and unassigned 17:05:23 o/ 17:05:27 hello gaba 17:05:47 looks like the ticket is being debugged though 17:05:48 We should get somebody assigned to tor#40312 for 0.4.7 17:06:12 i just took that one before the meeting began 17:07:01 great 17:07:28 for 046 and 045, I would really like us to have #40373 done this week. Is that possible? 17:07:35 it's just a log message, right? 17:07:52 we could do releases early next week instead if that's not possible 17:08:03 but no later, since we want to hit the TB release 17:08:38 tor#40373 17:09:00 I think it is quite simple, if asn can't, I can go for it 17:09:52 cool 17:09:59 that would rock, excellent 17:10:32 now that we are at it, is our https://gitlab.torproject.org/tpo/core/team/-/wikis/NetworkTeam/CoreTorReleases looking right? 17:11:24 I think so yeah 17:11:52 In 42 days we can throw out 0.4.4 17:12:06 woh 17:12:09 we're aiming to have 046 stable on Jun 15 17:12:33 the active release list will be short then 17:12:35 jep 17:12:38 yep* 17:12:40 I should move 047 up to 'current' since it's WIP 17:12:55 yeah, but is future right now. i am thinking of backport work too 17:13:35 One thing that came up at the S61 meeting that we should all think about is whether we delay 047 if needed for congestion control. 17:13:42 We're going to try not to, but we should keep the possibility in mind 17:14:04 Also, depending on arti funding, we should consider whether we want to reduce the frequency of Tor release series 17:14:06 is congestion control work also include the conflux work that dgoulet have started? 17:14:08 but that's also in the future 17:14:17 or is it the things mikeperry is focused on "purely" on actual CC 17:14:23 ahf: it includes packed-cells, but not necessarily conflux, afaik 17:14:26 ok! 17:14:34 I will try to use packed-cells in CC, yes 17:14:37 IIUC conflux is a stretch goal 17:14:41 yeah 17:14:45 i think mikeperry have said the CC parts are not going to be a huge change the last time we spoke about it? 17:14:54 hey mikeperry 17:15:05 yeah CC is relatively simple compared to conflux 17:16:00 in both tho, much tuning and experimentation will be required. I hope to get stuff working in shadow for CC as soon as I can 17:16:13 sweet. are you diving into that alone or are you paired up with someone? maybe i am asking question you already discussed in the other meeting, if yes, then i can just ask in after this meeting 17:16:25 cool! 17:16:38 nickm offered to help with negotiation and protover stuff 17:17:06 I'm glad to be a rubber-duck with the other stuff too. 17:17:16 I could also use help with figuring out how onion services announce support 17:17:20 :-) 17:17:36 announce support? 17:17:43 yeah in their descriptors 17:17:49 ah, that they support it 17:17:53 makes sense 17:18:05 if that's the only hard part I can probably figure that out 17:18:08 i could imagine david or george having opinions about that too 17:18:14 yeah 17:18:26 super exciting with this 17:18:31 yeah shouldn't be hard to do 17:18:57 okay, we have a pretty big backlog of backport tickets. i think we missed march with it since i think dgoulet and i were supposed to look at it right after the time where i went MIA 17:19:06 and i tink we missed our first month meeting in april 17:19:16 there are some bits in both cases wrt what hapens if the endpoints have consensuses that advertise different params, because of a recent consensus change that both endpoints may not share 17:19:27 i *think* i might be annoying and say we spend up to half of our thursday meeting on that :-) 17:19:34 unless something more urgent suddenly appears in the week 17:19:49 that list sucks more and more the longer we wait with doing it 17:19:57 fine by me 17:20:28 nickm: is this OK with you too? 17:20:40 thinking for a second... 17:20:51 I think that's ok. 17:20:58 ok, if something else comes up then this is low priority 17:21:16 except that if we find significant 045 backports on thursday that might make a friday release harder. But I hope we won't. 17:21:27 What could go wrong, thursday is probably fine :) 17:21:54 oh, i have a short informal announcement to add, let me know when it's time 17:22:17 ugh, that is true... thursday is a bit late for that 17:22:40 nickm: tomorrow you and i could put our heads together around this time and maybe go over the list quickly and see if something impacts releases on friday? 17:22:45 i only hvae one meeting tomorrow 17:22:50 ok, same here 17:22:59 shall we do that? 17 UTC ? 17:23:10 sure,, 1700 UTC works fine 17:23:13 excellent 17:23:55 i don't see any new team issues, and i think the ones we talked about a while ago is still not in a state where we are going to dive into them with all the other things going on right now 17:24:17 ok, and nothing new from other teams it seems! 17:24:19 OK! 17:24:21 that was the big list 17:24:36 [3 May] Tor archaeology: Nick and Roger are working on identifying "implicit knowledge" in the codebase and updating documents to reflect it. Keep an eye open! 17:24:40 seems to be from nickm? 17:24:44 yup, just added 17:25:07 what "documents" ? 17:25:09 which document might a question to ask? 17:25:10 last thursday Roger and I went over circuitbuild.c and made a huge list of stuff 17:25:20 cool! 17:25:29 we're going to try to turn it into tickets, spec patches, etc 17:25:53 I'm mentioning here so that if you see anything labeled "tor archaeology" you'll know what we mean 17:26:02 ok 17:26:17 nice 17:26:27 this is something roger have been so interested in for so long that i can remember when we have talked about alternative tor clients 17:26:33 super cool 17:26:54 yeah, I am hoping that this results in actual progress being made :) 17:27:03 ya 17:27:21 my motivation is that arti is now, so having some of this info now is better than all of it in 5 years or whenever 17:27:24 ;) 17:27:28 the CC and 0.4.7 we already went over right? i think the agenda item was discussed earlier 17:27:30 yes, for sure 17:27:50 are you documenting this in some public place or it will be only in tickets? 17:27:50 nickm: very good idea 17:29:16 gaba: it will be tickets or spec patches. 17:29:24 all of it will be in the spec eventually. 17:29:34 ahf: I thinnk that's right 17:29:50 gaba: at some point we should do a spec rewrite for clarity 17:30:13 sounds good 17:30:15 amsi - a markdown spec implementation 17:31:09 2021-05-03, nickm, let's plan our releases this week. 17:31:22 i am unsure what this means and who it is from? it is a different color than the others 17:31:25 I think we got to that -- I mentioned 045 and 046 releases happening 17:31:27 that's me 17:31:28 ok! 17:31:39 then i don't think we have anything left in the meeting 17:31:54 (let's try to get stuff for 046 and 045 reviewed and merged on the early side if we can) 17:32:28 yes! 17:32:47 sounds good folks, this was a bit longer as our first monday of hte month meetings always are 17:32:58 does anybody else have anything else we need to talk about? 17:33:03 * dgoulet is good 17:33:14 * nickm is fine 17:33:19 everything ok 17:33:40 👍️ 17:34:40 thanks folks! 17:34:42 #endmeeting