17:59:15 #startmeeting weekly network team meeting, 10 Dec 2018 17:59:15 Meeting started Mon Dec 10 17:59:15 2018 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:59:15 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:59:33 This might be the 2nd-to-last meeting of the year 17:59:56 unless we reeeeeealy have to meet on the 24th or the 31st, which maybe would not be fun 18:00:08 hello hello! 18:00:17 \o/ 18:00:18 hello 18:00:19 sorry for running off early last week; I hope folks are well 18:00:52 i think my family would be sad if we had meeting on the 24th :-P 18:01:04 np, hope the moz all hands was good! 18:01:21 Pretty good, yeah. We had some good discussions and presentations. 18:01:29 so... let's start with the roadmap 18:01:35 yep 18:01:50 looking at the "0.4.0 merge window opens" section, I see only 5 must-do sponsor8 items. 18:01:59 and they're all under "doing" 18:02:00 so that's good! 18:02:26 I see a bunch of stuff listed on "TODO" that we are not likely to get done for Jan 15. That's not great but not unexpectedc 18:03:04 I expect to wrap up my #28019 work rsn with the merge of #28624, and then I'll be done with my stuff 18:03:10 and my primary goal will be h 18:03:19 and my primary goal will become helping anybody who want sh 18:03:21 wants help 18:03:30 How are people doing with their DOING issues? 18:03:40 * nickm .oO(my secondary goal will be not hitting enter every time I try to type "h") 18:03:46 i could use some help with a circular dependency issue and integration (making the api ready i guess?) for the dormant code 18:03:56 things are happening in wtf-pad land. been fighting with fixups and CI stuff. done good progress on both. still have a bunch of fixups to do, and 32-bit appveyor is broken. will fix these this week. 18:04:00 and thanks a lot for the help already with the k/v parser work, nickm 18:04:09 mikeperry has tkaen over the unittest side of #28142. not sure how that's going. 18:04:45 asn: I just fixed a clang-specific test bug friday. that took a while 18:04:55 nice 18:04:58 yeah i saw that 18:05:01 asn: going to get to write more tests this week 18:05:15 ahf: the .may_include issue or sth else? 18:05:28 yeah, the may_include issue 18:05:41 ahf: I can try to do something there 18:05:53 that would be helpful 18:05:57 would be very nice, dgoulet 18:06:03 ahf: that's for the PT thing? I can try 18:06:19 nickm: i think dgoulet is up for it too ^^ 18:06:29 ok. Let me know if you want to pull me in 18:06:32 thanks dgoulet ahf! 18:06:38 I had a really good time helping with #28755 last week, and I'd love to do more tickets for people 18:07:05 that is _very_ useful and also thanks for the spec feedback there, i had a talk with dcf about it afterwards about finding some common ground there 18:07:08 ahf: that is in the #25502, right? 18:07:21 and i think our conclusion is that the PT world *will* find the STATUS API useful, they would like to see some examples of it 18:07:25 gaba: yep, spot on 18:07:36 ahf: speaking of, I have on my list to spec out that ^ 18:07:43 ahf: so if you have a branch or anything, I can be useful there 18:07:53 ahf: (or even just an idea, I can put words into a .txt :) 18:08:18 dgoulet: yep, i spend some of friday looking at meek + snowflake code (non-connect based PT's after the conversation with david thursday evening) and gonna try to write down some notes about my findings 18:08:25 so we can flesh out some initial K/V pairs that are useful 18:10:01 we ready to look at reviews? It seems that we don't have a lot 18:10:14 in general, s8 should still get priority over review. 18:10:23 though of course reviewing s8 stuff counts as s8 18:10:32 dgoulet: thanks for taking on #28624 18:10:42 everybody: I can review more stuff if you want to offload 18:11:16 mikeperry: btw perhaps we can close #26633 now? it's on the roadmap and perhaps it doesn't need to be open anyumore? 18:11:32 yeah I think so. I only left it open for you to have a chance to agree 18:11:36 ack 18:12:00 closish 18:12:07 \o/ 18:12:21 there was another one too.. 18:12:39 only #28142 and #28632 remaining as MUST for V 18:12:41 #28633 18:13:09 mikeperry: right 18:13:13 mikeperry: i want to do a review on that 18:13:18 about things missing for s8 I put tickets into a temporary pad to understand what is missing and wha tpeople are working on https://pad.riseup.net/p/qPH3xF4JaHOB It is mostly for me to deal with trac 18:13:18 based on ur comments on the other ticket 18:13:21 and i will close irght after 18:14:23 done with roadmap discussion for V 18:14:54 asn, mikeperry: [side note, a presentation about WTF-PAD to the net team in Brussels would be _amazing_ (like the one Nick did about Guards in Wilmington :)] 18:15:04 yes def 18:15:45 +1 18:16:11 yeah. we should also produce some docs on how to make padding machines 18:16:22 +1 18:16:29 those might be specs, or might just be researcher docs 18:16:53 on rotations: I see ahf on triage and catalyst on CI/Coverity. 18:17:02 ack 18:17:17 you're both pretty busy with sponsor8 stuff; would either of you like to switch with me? As noted I'm underscheduled this week 18:17:38 I can take on triage tbh 18:18:01 mikeperry: agreed 18:18:10 dgoulet: that would be helpful! 18:18:11 * catalyst can hand off CI+Coverity 18:18:12 (as I'm CI next week) 18:18:37 we can split it even, i can do mornings here and you can do afternoon at your place? 18:18:50 nickm: I am still feeling a bit distracted by lots of wtf-pad loose ends; if you want to take whatever subset of my reviews that would be helpful, 18:18:58 ahf: no worries, I'll take the full task :P 18:19:04 catalyst's update on s8-errors and roadmap is in their status in the pad 18:19:09 cool! 18:19:32 ok, I'-ll take CI/coverity this week 18:19:34 catalyst: ^ 18:19:39 nickm: thanks! 18:19:49 ok, I'll eat some of your reviews. are any of them s8 items? 18:20:47 * catalyst can do a quick look at the revised #28007 but someone who's set up do the actual scan-build run should smoke test the changed script 18:21:43 nickm: none are sponsor8 I think. just ipv6 and doc fixes 18:21:58 some of them are swbws from last week 18:22:08 fwiw 18:22:15 catalyst: okay, that's me who has done the scan-build before. If you put it in merge_ready i'll try out the changes 18:22:22 nickm: thanks! 18:24:22 so... on to discussion? 18:25:08 Please people remember to read the announcements. I added a few things there. One of them is the pad for doing the report on Q4 for sponsor 8. 18:25:12 ahf: what do we need to figure out w snowflake today? Is that still on the list? 18:25:20 oops, I will correct the link 18:25:38 i think so, i ended up dropping things a bit on the floor to get moving after your review on #28179, but gaba had a very good list of priorities 18:25:43 gaba: the dates are Sep 1 through now? 18:25:44 let me just put them into the doc 18:26:07 ahf: anything we need to figure out before end-of-year with that? 18:26:31 yes, we should get started with the testing infrastructure that was discussed at the meeting, if possible 18:26:36 yes nickm 18:27:15 do we need people on that? 18:27:38 and i think we also need to get more people to try it out by now. right now dcf and i are talking a bit about #28726 which is hindering you to run your own snowflake network i think 18:27:48 juga had tried it out too and reported the same experience as i have seen 18:28:08 i think teor expressed interest in looking at integrating PT/snowflake tests with chutney 18:28:15 if they have time 18:28:38 hm, ok 18:28:40 I understood they will not have time for it in December. 18:28:45 ok 18:28:51 then maybe it's not worth doing this month 18:29:44 until i'm done with the s8 tasks i'm also not gonna be able to make much progress on these things, but being able to run a isolated snowflake test network is still my current goal 18:30:57 nickm: the spec is something you can start this month? 18:31:21 hrm. maybe 18:31:23 and get ahf on board when he is done with s8? 18:31:28 I can try, but I might need help from ahf 18:31:33 we'll see! 18:31:38 ok 18:31:43 that could work 18:32:11 ahf: if you could send me a couple of links for where you think I should start investigating to write a spec, that would be great 18:32:40 i can do that. you want an email? just to be clear: this is the facilitator? 18:32:44 Next discussion item is ... first 2019 meeting on 8 Jan at the later time? 18:32:49 ahf: I don't know what I'm supposed to spec 18:32:52 email is good 18:32:58 ok, will write an email 18:33:01 (the time is okay with me) 18:33:18 ok with me 18:33:30 hearing no objections, I say it's scheduled. 18:33:44 ack 18:33:56 we talked about reviews already and sponsor8 already 18:34:33 for the RC blockers -- I'm going to go over them again, but I would love it if anybody else could take a quick glance over the 035 tickets to see if any of them, if unfixed, would mean that we shouldn't call the next release a release candidte 18:34:57 let me know if you'd rather do that together; otherwise just tag with 035-rc-blocker or 035-rc-blocker? 18:35:10 k 18:35:16 For the release schedule: we're obviously not going to get 035 final out this month, so I'll try to revise. 18:35:35 Has everybody looked at Teor's plans for the proposed-ticket process on the pad https://pad.riseup.net/p/network-team-triage-2018 ? 18:36:46 sounds plausible and worth trying 18:36:58 https://pad.riseup.net/p/network-team-triage-2018 18:37:03 ok 18:37:13 Let's plan to start this in January? 18:37:28 sounds good 18:37:38 seems reasonable. what does the tricket triage person do for milestone assignment though? 18:37:53 catalyst: I don't know. Maybe add an open questions section? 18:38:39 One thing we'll have to do to make this work is learn not to work on (most?) non-roadmapped tickets that aren't proposed & discussed 18:38:47 that might be annoying for us, but we'll have to learn to do it 18:39:06 I think right now the reason we don't use our current process so much is that when somebody believes strongly in something, they often just do it 18:39:16 right ^ 18:39:19 We can tune this process as we go 18:39:21 makes sense 18:39:30 sometimes it can take some hours to understand an issue and thus fixing it right away is the way to go 18:39:40 but yeah that is easy to deal with as we go, agree 18:40:26 and if you are working on something that is not roadmapped then we should have a way to track that as unplanned work somewhere 18:40:38 at least add it to the status in irc 18:42:14 right 18:42:36 now it's time to hunt boldface stuff! 18:42:45 I've already raised my items :) 18:42:55 catalyst has questions 18:43:21 feedback on #28731 would probably be better after the meeting 18:44:14 not seeing any more boldface, but I do see some "help with" items listed 18:44:35 catalyst: wrt bootstrap knowledge, I'd say "yes" but only after we have s8 done 18:45:11 nickm: ok. i should probably write down more of it than i currently have, just to keep better track for ongoing sponsor8 work 18:45:20 taking notes is great 18:46:04 anybody want to discuss anything else for this week? Please remember to read the whole pad, especially including announcements and "want help with" sections 18:47:17 hearing no new topics, let's say that the meeting is done 18:47:20 thanks, everyone! 18:47:23 thx! 18:47:23 thanks 18:47:24 thanks! 18:47:27 o/ 18:48:11 peace all! 18:48:13 #endmeeting