23:00:11 #startmeeting network team meeting, 5 March 23:00:11 Meeting started Tue Mar 5 23:00:11 2019 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 23:00:11 Useful Commands: #action #agreed #help #info #idea #link #topic. 23:00:13 hi folks! 23:00:22 Our pad is at https://pad.riseup.net/p/tor-netteam-2019.1-keep 23:00:24 who's around? 23:00:42 o/ 23:01:17 * teor +1 23:01:32 +1 23:01:38 hello, catalyst, teor, gaba! 23:01:44 hi, caw! 23:02:31 soooo.... we start with the kanban. are we all doing something roadmapped? 23:02:51 are we within the points estimates, or do we have to revise? 23:03:23 * nickm did some ci+chutney stuff, but won't know the size of next steps till I can confer with teor . No hurry there. I should find another thing to do 23:04:22 gaba: when I'm looking for next-thing-to-do, do I look at backlog or icebox? 23:04:56 I think I'm still within the points estimates, but I won't know until I split the branches I have right now into the correct tickets. They depend on each other. 23:05:25 make sense 23:05:29 *makes sense 23:06:22 Oh, gaba, you added a comment to some tickets on the kanban, but I don't get emails or notifications for comments. 23:06:39 * arma1 is around if needed 23:08:02 fwiw, gaba said she might be somewhat distracted right now; one of her kids has a birthday 23:08:12 if there's nothing else we can do on the kanban, let's move on? 23:08:21 I will add it to the meeting pad instead. 23:08:26 ok 23:08:35 hey, sorry 23:08:44 * ahf read backlog 23:09:00 needs_review tickets are next; I'm nearly done with mine for this week though the last one (#28288) is a big one. Please let me know if you need to offload 23:09:07 err not that 23:09:12 #26288 23:09:16 that's the big one 23:09:34 cool, gonna be nice to get in 23:10:03 any other stuff I could/should review for anybody? Please feel free to just change reviewer if so. keep in mind I can't review my own patches :) 23:10:16 or others might be able to take things too 23:10:43 rotations this week are nickm on triage (oh hey!), ahf on ci (oh hey!) 23:10:51 cool 23:11:18 is the appveyor changes done? or are there some tickets left for review there? i'm already almost done with my weekly reviews so i could takeover those related to the Ci role 23:11:36 I don't know what tickets those are. teor? 23:11:36 * catalyst thought they already set the required ticket to merge_ready 23:11:42 \o/ 23:11:43 cool 23:11:52 oh hey, there they are 23:12:43 I'll probably merge them to 0.4.0 and later in the morning; teor can do the backport per stable maintainer policy 23:12:52 Thanks! 23:13:06 thanks for looking into the speed part of appveyor btw 23:13:35 now let's see announcements: the one to discuss is US DST. Usually we move our meetings when US DST hits. Shall we do that again this time? I'm cool if so 23:13:41 i'm also cool if not 23:13:57 is tbb-team doing the same again? 23:14:02 I do not know 23:14:05 sorry I lost internet. I'm back 23:14:24 i'm okay with either, i think it would be nice to have it at the same time all year around, but i think it might get early/late for other people if we don't move it 23:14:40 I do not want to move this meeting, if that's ok 23:14:47 And the patch party 23:14:53 * catalyst was asking about the Monday meetings 23:14:54 gaba: welcome back; pending questions for you above :) 23:14:58 nickm: about the next thing to do is the backlog. Each month we should get stuff from the icebox into backlog and review priorities. 23:15:11 teor: the comments were mostly for me 23:15:45 teor: yeah, i'd prefer not to move this one either 23:16:17 okay; that makes it shift for me, but that's okay. 23:16:30 so we keep this one at 2300 UTC but move the monday one with US DST? 23:16:53 us dst changes in .. 3-4 days'ish? 23:17:01 oh hey we already document moving the Monday meeting with US DST 23:17:04 https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkTeam 23:17:07 I think shifting this meeting later is bad for Europe? 23:17:43 Ha, yeah, we had this conversation 6 or 12 months ago: 23:17:44 The primary meeting will track US daylight saving time. The Patch Party will not change with daylight saving time. The optional catch-up will track European daylight saving time. 23:17:52 it would become 1 in the night for me i think and 2 for asn 23:17:54 :) 23:18:05 ok then we should just send out reminders and update the schedule part of the meeting pad? 23:18:08 teor: oh. we've converged on the policy yet again 23:18:14 :-) 23:18:15 catalyst: sounds good to me 23:18:29 Convergence is faster every time. 23:18:51 :-) 23:19:32 on to discussion stuff -- I am +1 on the idea of skipping people who are on leave, and giving away reviews of everybody who is on leave 23:20:06 in fact, I suggest that giving away reviews of people on leave should be default: you should have to explicitly say "no I want to stay reviewer on this; wait till I come back" if you want to do that 23:20:10 does that sound ok? 23:20:26 yes, i think that would be nice to know what the policy in case ones forgets so it isn't in the back of the head 23:22:18 +1 23:22:21 ok, other question /thing is that we need to prioritize sbws 23:22:41 juga is wrapping up and it would be good to close that before they leave 23:22:43 this is the last month juga is paid to work on sbws, so they need our reviews asap 23:23:12 we should remind everybody who is reviewer on an sbws ticket about this... 23:23:21 ... and make sure every sbws ticket has a reviewer 23:23:31 gaba: any more we can do here? 23:23:56 not that i think of 23:24:09 ok 23:24:15 nickm: should we pick up new sbws tickets to review before the weekly assignments? 23:24:59 yes, we should make sure we don't delay for up to 2 weeks 23:25:02 if there are any unreviewed sbws tickets and somebody has extra time, I'd _suggest_ yes? But I'd like to know what dgoulet and asn say before I call that official 23:25:08 (1 week to assign, 1 week to review) 23:26:24 ok, let's try to add sbws reviewers early 23:26:40 I suggest that juga adds reviewers when they finish a ticket 23:26:48 teor: +1 23:27:05 yeah, could speed things up 23:27:10 teor: or asks for somebody to add reviewers, whatever juga feels comfortable with. 23:27:21 teor already answered my question about recommended versions; anybody else have thoughts there? 23:27:22 Sure. Whatever works for them. 23:27:26 right now i see #28864 unassigned? 23:28:15 I'll review that, if I can figure out AsyncResult 23:29:04 thanks, catalyst 23:29:26 anything on the recommended versions? if not, on to teor's question about blockers? 23:30:39 teor: my own thought there would be to ask "network-team" on IRC , and make it clear that there's a blocker, and about how hard it is 23:31:19 when I do that, somebody usually takes it on within the day. If not, it's usually because everybody thinks it will be too hard, and I have to ask again 23:31:28 Anybody have another method? 23:32:22 does everybody have network-team added as a highlight in their irc client? otherwise the other option is to highlight individual nicks 23:32:27 where's the blocker question? 23:32:38 i also didn't see the blocker question 23:32:40 catalyst: under "getting help" 23:32:44 ahhh 23:32:48 "what's the best way to get other people to help me with blockers" 23:33:22 oh, i was looking under teor's section 23:33:59 how about if nobody responds in a day or so, assign the reviewer yourself? and ping them on IRC? 23:34:08 +1 catalyst 23:34:30 huh. I am not sure how that will work, but let's try it for a few weeks and see how it goes? 23:34:32 * catalyst thinks people should also feel free to hand off such an assignment 23:34:46 let's do that judiciously and make sure we don't overuse it :) 23:35:13 hopefully these will be rare 23:36:38 Yeah, I tried network-team, and waited for 5 days. 23:36:48 Normally I would assign the review to the CI rotation, but that was me. 23:36:49 wrt ipv6-only network failures: I think right now our problem is that we don't think we have applicable funding. Maybe we should take this on as an unfunded task, but our procedure for doing that doesn't seem to work so well 23:37:10 teor: also try again every 12-24 hours. That's what I do. :) 23:37:23 teor: I emphasize how important the bug is, and how easy the review is 23:37:34 Yeah, we need a better process for this. Let's add it to the list of process problems. 23:37:47 yep ^ 23:38:09 gaba: any thoughts on the "when can we spend some time on getting tor clients working on ipv6-only networks?" question? 23:38:46 nickm: I would like us to allocate a "sponsor" in each roadmap for unfunded tasks, then assign tasks to that sponsor. 23:38:57 if we think is high priority then we should look at reserve time to do it in not sponsored time 23:39:26 Of course, that requires us to know the tasks in advance. And have our sponsored tasks completed in a reasonable time. 23:39:45 yes, it is tricky as we already have a lot for sponsored tasks 23:40:27 That's not quite what I mean. 23:41:09 I feel like we are already overloaded. I don't want to add another set of tasks to that overload. 23:41:31 I want to reserve time for urgent tasks, and decrease what we do for sponsored tasks. 23:42:39 (done) 23:43:28 gaba: any thoughts? 23:43:58 no thoughts right now about this, sorry :/ . I agree that it would be good to have more time for non-sponsor tasks 23:44:39 does anybody know how assigning people to 20%-time on "do what you think is important and fun" helps with this? I gather that it has been less than perfect at google, but many things have. 23:44:41 but I think it is not possible right now. Maybe something to consider for our next roadmap 23:45:32 I would love 20% time. But I feel like I already have more than 5 days of work to do every week. 23:46:04 Unless we want to cut 1/3 off our current roadmap? 23:46:34 I have a suggestion for the next few weeks: I don't know how urgent automatic IPv6 is. 23:46:37 the issue would be to sponsor the things that are fun and important to work on 23:46:53 hm. I think we should talk some time about expected time consumption? I'm comfortably within 5 days of work, which either means I'm slacking, I'm underallocated, or I have a slacker's idea about how much I should do each week 23:46:59 Likely some of each :/ 23:47:05 one could do a variant of 20% time where some people who have a project that we all find important, but don't have a sponsor for right now, get to work on that? 23:47:42 gaba: I disagree. We have funding that is not tied to sponsors. The network team should get some of that funding. 23:47:50 yes, I know indiegogo was doing something similar like ahf is mentioning 23:47:57 * specific sponsored work 23:48:08 teor: for that we need more specific sponsored funding for the other teams 23:48:14 the grants team is working on that 23:48:15 or figure out how to charge some of this work to "overhead"? 23:48:23 to get sponsor project for metrics and tbb 23:48:48 i _really_ like my current sponsor assigned tasks, so i'd prefer not to work on non-sponsored stuff, but i think we have projects that would be really good to get done (ipv6? pqc handshake? etc) and i'd love if some other people on the network team could get 20% time to do that 23:49:11 gaba: or those teams could do less 23:49:21 +1 ahf 23:49:38 if there is anything that really needs to be done and we do not have sponsor for we should look for a way to complete it 23:49:46 is that the case with IPv6? 23:49:47 teor: one problem we have right now is that we have enough specific funding to fund our team's salaries, but some other teams over the last year have not. 23:50:11 catalyst: that is what we are looking into doing right now for metrics 23:50:24 overhead for metrics as it is very hard to get sponsors for that work 23:50:42 I understand. And I understand that we are locked in to other teams' staffing, and the time to decide not to hire people has passed. 23:51:14 isabela will hold a tea time to talk about several things of the organization. maybe this could be one of the topics/questions for that 23:51:31 It would be great if we had a defined minimal mission that we needed to do regardless of funding 23:51:56 or maybe not 23:51:58 gaba: I can't go to the teatime, because it is scheduled for Americans/Europeans. 23:53:13 If we are going to schedule meetings for everyone, then we usually need to schedule two meetings. How can we make sure everyone knows that? 23:53:34 So, there is a plan that the next tea-time needs to be shifted from this one for timezones 23:53:42 At least, that's what isabela told me 23:53:51 It's likely that this is something she does know 23:54:15 yes 23:54:21 they will rotate for each time for what I understand 23:55:14 have this tea time thing been announced already or are we talking about something that is coming in the future? 23:55:27 I was looking for the email, but I can't find it right now. 23:55:50 mmm,sorry, maybe it has not been announced yet :/ 23:56:03 1) organizing ED tea time for March 21st, Thursday, 1400 UTC time 23:56:08 sat vegas team email :-) 23:56:10 last* 23:56:15 ok 23:57:22 Anything else from the meeting today? We are in the hour. 23:58:40 * ahf is good 23:59:07 teor: if you want we can have other moment to discuss this specifically thing about sponsors and non-sponsor work. We need to set priorities if there are some that are not contemplated in the sponsor work already. 23:59:17 * nickm is ok 23:59:54 I am done 00:00:27 okay. I'll call the meeting then. but if anybody wants to talk about something some more, I'll stick around #tor-dev a few more minutes 00:00:30 peace, all! 00:00:32 #endmeeting