17:59:27 #startmeeting weekly network team meeting, 26 Nov 17:59:27 Meeting started Mon Nov 26 17:59:27 2018 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:59:27 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:59:29 Hi all! 17:59:42 pad is still at https://pad.riseup.net/p/tor-netteam-2018.1-keep 17:59:46 hello 17:59:54 hello 18:00:00 it's almost december! How are folks doing? 18:00:51 gaba is still on thanksgiving time, so let's hope she is taking it easy 18:00:56 wrt roadmap -- I saw a bunch of bootstrap progress last week. Looks cool! 18:01:05 any blockers on roadmapped things? 18:01:34 everybody have something roadmapped to work on? 18:01:49 yep 18:01:58 still working on wtf-pad. crunching through the todo list. 18:02:08 i managed to integrate riastradh's probability distribution code into wtf-pad branch. 18:02:26 it was lots of work, but i think the code is more mature, accurate and well tested. 18:02:46 this week i will continue crunching the todo list by writing tests. hopefully mike will also come back and help me. 18:02:48 asn, mikeperry: do you two want another round of review from me there? I realized that there were some things in the previous branch that I don't remember, so maybe I didn't read closely enough 18:03:20 maybe. possibly at the end of the week? 18:03:22 ah, no mikeperry today either 18:03:38 i will let you know. i still need to squash riastradh's stuff pending his review. 18:03:41 asn: okay. Make a note to remind me on thu/fri? I will be at mozilla meeting next week, so preparing in advance will help 18:03:49 ack 18:03:51 asn: also I'll have a couple of questions for you after this meeting about the branch 18:03:53 nickm: will you have time to give #28179 a look this week when david is done with his review? i'd very much like a review from you especially on the windows code 18:03:59 i think right now i'm the only one who has run that code 18:04:11 ahf: ack 18:04:22 nickm: i have to bounce after this meeting. perhaps write it on trac, or send email, or write on irc and will reply later. 18:04:28 asn: ok 18:04:37 nickm: hope that works for you. 18:04:43 asn: no problem 18:05:00 should I take an early look at any other stuff? 18:05:11 catalyst: I see that you have some preliminary branches; want comment on them? 18:05:23 nickm: sure! it's still kind of rough 18:05:53 ok, what should I look at? 18:06:29 the stuff i've pushed to my orconn-tracker branch is some preliminary refactoring work. please let me know if anything seems awkward or horrible to you 18:06:58 will do 18:07:10 thanks! 18:07:24 let's see, that leaves dgoulet? 18:07:35 ? 18:07:43 #28179 is my world 18:07:48 ah, ok 18:07:52 Sounds like that's roadmap 18:08:06 on to review assignments.... 18:08:44 looks like I've got no reviews assigned this weel? 18:08:56 yep 18:09:06 it was basically one per person, max two. 18:09:11 anybody want to offload anything to me? 18:09:14 and we opted for not putting you nickm 18:09:17 ok 18:09:49 but if anybody is going to block sponsor8 stuff on reviews, I'd prefer that you consider offloading the reviews to me if you want. 18:09:58 everybody okay with their current load? 18:10:16 yes 18:11:05 ok. please ping me or the rest of the team if you want to offload, or if you have any reviews that are at risk of distracting you from s8 18:11:14 rotations! 18:11:26 I see nickm on triage, ahf on ci+coverity 18:11:30 ack 18:12:36 wrt announcements: There's the snowflake kickoff meeting tomorrow, in about 29 hours. 18:12:59 let's all stop by if we have the time, so that it isn't just ahf telling me about snowflake :) 18:13:15 :-) 18:13:57 on to discussion issues? 18:14:17 let's sort in rough order of how long this stuff will take... 18:14:38 Ahf reminds us about the snowflake meeting and wants folks to read a page on the wiki first 18:15:00 asn: mikeperry wants you and me to figure out his must-do items for this week. Want to talk about that tomorrow, since he won't be back till weds? 18:15:42 nickm: i think that was from last week fwiw 18:15:54 but yes we can talk about it tomorrow 18:16:03 asn: ok 18:16:07 figure out the remaining steps for the branch 18:16:09 i'll doublecheck 18:17:00 Here's another fast-ish one: I need people to tag any tickets that should block an 035-rc with "035-rc-blocker". 18:17:22 I think i had them all identified last week, and I'll look for new ones, but it's kinda important to know what they are 18:17:40 Anybody want to discuss any of those? 18:18:06 Also, gaba asks whether anybody wants any bootstrap reporting help. I think we already covered that, but please jump in if not 18:19:14 Nothing on those items? If so, let's move on to the review document thing. Does that maek sense to everybody right now? 18:19:20 Could somebody summarize what we're voting on? 18:20:12 i was not aware there is some sort of voting situation going on in that doc 18:20:25 it's on the discussion section -- there's a question with 2 options right now 18:20:30 someone (gaba?) asked us to vote 18:20:56 I think that's gaba's color on the pad 18:21:34 ok for (2) does assigned reviewer following the ticket until it's done mean the reviewer should update the review assignments sheet too? 18:21:53 I don't know. Any thoughts, anyone? 18:21:57 no 18:22:21 no thoughts or no it doesn't mean that? 18:22:23 the review assignements sheet is just there for pretty printing and helping anyone that wants to use it has a checklist 18:22:34 but Trac is the ground thruth 18:23:28 the spreadsheet is a better visualization of who's reviewed how much than you can get from trac though 18:24:30 can anybody summarize what we're voting on, or does only gaba understand? 18:24:58 that pad is _not_ in a state to vote on I think, I don't see a concrete proposition but rather more comments and brainstorm 18:25:24 i see two options to vote on. are other people not seeing those? 18:25:25 There's a pair of options under "TO DISCUSS AND VOTE" on that pad -- I think that's maybe what gaba had in mind? 18:25:43 oh the 1) and 2) in the second section? 18:26:31 I think? 18:26:41 my impression is that we follow 2 right now? 18:26:43 yeah ok 18:26:47 those are the only things labeled as needing a vote 18:26:56 when i get assigned i either need to ask someone to take over it but otherwise it's mine until it's approved/rejected? 18:26:57 there was some confusion last week 18:27:01 I think we do a mixture 18:28:10 I added a #3 that I think is more like what we do now? 18:28:17 (is this indeed more like what we do now)?) 18:28:41 not sure no 18:29:03 nickm: if the author dissapear while the ticket in needs_revision, then it is anybody to pick it off for fixes or new branch 18:29:27 but the Reviewer still stays assigned so is 3) about "then the Reviewer can be removed" ? 18:29:34 right, but I'm asking A: if the author gets back to it 3 months later, does it get the same reviewer? 18:29:35 for anyone to become the new reviewer* 18:29:52 B: if somebody else takes over writing the ticket, does it get the same reviewer? 18:30:36 * catalyst would be ok with (2) only if (a) review assignments account for the additional load and (b) reviewer updates the review assignments spreadsheet when changing a ticket back to needs_review 18:30:37 yes i can also see how tracking the ticket after N months might be hard. i think i remember tickets that I have been reviewer on, but sometimes maybe i forget myself. 18:31:46 I have a request for the review assignment spreadsheet. Could/should we have a column for estimated/actual review points? 18:31:58 ideally, to be filled in by the reviewer 18:32:01 ok wait a sec, I want to take a step back here 18:32:08 basically i see (2) as asking reviewers to periodically review their assignments to change the ticket state accordingly if needed (because dgoulet and asn don't want to do that?) 18:32:50 the spreadsheet is _not_ what should be looked at after our weekly meeting as in Trac is where the real status of everything is at... if a ticket changes 3 times Reviewer on Tues. and then on Friday you look at the spreadsheet, it is the wrong idea imo 18:33:00 it makes _no_ sense to track the same thing in two different places 18:33:27 sooo if we want the spreadsheet to be the *main* thing then we need to take a very different approach 18:33:40 that makes sense in theory. in practice, the review spreadsheet is laid out to function as a shared scorecard 18:33:41 I think the spreadsheet tracks something that trac doesn't 18:34:03 the spreadsheet doesn't track reviews other than those assigned at start-of-week 18:34:08 err, arg 18:34:18 the spreadsheet has the affordance of being easily used as a scorecard, so whether we intend it to or not, people will do so 18:34:19 the spreadsheet tracks which reviews are assigned at start-of-week 18:34:40 which we don't have otherwise 18:35:28 right so this is why we do it :) 18:36:03 but then _after_ that, the spreadsheet is really about a way for everyone to either use it as a nice todo list, update it, do whatever but ultimately Trac is where everything is modified 18:36:17 so if we modify the spreadsheet as we go over the week then we loose what you said nickm 18:36:46 hmmm 18:37:17 it sounds like we want a different set of things from this review / review-assignment process 18:37:51 maybe we should focus the "reviewer role" document for now on the parts we agree, and enumerate additional desired things separately, and try to get those somehow 18:37:55 how would that be? 18:38:24 yeah I do see those _very_ different but seems we keep circuling back to the spreadsheet in the discussion above ^ :S 18:38:31 so I'm thinking it is kind of related for some? 18:38:42 it's related for me, because it affects my choice 18:39:42 We have 20 min left in this meeting, and more things to talk about. How do we feel about my proposal for moving forward with discussing this via a summary of what we agree on, plus work on answering the harder questions, as independent things? 18:40:17 catalyst: I'm not sure I understand what affects your choices ? :S 18:41:11 for me, (2) means doing additional work. i wasn't regularly rechecking my assigned-review tickets if they weren't in the spreadsheet 18:43:08 and this additional work isn't visualized the way the review assignment spreadsheet is, so it's kind of invisible 18:43:35 there are choices involved about how we make work visible 18:43:46 * nickm is trying to answer teor's questions in the discussion question, but if anybody else has thoughts there, it would probably be helpful to them. 18:43:54 I'm very confused, you can't see the ticket that you are a Reviewer for ? 18:44:12 catalyst: as in over the week goes? 18:44:20 err during the week* 18:44:45 i'm saying people are going to look at the spreadsheet as a scorecard, and if i'm not updating it, i'm possibly doing work reviewing tickets that doesn't get visualized there 18:46:05 catalyst: ok so it is about tracking the work that sheet? 18:47:27 i might be ok with flipping a ticket back to needs_review when needed and then deferring it until the next round of review assignments come around if it's not urgent 18:48:47 ooook we are clearly in a different place about reviews within this team so yeah this needs much clarification :) :) 18:50:08 dgoulet, catalyst: do you think it would help if we summarize what we agree on and what we don't, in order to move forward? 18:51:16 thing is that the Reviewer role and then the process of reviewing over a week (that is a reviewer assignement period) seems very different between us all ... so seems *both* needs to be defined somehow where the pad for me is about the role only for now 18:51:26 but we have 8 min left so we can move on... 18:51:38 (as in try to address this later) 18:52:03 ok. What other discussion things do we have? 18:52:38 Let's think about teor's "Should we add more testing time to our funding applications" 18:53:44 When I work on the timelines I usually try to add a bunch of testing time... 18:53:56 ... but we usually start development later than planned, because we're busy with other stuff 18:54:04 is that something we can do usually? 18:54:04 I wonder if we need even more testing time, or something else 18:54:13 ahf: I don't know 18:54:23 maybe we should talk about this with gaba when she's back too 18:54:24 we might not account enough for maintenance overhead when writing grants, maybe? 18:54:32 that's possible! 18:56:43 okay, I think we're out of time. Let's keep talking about this stuff more, on IRC and the ML and etcc. 18:57:05 last thing in case anybody didn't see: I'm not here next week, I'll be at the mozilla thing 18:57:18 See you online everybody! 18:57:23 o/ 18:57:32 #endmeeting