16:58:54 #startmeeting Network team meeting, 26 Mar 2018 16:58:54 Meeting started Mon Mar 26 16:58:54 2018 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:58:54 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:59:05 hello 16:59:11 hello, everyone! 16:59:11 hi all ! 16:59:20 our meeting pad is at https://pad.riseup.net/p/q7WG8JlWWQZd 16:59:22 hello 16:59:41 hi 16:59:48 I see updates from nickm, dgoulet, isabela, mikeperry, ahf, juga, and catalyst 17:00:08 asn, isis: ping? 17:00:43 the idea is that we ackknowledge the tickets we have *and* our rotation roles at the beginning of the meetings, right? 17:00:49 we're going to be doing a somewhat modified format this week, in conformance with our changes... 17:01:11 ahf: I think so. isabela: is that right? 17:01:27 pong 17:01:46 hi! welcome! happy monday! 17:02:02 hey hey hey 17:02:54 while we're getting the last couple of updates, let's start by looking at the roadmap 17:03:06 isabela: you can take over for this part if you want 17:03:11 oi 17:03:20 sorry folks! 17:03:23 on our plan for right now is: fix must-fix bugs in 0.3.3 17:03:25 :) 17:03:35 that includes fixing them all and reviewing the code 17:03:37 yes, i would like everyone to look at the queries on the pad 17:03:50 there are around 30 tickets there (if I am not mistaken) 17:04:05 and as we discussed in rome and following our roadmap 17:04:12 this should be priority right now 17:04:24 around! 17:04:26 * arma1 lurks and observes 17:04:42 so my question to ppl is: do you think what is in the query can be done in the next 2 weeks (starting today) 17:05:00 * dmr also lurks :) 17:05:04 * asn posted review 17:05:06 because this will be critical to assess how much we can do for what we are allocating in our roadmap 17:05:16 * asn ugh posted progress report 17:05:19 * antonela holds popcorn 17:05:22 hi asn ! 17:05:42 * pastly finally posted his update 17:05:47 isabela: are you talking specifially about the ones in "033-included-2018" ? 17:05:50 isabela: i'm sorry, which queries? the 0.3.3 ones? 17:05:57 at the pad 17:06:00 the one nickm is saying 17:06:08 I think she means the first of the two very long trac URLs in the pad 17:06:16 yes 17:06:30 ack 17:06:39 i would like the team to ack those tickets and ack their priority for this week and the next one 17:06:55 so, I think that it would actually be safe if we can't fix the pending sandbox stuff in 0.3.3. I plan to spendno more than a day on those two items 17:06:58 and it's got an 'owner' column so you can see which ones everybody thinks are yours :) 17:07:03 and either succeed or fail 17:07:18 There are 4 tickets in "new" with no owner 17:07:39 i can take one or more of those; i only have one right now 17:08:00 yes, this should be priority over other stuff 17:08:02 there are also the needs_information ones. If we can't figure those out, we have to defer 17:08:12 * ahf takes #25245 17:08:36 thanks ahf 17:09:01 arma1: I have a suspicion that you maybe know what part of the code to patch to fix #25061 . If you don't have time to fix it yourself, could you write a quick note to the person who will do the fix? 17:09:03 wait, there are 10 more in 033 milestone 17:09:11 (that don't have "033-included..." 17:09:15 dgoulet: yes 17:09:16 yes 17:09:24 dgoulet: some of them are not yet removed but we are proposing to be removed 17:09:25 those did not make it in during the first round of triage 17:09:33 i am not prioritizing ti now for this discussion though 17:09:36 they are currently proposed-for-removal if they have 033-removed 17:09:47 i can take a look at #25347 17:09:48 because i want to make sure the things we are sure are in are taking care of first 17:09:57 asn: cool! 17:10:01 is this for next monday? 17:10:16 ahf, asn: If you get stuck on tose, please pull others in to help 17:10:18 nickm, isabela: ok filtering on 033-removed, 27 tickest exists, I put one I think in 033 today as a must because regression 17:10:19 *those 17:10:27 so I think we don't have the right filtering rule or I messed up :) 17:10:29 asn: we want all in this query to be done by april 6 17:10:42 ideally sooner if possible 17:10:44 dgoulet: just use the query from the pad 17:10:52 nickm: yep, i might pull others in for help with the one i have that is in needs_info 17:10:53 so we can stabilize 0.3.3 17:10:55 isabela: I did but its missing 2 tickets :) 17:11:10 ah 17:11:23 which are these two must-fix tickets? 17:11:25 2 tickets got added since the 033 triage basically 17:11:35 i have the impression that #25347 will be a hard one. i'll take a look anyway since it's guard stuff. 17:11:37 before friday the 6th of april? why friday? 17:11:42 catalyst mikeperry isis please take a look at those too 17:11:54 our stable release date is 17 april 17:12:09 sorry what's the plan for 033-removed-2018? 17:12:10 ahf: i will be chekcing this friday march 30th how things are going 17:12:22 the 033-included-2018 query? 17:12:27 nickm: #25616 and #24795 it appears 17:12:33 nickm: yes i will try to post a hint to #25061 17:12:36 ok, so we use monday to friday for this and then we'll get next round on the monday meeting 17:12:41 that sounds sensible to me 17:12:52 * ahf keeps thinking in monday to monday periods 17:12:52 cool 17:13:04 #24795 can stay in, but needs_information. I can't reproduce it 17:13:18 yeah the goal here is to make sure all this can be done asap so we are good for the release 17:13:24 after that we can move on to the rest of the roadmap 17:13:44 ok 17:14:08 dgoulet: #25616 isn't listed because it wasn't triaged: you put it into the milestone after my most recent triage pass 17:14:23 nickm: yes that is what I said, I shouldn't have? 17:14:39 no; you can put it in -- it just doesn't count as a triaged ticket because it hasn't been triaged 17:14:57 I'll mark it triaged now 17:15:17 there 17:15:21 * asn takes ownership of #25347 17:15:38 asn: ty 17:15:47 two tickets left 17:15:55 isis: are you happy as owner of #25517 ? 17:16:37 I can do #25061 if arma1's suggestion isn't too hard, but maybe somebody else could take it too :) 17:17:07 * catalyst will take #25961 for now 17:17:26 er, #25061 17:17:38 cool; thanks! 17:17:55 remember to assign the tickets you are taking to yourselves. 17:18:05 i expect my suggestion will be "find the place in that code that calls that thing, and only call it if the circuit is an origin circuit" 17:18:26 arma1: I think that by expanding "that thing" you will be a helpful person. 17:18:32 ^^ what nickm says! :) 17:18:44 I can take #25616 with some asn help for review 17:18:49 everyone: Does anybody have any tickets assigned to them for review or for fixing in 0.3.3 that you think you won't be able to do? 17:18:56 awesome, thanks! 17:19:36 do these big trac queries have a 'reviewer' column displayed? if not, should we make them have one? 17:19:55 next time sure 17:19:57 * isabela reminds folks on how important it is to asses what you can do or not and let us know, so we can plan accordingly 17:20:02 arma1: or just use https://trac.torproject.org/projects/tor/query?status=accepted&status=assigned&status=merge_ready&status=needs_information&status=needs_review&status=needs_revision&status=new&status=reopened&keywords=~033-included-2018&milestone=Tor%3A+0.3.3.x-final&group=status&col=id&col=summary&col=owner&col=type&col=priority&col=component&col=version&col=reviewer&order=priority 17:20:25 nickm: i'm fine with doing #25517, and was planning to do the same fix in both languages as in in the patch on the security list, if people are okay with that 17:20:45 k 17:20:46 cool; I'll re-read the message and assign to you 17:20:46 nickm: is there a convenient time to have that done by for a point release? 17:21:09 I'm hoping to do an alpha this week and another one in about 2 weeks 17:21:10 oh whew. dgoulet signed up armadev as the reviewer for 25226, not arma. great then. :) 17:21:23 the fix shouldn't take more than half an hour 17:21:24 :-D 17:21:29 arma1: oh did I lol... arma is on trac ok ... 17:21:31 great 17:22:19 so the idea is now: get this list down to be as short as possible, once it's shorter/empty we continue with roadmap stuff? 17:22:20 (me wonders if we are moving to the tickets review part of the meeting) 17:22:30 ahf: yes 17:22:34 cool, i like that 17:22:37 yes, by fixing issues 17:22:43 yep 17:23:11 (everyone, please prioritize this stuff so we can get 0.3.3.x out on time. I know features and design things are more fun, but please remember that nothing ever lands until we debug it and get it stable.) 17:23:19 ack 17:23:35 * ahf nods 17:23:38 it would be great if we finished this all asap, and got to the 0.3.4.x stuff ahead of time 17:23:51 alright, so to the reviews assigned for the week - does dgoulet or asn wants to highlight anything? 17:24:20 if someone wants more reviews this week, please let us know 17:24:31 same for the opposite 17:24:48 does anyone has any questiosn to dgoulet or asn ? 17:25:01 also if anyone has any super important tikckets in needs_review, please let us know 17:25:05 so that perhaps we do a swap or sth 17:25:22 also, isis did you figure out your review situation? 17:25:33 to add to this ^, if you can find a reviewer also for your important patch, that is also very nice :) 17:25:38 isis: where you are assigned as the reviewer on a ticket where you wrote the code? 17:25:39 isis: i've left some comments for you on #25425 btw 17:25:41 but else, we can assign based on weekly tasks 17:25:57 asn, isis: I think that was the crypto.c splitting one? 17:25:59 yes 17:26:04 asn, isis: If so, I've taken the review, or I mean to. 17:26:08 ahf, asn, catalyst, dgoulet: you are listed for rotations this week, https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkTeam/TeamRotations 17:26:17 yep 17:26:23 ack 17:26:26 ack 17:26:28 nickm: ah great! 17:27:09 catalyst: coverity seems like it could be busted right now; I'd be happy to help debug it though 17:27:34 (or it might just be messed up. It says "In-queue. Your build is in the queue to be analyzed. There are 101 builds ahead of it.") 17:27:34 nickm: isis: ack on the review swap with isis. will you reflect the spreadsheet, or you want me to do it? 17:27:42 could you edit the spreadsheet? 17:27:46 sounds good 17:27:52 nickm: i recall they sent out mail about some operational issue that took them down for a while? maybe an intrusion or leak? 17:27:53 also I'm going to prioritize 0.3.3 review over others 17:28:13 nickm: you have many reviews for htis week!!! 17:28:21 nickm: you have 5 reviews 17:28:21 asn: fewer than usual tho 17:28:35 that's small compared to what I'm used to :) 17:28:58 dgoulet: did you add #25605 to nick? 17:29:06 catalyst: Yeah, something like that. They say they last analyzed tor on Feb 3 17:29:17 asn: I did, check sheet ;) 17:29:21 let's work on making it smaller still over time. that way nickm can eventually get to bigger picture stuff, and also more people get more experience reviewing. 17:29:29 dgoulet: perhaps we should take it back for this week? given that he has too many? 17:29:30 nickm: thanks! i can trade you for something, if you like 17:29:38 catalyst: let's see if we can fix the problem, but not worry about it too much if not. 17:29:41 ah or isis trades with nick 17:29:55 isis: ask me again once you have all your 033 tickets and needs_review stuff done :) 17:29:59 sure whatever works for that ticket 17:30:05 nickm: we can talk after the meeting about that to keep the noise down here :) 17:30:06 nickm: ack 17:30:12 catalyst: good idea. 17:30:18 ok 17:30:28 we're 30 minutes in and have done a huge amount of scrambling 17:30:39 I'm pretty impressed at everybody following through on the "/me status: " updates too 17:30:54 we probably can't hit everything else today... 17:31:12 i like it, i think it gives a great overview of what is happening 17:31:24 i also like it 17:31:27 one thing that's important: even though we aren't working on it yet, we should all make sure we're talking with the people we're collaborating with some time over the next couple of weeks 17:31:52 yeah, the status updates are pretty good 17:31:53 so, if you are listed as working with somebody on some roadmapped item for 0.3.4.x, please make sure you set up a time to all talk together for 15-30 minutes 17:32:28 oh right, meeting tomorrow for nickm, me, dgoulet(, and catalyst?) 17:32:39 isis: yes! for wide create chat 17:32:48 isis: yes! 17:32:50 and i'm talking with ahf about idling APIs after this meeting 17:32:57 :-) 17:32:57 great! :) 17:33:00 going through other people's questions ... 17:33:11 juga: hi! You are totally welcome here! 17:33:32 Yes, please feel free to step by our meetings; and ask us for whatever support would be most helpful 17:33:48 nickm: thanks 17:34:03 * asn still hasn't read status updates 17:34:05 I think your questions would be welcome in #tor-dev and tor-dev@ 17:34:06 nickm, isis: just as a quick primer... the "/me status: <...>" updates is a recent informal mechanism in #tor-dev to keep each other posted on what one is diving into? 17:34:11 question: folks mentioned apr 6 as the "finish by then for sure" date. that's two weeks from now (well almost). the idea is to re-triage next week at the mid-way-through mark? 17:34:32 (/me wants to comment that we wont take 30min with roadmap review and tickets review ackownlegments all the time, this was our first time doing this and included important 033 assigments. So it longer but that is fine since this is very important for all the team) 17:34:34 mikeperry: the two-guards thing seems to be making progress, but I haven't been as active in that thread as other people. what do the others think? 17:35:13 dmr: yes, that's the idea. It's meant to help counteract our timezone spread and give us more visibility to one another 17:35:45 nickm: fyi I didn't see anythig that jumped out at me that was not yet done, but I can take review of #21394 off your plate, since I have touched a bunch of related stuff 17:35:49 nickm: cool, thanks! :) (and cool plan - hope it works out!) 17:36:04 mikeperry: cool ; please do! I'd greatly appreciate that 17:36:41 mikeperry: you are also listed as reviewer on #24456 and #25055 in 0.3.3.x 17:37:20 nickm: yeah I spoke with asn about #24456. basically just need to update the ticket. I also started looking at #25055 this morning 17:38:01 great 17:38:18 nickm: wrt two guards, I am not sure what others think other than people on the thread. asn and dgoulet seem convinced now. asn wants a short proposal 17:38:56 asn, dgoulet: you agree? 17:39:03 I'm getting convinced yes, I do agree with a proposal here so we clearly document the why and when 17:39:05 if so, yes, a short proposal is appropriate. 17:39:17 let's do that at a _lower_ priority than finishing 0.3.3 though 17:39:28 basically we're in the rock&hardplace position in that in order to *actually* use only one guard all the time, we need to abandon Tor's path restrictions entirely 17:39:30 yes im also getting convinced. i still havent done the math, on hwo the sybil probabilities increase when we go from one to two guards. 17:39:36 (that's assuming that a short proposal is also an easy-to-write one) 17:39:40 which I think is not as good of an idea as having two guards 17:40:33 mikeperry: +1 on that 17:40:40 I think pretending that Tor could use one guard and comparing that to two guards is not as helpful without that context (of needing to abandon our restrictions to do so) 17:40:49 but yes short proposal shouldn't take too long considering the data in the email thread :) 17:41:22 to me it comes down to "how safe are you under passive observer" and "how safe are you under active attack" 17:41:31 asn, dgoulet, mikeperry: Are you fine taking the "write a proposal" task there? Targetting _after_ all the 033-must stuff is done 17:41:31 and so far they seem to have different answers for 1 guard vs 2 guards 17:41:34 ? 17:42:06 yes. i'd like to work on that. 17:42:09 and agreed about priority 17:42:14 not sure if we should roadmap it. 17:42:22 very much a reviewer on this, not an author so I'll let it to mikeperry/asn 17:42:52 i'm going to close #25508 as a duplicate of #24660 and copy over missing tags/cc/sponsor fields 17:43:09 isabela: should we put "write this 2-guard proposal" on the roadmap? 17:43:11 if it's not on the roadmap currently, it seems clear that it should get put on the roadmap, as part of the "put stuff on roadmap consensually" process? 17:43:16 nickm: yes 17:43:28 we can put it as 'R' task under SponsorV 17:43:31 +1 17:43:31 isis: if you do that, make sure to update the roadmap so it no longer points to #25508 17:43:42 ok. who updates the roadmap? 17:43:50 * asn doing it 17:43:55 tx asn 17:44:33 added a basic item on roadmap. we should make ticket etc. 17:44:46 arma1: the problem is that passive observers don't have to participate in this particular active attack.. they can remain passive and still get all the info the adversary needs 17:45:10 pfew, that's a bunch of stuff we did 17:45:14 asn: ack 17:45:16 any more for this week's meeting? 17:45:17 mikeperry: well sure but if nobody is doing the active attack on you right now, then one guard is better. but this is a matter for the proposal, not this meeting. :) 17:45:56 nickm, isabela: i had a question from above: folks mentioned apr 6 as the "finish by then for sure" date. that's two weeks from now (well almost). the idea is to re-triage next week at the mid-way-through mark? 17:46:05 k, updating the roadmap to point to #24660 17:46:12 we will be looking at the roadmap and the assigned tickets every week 17:46:34 triage can always be reconsidered... 17:46:38 arma1: what nickm said plus that I plan on checking the progress of things this friday before next monday 17:46:51 isabela: note that many people are away on friday 17:46:59 sure 17:47:15 great 17:47:21 i assume she means "progress as described on ticket" 17:47:28 because no ticket updates = no progress :) 17:47:35 exactly 17:48:00 oh, one more discussion topic: kat5 has asked us to try to list volunteers whom Tor should send T-shirts. Let's set up a not-public pad on network-team@ to talk about that, so we aren't discussing everybody in public? 17:48:16 yes 17:48:44 oh, yes 17:48:54 i will uh, erase my suggestion for now 17:49:05 good call 17:49:20 already done :) 17:49:27 ah thanks :) 17:49:57 ok. if there's nothing else for this week's discussion... I suggest that we call the meeting over ... 17:50:03 but don't leave the internet right away... 17:50:21 instead, have a look at everybody else's updates if you haven't done so already, and see if anybody has questions for you! 17:50:46 * isabela happy we had our first meeting in this format 17:50:53 lots of things done 17:50:59 yes, i like it too 17:51:09 asn: ahf: i don't think i got assigned any more reviews? is that intentional? 17:51:24 i'm not assigning reviews: dgoulet ^^ 17:51:24 catalyst and others had i think a fruitful discussion about renaming the community hero. community guide and community advocate were the options that floated to the top i think. 17:52:00 isis: it is, the list we each have now is the list from Rome and it "starts" this week as in by next Monday, they should be reviewed :) 17:52:38 ah okay, i'll work on my needs_revision stuff and then maybe thursday or so see if i can steal tickets from someone else 17:53:58 okay, thanks everyone! 17:54:00 #endmeeting