16:59:47 <nickm> #startmeeting weekly network team meeting, 21 May!
16:59:47 <MeetBot> Meeting started Mon May 21 16:59:47 2018 UTC.  The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:59:47 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:59:48 <haxxpop> Usually I wait for the pad url when the meeting starts Haha
17:00:08 <asn> hello
17:00:14 <haxxpop> hello everyone!
17:00:14 <nickm> the much-discussed pad is at https://pad.riseup.net/p/zokuBZL3bEzK
17:01:20 * asn reading updates
17:01:35 * arma2 pops in and is nearby but doesn't have much eyeball (so let me know if you need something from me and then be a bit patient)
17:01:39 <nickm> first announcement is: we won't have any online meetings next week; monday is a US holiday, and nearly everybody is traveling on Tuesday
17:02:02 <ahf> cool
17:02:27 <nickm> this week we have asn on bug triage, catalyst on community engagement, dgoulet on coverity, and isis on CI.
17:02:43 <nickm> (per rotations at https://trac.torproject.org/projects/tor/wiki/org/teams/NetworkTeam/TeamRotations )
17:02:54 <isabela> o/
17:02:56 <nickm> Let's start with our regular roadmap checkin!
17:02:58 <nickm> hi isabela !
17:03:46 <nickm> None of our remaining march/april items are 034 blockers, though we should really close them out.
17:04:02 <nickm> catalyst: you and I should come up with something for #25507
17:04:03 <isabela> nickm: should we move them to may ?
17:04:20 <nickm> we should move #25507 to may.
17:04:55 <nickm> done
17:05:06 <catalyst> nickm: yeah #25507 can move to may
17:05:12 <nickm> for #25510, we should split it -- I'll do that after the meeting
17:05:25 <nickm> ahf: what should we do with #25496?
17:05:52 <ahf> move it to may i think. i've been running with it for testing on my own device and there it looks fine. gonna talk with nathan/hans about it again
17:05:58 <nickm> ok
17:06:15 <nickm> is there going to be any further action there, or do we just say "yes, it works" and close the ticket?
17:06:37 <ahf> i was gonna close it once there was a release out with it
17:07:10 <nickm> ok
17:07:40 <nickm> so it seems we'll actually have our may15 roadmap closed out
17:07:43 <isis> o/ sorry i am late
17:07:53 <nickm> our may-jun items are mostly "continued"s
17:07:53 <isis> i moved into a new house
17:07:59 <nickm> isis: congratulations!
17:08:21 <nickm> isis: remember you're on CI this week :)
17:08:25 <isis> and the internet is not hooked up yet, so i am working i guess from coworking spaces/cafés
17:08:50 <nickm> also please try to get your updates into the pad before the meeting starts
17:09:30 <isis> if there is something super urgent at any point, signal is definitely the fastest way to reach me this week
17:09:36 <nickm> let's all take a minute to talk to the people we're working with on may..jun things about what we're going to do on them.
17:09:52 <nickm> (if we need to coordinate)
17:10:12 <nickm> catalyst: let's take 5 minutes after this meeting to make sure we progress on #25507 this week?
17:10:22 <catalyst> nickm: sure
17:10:24 <ahf> i've got far with #25497 but i'm alone on it, so nobody to chat with there :-)
17:10:33 <ahf> gonna coordinate with the guardian project there
17:11:04 <nickm> hm. anybody know where dgoulet is today?
17:11:35 <asn> i think he is away
17:11:40 <asn> apparently there is public holiday in ca
17:11:42 <nickm> ah
17:11:54 <dmr> dgoulet was away the end of last week, right? I guess still away
17:12:06 <nickm> it's... victoria day?
17:12:08 <arma2> many canada people did a four day weekend
17:12:12 <nickm> makes sense
17:12:23 <ahf> i think he says in his nsec email that he is away for some of today too
17:12:38 <nickm> ok, I'll talk to dgoulet later this week about continuing #25550 stuff
17:12:41 <nickm> err
17:12:44 <nickm> #25500
17:13:22 <nickm> on to review assignments?
17:13:32 <arma2> if anybody, while looking at their stuff, runs across tickets that i am supposed to be on, please let me know. :) (i'm not so good at 'skim lots of things' currently)
17:13:40 <nickm> I see a bunch of stuff assigned, approx 1-3 per person...
17:13:59 <nickm> arma2: what's the best way for people to let you know?
17:14:47 <nickm> asn: Did you see the discussion in #tor-dev earlier today about assigning reviews for volunteer patches even if they're for 0.3.5?
17:15:02 <asn> hmm i think i missed that
17:15:18 <asn> that said, we assign reviewers for all tickets in needs_review
17:15:24 <asn> regardless of milestone
17:15:30 <nickm> asn: okay.  I was musing that maybe we should also be reviewing 035 tickets if they are patches from volunteers.
17:15:34 <nickm> Great!
17:15:41 <nickm> then you're already doing that.
17:15:46 <nickm> good to know :)
17:15:54 <asn> :)
17:16:36 <nickm> okay, so standard announcements:
17:16:44 <nickm> no online team meetings next week
17:16:49 <nickm> instead there's a hackfest in seattle
17:17:06 <nickm> Please see the hackfest pad for planning there
17:17:30 <nickm> IMO the most important stuff we can do this week, after code review and bugfixing, is planning for next week and making sure we show up rested
17:17:45 <dmr> nickm: said pad is still private, right?
17:17:49 <nickm> Everybody make sure to look at the list of stuff and see if you want to check in with asn about any of it, too: he isn't going to be able to come
17:18:03 <nickm> Does anybody object if I post the hackfest pad publicly?
17:18:22 <catalyst> it's findable via a chain of public links already, i think
17:18:46 <Samdney> yeap, hehe...
17:18:48 <ahf> i think the pad have been shown in meetings here before and in minutes
17:18:51 <nickm> It is, but I don't like the reasoning that whatever is findable okay to advertise without asking
17:18:58 <nickm> Though I think this is okay to post
17:19:18 <nickm> Last chance to object....
17:19:39 <nickm> okay, it's in the notes.
17:19:39 <arma2> nickm: here during this irc meeting is one option. or email is a good backup option.
17:19:41 <nickm> https://pad.riseup.net/p/0RxS2ZcRe9Eb
17:21:25 <isabela> i will add the links for 1:1:1 schedule
17:21:31 <nickm> I already did
17:22:05 <isabela> i plan on adding link to pads of the things i will be organizing prep-work this week (noted on my update of this meeting pad (not the seattle one))
17:22:08 <isabela> nickm: ! tx
17:23:01 <nickm> Does that cover all our items up to (but not including) discussions?  Anything you want to add, isabela ?
17:23:38 <isabela> nope
17:23:55 <nickm> okay.  Please keep an eye on the network-team mailing list for any updates about Seattle stuff
17:24:10 <nickm> on to discussion
17:24:22 <ahf> ack
17:24:43 <nickm> catalyst: yes, I'm okay with merging the appveyor branch as-is, and chasing the remaining heisenbug if it turns up too often in practice
17:24:58 <nickm> if you think that's okay
17:25:14 <nickm> My big question is: can I release 0.3.3.6 this week?
17:26:08 <catalyst> nickm: i'm ok with that. i would like to look at a squashed version of the appveyor branch with logically organized commits
17:26:30 <nickm> catalyst: okay.  I'll have that to you before this time tomorrow
17:26:38 <catalyst> nickm: thanks!
17:28:02 <nickm> isis, arma2: I'm interested in particular in your thoughts about whether I should do TROVE-2018-005 in 0.3.3.6, if I put 0.3.3.6 out in the next day or 2
17:28:46 <catalyst> do we want tbb-dev to look at #25549 to see if it reasonably resembles how they build tor.exe for Tor Browser on Windows?
17:29:42 <nickm> I think we could do that as a followup
17:29:55 <nickm> their build process is ... big
17:30:36 <isis> nickm: i think it should be fine to include? i ran it under chutney for a while and didn't notice any issues, but also i didn't test all that well since i didn't write a poc for the vuln
17:30:38 <arma2> nickm: that's the trove with the dir auth fix? the one i'm testing on moria1? putting it in 0.3.3 is fine by me. most dir auths won't move to 0.3.3 immediately, but that's fine i think. we can react if somebody starts using it on older versions.
17:31:00 <arma2> i guess yea, another risk is that the fix doesn't really fix it. but, i think in that case we can react too.
17:31:09 <nickm> arma2: okay, so you're okay with an 0.3.3 release with it, and a possible bckport?
17:31:14 <nickm> it's "just" a dos :)
17:31:27 <arma2> yep. put in the fix and move on and hope you never need to go back to it, i say :)
17:31:40 <nickm> if you're both okay with that, great!
17:31:46 <isis> i think we discussed backporting it to 0.2.7? should i do that?
17:32:07 <nickm> isis: 0.2.9 is the earliest we would consider at this point... and hm
17:32:12 <nickm> here's what we should do
17:32:18 <arma2> i'm mildly inclined to put it into 0.3.3 and keep it around for a backport if we learn that we need one
17:32:24 <nickm> arma2: could you email the dirauth ml and let them know this is coming?
17:32:40 <arma2> yes
17:32:47 <nickm> today ideally :)
17:32:54 <nickm> isis: How long would the backport take?
17:33:11 <nickm> If less than an hour, then yes.  Otherwise, let's be ready to do the backport, but hold off on it.
17:33:16 <nickm> sound ok?
17:33:50 <arma2> no dir auths run 0.2.9 btw
17:33:57 <nickm> then forget that
17:33:57 <nickm> w
17:34:04 <nickm> what is the oldest version any dirauths run?
17:34:20 <arma2> 0.3.2.x
17:34:28 <arma2> there's one called Bifroest that is on 0.3.1 for some reason
17:34:35 <arma2> but it doesn't handle these documents i think maybe
17:34:38 <asn> seems to be 031
17:34:42 <asn> bifroest0.3.1.9
17:34:45 <asn> from consensus health
17:34:55 <asn> ah arma2
17:34:56 <arma2> they're all on 0.3.2.10 except moria1 which is on master
17:35:01 <nickm> ok
17:35:36 <nickm> isis: so what would be most useful at this point would be an 0.3.2 backport, if it takes less than an hour
17:35:40 <nickm> what do you think there?
17:36:11 <isis> that should be easy to do
17:36:19 <nickm> and if not, I'll just take the latest 0.3.3 patches again please :)
17:36:20 <nickm> ok
17:36:23 <nickm> please assume t
17:36:27 <nickm> arg
17:36:37 <nickm> please assume that my gpg setup is now 100% command-line :)
17:36:59 <isis> bifröst will need an upgrade
17:37:21 <isis> nickm: i'll do the encrypted tarball thing again, does that work?
17:37:29 <dmr> nickm: because of efail, or for other reasons?
17:37:37 <nickm> works for me :)
17:38:38 <nickm> dmr: complicated other reasons
17:38:51 <Samdney> what is the best way to get in touch with komlo? want to talk with her about multi threaded crypto stuff. left a msg at irc and wait, or? she isn't often online at irc, I think. :)
17:39:12 <nickm> dmr: though dmr is a nice told-you-so
17:39:15 <nickm> err,
17:39:25 <nickm> dmr: though efail is a nice told-you-so
17:39:48 <nickm> Samdney: IRC is usually pretty good.  If she's busy, it's probably best to let her work. She's a volunteer, after all :)
17:40:31 <dmr> nickm: gotcha :) - and slightly glad but sad that _I'm_ not a nice told-you-so ;)
17:40:50 <Samdney> thanks nickm. it's only a bit hard for me to get in touch with her, hehe .. :)
17:41:26 <nickm> last discussion thing is catalyst 's note on architecture pre-discussion.
17:41:30 <dmr> Samdney: just a suggestion: ask komlo what her preferred method of coordination is
17:41:33 <nickm> want to say anything there catalyst ?
17:41:35 <nickm> o
17:41:50 <nickm> oh, and I'm away on Friday&Monday, and isis is moving and prefers signal.
17:42:01 <ahf> i'm traveling monday next week
17:42:06 <nickm> ack
17:42:12 <Samdney> if i can reach she on irc, i will ask, dmr ;)
17:42:49 <catalyst> re architecture: i was thinking people who are interested in writing down some architecture guidelines might want to meet up during or before this week's patch party
17:43:27 <nickm> I'm happy to carry over architecture stuff from Tuesday evening to the next morning, so everybody has a chance to discuss regardless of timezone
17:43:48 <catalyst> i think nickm teor komlo isis are the people i usually think of as liking to think abaout architecture. we can also continue the discussion in Seattle
17:43:51 <ahf> what are the architecture guidelines here?
17:44:12 <nickm> I think we might not get teor4 : they're busy with a paper, and about to be decompressing afterwards.  But we'll see
17:45:01 <arma2> right -- guidelines for what?
17:45:11 <nickm> ahf: trying to figure out how to make tor less spaghetti-ish basically
17:45:15 <ahf> ack!
17:45:28 <nickm> like, we have all these modules on a chart we came up with in Rome
17:45:35 <ahf> ah, cool
17:45:35 <catalyst> ahf: high level things to think about when making choices about code, especially "how we can make choices to make stuff more maintainable and reduce the rate of increasing technical debt"
17:45:45 <nickm> but how do we get from where we are now to a place where they fit together nicely
17:46:01 <ahf> yep
17:46:01 <nickm> that kind of thing
17:46:25 <catalyst> like not a "burn everything down and start fresh" but "what does our more perfect software look like and how do we incrementally make choices to get us closer"
17:46:39 <epsilon-nFA> sorry if this is not the right time but : Is the main goal to move the whole code base to rust?
17:46:44 <nickm> kind of like how we've done in some other areas in the past
17:47:00 <nickm> epsilon-nFA: to be determined.  Maybe.
17:47:20 <nickm> epsilon-nFA: We won't know if we want to move the whole codebase to rust until we've moved some of it to rust :)
17:47:33 <nickm> and that part is still a work in progress
17:48:15 <epsilon-nFA> nickm: copy that
17:48:33 <nickm> examples of other places where we've done things like this in the past: migrating to improved testing; trying to have fewer fatal assertions; using trunnel; etc
17:49:29 <nickm> okay, any other announcements or topics for this week?
17:49:45 <nickm> if not, I'll thank everybody and declare the meeting over!
17:49:56 * nickm waits 30 sec...
17:50:51 <epsilon-nFA> leave
17:50:55 <nickm> ok, thanks everybody!
17:50:58 <asn> thanks :)
17:50:59 <nickm> #endmeeting