13:29:20 #startmeeting network team 13:29:20 Meeting started Wed Feb 17 13:29:20 2016 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:29:20 Useful Commands: #action #agreed #help #info #idea #link #topic. 13:29:34 i can start (?) 13:29:36 hi all! It's another wednesday morning, I've just rolled out of bed, my breakfast is on the stove, and let's talk about tor! 13:29:42 go for it, asn_ ! 13:29:45 hello there 13:29:47 hi 13:29:58 during the past week, i worked on prop#250 and the excellent review of teor. 13:30:12 i think yesterday we fixed all the issues. today i'm going to review dgoulet's changes and hopefully put it back in needs_review 13:30:32 i was also kept busy by olabini & team who are working on the new guard algorithms. 13:30:36 they've been doing lots of progress. 13:30:49 they have now specified a new algorithm, and working on simulating it. 13:30:55 soon they are going to start implementing it in tor. 13:31:27 i'll try organize a small meeting with them so that they can brief us up on the project, and also so that we can show them how they can implement it in tor the best way possible. 13:31:51 i also setup karsten's #8786 on my bridge to test it. 13:32:06 some problems occured in those stats, but they will be fixed soon i believe. 13:32:13 moin 13:32:14 that's that from me. i've also not written my january status report yet. 13:32:51 are the thoughtworks people coming to valencia? 13:32:57 olabini and rjunior are 13:33:02 could we organise a session for this? 13:33:10 yes, i guess we will. 13:33:22 however, helping them now is also important, because they wont have much wokring time after valencia 13:33:33 i think the project stops end of feb/beginning of march. 13:33:39 *their project 13:33:45 i would be very curious to see how my algorithm was changed/improved :) 13:33:54 oh, lame, we don't get to keep them forever? 13:34:04 unfortunately not. 13:34:43 (next?) 13:34:50 hi! 13:35:15 I've been reviewing, merging, debugging, etc. Organizing review meetings, dealing w my kid being off school for the week, etc. 13:35:39 I havent seen too much from the MIT students yet; I meet with them again tomorrow . Hopefull they've made nonzero progress. 13:36:28 I'm trying to get some improved PT API ideas done with Ben Schwartz before the valencia PT meeting. 13:36:46 and I am really hoping to have long enough to concentrate in order to fix the ed25519 voting bug. 13:37:17 also hm, I set this up in a moment of weakness: https://people.torproject.org/~nickm/tor-auto/ 13:37:22 It might be useful to somebody else too. 13:37:42 I think that's it for me? 13:38:00 hi! 13:38:06 i've implemented a poc for phase 0 of https://gitweb.torproject.org/user/linus/torspec.git/tree/proposals/ideas/xxx-tor-consensus-transparency.txt?h=tct 13:38:25 i have a patch making tor, when acting in the capacity of a directory authority, submit consensus documents to a (one, hardcoded) log over http over plain internets 13:38:58 the other end of that rope is a public append-only externaly verifiable called "gaol" 13:39:15 next steps will be to 1. make the log less unhappy about such large unchunked HTTP POSTs 13:39:35 2. make tor submit over tor instead, hopefully to hs (gaol does have a functional .onion) 13:40:06 3. put together some poc tools for auditing and monitoring the log 13:40:24 4. get the proposal to a state more ready for review 13:40:34 that's it. thanks for reading. 13:40:53 whoa, maybe that proposal should get a number? 13:41:00 that'd be grand 13:41:27 i'll "finish" it and send it to tor-dev. unless there's another process you prefer? 13:41:34 hi 13:41:52 ln5: works for me 13:42:01 hi Yawning ! 13:42:08 jetlag sucks 13:42:26 my cat has more furniture than I do because ikea sucks 13:42:37 but I have internet so hacking will resume 13:42:57 I am also trying to figure out how corporate law works in this state 13:43:24 my plans re hacking are getting that multithreaded branch polished 13:43:31 unless there's more pressing things 13:43:41 nickm: should I look at the pt api ideas? 13:43:56 because last I checked, stuff other people wants does not reflect what I think tor wants 13:44:20 (also lol 5am) 13:44:35 :) 13:45:06 Yawning: yes, and nothing's totally specified yet, so maybe instead we should chat for a little while about how I think stuff is going 13:45:33 possibly later today? 13:45:42 uh 13:45:47 or another day? 13:45:51 depends on how my sleep schedule works 13:45:54 today shoud be fine 13:46:11 ok. then let's try to intersect, but don't worry if we don't intersect till tomorrow or something. 13:46:47 is what I think I should be hacking on 13:46:52 somethign that reflects reality 13:46:57 or should I poke at something else 13:47:01 as for more pressing/less pressing -- getting stuff mergeable is helpful, but I hope everybody will have time to chip in a little code review too before EOM. 13:47:11 yeah ok 13:47:13 * nickm looks at the tracker 13:47:47 cc a bunch of stuff to me? 13:47:52 ug wth is trac down for anybody else? 13:47:56 I've been meaning to look over your rng branch 13:48:10 that one could probably wait for 029 13:48:15 yes 13:48:20 trac is kapput 13:48:24 hi meeting! sorry overslept :( 13:48:45 also for everybody: once trac is happy again, check for stuff that's owner==you, milestone==028, and correct that if it's wrong? 13:50:56 hi dgoulet ! 13:51:09 okay, who's next? 13:52:18 dgoulet: this could be you? 13:52:58 trac seems back 13:53:21 sure 13:53:38 give me 1 min, I'll change computer 13:53:45 ln5: not for me yet 13:54:00 ^ 13:54:30 nickm: i use this thing called tor. it takes me there. from switzerland, apparently. 13:54:38 (anybody else around who has an update?) 13:54:43 here 13:54:56 hi 13:55:28 i can give an update 13:55:52 i worked more on revising my #7144 branch as per nickm's review 13:56:32 but i have this strange problem that the tests don't pass, when i know that they passed when i originally pushed the branch… 13:56:41 so i need to dig into why a bit more 13:57:29 i suspect that i did something wrong somewhere in the several rebasings that i did, but i still have all the intermediate branches so it should be fine 13:58:17 ok. Some of the newer tests can be a bit fragile, so I wouldn't be too surprised. If you post what's failing here, it might be something others have run into ... especially if you do it when teor1 is around 13:58:55 it is my own tests for test_loose.c which are failing 13:59:03 ah, weird. 13:59:34 just for expectations management: there are 5 regular working days left in the month, plus one travel day, plus a weekend and a few dev meeting days. 13:59:56 Those are the days I have left allocated to fix bugs and get stuff reviewed and merged before the freeze begins. 14:00:14 So probably a lot of great branches are going into early 0.2.9 rather than later 0.2.8. 14:00:15 an assert is being hit that the underlying or_circ->p_chan exists… which i know i hit this problem when making the tests months ago, and then i solved it somehow, so basically i need to dig into why i'm having the same issue again 14:00:24 ok 14:00:35 so #16943 took most of my time with two large review from teor, I also did some ticket triage for hs. I've reviewed also I think last week RSOS ticket and need to ack the fixes today. I did some roadmapping for prop#224 as well as some new tickets about it. Battled with malicious HSDirs. Finally, I'll need some help with #15618 from someone that could have a better idea of that code 14:00:37 -- 14:01:34 okay, i'll aim to have #7144 revised before the dev meeting, and if not, or we don't have time to get to it, 0.2.9 is okay too 14:02:59 ok, who's next 14:03:03 ? 14:03:06 isabela: ? 14:03:11 Hi there, I have been working on preping things for valencia 14:03:30 for networks team I did some documentation of 028 to help with the releases discussion 14:03:51 https://storm.torproject.org/shared/x2v1GLIbNB2MfFSU2Enyy5Mf6OvRlOEOaX7xDUmzhcN 14:04:13 and I was wondering if we can cross over the things we did from the roadmap we created in berlin 14:04:45 do we have a locally cached copy of those pages? Right now Trac seems to be down for me 14:04:47 i am wondering about 029 triage in valencia 14:04:56 if we should wait for 028 freeze before doing anythign 14:05:28 I have the roadmap and the release stuff 14:06:31 ok. so probably the best way to do the roadmap IMO would be to throw it in a spreadsheet and mark everything with "when was it done" and "is it done" maybe with a few notes? 14:06:38 or is there a better way? 14:07:23 I thought of crossing over what we did, and mark what was left with 2016 high, 2016 medium and 2016 low (as per priority) 14:07:42 and that would be for us to evaluate if we will take it to our planning in valencia or not 14:08:35 and now trac's up for me again 14:08:43 great 14:09:05 that sounds look a good idea; can folks stick around to look at it after this meeting? 14:10:01 uh sure 14:10:25 ok. other discussion topics for today? 14:12:13 going once... 14:12:18 going twice... 14:12:33 #endmeeting