13:31:07 <nickm> #startmeeting 13:31:07 <MeetBot> Meeting started Wed Jan 27 13:31:07 2016 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:31:07 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic. 13:31:11 <nickm> I think it's dev meeting time 13:31:54 <nickm> Who's here? 13:31:55 <Yawning> *wakes up briefly* 13:32:14 <nickm> hi Yawning 13:32:39 <mikeperry> hello. by some miracle I am still in this timezone, barely 13:33:20 <nickm> hi all! 13:33:34 <nickm> athena: asn_: dgoulet: isis: etc: etc: etc: hi! 13:33:37 <nickm> you awake/around? 13:33:55 <isis> hey hey 13:34:43 <nickm> hi! 13:34:44 <isis> fwiw, i can do the meeting later, if it's better for people in other timezones 13:34:48 <nickm> I'll start w my status 13:34:59 <isis> but that can be a topic for the mailing list 13:35:27 <nickm> I'm spending this week braindumping my institutional knowledge on shari, and I'm getting patches reviewed slowly but surely, and I'm happy w how proposal review is going 13:35:32 <nickm> my jan deliverables are soooo behind. 13:35:39 <nickm> isabela: ^ 13:35:52 <nickm> not thrilled about all of that, but I've been working hard all month and I know I've been working on important stuff 13:37:13 <nickm> I hope that between now and next week (which will be feb) I get som serious stuff done, I get isabela's questions answered, and I get reviews more done. 13:37:18 <nickm> who's next? 13:37:22 <Yawning> hi 13:37:31 <Yawning> pushed a new obfs4proxy release 13:37:38 <Yawning> contributed to discussion on a bunch of bugs 13:37:43 <Yawning> continued to work on obfs5 13:37:56 <Yawning> busy turning my life upside down again by moving 13:38:17 <Yawning> ps: airliners charge stupid amounts to fly with a cat in the hold 13:39:05 <nickm> ow 13:39:26 <isis> Yawning: does the obfs4proxy release fix the problem with the elligator2 encoding of the first step of the handshake? 13:39:32 <Yawning> no 13:39:40 <Yawning> can't without breaking compatibility 13:39:43 <Yawning> obfs5 will fix it 13:39:46 <isis> because it wouldn't be backwards-compat 13:39:52 <isis> ah right 13:39:53 <Yawning> yah 13:39:59 <isis> ugh lame 13:40:05 <Yawning> was more fixing a dumb bug when I added the meek client 13:40:07 <Yawning> and tagging 13:40:18 <Yawning> obfs5 is getting there 13:40:26 <isis> \o/ 13:40:39 <isis> let me know if you want help with that, it sounds fun 13:40:42 <Yawning> assuming people are ok with the spec being nothing but "I am Groot", I'll have something that runs shortly 13:41:05 <nickm> oh fyi i talked with ben from google (the uproxy guy we met at that meeting in DC) last week. We've got some ideas about possible ways to make PTs more useful for otherss without uprooting the usefuls stuff that works today 13:41:08 <nickm> *useful 13:41:20 <nickm> I'm supposed to write up all our notes and drafts by Jan 25 13:41:26 <Yawning> unless someone tells me otherwise 13:41:44 <Yawning> my view on what other people want will remain "not my circus, not my monkeys" 13:41:45 <nickm> I'd like the spec to be real at some point, but it's fine to do code and then finish the spec to match 13:41:56 <nickm> if that's not a huge pile of crud 13:42:24 <Yawning> yeah the only annoying parts to spec are probably going to be the padding stuff 13:42:36 <Yawning> the rest is straight forward 13:42:45 <Yawning> since I invented less crypto this time around 13:44:06 <athena> hello meeting! 13:45:40 <nickm> hi athena ! 13:45:59 <Yawning> let me know if other things should take priority 13:46:13 <Yawning> but next month will be all sorts of nuts due to moving etc 13:46:41 <nickm> oh btw: Currently planning the soft feature freeze for 028 on 1 Mar, and the hard feature freeze on 15 Mar. 13:46:51 <Yawning> ooof 13:46:54 <Yawning> ok 13:47:01 <nickm> a six month schedule means that we're always freezing too fast 13:47:07 <nickm> but that it's never too long to the next release 13:47:09 <nickm> imo 13:48:12 <nickm> Yawning: any more? 13:48:13 <nickm> anybody else? 13:49:07 <mikeperry> I have been thinking about prop#247 stuff. I am going to write one more mail to tor-dev on it, and then switch back to #16861 and children 13:50:08 <athena> i got dragged into thinking about #18116 and concluded it was almost certainly introduced by that periodic tasks thing 13:52:18 <nickm> ok. how do we fix it? I don't want to revert periodic tasks 13:53:52 <athena> oh, we won't have to - still working on figuring that out; need to look into exactly what order things happen a shutdown time 13:54:43 <nickm> ok cool 13:54:54 <nickm> grab me any time if you want a rubber duckie to brainstorm with 13:55:02 <nickm> (everybody knows about rubber duckies and brainstorming, right?) 13:55:15 <Yawning> I have my cat for that 13:55:24 <athena> no? 13:55:30 <Yawning> It's a rubber duckie that goes *CHOMP* 13:57:04 <asn_> hello there. this week i did prop247. also attended two proposal reading groups. 13:57:10 <asn_> also launched tor-onions the mailing list 13:57:11 <nickm> in brief: talking to an inanimate object to help engage different parts of your brain in solving a problem. Explain the problem to a rubber duck. sometimes it helps 13:57:30 <Yawning> (I tried using plush toys but the cat would steal them) 13:57:31 <asn_> and also accepted three bugs from guido via hackerone 13:57:42 <athena> right, okay 13:58:39 <nickm> Make sure to thank guido btw :) 14:00:47 <asn_> ack 14:00:54 <asn_> sending you the file you asked now 14:00:57 <nickm> thanks 14:00:58 <asn_> you can get it if you log in hjackerone 14:01:03 <asn_> but i will send it to you in any case 14:03:07 <nickm> anybody else with a status? 14:03:11 <isis> this week i spent some time working on the guardsim stuff, and the algorithm is working even better now (under the flaky/censored network conditions) 14:03:16 <nickm> woo 14:03:21 <nickm> are we starting to get numbers ? 14:03:37 <nickm> so very glad we are making progress there. 14:03:46 <isis> (this is for #17261 and prop#241 and prop#259) 14:03:55 <nickm> ack 14:04:21 <isis> yeah the numbers are now roughly 1% circuit success rate for prop#241 for a flaky network 14:04:32 <isis> versus 80% success for prop#259 14:04:49 <isis> and about the same for fascist firewall networks 14:05:24 <isis> i am committing code right now because i am a bad person who stays up too late hacking and then falls asleep before committing stuff :/ 14:05:46 <nickm> ok. i'll be curious to find out what it does for the attack cases. 14:06:09 <kernelcorn> isis: in case you are sleep deprived, remember to run "git push" :) 14:06:36 <Yawning> --force makes all the problems go away and lets you get to bed faster 14:06:51 <isis> nickm: the logic is a bit spaghetti-ish… i am not quite sure how to abstract it better 14:07:02 <isis> but it works :) 14:07:39 <nickm> looking forward to checking it out 14:07:45 <asn_> isis: btw, i suggested to the thoughtworks poeple that the entry guard simulation task might be a nice project for them. 14:08:13 <asn_> isis: i see you are activelyy working on it. 14:08:29 <asn_> isis: do you think my suggestion was a bad idea? 14:11:14 <asn_> (sent an email about this: https://lists.torproject.org/pipermail/tor-dev/2016-January/010324.html) 14:11:48 <asn_> isis: i suggested that btw because during the meeting last week, the simulation project did not get any volunteers. so i thought it would be appropriate for ola bini and team. 14:12:06 <mikeperry> asn_: maybe they could take the txtorcon prop#247 implementation instead? it will require some tor patches, txtorcon patches, and a couple different implementation + testing scripts 14:12:25 <mikeperry> and I may have enough other things to do 14:12:33 <asn_> maybe yes 14:12:56 <asn_> but i have already suggested the other project... 14:13:06 <asn_> not sure if we should contet switch them now, although it's pretty early since they start next week. 14:13:11 <asn_> let's see what isis says about the prop259 project 14:14:02 <isis> asn_: no, that sounds great if they want to take it 14:14:09 <asn_> fantastic 14:14:18 <isis> it's almost done, i think, but could use more eyes/ideas/testing 14:14:38 <asn_> isis: would you like to send an email with the current state of the project, and what needs to happen next? 14:14:51 <isis> tbh, i was a bit stressed about it, since i also have two weeks of travel starting saturday, have my OTF work on bridgedb to do, and have #7144 to revise 14:15:00 <isis> asn_: sure, will do 14:15:01 <asn_> great 14:15:04 <asn_> thx 14:15:10 <asn_> maybe answer to https://lists.torproject.org/pipermail/tor-dev/2016-January/010324.html 14:15:13 <Yawning> (is everyone traveling or something?) 14:15:24 <nickm> i'm at home till i leave for the dev meeting 14:15:25 <asn_> isis: you going fosdem? 14:16:36 <nickm> any more statuseses? 14:18:02 <isis> asn_: nope, i am going to nijmegen to visit with peter schwabe and his team 14:18:18 <isis> asn_: i am never going to fosdem ever again full stop. 14:19:34 <nickm> when you see peter, ask him about rumors of an even faster newhope? 14:19:45 <nickm> also say hi and have fun 14:19:56 <athena> isis: what have you got against fosdem? 14:20:19 <Yawning> "it's filled with nerds" 14:20:39 <Yawning> my apartment is filled with nerds too :( 14:20:50 <isis> athena: in 2012, jake dumped his fosdem talk on me with two days notice, no slides, no advice, and "forgot" to mention that it was in the main hall with ~3000 people 14:21:42 <Yawning> isis: I have a recurring nightmare that starts outsimilarly to that 14:21:46 <isis> also i asked for no photography because i was about to fly to china for the first time, and some asshole reporter put a picture of my face in linux magazine 14:22:01 <isis> fuck fosdem 14:22:16 <Yawning> then the audience sprouts orfices, tentacles, and pseudopods from locations commononly thought to be unnatural, and tries to kill me 14:22:19 <Yawning> as I give my talk 14:23:46 <nickm> isis: let me know when i should merge your guardsim 14:23:49 <isis> also, at the fosdem speakers' dinner, "saint peter" (bro dude who invented XMPP) was like "oh you're isis from the tor project, do you know who i am?" and i said nope and then basically this ended in this dude screaming at me over the table "WE FUCKING WON!! YOU HEAR ME? WE, XMPP, WE FUCKING WON!! YOU!! YOU PRIVACY PEOPLE, YOU FUCKING LOST!!" and banging on the table 14:24:06 <isis> lol 14:24:34 <nickm> note to self: never declare victory. 14:24:42 <nickm> ok, so, discussion topics? 14:24:53 <Yawning> the way you describe it, maybe tentacles are preferable to going to conferences 14:24:57 <nickm> maybe it's time to nominate another set of proposals for discussion? 14:25:33 <isis> the versioning stuff probably needs more review 14:25:35 <dgoulet> isis: omg! 14:25:46 <athena> isis: ouch :( 14:25:47 <nickm> dgoulet: ho hai 14:25:50 <nickm> *oh hai 14:25:51 <mikeperry> I am confused by asn_'s latest simulation results for prop#247, butmaybe that is a post-meting discussion 14:26:05 <nickm> dgoulet: got a status? :) 14:26:24 <athena> Yawning: audiences are already monstrous egregores occupying thousands of distinct bodies simultaneously. get out while you still can. 14:26:33 <asn_> mikeperry: the latest results should be more correct than the old ones. 14:26:40 <asn_> mikeperry: i fixed some bugs. 14:26:45 <dgoulet> nickm: yes! sorry for being late 14:26:48 <mikeperry> asn_: since yesterday? 14:27:14 <nickm> still on the queue are the proposals previously scheduled for friday: prop#252, prop#260, and prop#246 14:27:20 <mikeperry> asn_: cause I don't understand how uniform coud be so close to the MinXX dist. according to the day-resolution calculation, it should be substantially different for the CDF 14:27:25 <nickm> I hear a nomination from isis for prop#264 ? 14:27:48 <nickm> any more nominations ? No more than one each please :) 14:27:54 <nickm> we'll get through everything sooner or later. 14:27:55 <dgoulet> status: lots of proposal reading last week which made me site down quietly and think about them a lot. Few little-t tor things this week, I'm currently working on fixing prop#250 code after discussion yesterday 14:27:56 <asn_> mikeperry: ye, i was also surprised. 14:28:07 <asn_> mikeperry: the minxx distr seems very close to what you had calculated in the prop. 14:28:17 <asn_> mikeperry: the CDF i mean. 14:28:24 <asn_> mikeperry: there might be some bugs still.... 14:28:45 <dgoulet> also announcement: I'm leaving this country tomorrow morning for Fosdem (sorry isis) then back to my homeland on feb 2nd so expect me being a bit afk 14:29:09 <dgoulet> and I did some progress on tor-genkey also. 14:29:10 <dgoulet> done 14:29:11 <isabela> hi 14:29:25 <asn_> nickm: there is that prop224. but im not sure how such a reading group would go. 14:29:55 <isis> nickm: maybe prop#257 in a different meeting to #prop264 14:30:05 <nickm> ack 14:30:14 <nickm> noted + noted . 14:30:26 <nickm> anybody else? 14:30:30 <isis> dgoulet: oh no worries, have fun! (i'm not opposed to others going at all!) 14:30:58 <isis> uh… let me glance at proposals/ quickly 14:31:27 <asn_> mikeperry: this is the code https://gitweb.torproject.org/user/asn/hax.git/tree/simulation.py?h=prop247_simulation_draft#n243 14:31:40 <asn_> mikeperry: that's like one run of the simulation 14:32:07 <asn_> (btw in my simulation i rotate the third-layer guard when the secnd-layer guard rotates) 14:32:17 <asn_> (so when the second layer guard rotates, it takes the third layer guard with it) 14:32:36 <mikeperry> oh hrm. I believe I made them fully independent 14:32:57 <isis> nickm: we could do a new crypto review and look at prop#261 prop#262 and prop#263 14:32:59 <asn_> (just_rotated_second_guard on the code) 14:33:32 <nickm> let's skip prop#261 for not; I expect something better soon. 14:33:34 <nickm> for now 14:33:49 <Yawning> is 263 the ntru thing? 14:33:55 <isis> nickm: oh… neat! what what what is it 14:33:56 <Yawning> I have corrections for it 14:34:01 <Yawning> but was waiting for newerhope or whatever 14:34:10 <isis> lol "newerhope" 14:34:16 <nickm> I expect dan to get me more hhfhfh ideas rsn 14:34:17 <isis> i hope it's named that 14:34:21 <mikeperry> asn_: that will certainly skew it more towards the uniform direction and away from what I calculated.. 14:34:45 <isis> ah right, i forgot about the hhfhffhhfnnfhfnf thing 14:34:58 <nickm> "hufflepuff" 14:35:03 <isis> :) 14:35:06 <Yawning> I no longer feel bad about naming things obfs<blah> 14:35:48 <isis> yeah, i think djb wins the obfsucated pronunciation contest for now 14:35:53 <nickm> ok, that will carry us for a week or two. 14:36:02 <nickm> I'll pick times, send mail, let people complain, move stuff. 14:36:13 <isis> cool, thanks, nickm 14:36:21 <nickm> teor, dgoulet: I could use info on what times you are free monday that intesect with my times; see tor-dev email. :) 14:36:31 <nickm> any more topics for the meeting? 14:36:36 <nickm> I'll be online after. 14:37:27 <nickm> #endmeeting