13:32:37 #startmeeting 13:32:37 Meeting started Wed Apr 8 13:32:37 2015 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:32:37 Useful Commands: #action #agreed #help #info #idea #link #topic. 13:32:45 hi all, it's another lovely Tor dev meeting. 13:32:52 who's here for the meeting this morning? 13:32:53 Bonjour Meeting! 13:33:00 hi meeting 13:33:03 Bonjour David! 13:33:06 hi athena ! 13:33:13 athena: hello! 13:33:17 I wonder if isabela is around or if she's still sick 13:34:11 yawning doesn't seem to be online 13:34:25 well, if it's just us 3 it's just us 3. Sebastian checked in before the meeting. 13:34:30 let's start with status updates? 13:35:06 anybody? :) 13:35:07 much security update, sorry for yesterday but now I'm better so working on all ticket that you reviewed nickm 13:35:18 great 13:35:34 sponsorR meeting next week in DC so I'll be much more R than little-t tor in the coming days 13:35:39 I lost a lot of last week to security stuff, to running around trying to finish end-of-month things, and etc 13:35:51 topic to discuss during discussion time: how can we help that meeting go great for you? 13:36:06 (and is the sponsorR meeting this week on Friday?) 13:36:22 nickm: Monday to Friday next week, I'm on a plane Sunday for that 13:36:35 err, I mean, the IRC meeting 13:36:40 that we didn't have yesterday 13:36:42 ah yes on Friday 13:36:47 what time? 13:36:55 12h00 #tor-dev - SponsorR meeting 13:37:04 nickm: your time ^ ;) 13:37:06 thanks! 13:37:51 athena: how's it been with you? what are you up to now? 13:39:01 long code review is long but progressing; that homotopy type theory workshop last week and trying to look for a new apartment sucked up rather more of my time than i thought it would though :/ 13:40:40 As for me, I've been scrambling to do lots of tiny little fiddly things for a while, and am just crawling out and looking for bigger things to do 13:41:00 Glad I got my sponsorS stuff done; got to look at april 15 stuff though 13:41:15 hmmm. 13:41:44 so, since there are just 3 of us, let's figure out discussion topics. 13:41:54 I think the biggest one is, "what should we all be working on for the next week" 13:42:05 probably we should help the sponsorR crew with whatever they need, yeah? 13:42:18 I hear from Roger that blowing the sponsor's socks off with our brilliance might be a fine idea. 13:42:29 ah yes always :) 13:42:44 dgoulet: my current SponsorR item is reviewing a big pile of SMC codebases for Aaron J. 13:42:49 any thing else that I could do? 13:42:51 or athena? 13:43:04 (this week or beyond?) 13:43:05 I can quickly explain a bit the state of this sponsor 13:43:30 ok, but will this terminate in "how you can help this week is by doing______" ? ;) 13:44:00 #tor 13:44:02 basically April meeting is "routine" and by that I mean we are going to collaborate and stir the ship there but the real coup must be in July so by next week, not much urgent needs to be done 13:44:14 wrong chat, apologies 13:44:21 but I would say going towards a decision on SMC is an important one 13:44:22 np 13:45:07 ok 13:45:20 nickm: my 14847 that you reviewed is also a cool one to have upstream soon thus I'm working on it now because it affects a big part of R that is hs health 13:45:34 nickm: but that's clearly underway after the review of yesterday so ++ 13:45:42 the more you can say about the testing strategy of the ticket, the better. 13:46:36 s/of/on/ 13:46:43 sure good 13:46:56 nickm: also I have a topic of dicussion maybe but that's not really "what to do next week" 13:47:02 okay 13:47:05 go for it 13:48:00 on HS subject, we ended up trying to figure out DoS (and still are with new warning we find on some HS), very difficult to reproduce and we actually I think really need a better way to network test stuff 13:48:27 dgoulet: sounds good; what features do you want on it? 13:48:36 that is have a self contained network somewhere that is not only chutney running on loopback but with actual latency and bw limitation and stuff 13:48:40 We've got a sponsor S item to improve network testing any way we can. 13:49:10 right 13:49:11 dgoulet: are you thinking of https://github.com/nsec/the-internet , or shadow, or what? 13:49:37 nickm: I'm actually thinking about deploying something like the-internet yes, that's a possible solution 13:49:43 but not shadow in this case 13:49:56 chutney2 + the-internet? :) 13:50:11 nickm: so I was wondering what's the status of your chutney2 about maybe addressing this? 13:50:14 ah yes voil :) 13:50:25 I've been doing the spec/hack/spec cycle 13:50:46 I wrote a decent spec, then did a thought experiment to see whether I thought it was what I wanted, and decided that it was actually terrible 13:50:51 so I'm putting another together 13:51:19 as for addressing this, what's involved in setting this up? Does one need root? Is it a set of commands to run? etc 13:51:58 the main goal of chutney2 is flexibility without code-bloat 13:52:03 nickm: yeah full root because "the-internet" requires tc for rate limiting and lxc 13:52:12 so this seems in-scope for that 13:52:43 ok cool, glad to see there is a sponsor for that :) 13:52:44 How would somebody usually set it up? Set up the-internet as root, and then control the various server processes from userspace? 13:52:59 we can make some of this work joint R/S. 13:53:09 R/S/U even 13:53:10 nickm: it's all lxc base , there is a map of BGP router and links between them 13:53:25 since U cares about resisting DoS, R cares about HS, and S cares about testing. :) 13:53:26 nickm: right I'm not assigned on S/U but 100% we can put some in R 13:53:51 (the-internet has to be one the worst names for a program ever) 13:54:08 lol 13:54:38 nickm: fun fact, since we made it public, akamai is using it in some teams for internet simulation for their stuff :) 13:55:42 athena, dgoulet: on the SMC angle: do you know by any chance whether anybody who is working on practical SMC happens to be in my area? Harvard, MIT, BU, Yale... 13:55:43 nickm: ok so once you have chutney2 spec ready for a public mailing list, I'll read it :) 13:55:58 nickm: no idea on that :S ... Aaron might be a good person to ask that 13:56:25 ok 13:56:48 athena: so, let's talk about your stuff. Right now you're focusing on code review, is that right? 13:57:03 Is there anything that the rest of us can do to help there, or to make sure you're not getting stuck too hard? 13:57:14 Would it help to take a break and have some software tasks to do as well? 13:57:25 (I know that I can't do review for more than an hour or two without wanting to bite somebody) 13:58:31 nickm: yeah, if you have some nice little ticketsto toss at me that might help lubricate things 14:00:06 athena: ok, here's what algorithm I'd use I'd go over to https://docs.google.com/spreadsheets/d/1H_kAcQQ-NpGpKWG9RlF3w-NhruKKE9EZJcRaXBQGVy8/edit#gid=499528971 , look at the stuff in the first list that isn't triaged out, with a sponsor (ideally R or U or S, roughly in proportion to your funding), and for little tickets, look for the ones that have size == small. 14:00:47 does that soudn like a plausible start, or would you like me to try to pick some? 14:02:35 sure, that's fine 14:03:00 ok. Want to grab a couple some time today and assign them to yourself? 14:03:05 yeah 14:03:18 Great; let me know if you've got any questions or if it gets overwhelming 14:04:21 I'll be online, on and off, for the next 6 hours. 14:04:28 anything else for us to talk about today? 14:04:32 anything else I should be working on? 14:05:01 nothing crazy on my side for you :) 14:05:44 sounds like an #endmeeting? 14:05:48 * nickm waits another couple of minutes 14:06:05 * nickm sends an email to isabela saying "hey would you like somebody to input the triage outcomes into trac"? 14:09:16 #endmeeting