17:00:48 #startmeeting 17:00:48 Meeting started Mon Dec 5 17:00:48 2016 UTC. The chair is nickm. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:48 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:00:56 hello network team people! 17:01:34 it's good to see folks around 17:01:36 status time! 17:01:47 hi 17:01:52 hello 17:01:57 * nickm spent last week finishing the prop#271 code. Then releasing more tors. 17:02:21 In my upcoming week, I'll try to do more review & merging; try to fix all the issues that will show up in my code during review... 17:02:37 ...try to advance mikeperry's netflow branch towards a merge... 17:02:51 ...try to release 0.2.8.11 if appropriate (and I think it is) 17:03:05 ...and try to figure out what I'm working on next. 17:03:09 I'm keen to fuzz some. 17:03:13 that's all for me! 17:04:43 can go next 17:04:46 here we go: 17:04:47 Hello this has been a busy week that involves lots of travelling. During the week I met with Aaron Johnson, 17:04:50 David, Roger and Chelsea and worked on various things Tor. During this week, I 17:04:53 also merged the client auth torspec branch that was hanging around, and also 17:04:55 updated the proposal-status.txt metatext in torspec: 17:04:58 https://lists.torproject.org/pipermail/tor-dev/2016-December/011709.html 17:05:00 I also sent an email about restarting the proposal reading groups that was well 17:05:03 received, but I have had no time to schedule anything: 17:05:05 https://lists.torproject.org/pipermail/tor-dev/2016-November/011703.html 17:05:08 I also wrote the Ahmia blog post for the fundraising campaign. 17:05:10 I have tons of *big* things to do like revise the ESTABLISH_INTRO branch, 17:05:13 revise HSDir branch, and review the guard stuff. I'm currently prioritizing the 17:05:15 latter, and I'm not really sure when I'll get to the rest. A bit stressed. I'm 17:05:18 also going for vacations on Thursday and will be back Monday. I will be 17:05:20 internet connected during vacations but not super active. That's that. 17:05:23 [3$ 17:05:45 The review is going OK so far, but there are _tons_ of stuff and behaviors to test. I'm not sure how much I will manage to do before mid-December, if I also want to fit the prop224 for the 0.3.0 deadline. 17:05:55 that's that! 17:06:09 hello world. last week i spent keeping some funders happy. also the month of december, as asn says, is going to be exciting with the blog post treadmill. 17:06:33 i actually wrote half of a proposal on the train though -- "authenticating sendme cells to mitigate bandwidth attacks" 17:06:41 maybe i'll write the other half on my airplane tomorrow 17:06:42 . 17:06:57 armadev: neat! 17:07:01 * dgoulet can go 17:07:14 Last week for me was mostly Sponsor R quaterly meeting. Spent one full day with asn on working out prop224 open questions and issues. I will be following up with those on tor-dev@ this afternoon. 17:07:17 I just finished #15056 review. I'll be working this week on #20657 and finishing unit tests (finally) on #20029. Will also take some time on review group 13. 17:07:26 -- 17:07:50 * isabela can go 17:07:58 when you say "the review is going okay so far" do you mean prop#271 ? 17:08:02 yes nickm 17:08:24 hi there, i got some load of work last week that had to be prioritize since it was time sensitive stuff (fundraising and drl propposal) 17:08:32 but i did worked on organizing our wiki page - i did it offline and will be updating the week today with it - will email the list and we can revert it if ppl hate or improve it if ppl kinda like but it could be done better :) 17:08:43 I'm closing the loop on sponsor u report too 17:08:49 and I am leaving thursday for short vacation and will be back on tuesday / will miss next monday meeting 17:08:52 done 17:11:02 dgoulet: sounds like I'm revising #15056 soon. Am I in for a lot of work there? 17:11:26 nickm: nope 17:11:44 nickm: no show stopper so it shouldn't be much work to revision 17:12:18 asn: any way I can help with the prop#271 thing? I can put in a bunch more testing and r&d time, I think. 17:12:59 I am here. I saw your mail nickm. I need to page all that stuff back in again. maybe later this week. happy to discuss things whenever, though 17:13:13 (re netflow padding) 17:13:26 mikeperry: cool! What I most want there is for Roger to tell me "I read it and understood it and mike explained it to me. Seems ok." 17:13:33 I'm working through the other stuff as well. 17:13:40 sorry for taking so long with this one 17:14:37 nickm: hmm, more testing would definitely help. With Chelsea, we made a list of scenarios that should be tested, and there is not enough time to do all of them this week. 17:15:00 nickm: you can also help by answering a few questions on IRC. I have 2-3 of them. 17:15:01 a list would help! 17:15:10 asn: glad to! 17:15:42 nickm: ok perhaps after the meeting! 17:15:43 asn: nickm: maybe we can put that QA list somewhere and track who has tested what/with which results? 17:15:54 chelseakomlo: that could be smart yes. 17:15:57 let me make a pad 17:15:57 chelseakomlo: sounds good! Maybe a wiki page, linked from the ticket? 17:16:12 hm a wiki page 17:16:13 ok 17:18:46 mikeperry: I don't know if you have time, but if you _do_ have time to look at #15056, you'd have some good feedback there too 17:18:49 any more updates? 17:19:01 short update 17:19:39 go for it! 17:20:13 i will help with guard QA, and continue reviewing code as well 17:20:42 nickm: i sent round 2 of object diagrams, and have some more arch diagrams from conversations with dgoulet and asn that i will create 17:20:47 that's all for me :) 17:20:53 cool. I'll try to look at those rsn! 17:21:07 i put notes of testing scenarios here: https://trac.torproject.org/projects/tor/wiki/doc/NewGuardAlgorithmTesting 17:21:20 I'm still going through the code so I have not had time to actually do testing. 17:21:35 ok. 17:21:56 one other thing that would help is to be more explicit about how stuff is tested 17:22:24 you mean isntead of "Test guard lifetime" I should specify the entire testing procedure? 17:22:37 yes indeed that would be helpful. i have not had time to think about that either. 17:22:47 yep agreed 17:23:04 not necessarily the whole procedure, but like ... "Start with clock a year ago. Stop. Restart with new clock. Verify that new guards get used" 17:23:09 (actually that wouldn't work) 17:23:20 dgoulet: btw, I will be busy with the guard branch for the next days. unclear when I'll get to the ESTABLISH_INTRO branch. probs late next week. 17:23:23 (but just the basic idea of "how do we do that") 17:23:49 any more updates, or are we in discussion? 17:24:23 let's say we're in discussion! 17:24:45 nickm: ok, I will try to give #15056 a look. I have a long list of tiny nagging things to ge through first, though 17:24:48 asn: feel free to break from guards if you need to go back to the ESTABLISH_INTRO branch. 17:24:51 mikeperry: sure, no worries. 17:25:25 nickm: both of those tasks are big and lined up for 0.3.0 so I'm not sure how to prioritize. 17:25:27 Here's a question for folks: would it help if I resusrrected my "tor-next" branch -- that is, a tor branch that has all of the big pending stuff merged on it already, for testing? 17:25:33 i will go with guards for now. 17:26:03 like, would folks run that as their tor / on the test network / etc? 17:26:24 (if not, there's not a huge amount of point.) 17:26:29 asn: you want me to take it over addressing nickm review? I don't mind else it's fine by next week 17:26:50 dgoulet: do you have any other tasks lined up for 0.3.0? 17:27:10 asn: many :) but I want that branch to be merged as #20029 would benefit from it as I'm based on it 17:27:37 right 17:28:08 your 0.3.0 prop224 task is #20029, right? or there is more? 17:28:35 that one then emails then service side #20657 17:28:49 and some tickets around prop224 that are pending 17:28:52 service side is not 0.3.0, right? 17:29:03 ah right no for 030 it's relay cells 17:29:06 ack 17:29:14 and some tickets with little code to make on them 17:29:18 BTW, I don't know for sure what I'm on for the rest of the month. If you'd like some SponsorR hacking from me, I can try to help there too maybe 17:29:24 unless something comes up, which it might. :) 17:29:32 in this case, if you finish #20029, and I'm still stuck in guard land, then maybe you can take ESTABLISH_INTRO as well. 17:29:46 but let's be in contact about that. 17:29:46 asn: ok I'll let you know if I mutex_lock() 17:29:50 ack 17:30:09 nickm: so far, review/merge is what we'll need for 030 :) 17:30:44 okay; I'll try to be fast 17:31:04 nickm: btw, I'm running #15056 on my public relay 17:31:07 any other discussion topics? Anyone want to talk about fuzzing, or should I wait for teor to be around? 17:31:12 dgoulet: neat 17:31:47 i have two questions wrt guard branch. not sure if it fits to discussion or post-meeting. 17:32:07 asn: let's do immediately post-meeting if we wrap up soon?: 17:32:11 ack 17:32:24 I'm good for this meeting 17:32:26 dgoulet: the stuff for testing I had in mind was #15056 + #19877 + #16861 . 17:32:30 going once! 17:32:38 going twice! 17:32:45 nickm: once merged, we can definitely at least deploy that in the test network 17:32:52 if there are no more discussion topics.... 17:33:07 ...then it will have been a short meeting.... 17:33:10 ...which is nice! 17:33:31 #endmeeting