17:00:25 #startmeeting anti-censorship weekly checkin 2019-09-26 17:00:25 Meeting started Thu Sep 26 17:00:25 2019 UTC. The chair is phw. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:25 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:00:28 hi everyone 17:00:35 hi! 17:00:36 hi 17:00:36 hello 17:00:37 hi 17:00:44 here's our meeting pad: https://pad.riseup.net/p/tor-censorship-2019-keep 17:00:55 hello 17:01:02 hihi, will be lurking again 17:01:09 cohosh: the first announcement is yours, right? 17:01:14 oh yep 17:01:25 (now with obligatory exclamation mark) 17:01:38 GeKo reviewed the patches for using the pion branch of snowflake and wants to put it in the upcoming alpha release 17:01:47 great work 17:01:50 thanks for adding that (!) 17:02:04 that's a big milestone, nice work 17:02:13 indeed! 17:02:18 thanks for the help with those scripts dcf1 17:02:59 we can close a bunch of tickets now 17:03:08 but i'll wait to for when we merge that branch to master 17:03:46 sadly I have been pretty much leeting tickets pile up in my inbox for the last week, and I will probably continue to ignore them for at least a few more days 17:04:05 "leeting" typo not intended, but appreciated 17:04:05 ok 17:04:14 lol 17:04:17 lol 17:04:20 we're not blocked on it right now 17:04:23 :) 17:05:53 17:05:55 let's talk about the contribution guidelines next (#31847) while we're still on the topic of snowflake? 17:06:01 ah sure 17:06:20 so sah (who might not be here) filed a linting related ticket 17:06:30 and made some patches 17:06:45 #31794 17:07:06 there were a few pieces of feedback that would apply to all patches and it would be nice to have them in our contributing guidelines 17:07:12 like how to structure commit messages, 17:07:21 keeping changes relevant to the ticket, 17:07:31 how to submit a PR in the first place, 17:07:32 etc 17:07:50 esp. since we are getting more people contributing to snowflake recently 17:08:12 i would like to add the linter sah used to the guidelines as well 17:09:30 thanks for shepherding this case, cohosh 17:09:37 bridgedb's hacking.md can provide some inspiration: https://github.com/NullHypothesis/bridgedb/blob/master/doc/HACKING.md 17:09:42 np 17:09:51 thanks phw, i'll take a look 17:09:53 these kinds of tickets are sometimes hard for me to deal with because they often provide little value relative to the amount of work they require 17:10:20 yeah i'm hoping some contributing guidelines can also cut down a bit on reviewer work 17:10:21 but sometimes good developers use a ticket like this one to get themselves going when they don't know where else to start 17:10:28 so sometimes the investment pays off 17:10:47 yep i think they are interested in tackling some code coverage issues next 17:10:55 which is nice because i've let that fall down on my priority queue 17:11:58 are we tagging snowflake tickets with 'easy' on trac? it would help with pointing volunteers at good tickets to start with 17:12:25 some are tagged that way, but i'm not consistent with remembering to do that 17:12:41 i might go through them and tag some as easy or starter this week 17:13:03 i have a pretty good mental map of all the snowflake tickets at this point 17:13:18 ok, nice 17:14:48 are we done with the volunteering part? 17:14:54 i think so 17:15:05 gaba, let's talk about sponsor28 and 30 next? 17:15:12 ok 17:15:12 I'm not sure about some of these reported errors, for example the "shadow" ones look like deliberate narrowing of variable scope, and probably should not be changed. 17:15:20 (I'll write on the ticket) 17:15:26 thanks dcf1 17:16:12 s30 officially started and there are a few objectives that need tickets. There is a master ticket for objective 2. We could later talk about how to go into creating the rest of tickets for it. 17:17:03 what is the timeline for s30? are we going to follow the original proposal timeframes? 17:17:26 and, will this meeting be the place where we will discuss progress on overlapping tasks? 17:17:27 yes. It started on September 20 by the contract. 17:17:40 I can share the timeline with you once nextcloud is back up 17:17:52 And share it with everybody in the meeting in the next couple of weeks 17:20:45 ok, thanks gaba 17:21:16 so for sponsor 28 phw is going to be liason, right? 17:21:28 we can talk later with phw on the rest of the activities 17:21:33 sorry. I meant sponsor 30 17:21:41 gaba: yes 17:23:01 next is an update on sponsor 28. We we re doing updates a few months ago and then we stopped. It would be good to give a general update here about it. 17:23:52 on my side, i worked almost exclusively on #30716. it's a lot of work and happens to be our first deliverable 17:24:27 basically, we want to improve obfs4's flow obfuscation and find a way to "regulate" per-packet entropy, so a high-entropy filter cannot easily block the protocol 17:24:40 i have a prototype for the former and nothing for the latter 17:25:06 phw: nice, is there help you'd like with that? 17:25:40 nice 17:25:49 cohosh: yes, plenty 17:26:38 in particular regarding the entropy part -- it also involves some research and potentially significant changes to the protocol 17:27:40 in parallel, we're following the work our georgetown colleagues are doing on evaluating obfs4 17:27:54 that is, throwing classifiers at it and seeing how well they can spot it 17:28:12 ideally, this should inform the flow obfuscator that we're working on 17:28:31 that's it from my side 17:28:48 for me, pretty much all of the snowflake work falls under sponsor 28 17:28:52 but it's a later deliverable 17:28:54 thanks 17:29:08 yes 17:29:44 i think snowflake is important to work on now since we're going to be getting more users and the quality of connections still aren't great 17:30:00 but if we're under time pressure, i can spend some time helping out with the obfs4 successor 17:30:35 i'm almost done with #29206 which should really help with a few improvements 17:31:11 yes, we need to take another look at our capacity planning 17:31:29 i think only 10% (or was it 20?) of my time are allocated for sponsor 28 17:31:35 this ticket ^^ has taken a lot longer than i expected 17:31:51 phw: ok, sounds good 17:32:04 at the roadmapping i pretty much just roadmapped snowflake tickets 17:32:43 but that left you with the obfs4 (sharknado?) stuff and the sponsor 30 work 17:33:36 right, and we haven't even started looking at obfs4's high-entropy issues 17:33:44 phw, cohosh: reality is that we have you two full time for s28 and 30. We can rearrange on how work is divided. 17:34:14 gaba: ok 17:34:43 right, gaba. i'd like to take a look at upcoming deliverables and figure out what we should prioritise in the near future. 17:35:18 i have the timeline for s30 but we had one for s28 too that we should look at. maybe later in our sync 17:35:35 yes, sounds good 17:37:01 the next item on our agenda is our september report 17:37:10 can y'all please add your monthly highlights to this pad? https://pad.riseup.net/p/B8Um_zirxIsHD84D6iR_ 17:38:07 ok, time to take a look at our 'needs help with' sections 17:38:45 * phw wonders if there's something new on hiro's gettor front 17:39:28 other than that, we have #30830, #31843, soon #29206, and #28942 for cohosh 17:39:37 #31391 for arlolra 17:39:50 #29206 and #28942 are for dcf1 when he has time 17:40:02 i can take #31391 17:40:45 phw: could you look at #30830 and #31843? they are small changes to snowflake 17:41:05 cohosh: yes, will do 17:42:00 thanks! 17:42:41 ok, we went through our agenda for today. anything else on anyone's mind? 17:44:02 let's wrap it up for today 17:44:05 #endmeeting