17:59:13 <phw> #startmeeting anti-censorship meeting
17:59:13 <MeetBot> Meeting started Thu Apr  9 17:59:13 2020 UTC.  The chair is phw. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:59:13 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:59:18 <phw> hello everybody
17:59:25 <phw> here's today's meeting pad: https://pad.riseup.net/p/tor-anti-censorship-keep
17:59:25 <cohosh> hi
17:59:56 <juggy> yo
17:59:59 <agix> hi
18:00:39 <phw> not much on the agenda for today but the world is waiting for our march 2020 report and i'd like to ask you to add your accomplishments to our report: https://pad.riseup.net/p/GDSSo-bqhi_dE1oJENyP
18:01:07 <thymbahutymba> Hi
18:01:58 <phw> any spontaneous discussion items for today?
18:02:37 <phw> let's move on to our 'needs help with' sections
18:03:30 <phw> juggy has a question about moat
18:03:49 <juggy> Ya
18:03:54 <phw> ...and arlo's item seems to be from last week
18:03:57 <dcf1> juggy: this might be a good question for the Tor Browser team
18:04:21 <juggy> ah ok I'll drop a message in their IRC
18:04:21 <dcf1> GeKo, mcs, brade on #tor-dev are names I would think of first.
18:04:52 <dcf1> I'm not sure what the current state of Tor Launcher is wrt XUL, but it's a good question.
18:05:01 <cohosh> sysrqb too
18:05:03 <gaba> hi
18:05:07 <cohosh> hi gaba!
18:05:25 * catalyst is here
18:06:56 <phw> does anyone else have something to talk about?
18:07:08 <dcf1> juggy: you can check what versions of things are currently used and where their upstream repositories are, from the tor-browser-build repository
18:07:20 <dcf1> https://gitweb.torproject.org/builders/tor-browser-build.git/tree/projects/tor-launcher/config is the file for current tor-launcher
18:07:49 <juggy> ok, thanks!
18:07:56 <dcf1> So I would guess that whatever is at https://gitweb.torproject.org/tor-launcher.git 0.2.21.5 is current
18:09:17 <dcf1> cohosh: thanks for your work on #25595/#25596/#33666
18:09:29 <cohosh> ah yeah
18:09:33 <cohosh> thanks for those scripts
18:09:42 <cohosh> i didn't realize how bad the situation was
18:09:53 <cohosh> because i'm getting an 85% success rate at home
18:10:24 <cohosh> but yeah depending on the type of NAT you have, it sometimes takes a very long time to be matched with a snowflake that has a compatable NAT
18:10:42 <cohosh> i have some ideas on solutions that I'll work on next week
18:10:46 <dcf1> I suspect amiableclarity was in a situation like that when they reported a long connection time
18:10:56 <cohosh> yep sounds like it
18:14:01 <cohosh> i'm glad we looked at this more closely because simply asking for 3 snowflakes will not help a 25% success rate much
18:15:02 <dcf1> well, it makes 1-0.75 into 1-0.75^3 (= 0.58)
18:15:12 <cohosh> true
18:15:59 <cohosh> and it's less bad now that turbotunnel lets us recover from getting all bad snowflakes
18:16:07 <cohosh> but still, good to know
18:19:04 * phw waits another 3 minutes to close the meeting if there's nothing else
18:20:35 <cjb> If someone has a link to somewhere I could read about the state of replay attacks against shadow proxies, would be curious to learn :)
18:20:42 <cjb> (can happen outside the meeting of course)
18:22:03 <cohosh> cjb: for shadowsocks?
18:22:39 <cjb> yeah, and obfs4 etc
18:23:13 <dcf1> I can help with some recent shadowsocks knowledge
18:23:56 <cjb> ah, I think I just found some code -- looks like obfs4 just stores previous received handshakes and ignores repetitions, that'd do it
18:24:27 <dcf1> yeah obfs4's replay protection is actually specified in its spec; shadowsocks it depends on the implementation
18:24:35 <cjb> hopefully it does it in a way indistinguishable from the normal timeout
18:24:57 <dcf1> The ScrambleSuit paper also talks about replays
18:29:04 <cjb> thanks! will check it out.
18:32:21 <phw> my three-minute timer is over, so let's wrap it up for today
18:32:29 <phw> #endmeeting