16:00:15 <cohosh> #startmeeting tor anti-censorship meeting
16:00:15 <MeetBot> Meeting started Thu Sep  9 16:00:15 2021 UTC.  The chair is cohosh. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:15 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:00:18 <cohosh> hi everyone!
16:00:21 <cohosh> welcome :)
16:00:25 <meskio> hello
16:00:35 <dcf1> hallo
16:00:36 <cohosh> here is our meeting pad: https://pad.riseup.net/p/tor-anti-censorship-keep
16:00:51 <cohosh> feel free to add what you've been working on and put items on the agenda
16:05:23 <cohosh> ... or just jump in with discussion since it seems like a quiet day
16:07:02 <dcf1> Is there consensus as to what's going wrong with https://gitlab.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake/-/issues/40061
16:07:38 <dcf1> 2 people went to the trouble of making a gitlab account and finding an existing ticket
16:07:49 <dcf1> is it just a case of them getting unlucky with proxies?
16:08:02 <dcf1> I was thinking to ask them to enable the snowflake.log
16:08:02 <cohosh> hmm, i'm not sure
16:09:03 <cohosh> my suspicion is that they are getting unlucky with proxies and then the client times out
16:09:11 <cohosh> but enabling snowflake.log would tell us more
16:10:14 <cohosh> i think that's a good next step
16:10:25 <dcf1> ok, I'll say that in the ticket
16:10:40 <meskio> but if the client does timeout should it reconnect to another proxy?
16:12:22 <meskio> anyway, let's see if they can provide logs
16:12:39 <cohosh> meskio: right, this is if the SOCKS connection times out
16:12:50 <cohosh> or the tor circuit timeout is reached
16:13:29 <cohosh> which happens if it takes > 2 mins to get a working proxy
16:14:02 <cohosh> looking at the broker metrics, we are very short on unrestricted proxies still
16:16:10 <meskio> $Cif we are done with this topic I have another short one
16:16:35 <cohosh> meskio: ok! go for it :)
16:17:01 <meskio> two weeks ago a person asked in tor-relays about rotating IPs of bridges
16:17:22 <meskio> if it was ok to rotate their VM IPs once per year or it was dependent of the pool the bridge is
16:17:41 <meskio> I guess we are happy with IP rotation as long as they are not in the private human handled pool
16:17:58 <meskio> but AFAIK no "random" bridge gets there, only manually added ones, isn't it?
16:18:07 <meskio> I can answer, but I was not 100% about it
16:19:02 <dcf1> https://lists.torproject.org/pipermail/tor-relays/2021-August/019788.html
16:19:11 <meskio> that one :)
16:19:41 <cohosh> yeah that should be okay in the sense that if they rotate their IP, BridgeDB will stop handing out the old IP and hand out the new one instead
16:19:41 <dcf1> I don't know how the pool assignment works
16:20:02 <cohosh> pool assignment is done based on a hash of the fingerprint
16:21:18 <meskio> cool, I can answer to them
16:21:43 <cohosh> in terms of usefulness, i don't have good numbers here. there are a few places we know of that enumerate BridgeDB bridges and getting fresh IP addresses might help users in those locations by providing them with a bridge before it is discovered and blocked
16:26:58 <cohosh> i guess bridge operators can also specify which pool they want to be in
16:27:52 <cohosh> anything else for today?
16:28:22 <meskio> not from my side
16:30:03 <cohosh> okay i can end the meeting here
16:30:11 <cohosh> #endmeeting