16:00:31 #startmeeting tor anti-censorship meeting 16:00:31 Meeting started Thu May 21 16:00:31 2026 UTC. The chair is Shelikhoo[mds]. Information about MeetBot at https://wiki.debian.org/MeetBot. 16:00:31 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:00:31 here is our meeting pad: https://pad.riseup.net/p/r.9574e996bb9c0266213d38b91b56c469 16:00:31 editable link available on request 16:00:32 hi~ 16:00:36 hello 16:00:47 hihi o/ 16:02:06 please go ahead to add discussion topics to the pad 16:03:51 maybe we have a quiet meeting today 16:04:13 yeah, in theory we could pick a reading group if there is no other topics 16:04:26 but cohosh is not here... 16:05:03 so maybe as we all busy working on our assignment on signaling channel, there isn't a lot to discuss if everything is proceeding as planned 16:05:47 but I will wait 2 more minutes for anyway last minute topics or anything we wants to discuss in this meeting. 16:05:49 there's a pets paper that might be good for reading group 16:06:02 which one? 16:06:30 "Evaluating connection migration based QUIC censorship circumvention" 16:06:53 just saw it in censorbib 16:06:53 https://petsymposium.org/popets/2026/popets-2026-0014.pdf 16:07:38 we could pick it up for June 11th (I'll be AFK June 4) 16:07:41 that's the one 16:07:54 Seungju Lee was at the last online FOCI 16:08:22 I thin this paper looks nice 16:08:30 I think this paper looks nice 16:08:45 I haven't read it, but my impression is that it's about tunneling just the handshake (incl. SNI) over a separate channel, then gluing it to a standard QUIC UDP connection for the remainder, using connecting migration 16:09:37 interesting idea 16:10:47 yeah, this is very impressive as this can be done in user mode without special privilege 16:11:18 if we wants to do similar thing to a tcp based tls connection, then it would require raw socket 16:11:37 or turbo tunnel, at the cost of additional overhead 16:11:41 which won't work on restricted device like locked smartphone 16:12:17 there is another little-known paper I read recently, called DNS-Morph, that does something similar, but specialized to the handshakes of obfs4 and scramblesuit 16:12:29 I'm going to post a summary of DNS-Morph to net4people pretty soon 16:12:38 dcf1: you are right. I think it is more important to use this to run domain fronting on host that don't allow it 16:13:00 so if a cdn does not allow domain fronting, but allows quic 16:13:21 this can be used to create a data channel over a signaling channel 16:13:39 this can be used to create a data channel using a signaling channel 16:13:51 and a quic allowed domain fronting target 16:14:50 but I think we can save more discussion to the reading group once we have read the paper. EOF 16:15:33 yeah, let's discuss it then 16:15:42 are we happy with the proposed date? 16:16:21 it should work for me 16:16:46 we can maybe confirm the day again next week 16:16:51 as cohosh isn't here 16:16:57 sure, lets check with her 16:17:35 oh yeah, the 11th should be fine for me too 16:17:37 yes. let me know if there is any more topic we wants to discuss today 16:18:30 nothing from me 16:18:45 Thanks everyone! have a nice rest of the day and weekend! Let's call it a meeting. 16:18:45 #endmeeting