16:00:08 <meskio[mds]> #startmeeting tor anti-censorship meeting
16:00:08 <MeetBot> Meeting started Thu Apr 23 16:00:08 2026 UTC.  The chair is meskio[mds]. Information about MeetBot at https://wiki.debian.org/MeetBot.
16:00:08 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:00:13 <meskio[mds]> hello everybody
16:00:19 <meskio[mds]> here is our meeting pad: https://pad.riseup.net/p/r.9574e996bb9c0266213d38b91b56c469
16:00:21 <M4i_un[mds]> hello
16:00:22 <meskio[mds]> ask me in private to give you the link of the pad to be able to edit it if you don't have it
16:00:22 <theodorsm> hi:3
16:00:26 <cohosh> hi
16:00:29 <meskio[mds]> I'll wait few minutes for everybody to add you've been working on and put items on the agenda
16:01:35 <Shelikhoo[mds]> hi~hi~
16:03:51 <meskio[mds]> ok, I guess we can start
16:03:57 <meskio[mds]> we have one discussion point:
16:04:07 <meskio[mds]> How to Integrate UAT with Tor
16:04:13 <meskio[mds]> 4i_un: is it you?
16:04:24 <M4i_un[mds]> yees
16:04:48 <M4i_un[mds]> Currently, existing PTs (such as obfs4 and Snowflake) are based on Go, and packaging pipelines for deploying them to the Tor Browser and bridges are already in place. However, there is no pipeline for integrating Rust-based PTs into Tor. Therefore, I would like to ask about the following:
16:04:48 <M4i_un[mds]> At this point, I’ve created the following CI checks (can everyone view them?). To be honest, I’m stuck because I’m not sure how to proceed with building the pipeline. I believe the integration process requires several steps, such as recruiting a Test Relay operator and conducting production-level testing. I would appreciate any advice on how to move forward with these plans.
16:04:48 <M4i_un[mds]> https://gitlab.torproject.org/41_un/uat/-/merge_requests/2
16:05:46 <M4i_un[mds]> Also, back in February, I published the repository on GitLab and created an issue, and I received the following review from Cecylia Bocovich@cohosh:... (full message at <https://matrix.debian.social/ircbridge/media/v1/media/download/AdNdLMyEwdqx2-EVYAhWZAzIx_JBT5dLYGS4CJqB4YbhMYOKIDgu44hWlTSZoUGJxA5Doj6sIkOP_uXuLwuVmgpCed3yObrwAG1hdHJpeC5kZWJpYW4uc29jaWFsL2JSaUZ3WEpUUlZVWklPdnF3a2JkVkpmRQ>)
16:06:06 <M4i_un[mds]> I’d like to discuss this with everyone!
16:07:38 <meskio[mds]> the merge_request link gives me a 404
16:07:46 <meskio[mds]> I can see the repo, not sure why the link doesn't work
16:07:48 <theodorsm> Same here
16:07:57 <cohosh> it might be private
16:07:58 <theodorsm> Do we have the permissions to see MRs?
16:09:22 <Shelikhoo[mds]> yeah, same here, the repo is visible, but the merge requests are not visible to me
16:10:20 <cohosh> M4i_un[mds]: thanks for your continued work on this
16:10:53 <cohosh> as far is integrating with Tor, there's no pipeline because it's not really well defined what that means
16:11:03 <Shelikhoo[mds]> yes, thanks for your work 4i_un !
16:11:18 <cohosh> as you've probably seen with testing, you can already use your PT with Tor on the command-line
16:11:54 <M4i_un[mds]> I'm looking at the settings, and the “Visibility,” “Project Features,” and ‘Permissions’ sections are set to “Public.”
16:12:08 <cohosh> further steps for a PT could be distributing bridges with rdsys, and integrating it with client applications
16:14:04 <cohosh> also: adding support and setup pages on the tor project website
16:14:26 <cohosh> but, deciding whether and when to move forward on these is a collective decision making process that takes time
16:14:53 <cohosh> a first step is for us to take a look at the code and protocol
16:16:33 <M4i_un[mds]> <M4i_un[mds]> "Currently, existing PTs (such as..." <- I opened it in an incognito tab, but if you're logged into GitLab, you should be able to view it, so please take a look!
16:16:56 <M4i_un[mds]> I opened it in an incognito tab, but if you're logged into GitLab, you should be able to view it, so please take a look
16:17:59 <cohosh> thanks M4i_un[mds] i can see the MR now :)
16:18:22 <Shelikhoo[mds]> Yeah, usually adding a PT will take a while and there are a lot more difficulty than just get it to run. But I think we should have a round review of the protocol soon
16:18:37 <M4i_un[mds]> cohosh: Yes, I believe that when it comes to the code, we need to review it for potential security risks and ensure that it has been implemented according to the specifications.
16:18:46 <cohosh> okay for next steps, i can offer to do a review of the code
16:18:59 <cohosh> a big review like this will take time and i'll be afk a few days next week
16:19:29 <cohosh> we'll also discuss as a team next week about when we have capacity to look at it
16:20:14 <Shelikhoo[mds]> I agree we can discuss about it in our next team meeting
16:20:19 <M4i_un[mds]> cohosh: Understood. Thank you very much for your cooperation.
16:20:25 <cohosh> thanks for your patience M4i_un[mds] and i'll also suggest in the meantime running your own bridge and try using the client yourself for daily browsing tasks
16:20:52 <cohosh> you can hack tor browser to use a custom PT, i'll add some instructions to https://gitlab.torproject.org/tpo/anti-censorship/team/-/work_items/178
16:21:26 <M4i_un[mds]> Understood.
16:22:52 <cohosh> i'm looking forward to taking a look!
16:23:04 <meskio[mds]> great, I guess next step is clear
16:23:12 <meskio[mds]> thanks 4i_un , sorry things are slow
16:23:13 <Shelikhoo[mds]> nice!
16:23:44 <meskio[mds]> anywthing else on this topic?
16:23:46 <M4i_un[mds]> No, no—I’m not very good at English either, so this actually helps me a lot!
16:23:55 <meskio[mds]> :)
16:25:11 <Shelikhoo[mds]> EOF on this topic from me
16:25:27 <M4i_un[mds]> me too
16:25:31 <meskio[mds]> anything else to discuss today?
16:25:31 <M4i_un[mds]> thanks everyone
16:25:45 <GeKo> meskio[mds]: i have just an announcement
16:25:45 <Shelikhoo[mds]> nothing from me
16:25:48 <GeKo> question
16:25:55 <GeKo> (hi!)
16:25:56 <meskio[mds]> sure, go ahead
16:26:03 <cohosh> hi GeKo!
16:26:19 <GeKo> so, we have this project (p183) where we work on anomaly detection in the tor network
16:26:21 <GeKo> cohosh: hi!
16:26:33 <GeKo> i wonder whether that would be something useful to talk about here next week or so
16:26:44 <GeKo> maybe there is some interest and ideas/thoughts about it
16:27:14 <GeKo> one thing we did is taking wright's paper from 2018 and porting their tool over to python
16:27:18 <GeKo> https://censorbib.nymity.ch/pdf/Wright2018a.pdf
16:27:37 <cohosh> oh cool
16:27:39 <meskio[mds]> sure, I mean if you want it to do it now we have some time and talk about it
16:27:45 <GeKo> joss gave the original r code to us and juga did the heavy-lifting to have it in python
16:28:08 <GeKo> https://gitlab.torproject.org/tpo/network-health/tor_anomalies
16:28:09 <meskio[mds]> nice
16:28:32 <GeKo> well, i don't have much time now but could prep a bit for next week
16:28:47 <GeKo> and you folks could think about questions/ideas in the mean time
16:28:53 <meskio[mds]> sure, and might be good to have some time to look into it
16:28:54 <GeKo> or just get excited :)
16:28:54 <meskio[mds]> thanks
16:28:55 <cohosh> yeah, we could also read up on the paper and come prepared with questions
16:29:14 <cohosh> is there a public link to the planned project work?
16:29:44 <GeKo> it's actually ongoing for a bit and we are now at the stage where we can do actual analysis, using e.g. wright's tool
16:29:56 <GeKo> let me give you a proper link
16:30:12 <GeKo> https://gitlab.torproject.org/groups/tpo/network-health/-/work_items/1 is the overall epic
16:30:23 <GeKo> and then you can drill down
16:30:50 <cohosh> awesome
16:30:54 <cohosh> thanks!
16:31:00 <GeKo> sure
16:31:06 <Shelikhoo[mds]> nice! thanks!
16:31:25 <GeKo> i'll come better prepared next week and we can talk about stuff that interests you
16:31:47 <meskio[mds]> that sounds great, thanks for the heads up
16:32:02 <GeKo> you are welcome!
16:33:48 <meskio[mds]> ok, anything else for today?
16:34:18 <theodorsm> pion/webrtc with boringssl: https://github.com/zshang-oai/webrtc-boringssl
16:34:24 <theodorsm> People over at openai thought that our efforts to implement DTLS 1.3 in pion was too slow, so they forked webrtc and replaced pion/dtls with boringssl (used in Chrome)
16:34:29 <theodorsm> If the DTLS traffic in Snowflake gets blocked by advanced fingerprinting in the future, we have a drop-in replacement for DTLS.
16:34:32 <theodorsm> Might be hard to deploy to Snowflake clients, but should be easy with standalone proxies.
16:34:36 <theodorsm> I want to play a bit with it when I have time in the future.
16:34:40 <theodorsm> Just wanted mention the existence of this library, in case we ever need it.
16:35:04 <Shelikhoo[mds]> nice! thanks for sharing
16:35:18 <meskio[mds]> ohh, pretty cool
16:36:10 <theodorsm> That's all from me:)
16:37:16 <meskio[mds]> great, I guess I will close the meeting here
16:37:19 <meskio[mds]> #endmeeting