16:00:22 #startmeeting tor anti-censorship meeting 16:00:22 Meeting started Thu Jul 24 16:00:22 2025 UTC. The chair is onyinyang. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:22 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:00:22 hello everyone! 16:00:22 here is our meeting pad: [https://pad.riseup.net/p/r.9574e996bb9c0266213d38b91b56c469](https://pad.riseup.net/p/r.9574e996bb9c0266213d38b91b56c469) 16:00:31 hello 16:01:00 oh hey, i'm crashing this meeting 🙂 16:01:06 hi~hi~ 16:01:10 hi 16:01:13 dan_b, welcome! :d 16:01:17 :D 16:01:36 dan_b: I hope you brought drinks 16:01:40 hahaha 16:01:56 our agenda is pretty empty but I'll give everyone a few minutes in case there are any late comers :) 16:02:04 thanks dan_b, here is your crash link: chrome://crash 16:02:16 oh interesting, is the pad locked on whom can edit? 16:02:56 yes, I can send you a link if you need to edit the pad 16:03:09 we've being getting trolls on the pad 16:03:17 oph 16:03:18 we share the link on private, I just sent it to you 16:03:22 cool 16:03:25 onyinyang: i have assigned you as an reviewer of Add Domain Fronting Test Support to probeobserver https://gitlab.torproject.org/tpo/anti-censorship/connectivity-measurement/probeobserver/-/merge_requests/8 16:03:43 I can reassign it if you like 16:03:45 shelikhoo, ok, I'll take a look 16:03:48 or wants to 16:03:54 feel free to assign it someone else 16:06:13 ok 16:06:57 maybe we can talk about dan_b's TorVPN topic 16:07:06 alright! 16:07:07 so 16:07:17 we're planning a 3 pronged distribution 16:07:30 torproject.org download, google play store, and fdroid 16:07:38 so the matter of signing apps arose 16:07:43 and it's complex 16:07:54 https://gitlab.torproject.org/tpo/applications/vpn/-/issues/244#note_3227570 16:08:01 here was the summary in gitlab 16:08:07 but basically 16:08:24 when apps from different distribution channels are signed with the same key, you can like "side grade" them? 16:08:41 so if we shared signing keys with google, and you initially got tor vpn from play store 16:08:48 but then say your internet was getting blocked 16:08:58 and Tor gave you a new experimental copy via another channel 16:09:06 if they shared signing keys, they could just install that on top 16:09:18 if they dont share signing keys, they'd have to uninstall, lose settings, install the new one 16:09:27 nice, so we can also distribute it over gettor and other channels :) 16:09:39 the problem with google is they demnand to do the signing for what they distribute 16:09:45 which means giving them a copy of the keys 16:09:46 Do they have to share the same appid? 16:09:51 we in the meeting leaned away from that 16:09:59 can we have something like org.tor.googlever 16:09:59 wow, so they have the private key? 16:10:02 and picked a stratgy of let google do the signing itself with their own keys 16:10:11 org.tor.apkrelease 16:10:14 and we then distribute on torproject.org one signed with our own 16:10:24 org.tor.fdroidver 16:10:34 and also fdroid has a mechanism where if they confirm reproducibility, they'll distribute the one we publish, so signed with our key 16:10:43 so they cannot upgrade to each other 16:10:51 so this means that we'll be able to distribute side-grade options with fdroid, but not google play store 16:11:01 but, user can install each version of them independently 16:11:10 they all have the same appid 16:11:11 so no 16:11:16 as it currently stands 16:11:30 we'd not have that option for folks who get their tor vpn from play store 16:11:45 the meeting had no representation from anti censorship or community 16:12:03 so we all focused more on the uncomfort and potential risk of sharing signing keys with google 16:12:14 I understand 16:12:17 and decided that "we" could live with this side effect 16:12:21 If I were to publish such an app, I would make app from different distribution channel a "different" app 16:12:25 but we wanted to actually check in with anticensorship 16:12:36 incase you all wanted to flag that as a bigger concern than we could see 16:12:51 with different appid, so that they won't conflict with each other 16:12:55 and or just to inform so we can strategize accordingly for launch and after and docs and etc 16:13:17 I guess this problem already exist with TBA, isn't it? 16:13:29 TBA is a grandfathered in case 16:13:33 we do distribute TBA over gettor, and I haven't heard complains about upgrades 16:13:40 it's "old" and so we can actaully push our self signed .apks 16:13:59 ohhh, I see, is a new rule for google play to demand the private key 16:14:03 that has other limits, there's a hard size limit for that legacy path (in part cus they want ppl to upgrade to the new path) 16:14:05 yes! 16:14:08 yes 16:14:15 google wants your private key 16:14:31 that is fucked up 16:14:34 and on android as everything grows from firefox android to the PTs it's harder, there may come a time we have to revisit that 16:14:49 meskio: yes, those words and others were used in the meeting as well 16:14:59 :) 16:15:07 but like, you could just generate a new one for that app... but yeah. its just... not great 16:15:48 shelikhoo: that is an interesting idea, i will make a note to investigate to see how easy it would be to use a seperate appid for google 16:16:05 then they could be installed side by side, but the settings wouldnt migrate over 16:16:26 does that seem like it would still be a minor benifit to pursue? 16:16:50 dan_b: alternatively, we can make backup settings really easy 16:16:58 like just 2 click away 16:17:00 yes, that was also raised to investigate 16:17:19 but the initial launch of the beta is locked in already for next month so it might be a fast follow and not launch item 16:17:35 sounds reasonable 16:17:45 https://gitlab.torproject.org/tpo/applications/vpn/-/issues/319 16:17:54 yes... it could be used to move settings between different version of apps 16:18:00 was created yesterday as feedback and mitigation from the meeting on this topic 16:18:15 I can also comment on this issue later with more details of split appid idea 16:18:17 currently not tagged for the beta release 16:18:28 over 16:18:42 cool. I'll also do some research to see how feasible/easy it would be 16:18:53 i think we do something like that for TB for the alpha / nightly / release channel 16:18:58 I think is fine if users need to reinstall to ugprade, sadly users in censored areas are used to suffer some pain 16:19:07 haaaaa ok 16:19:37 thanks. and so yeah I was dispatched too just to get this on all your radars 🙂 16:19:48 thanks for bringing this up to us dan_b 16:20:00 interesting/horrifying 16:20:02 my pleasure! 16:20:03 ha yessssss 16:20:17 haaaaa thanks dan_b 16:20:18 ok that's all i have 16:20:19 dan_b: will TorVPN releases be advertized in the same json than TB releases? 16:20:30 oooh, which json? 16:20:48 I mean https://aus1.torproject.org/torbrowser/update_3/release/ 16:21:01 we use that in gettor 16:21:16 or something similar that we can use in gettor? 16:21:20 what is gettor? 16:21:33 that currently isnt on my radar 16:21:38 gettor is a way to get TorBrowser over email or telegram 16:21:42 ooooh 16:21:59 https://gitlab.torproject.org/tpo/applications/vpn/-/issues/244 16:22:02 the telegram one is very much used 16:22:04 here is the parent issue i'm working from 16:22:20 could you drop a note about gettor and the json file and how it's generated or populated 16:22:29 yes, I'll add a note about it 16:22:29 so I can track that and see if I can make that happen then? 16:22:34 awesome thanks!! 16:23:17 it's kinda of a mess of a ticket, and some sub tickets have belatedly been spun off to track tasks 🙂 but that's the "epic" i guess, and I can make new tasks from that later 16:23:45 cool, we can coordinate there and create issues as needed 16:23:57 we'll also need an issue in our side to add support for TorVPN in gettor 16:25:00 ok cool, can you link to the repo in your post (for tor-browser even tho we have/had a bunch of repos we tracked all the issues just in the TB one, not sure if your team does the same) 16:25:49 sure, I'll link the gettor repos (is actually two different things for email and telegram) 16:25:57 aaaah k 16:27:30 anything else to discuss on this topic? 16:27:35 not from me 16:27:36 eof from me 16:27:59 nope i'm good, thanks all! making comments on tickets now (shelikhoo I'm about to mention your seperate appid idea 🙂 ) 16:28:10 great! 16:28:13 ok, we don't have any other topics today, just some interesting links from FOCI and PETS last week 16:28:23 maybe we'd like to make a reading group for one of them? 16:28:35 Or anything else from the conferences last week that was interesting? 16:28:41 nice! I will have a look and maybe add some details 16:28:48 shelikhoo do you want to be tagged in the issue, and what's your gitlab handle if so? 16:28:55 @shelikhoo 16:28:59 please tag me 16:29:20 maybe we can discuss theodorsm paper in a reading group 16:29:31 I'm also interested in the FOCI's one about ECH 16:30:00 the ECH one is good 16:30:04 so many great papers 16:30:24 https://www.petsymposium.org/foci/2025/foci-2025-0016.pdf 16:30:33 ^- the ECH paper I talk about 16:31:05 CenPush was not in my radar, I should check it out, thanks 16:31:19 so what do we want to read as a reading group? 16:31:53 i'll second the ECH one, i think it's good to be aware of 16:32:03 ok, let's do that one 16:32:22 checking the calendar the next meeting that none of us is AFK is August 14 16:32:26 am I correct? 16:33:01 that should work for me 16:33:06 me too 16:33:25 should work for me 16:33:27 as well 16:34:15 nice, we have a date and a paper :) 16:34:41 good work everyone XD 16:34:49 I guess that's all for today 16:34:54 yep 16:35:12 thanks for the discussion! :) 16:35:13 #endmeeting