14:59:57 <pili> #startmeeting S27 01/14 14:59:57 <MeetBot> Meeting started Tue Jan 14 14:59:57 2020 UTC. The chair is pili. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59:57 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic. 15:00:21 <pili> here is the pad: https://pad.riseup.net/p/s27-meeting-keep 15:00:26 <pili> who's around today? :) 15:00:29 <brade> hi 15:00:45 * mcs is present and accounted for 15:00:53 <antonela> o/ 15:01:37 <sysrqb> o/ 15:02:00 <pili> welcome people :) 15:02:11 <acat> hi! 15:02:14 <pili> please add your updates as usual 15:03:46 * antonela pinging asn and dgoulet 15:03:57 <pili> thanks antonela 15:05:24 <dgoulet> hello 15:05:59 <pili> hi dgoulet 15:06:14 <pili> I'm guessing asn is still mainly literally afk ;) 15:06:21 <dgoulet> yes 15:08:59 <pili> ok, I think we may be able to start now 15:09:20 <pili> ah no, still waiting for ux and network team updates 15:09:49 <antonela> sorry got ping-distracted, filling the pad now 15:10:30 <dgoulet> pad link? 15:11:59 <pili> https://pad.riseup.net/p/s27-meeting-keep 15:12:03 <pili> sorry :) 15:12:56 <dgoulet> there 15:13:46 <pili> ok, let's start :) 15:14:10 <pili> I wanted to talk mainly about O2 today, but maybe first we should discuss a bit about O1 15:14:19 <pili> I spoke with asn last week about this 15:15:00 <pili> but maybe dgoulet has some more up to date info 15:15:01 <pili> I believe it's just the onionbalance work outstanding to complete O1? 15:15:23 <dgoulet> yes 15:15:26 <pili> I'm not sure if asn will complete it once he's back or whether dgoulet is going to take over now 15:15:57 <dgoulet> I took over part of the little-t tor stuff that remains. OBv3 works now but we need that part in tor to be merged within 043 15:16:02 <dgoulet> it should be done in the coming days 15:16:10 <dgoulet> than asn will finalize OBv3 first release 15:16:15 <dgoulet> to get it "out there" :) 15:17:00 <pili> cool, sounds good, thank you for the update dgoulet 15:17:01 <pili> anything else to share on O1 before we move on to O2? 15:17:41 <dgoulet> I think that is all, OB is really what we have left, couple tickets remains but they should be merged once asn comes back so nothing big 15:18:16 <pili> great, let's move on to O2 then :) 15:19:31 <pili> first I wanted to discuss where we are with O2A3 and review all the other child tickets under this activity: https://trac.torproject.org/projects/tor/ticket/30024 15:20:29 <pili> These are the current "-must" tickets: https://trac.torproject.org/projects/tor/query?status=accepted&status=assigned&status=merge_ready&status=needs_information&status=needs_review&status=needs_revision&status=new&status=reopened&sponsor=Sponsor27-must&parent=%2330024&order=priority 15:20:36 <sysrqb> we have a few child tickets :) :( 15:21:00 <pili> we should discuss which (if any) we might want to drop 15:21:15 <pili> #27502 15:21:19 <pili> #27590 15:21:26 <pili> and #30599 15:22:06 <antonela> yes, we have child tickets -- the circuit display related ones (UI specific) can be handled in parallel with #32645 15:22:31 <antonela> like #27590 15:23:44 <pili> so those 3 tickets are explicitly mentioned in the proposal 15:24:25 <sysrqb> hrm. 15:24:27 <pili> so we need to think about a) does it still make sense to do them b) can we achieve the same in any other way c) how long is it going to take us to do them 15:25:00 <pili> antonela: that's good to know 15:25:20 <antonela> do we need to have a deep discussion about them or? prioritizing v3 over v2 seems smart 15:25:45 <pili> we also mentioned: "Educate the user on how the redirection takes place, so that users get more familiar with the concept of onion services. We will employ appropriate “stop-and-learn” techniques to inform our users of the benefits of onion services and how they should be used." 15:26:22 <antonela> yes, we have it contemplated at the opt-in flow 15:26:32 <pili> and "Continue work on our Onion-Location proposal which gives the user the option to choose whether or not onion redirect should take place" 15:26:51 <antonela> that is basically all the work acat made 15:27:08 <pili> yup 15:27:11 <sysrqb> okay, let's see. 15:27:21 <pili> so we just need to formalise it by updating the proposal? 15:28:16 <pili> e.g all the discussion on #21952 works towards work on the onion-location proposal? 15:29:43 <pili> ok, so it sounds like it's possibly just #27502 and #30599 that are outstanding? 15:30:15 <sysrqb> and #27590 ? 15:30:37 <sysrqb> but #27590 and #30599 are related 15:31:00 <pili> ok 15:31:19 <pili> yeah, I wasn't sure if #27590 was covered by #32645 which is under O2A4 15:31:24 <pili> antonela? ^ 15:31:33 <sysrqb> ah 15:31:42 <antonela> i can work on 27590 15:32:08 <antonela> last time we discussed about it we got an agreement about to have the onion address at the circuit display, even if the url bar says something different 15:33:00 <antonela> in fact https://trac.torproject.org/projects/tor/ticket/30024#comment:4 has a proposal 15:33:01 * pili is reading #27590 15:33:02 <pili> antonela: ok 15:33:14 <antonela> i can update that ticket and move it forward, not sure what is the dev effort on it tho 15:33:17 <antonela> no idea 15:33:39 <pili> antonela: ok 15:33:50 <pili> sysrqb: we need to add this to our planning :) 15:34:21 <pili> because it does seem like there's more to #27590 than deciding on the UX 15:34:28 <pili> i.e there's some sort of bug? 15:35:55 <pili> so do we agree that we should still solve #27590 #30599 and #27502 ? 15:39:50 <sysrqb> sorry, internet dropped at a bad time :/ 15:39:59 <pili> oops :) 15:40:06 <pili> did you lose backlog? 15:40:24 <sysrqb> nope, i still have backlog :) 15:42:00 <sysrqb> for #27590, i hesitate saying i'm concerned about what information we should display 15:42:15 <pili> ok 15:42:54 <sysrqb> because we're re-using the logic for making connections to internet/clearnet domains 15:42:54 <antonela> which are your concerns sysrqb? 15:43:01 <antonela> right 15:43:08 <sysrqb> and most of the time, those use the same circuit 15:43:17 <sysrqb> it's not guaranteed, but it's usually okay 15:43:46 <sysrqb> but with onion services, each unique onion address uses a new circuit 15:44:13 <sysrqb> so displaying a single onion address in the circuit display path could be weird 15:44:29 <sysrqb> if the webpage uses resources on to twher onion addresses 15:44:44 <sysrqb> maybe lying about this for now is okay 15:45:01 <pili> right, or even resources on non-onion addresses, right? 15:45:12 <sysrqb> because we're already lying about which circuit is used for fetching some content on the clearnet, anyway 15:45:15 <sysrqb> yeah 15:45:16 <pili> and as antonela said, some of that is covered under O2A4 15:45:34 <pili> so maybe it no longer makes sense to do this and/or it needs more discussion/design 15:46:01 <pili> which is fine as long as we can justify it and provide some alternatives or progress towards a solution 15:46:09 <pili> instead of silently dropping it from the project :) 15:46:37 <sysrqb> right :) 15:46:37 <sysrqb> hrm 15:47:14 <sysrqb> we can think about how we can best spend our time over the next three months for accomplishing these 15:47:34 <sysrqb> i don't think we can/should solve these questions now 15:47:35 <sysrqb> :) 15:48:32 <sysrqb> like, recreating the circuit display is not a good use of our time right now 15:48:53 <sysrqb> but we can truncate the v3 onion address in the current display 15:49:00 <pili> right 15:49:22 <pili> so we can group #27590 and #30599 together as ones to keep thinking about 15:49:32 <sysrqb> yeah 15:49:52 <sysrqb> and i can chat with anto about which parts of these tickets we can focus on 15:49:54 <pili> and what about #27502? 15:50:07 <antonela> wefm 15:50:08 <antonela> wfm 15:50:24 <sysrqb> pili: #27502 confuses me :) 15:50:56 <sysrqb> but it seems like an easy patch 15:51:45 <sysrqb> but i don't know how/why "prioritize v3 over v2 onion addresses" is related to this 15:52:12 <sysrqb> i don't think any website is advertising both v3 and v2 onion addresses in their alt-svc 15:52:14 <pili> So Activity 3 is all about: "Integrate functionality that will tell a user if a website they are accessing has an onion service option" 15:53:00 <pili> and "Build the code necessary to optimally support this experience on Tor Browser such as prioritizing .onion addresses when Alt-Svc is used" is the explanation for #27502 15:53:01 <pili> not sure if that clarifies or not... 15:53:36 * antonela is watching https://perfectoid.space/onions.html 15:53:57 <pili> we only have 5 minutes left... 15:54:14 <pili> I really wanted to kick off O2A5 15:54:17 <mcs> comment:11 in that ticket suggests that v2 and v3 onions could be in the same alt-svc header but I am not sure if that is a “real world” scenario 15:54:49 <mcs> maybe just do the main thing and make sure the browser uses .onion when possible 15:55:11 <sysrqb> mcs: yeah, that is where i'm leaning 15:55:22 <pili> can we quickly jump on that one? I would like to a) discuss whether we know what is needed for this one b) who could work on it c) when they can start on this 15:55:34 <pili> mcs sysrqb sounds reasonable 15:55:35 <sysrqb> i think we can look at the necko code and find out how to prioritize .onion addresses 15:56:19 <sysrqb> tom also opened a bugzilla ticket for uplift, so we could reasonably write the patch for FF and then backport it 15:56:28 <sysrqb> okay, who can work on it... 15:57:28 <sysrqb> i think i'm probably the best person right now 15:57:41 <sysrqb> given that everyone else is overloaded 15:57:52 <sysrqb> i can delay other tickets i'd work on 15:58:34 <pili> ok :) 15:59:00 <pili> thanks sysrqb 15:59:23 <pili> lets leave it there then 15:59:38 <sysrqb> okay, sounds good, thakns pili 15:59:50 <sysrqb> i'll follow up with anto about the UI stuff 15:59:52 <pili> thanks everyone 15:59:58 <pili> #endmeeting