19:03:42 #startmeeting onion ux 19:03:42 Meeting started Wed Dec 20 19:03:42 2017 UTC. The chair is asn. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:03:42 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:03:58 what's up? 19:04:06 so ! 19:04:27 i start organizing things on the google doc 19:04:35 do we have tjr: geko: ? 19:05:01 i am not sure 19:05:08 * arthuredelstein is lurking. lurk lurk 19:05:13 GeKo might be traveling 19:06:37 ok we got #23247 and #21952 19:06:50 yes, so on the first one 19:06:53 wrt #21952, i posted a proposal on tor-dev 19:06:56 i updated the google doc 19:07:06 based on feedback from tjr 19:07:19 but then received some very good feedback from geko which i havent addressed yet 19:07:20 nice 19:07:39 (revised proposal: https://lists.torproject.org/pipermail/tor-dev/2017-December/012660.html) 19:08:13 here is the feedback from geko which i havent addressed: https://lists.torproject.org/pipermail/tor-dev/2017-December/012690.html 19:08:41 anyway and that's that for #21952 19:09:10 ok 19:09:11 now for #23247 19:09:16 * antonela is reading 19:09:18 so we can pick it up next year 19:09:27 yes 19:09:29 for that 19:09:46 https://docs.google.com/document/d/1KHkj2DpmFMB0mjHEfehD5ztY2L0lQzKNtZqct1TXbmg/edit 19:10:14 so i start to organize things at this doc 19:10:54 mixed content still something complicated 19:11:17 but i am going towards the .onion should be green because is better than just accessing http 19:11:47 aha 19:11:49 added some copies that are just to show, not final of course 19:11:52 based on https://trac.torproject.org/projects/tor/ticket/23247#comment:25 19:11:57 it says "Another vote, separate from that discussion, was a very strong 'no positive indicator for .onion' 19:11:57 " 19:13:00 not sure what tjr has to say about that 19:13:06 yeah 19:13:37 interesting that " 19:13:37 Onion with Self-Signed HTTPS 19:13:40 ehm 19:13:50 interesting that "onion with self-signed https" looks more sketchy than "onion with http" 19:13:59 on ur spreadsheet 19:15:16 seems like you could just make that look the same as onion with http 19:15:53 i thought of it but then i thought also that in this case, where the certificate would return it could also means a bad certificate 19:16:58 like being intercepted or something 19:17:06 i am not sure either i can be wrong 19:17:22 isabela: didnt get the logic 19:17:23 if you think should be changed we should do it then 19:17:36 can you reexplain? 19:18:00 asn: for instance on brave with a self signed certificate they say is insecure because of the certificate 'being bad' 19:18:11 or the risk of it 19:18:36 personally i dont see much use for a self-signed https cert, so i'm ok with discouraging it 19:18:38 because is self signed so that can be someone trying to intercept things 19:18:43 true 19:19:07 but in this case the interceptor can just setup an "onion with http" to get a nicer onion icon 19:19:52 but anyway 19:20:00 perhaps we should wait for tjr and geko to tell us their insights from all hands 19:20:03 Perhaps that is actually nicer. Using a self-signed cert is a cause for slight concern. 19:20:11 i was at the all hands 19:20:19 isabela: true 19:20:22 but i am not sure what other insights came from it 19:20:24 they seem to have some relevant discussions tho 19:20:30 based on comment:25 19:20:34 or wasnt on this conversation he mentioned 19:20:42 i spoke with design on it 19:20:48 they are not removing padlocks but making them nicer 19:20:59 ah 19:21:00 with rounded corners etc so those are old version of it 19:21:04 ha nice 19:21:06 ack 19:21:08 anywa 19:21:10 anyway 19:21:25 im personally fine with all the suggested stuff on the spreadsheet, also with the self-signed https one. i dont have strong opinion on that. 19:21:41 tjr's comment on "no positive indiicator" might be worth learning more about 19:21:54 but perhaps we are also overthinking 19:22:00 yeah, i dont know what conversations was that 19:22:20 well 19:22:37 i think that whatever we decide here we will want to do test 19:23:23 aha 19:23:30 isabela: +1 19:23:35 test how? 19:23:39 roll it out on an tbb alpha? 19:23:41 user reserach 19:24:00 i think we can launch it and continue working on top of it 19:24:20 i think right now there is nothing there explaining the real state of a .onion 19:24:24 right 19:25:00 is ok to have it out and do research and iterate on the top of it 19:25:11 look how browsers have iterate with it so far 19:25:20 i also met the user research person focus on privacy stuff 19:25:48 and i think we can exchange ideas on how to conduct these tests 19:25:58 ahja 19:26:07 i was showing them we are planing to do this with .onion states 19:26:18 i currently have no idea how to do this iterated user research thing 19:26:25 but it sounds interesting 19:26:30 yes it is! 19:26:42 yes 19:27:06 sorry I am here now 19:27:08 asn: our goal is to always do it with all our work or user facing things we build or change 19:27:11 Let me read backscroll 19:27:36 tjr: o/ 19:27:44 tjr: yoo 19:28:53 okay so as far as no positive indicator goes, that I think is the long term hope/plan for HTTPS indicators in general. 19:29:03 HTTPS with a DV cert gets no indicator at all 19:29:15 HTTP gets a Red Negative Insecure Indicator 19:29:18 (Who knows about EV) 19:29:23 interesitng 19:29:33 But it doesn't answer the question "What should TB do now" 19:30:11 perhaps we can do our planned thing, and adapt if/when we learn that this "no positive indicator" future is a good one? 19:30:24 Yea. 19:30:42 But what's the planned thing? Did you decide what to do about HTTP/HTTPS self-signed onion? 19:31:05 https://docs.google.com/document/d/1KHkj2DpmFMB0mjHEfehD5ztY2L0lQzKNtZqct1TXbmg/edit 19:31:22 isa's current suggestion is green onion for http, green onion with notification for self-signed 19:31:29 sounds good to me 19:31:57 Sure 19:31:58 most things sound good to me at this stage 19:32:04 Yea, ditto that 19:33:31 ok, so maybe we catch up with tb team on this next year when they are back from holiday stuff 19:33:39 aight 19:33:40 i plan on having a big sync with TB team on this and other ux work we are doing 19:33:45 ok great 19:33:49 should we write them a summary email? 19:33:49 so i might just add it to the list 19:33:52 or just wait for u to synch up with them? 19:34:29 i will organize the info about this ticket with the others we will be covering at that meeting 19:35:11 the goal is to coordinate stuff, since the work on our side needs to be done for them to implement, and once is implemented we need to organize the user testing etc w it 19:35:22 ack 19:35:29 and i will also update #21952 based on geko's comment 19:35:38 great 19:36:07 i think another sync with you and tjr (if available) would be helpful to review some of the copy i still need to write 19:36:19 like the stuff i start adding a the description column 19:36:20 I won't be around until Jan 3rd ish, sorry 19:36:30 yeah, i meant next year 19:37:02 sounds good 19:37:32 aight i think we done here then!! 19:37:36 scary we are talking about next year 19:37:41 but it's in 11 days so i guess that makes sense 19:37:48 :) 19:38:11 hehe 19:39:15 ok 19:39:21 closing this!!! 19:39:23 #endmeeting