17:59:25 #startmeeting 17:59:25 Meeting started Mon Jul 23 17:59:25 2018 UTC. The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:59:25 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:59:29 hi all! 17:59:33 * tjr waves 17:59:40 hello 17:59:56 hello 18:00:01 o/ 18:00:34 hi! 18:00:36 our pad is at https://storm.torproject.org/shared/tHoN4Ii7rLSjPE0OP4gydX4cMGadsXmRQNc-6lwru0N as usual, please update your items if you did not have done so already 18:00:48 and mark things as bold which you want to talk about 18:01:08 O/ 18:02:40 if we had a "wtf bug of the month" contest, #26514 would be a good candidate for it 18:02:48 (see: https://trac.torproject.org/projects/tor/ticket/26514#comment:7 for some details) 18:03:01 glad we are still in the alpha cycle for win64 :) 18:03:37 boklm: i think we don't want to get out another alpha this week but i hope maybe mid-week next week 18:03:43 would that work for you as well?# 18:03:47 s/#// 18:03:50 i saw that; that's an issue with the gcc-based library code that gets compiled in? e.g. it wouldn't affect FF? 18:04:15 i am inclined to think that's not affecting firefox but we don't know yet 18:04:20 GeKo: ok, that would work for me 18:04:47 boklm: good. do you have any eta for getting our test suites running? 18:04:52 like on linux first? 18:05:41 tjr: i have added some questions for you. no need to discuss them here i guess 18:05:50 I can try to get them running on linux next week 18:05:53 * mcs is here now 18:06:03 (sorry for running late) 18:06:08 but we might want to think about not wasting too much time/resources for the esr60 thing 18:06:12 o/ 18:06:32 boklm: that would be good, yes. 18:07:28 boklm: oh, and could you finally file the binutils bug this week? I have still hope that we can get that fixed with help from binutils devs so we can use a more recent binutils version 18:07:40 ok 18:08:37 sisbell: awesome work. looking forward to try some branches out, and give feedback 18:09:17 me too! 18:09:30 igt0: please see my sponsor8-related tbb-dev mail, in case i forgot something 18:09:37 cool, learned more about rust than I intended 18:09:45 :) 18:10:01 GeKo, yep, I already updated the document! 18:10:15 oh, great, thanks 18:10:39 mcs: could you put #25696 on your plate for this week? 18:11:10 i guess that's worth the main focus given that we might want to ship a new alpha next week 18:11:43 GeKo: Yes, it is on my mind but didn’t make it onto the pad. Sounds like a lot to get done by next week, but we will take a look. 18:12:00 * antonela is around 18:12:12 err: #26595 18:12:17 question, will we share about:tor between desktop and mobile? 18:12:22 #25695 18:12:52 That’s the one I thought you meant anyway. 18:13:01 just to be clear :) 18:13:03 me too 18:13:03 ha 18:13:22 antonela: yes, i think so. 18:13:46 cool, i made a TBA version too 18:13:49 do we have everything ready to start the coding part? 18:14:23 illustrations are not at the final version. We will iterate them this week. 18:14:33 but, everything else is ok i think 18:14:50 ok. could you add everything we need to the ticket? 18:14:52 both TB8 and TBA 18:15:02 yep 18:15:19 mcs: yes, the schedule is tight. if we miss the one for next week, then be it so. bu ti figured we should try at least 18:15:34 sorry, do you mean in terms of user flow? 18:16:16 what do you mean? 18:17:17 when you said 'everything we need', are you talking about user flow explainer? assets? svg/jpgs? 18:17:39 well, all the things we need for implementing the feature :) 18:17:44 so, i guess all of that 18:17:48 oh yes, cool 18:17:54 thanks :) 18:20:24 sysrqb: igt0: is there anything i can help with for mobile in particular? 18:20:39 i hope the dust stettles a bit once we have #26401 merged 18:20:56 and i plan to push all the other backports and bugfixes for mobile on top of it 18:21:19 but is there anything apart from that that comes to mind? 18:21:21 cool 18:21:49 if nothing right now, that's cool too :) 18:22:03 hrm. i think we may need help with torbutton and noscript interactions 18:22:13 but igt0 may be looking that already 18:22:18 *at that 18:23:01 sysrqb, yep, as part of the settings work. 18:23:07 i have it a TBA build running on my phone, and i don't think the default Tor Browser noscript settings are set 18:23:15 ah, excellent! thanks! :) 18:23:30 okay, ignore that 18:24:40 okay, aynthing else before moving forward to the discussions part? 18:24:44 *anything 18:25:39 just a reminder i have to be afk again between 8/2 and 8/12 which i am sorry for but as i said on the pad the planning for that was already fixed before mozilla moved the esr :( 18:26:07 okay, discussion time then i guess 18:26:15 tjr: you had some items it seems 18:26:37 Yea, maybe not discussion worthy, just some feedback from the Hackign doc 18:26:50 We can move on to github 18:27:08 i can see moving orbot but i tihnk orfox is still relevant for that page, no? 18:27:21 and we can remove it when TBA is released, i think 18:28:06 or are you concerned the build instructions are so large that it'll be better documented on another page? 18:28:12 (for orfox) 18:28:24 Sure. Mostly I was just thinking about 'What is discreet enought o go to another page to make this page shorter' 18:28:32 Debugging might be another example 18:28:53 i can see that, i'm not against that 18:29:07 makes the hacking page more readable 18:29:30 okay, yeah, we can tihnk about that 18:29:30 okay, let me file a ticket later and then we can come up with a plan and get it done 18:30:07 the github item 18:31:11 that's the reminder to think about isa's mail to tbb-dev in particular for those that were on vacation and/or did not read last weeks meeting notes 18:31:45 i think it's not only about code hosting but about review tools as well 18:32:43 yeah 18:32:54 I love github. I don't care for the centralization of course but yea... it's very easy to use and lends itself to very easily contributing pull requests for tiny improvements 18:32:55 do we as a team have an idea about what we want already? 18:33:01 And review comments 18:33:44 I think there are a couple of us already using GitHub in some capacity 18:34:14 so the idea would then be follow the network team model to have the canonical repo at our infra but do the interaction on github? 18:34:15 How does the flow work for tor (Network Team)? Do they mirror to github and manually merge PRs or ? 18:34:53 fetch branches from github, merge into git.tpo repo 18:35:07 sync commits back to github 18:35:17 right; I guess “manually” isn’t quite the right word 18:36:06 would gitlab work as well? 18:36:54 i assume a similar flow/process is possible, but i don't have any experience with it 18:37:21 (i also don't have any experience with this github process, but at least we're not the first team trying this with GH :) ) 18:37:22 I find GitHub to be more smooth but that may also be because I use GitLab very rarely 18:37:24 Probably. Not quite as slick of course; but my main desires are 1) Being able to view diffs in a pretty format and leave comments right there 2) Being able to submit a pull request from my branch in a web ui and 3) being able to edit files in my branch in-browser without needing to edit it locally 18:38:08 tjr: how do you use 3) ? 18:38:16 or, what do you use that for? 18:38:37 do we know why the oniongit experiement wasn't successful? 18:38:56 mcs: Exhibits a) and b) https://github.com/tomrittervg/tor-browser-build/commit/db075a1a936b8a4432f56bba3b69cde4bdc8e315 https://github.com/tomrittervg/tor-browser-build/commit/f374c28cb89fd091a84a17fcacf6e81e662cf957 18:39:04 i wonder if anyone on the network-team still uses oniongi 18:39:09 t 18:39:20 sysrqb: for me it was not successful because the workflow was awkward 18:39:29 that's not necessarily an oniongit issue 18:39:36 yeah 18:39:38 tjr: got it; thanks. So: docs or simple chnages ) 18:39:41 changes 18:39:43 but was very likely more an issue with the trac + oniongit thing 18:39:47 sysrqb: i dont think so 18:40:00 asn: kk, thanks :) 18:40:13 asn: hey, so anything regarding how the network-team is using github question? 18:40:22 are you happy with it? 18:40:27 GeKo: i havent found this question in the backlog 18:40:39 personally im happy with it 18:40:47 i havent used it much with TB i admit 18:40:53 18:35 <+mcs> How does the flow work for tor (Network Team)? Do they mirror to github 18:40:54 and manually merge PRs or ? 18:41:02 i know that dgoulet uses it with TB and it seems to work fine, but he is afraid that microsoft is gonna block tor at some point 18:41:18 eeehm 18:41:32 we keep the upstream on git.torproject.org 18:41:36 yes, that's an issue which worries me a bit as well 18:41:41 and we have this gh https://github.com/torproject/tor 18:41:44 (the one dgoulet has) 18:41:54 which gets somehow updated (maybe using a bot/git hook of some sorts) 18:42:09 from the git.tpo upstream 18:42:26 we use PRs just for review purposes 18:42:38 and they get closed automatically when the branch gets merged upstream 18:42:53 e.g. https://github.com/torproject/tor/pull/227 18:43:37 so basically the gh repo for tor is just there to be able to fork off it, and to be able to submit PRs 18:43:48 but it's not used as upstream in any capacity 18:43:51 and for doing CI 18:43:52 EOF 18:43:55 ? 18:43:59 yeah that too 18:44:07 and we all get CI for our personal branches too 18:44:15 i see 18:44:53 okay, i think that's some food for thought 18:44:54 right. i guess one major negative of this process is we could lose all review comments if we lose access to the torproject github account 18:45:03 yep 18:45:23 but worth considering 18:47:18 okay, i'll keep thinking about it a bit but plan to reply to the tbb-dev mail later this week 18:47:27 summarizing where we are at right now 18:47:52 feel free to chime in 18:48:44 i think we wait for arthur about the storm question as he brought it up 18:48:54 just a reminder for folks to think about it 18:49:14 but i guess we can make a decision or a plan next week when everyone is onboard again 18:49:23 so the sandboxing meeting 18:49:41 i plan to read over the thread on tbb-dev as preparation 18:49:51 sysrqb: is there anything else we could/should do? 18:50:19 i guess the agenda is starting to come up with an actionable plan for a sandboxed tor browser :) 18:50:34 and figuring out the pieces involved with it and what mozilla could help with 18:50:50 at least that's what i think would be worthwhile to do 18:50:56 yes, that is my goald for the meeting 18:51:11 1) what did we miss/what other mechanisms are available for us 18:51:32 tjr: do we have some mozilla folks that can make it? 18:51:39 2) how do we start designing/implementing this 18:52:43 I can follow up on the last email and send some high-level information and pointers, if that'll be helpful 18:52:53 i think it would, thanks 18:52:59 tor-dev@ announcement 18:53:00 okay 18:53:02 Hm. I had pointed the thread out to the sandboxing team last week and they didn't have much to say about it (especially after I reminded them Tor was interested in sandboxing the parent process). I had planned to report back after the meeting to them with what we discussed and see if they had any thoughts 18:53:29 well, that sounds fine 18:54:05 okay, do we have anything else for today? 18:54:20 it seems we are reaching our 60min... 18:54:32 I feel like there's a few paths forward we can pursue; and that we could use this meeting to enumerate them and then try to decide (not in the meeting) which is the overall best approach 18:54:45 yup 18:55:25 e.g. "We're gonne run Docker on every platform", "We're gonna use a container format that's the best option for each platform", "We're not going to use a container but we are going to sandbox the parent process using whatever platform-specific mechanisms are available to us..." 18:55:56 right. pros/cons for all of them 18:56:14 especially thinking about the requirements for tor browser 18:57:48 alright, let's wrap up then. thanks for the meeting everyone *baf* 18:57:51 #endmeeting