19:00:14 #startmeeting tor browser 19:00:14 Meeting started Mon Nov 28 19:00:14 2016 UTC. The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:14 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:00:20 hi all! 19:00:26 Hi! 19:00:32 let's get started with our weekly status updates 19:00:34 hi everyone! 19:00:41 who wants to go first today? 19:01:17 herro 19:01:48 hi! 19:01:59 so it seems it falls back to me again. then be it so 19:02:03 here is what i did: 19:03:07 i worked mainly on #20758 and #20352 and started looking into why /proc is needed in firefox 19:03:36 i plan to spend a bit more time on the latter this week and then turn my attention slowly towards the next release 19:04:11 there is number of smaller things that we might want to get fixed and we only have a bit more than a week 19:04:35 * isabela is around 19:04:37 if there is time left i hope to get back on track with the update for the torbrowser design doc 19:04:40 that's it for me 19:05:16 * arthuredelstein can go 19:05:27 This week I worked on #10281. I have a patch pretty much ready and am testing and hope to post today. 19:05:38 woot! 19:05:51 There is also a second approach I have a patch almost ready for as well. 19:06:02 That's what I will be doing this week, until Thursday when I start jury duty :( 19:06:07 amazing, that's good news! 19:06:17 :( 19:06:22 or should i say :) 19:06:28 ? 19:06:53 Frowny face re jury duty for sure 19:07:11 But we'll see -- could be over quick 19:07:28 Jury duty is awesome. No snark. 19:08:44 Takes me away from the job I love, though! 19:08:58 arthuredelstein: that was 6 weeks? 19:09:24 GeKo: Only if I am selected for the jury 19:09:42 Otherwise just a few days at most I guess. 19:09:48 aha! okay. 19:10:11 Though I have little clue at this point. 19:10:14 Wear a Tor tshirt to the courthouse? 19:10:28 arthurdelstein: good luck with that and let me know if we/i can do something for you 19:10:35 err arthuredelstein ^ 19:10:42 Believe it or not they have a dress code so I can't! 19:10:59 Yes, that is true around here too (dress code). 19:11:00 mcs: hehe 19:11:41 * mcs can give a report 19:11:48 Last week was a short week for Kathy and me due to the U.S. Thanksgiving holiday. 19:11:53 We finished looking at the sandboxed-tor-browser code for now (after we concluded that an OSX port of the launcher and control port filter parts will be a moderately large project). 19:11:59 We took a fresh look at #9675; there may be something we can do soon to improve error reporting in the face of Tor bootstrap failures. 19:12:05 We also helped with bug triage and code reviews. 19:12:10 This week we will work on changes to package the OSX Seatbelt profiles for the next Tor Browser alpha release (#20121). 19:12:14 If we finish that, we will work on #9675 and maybe #20083. 19:12:20 That’s all for us. 19:13:01 * boklm can go next 19:13:04 mcs: i wonder if you could prioritize #20761 19:13:12 instead of the latter two 19:14:07 GeKo: Yes, I forgot about that one. 19:14:09 i'd like to get the unix domain socket feature finished as good as we can i think and this issue is indeed a drawback 19:14:12 ok, cool. 19:14:35 This past week I helped with the Tor Messenger release, investigated the VM shutdown problem on #20352, fixed #20626 and started on #17380 19:14:44 This week I'm planning to review #20352, get a new release of the testsuite bundle fixing some issues, and continue on #17380 19:14:51 That's it for me 19:15:30 mcs: re #20083 19:15:33 I have a better idea 19:15:50 I think patching tor browser and the sandbox agent so that it works is better 19:16:24 (should I go?) 19:16:28 sure 19:16:34 I did the same thing I do every week, I worked on the sandbox 19:16:43 Yawning: OK, but that sounds harder ;) 19:16:49 poking at getting seccomp whitelist going 19:16:51 mcs: nah 19:17:14 "if it's in the sandbox, instead of running the update process, send a command to the agent over the tor control port" 19:17:27 "to have the agent check for the update" 19:17:49 Yawning: that makes sense. 19:17:52 since sandboxed firefox talks to my code, that it thinks is the control port 19:17:58 and not directly to tor 19:18:24 and I want the agent to do the update check in the background for 0.0.2 19:18:33 gluing those two together should be easy 19:18:58 also my most sincere appologies for the sandboxed-tor-browser code 19:19:04 it's... kind of complicated :( 19:19:39 fine with me 19:20:27 does that approach make sense? 19:20:41 sounds good to me and would be neat to have i think 19:20:54 was thinking the toggle for the behavior could be like TBB_SANDBOX=linux or whatever 19:21:22 I think the bds people probably want an option to hide the chrome though 19:21:33 and all the buttons 19:22:46 Maybe keep #20083 for that purpose and create new tickets for the Linux sandbox behavior we want? 19:22:59 yah 19:23:30 excellent this appears to work 19:23:42 sorry, adding an tor + obfs4 seccomp whitelist 19:24:02 so, when/how are we going to unleash this onto the world? 19:24:22 in about two weeks if all goes well 19:24:28 kk 19:24:36 I'll probably tag sometime this week 19:24:44 i think we can get the stuff into nightlies builds this week 19:24:50 so I can start working on the giant list of tickets of 0.0.2 improvements 19:25:05 yes, tagging sometime this week is totally fine 19:25:35 #20793 is a packaging thing 19:25:57 dunno where that should live 19:26:03 probably somewhere in the builder? 19:26:14 tor-browser-bundle 19:26:18 probably 19:26:33 that's Application/Tor Browser with a tbb-gitian tag nowadays 19:26:38 I also plan to write better documentation 19:26:51 okay. anybody else here for a status update? 19:27:19 Sure 19:28:06 o/ 19:28:12 hi! go ahead 19:28:19 I'm still working on the front-end of the Orfox slider 19:29:46 I have managed to determine a couple of routes we can take with it, I'm doubting if it's a good idea to go with some of the same techniques currently used by TorButton 19:30:30 Since XUL is now deprecated and all 19:31:29 yes 19:32:32 Apart from that, I have pushed the most promising prototype into a repository, and have some preview material and a basic UI implementation 19:32:46 neat 19:33:19 I could share some more details in the mobile channel, after the meeting 19:33:36 That's about it 19:33:41 i wonder whether we have a ticket in trac for that 19:33:50 it does not seems so. hm. 19:34:13 Synzvato: thanks for the preview material o/ 19:34:15 maybe i should create one and we can dump all the links etc. there and get a discussion going 19:34:36 i can do that - please continue the meeting and we can catch up on tor-mobile after the meeting 19:34:39 sounds good? 19:35:04 okay 19:35:09 That's true, would be great 19:35:39 ok 19:35:40 alright anybody else for a status update? 19:35:45 isabela: Yw :) 19:35:48 Sounds good 19:36:52 okay discussion time 19:37:09 there, obfs4proxy now uses a whitelist 19:37:14 we had a bit of discussion regarding the sandbox already 19:37:18 what else do we have? 19:37:28 (sorry) 19:37:34 np 19:37:58 * isabela wonders about the ooni tor browser discussion 19:38:09 i plan to think about work for the weeks and we may be able to have that discussion next week 19:38:16 *for the next weeks 19:38:42 i guess the next big step is getting 6.5 out end of january 19:39:24 so everyone is welcomed to think about that (what should get in, what should not get in etc.) 19:39:36 so that we can do a proper planning for it 19:39:52 apart from that ESR52 work is looming i guess 19:40:26 and i am particularly afraid of getting this for windows going 19:40:53 Yes, I was thinking of starting on the rebase once 10281 is done (time permitting) 19:41:01 i took me checking the status up the code wrt mingw-w64 as a weekend hobby for the past weeks 19:41:15 and it does not look good this time 19:41:35 mainly as jacek did not fix any of the tricky bugs as the tree was broken for him 19:41:39 for several months 19:41:57 so we might have a really hard time this time 19:42:13 :/ 19:42:34 anyway, yes, arthuredelstein, that might be a good idea 19:42:55 i'd like to have a rebased tor browser branch earlier this time 19:43:00 I remember someone mentioning trying to get cross-platform builds into Mozilla's try server system. Do you know if anyone is working on that? 19:43:31 GeKo: Yes, I agree we should try to rebase as early as possible. 19:43:34 we could share the workload if that's easier, like you are doing the upper half and i am doing the lower half of our patches or something 19:43:54 but i guess we might do fine this time due to all the patches that got upstreamed 19:44:13 arthuredelstein: re mingw-w64 builds on try, yes, that's in the works 19:44:28 i put it on rbarnes' list as a high prio item for us 19:44:30 It might make sense for Arthur to take the lead and then hand off difficult patches to other people. 19:44:35 GeKo: Aha, great. 19:44:47 mcs: yes, something like that 19:45:20 That sounds like a good plan to me. 19:45:37 arthuredelstein: but i think that would not mean mozilla is fixing all the issues themselves just that we would notice earlier 19:45:49 but it would be a start 19:46:09 I guess if it is considered a first-class platform, then maybe it would deter breakage? 19:46:23 yes, but that is a long way to go 19:46:28 I see. 19:47:09 anything else for today? 19:47:36 Something else we might consider with 52 19:47:50 is trying to start including the Orfox patches 19:48:01 once we have rebased the desktop patches, of course 19:48:14 yes, please, wehave #19675 19:48:40 and i am basically waiting for amoghbl1 or somebody else giving me a patchset or something 19:50:22 okay, thanks for the meeting then and good luck arthur *baf*! 19:50:25 #endmeeting