18:01:28 #startmeeting tbb-dev 18:01:28 Meeting started Mon May 18 18:01:28 2015 UTC. The chair is mikeperry. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:28 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:02:50 ok, let's get started 18:03:25 Last week, I wrote the blog posts for 4.5.1 and 5.0a1, commented on https://github.com/w3ctag/security-questionnaire/issues/7#issuecomment-101040335, spoke with some researchers about #1517, and sent a bunch of mail. 18:04:16 I think the releases went mostly smoothly, but I still need to review the latest batches of blog comments 18:04:29 This week, I swear I'm going to do some roadmapping, file a ticket to improve the #1517 defense, plan a bunch of travel, and start reviewing Firefox developer documents. 18:05:30 I think that's it for me. we can talk more about ff38 plans after updates 18:05:56 * mcs can go next 18:06:03 Last week, Kathy and I finished rebasing and testing our updater patches for ESR38 except for the MAR signing part (we will work on those next). 18:06:09 We also posted patches for #15984 and #16014 (reviews appreciated). 18:06:17 We did a code review for #13670. 18:06:24 We triaged some bugs including #16028 and #16029. 18:06:35 Later today we will make our ESR38 patches available to the team. 18:06:40 For the remainder of this week, we will continue with the ESR38 updater rebase work (MAR signing), 18:06:45 we will finish testing #16014, 18:06:49 and we will help with whatever else comes up. 18:06:54 Also, we will probably miss next week's meeting due to the U.S. Memorial Day holiday. 18:06:59 That's all for us. 18:07:10 oh right, I need to look at those update stats for you 18:07:27 Something else to add to your list. 18:07:31 I saw a lot of 206 partial downloads in the logs for #16028 18:08:33 If the update service is downloading in the background, it uses range-based downloads which means potentially lots of 206 responses (downloading in chunks) 18:08:51 So that could be normal. 18:10:07 here is what I did: 18:11:04 I helped with the releases and found soon after/just before releasing issues with the windows signing 18:11:29 I've filed #16011 be better prepared 18:11:33 for the future 18:11:46 then I worked on #15990 18:12:04 and there should be no blockers anymore for Windows and ESR 38 18:12:09 and for Linux neither 18:12:27 I'll work this week on the most fun part the OS X transition 18:13:37 additionally,last week I read some papers regarding tracking protection/browser hardening 18:14:22 I've seen there is a bunch of tickets that got opened in the wake of our rleases and I think I'll look over them + the blog comments 18:14:43 then I plan to do some reviews and we'll see what else comes up 18:14:52 that's it for me 18:16:16 * arthuredelstein can go 18:16:28 Last week I mainly worked on rebasing tor-browser.git patches to ESR38 (#15196). I have 42 patches rebased, and about 25 patches to go (not counting the updater-related patches Mark and Kathy are rebasing). 18:16:46 Also I posted a newly revised patch for bugzilla.mozilla.org/show_bug.cgi?id=418986 and I started looking into possible C++ patches for #14429. 18:17:03 This week I'll continue the rebasing task. 18:17:10 That's all for me! 18:18:44 I think boklm is not coming this week 18:18:56 (at least he indicated that last week) 18:19:33 ok. so I think we'll also move the meeting to Tuesday next week, because of the US holiday 18:20:32 fine with me 18:21:01 Tuesday for the meeting is fine for us. 18:21:16 me too 18:21:30 ok, great 18:22:22 mcs: do you have a sense yet how bad the updater and app dir layout changes are for osx? 18:23:07 (due to https://bugzilla.mozilla.org/show_bug.cgi?id=1047584) 18:23:23 We had to make some additional changes to the updater patches but things are working "in the lab" now. 18:23:59 The dir layour changes are extensive though. 18:24:04 layout changes 18:24:41 any idea if it will mess with our embedded profile + tor binary? 18:24:42 I think we will have to look at our dir layout more if we try to get Gatekeeper OS X signing working for Tor Browser. 18:24:54 I suppose we may not know that until we try to sign for Gatekeeper.. 18:25:31 I am not an expert, but Apple wants binaries only in certain subdirectories and other constraining things. 18:26:08 So Mozilla left only libraries and exes under Contents/MacOS and they moved everything else under Contents/Resources. 18:26:11 mikeperry: btw how is the planmoving forward on the gatekeeper certs we need from apple? 18:26:14 rehi friends 18:26:27 We of course have a lot of things scattered around. 18:27:33 GeKo: I was just wondering that.. I am not sure whose job that is. it is probably now either armadev+nickm, or mine. 18:27:49 mikeperry: it might be smart as a first step to have somebody with legal knowlegde look over the agreements we would need to sign 18:28:12 yeah 18:29:28 ok, I will also look into that this week 18:31:12 anything else? 18:31:47 Can we talk about the esr38 work a little? 18:32:02 (after satus reports) 18:32:08 sure 18:34:23 yeah, I imagine my review of the developer docs will cause a bunch of new tickets to be filed under https://trac.torproject.org/projects/tor/query?status=!closed&keywords=~ff38-esr 18:34:37 (*bangs head on keyboard* #16088) 18:36:01 it would be nice if others could also skim the developer docs, because I missed SharedWorkers last time 18:36:12 kernelcorn: Did you have any luck figuring out what was up with the stem installation issue? 18:37:01 mikeperry: yeah you asked me and I'll do it, too. 18:37:25 atagar: not yet, Python2 and Python3 still can't seem to find or load Stem for some reason. I tried "sudo pip install stem" and then set up virtualenvs and tried "pip install stem", same thing 18:37:40 hey mikeperry! amoghbl1 has an android build based on Tor's gecko-dev build now, but it isn't functional yet. we are working through it, but just wanted to give you a heads up he might have questions 18:37:41 under Python3, from stem.control import Controller 18:37:41 ImportError: No module named 'stem.control'; 'stem' is not a package 18:37:42 I agree that we should all take a look at the dev docs. Maybe a couple of people can make a first pass and then we can double check to make sure nothing was missed. 18:38:04 They used to provide a link to undocumented bugs right on those pages.. 18:40:04 n8fr8: just curious, why are you using gecko-dev and not, say esr38? 18:40:22 Dumq question: What's the URL for the dev docs? 18:41:35 arthuredelstein: https://developer.mozilla.org/en-US/Firefox/Releases/18 18:41:46 and you need to increment the version number accordingly 18:41:52 GeKo: thanks 18:42:30 DId Mozilla create an esr38 branch on gecko-dev yet? I do not see it. 18:42:51 #16090 for the developer doc review 18:43:15 GeKo: what i mean is we are building the android version of the repo from here: https://gitweb.torproject.org/tor-browser.git/ 18:43:18 mcs: hrmm.. it looks like not. hwine said he would, but I did only ask in IRC 18:43:23 all of this mozilla terminology still befuddles me... 18:43:38 i will get it straight soon 18:43:53 kernelcorn: If you provide '--verbose' to pip does it tell you where the files are being installed to? 18:44:01 before, we were building Orfox from Mozilla's repo, which some of the configuration and patches from tor-browser 18:44:12 but now we are building from the tor-browser repo directly 18:44:21 I figured it out atagar. I had the file named "stem.py". Apparently that's not a good idea. 18:44:28 ahhhhh! 18:44:35 yeah, that would do it :P 18:45:04 n8fr8: well, to put it another way: why using the developer branch which likeley include bugs/breakage for Tor Browser and not a stable one, like mozilla-release or something 18:45:20 *likely includes 18:47:06 well, we want to base ORfox on the changes made for Tor Browser, so I assume then we should work from a stable tag and not master/head 18:48:17 ah, ok that's fine I thought you were using mozilla's master branch 18:48:37 tor-browser.git is good :) 18:49:08 I'm currently using the tor-browser-31.7.0esr-5.0-1 18:49:17 branch of the tor-browser repo 18:50:28 what does "it' not functional yet" mean? 18:50:53 There seem to be some problems with the built apk that we're getting 18:51:15 we also have to change from using an internal/built-in tor instance, to relying on Orbot 18:51:23 interesting. so the build crashes or does not even start? 18:52:07 Oh, the build runs and creates the apk, but the apk doesn't run fine... 18:52:47 ok 18:52:49 armadev: Would you mind using your anti-spammer thing to scramble this guy's trac account? Yawning's been playing whack-a-mole with him but it's probably about time we disabled his login... 18:52:54 https://trac.torproject.org/projects/tor/ticket/16091 18:54:19 amoghbl1: you could try testing the branch without all the patches first to see if it is a patch issue 18:54:20 mcs: https://bugzil.la/1164217 is our bug for esr38 branch. hwine says it should happen tomorrowish, when canada comes back from holiday 18:54:33 mikeperry: Sounds good. 18:54:46 if things are still not working and you don't know why writing to tbb-dev might be a good idea 18:54:48 atagar: answered on #tor-project so we don't interrupt this meeting more than we have 18:54:51 amoghbl1: ^ 18:55:07 or you need to ping folks here 18:55:19 GeKo: i think we can build from mozilla's stock code, but amoghbl1 i agree it would be good to ensure that again 18:56:42 if mozilla's stock code works happy bisecting :) 18:58:35 ok. anything else pressing for FF38esr or other issues? 18:59:22 not from me. 19:01:02 alright, that should wrap it up for today. see you at the same time on Tuesday next week! 19:01:20 #endmeeting *baf*