18:00:05 <mikeperry> #startmeeting tbb-dev
18:00:05 <MeetBot> Meeting started Mon Jun  8 18:00:05 2015 UTC.  The chair is mikeperry. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:05 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:00:23 <mikeperry> hello everyone, let's get started
18:00:56 <mikeperry> I will kick things off with my reportback first
18:01:03 <mikeperry> Last week, my cycles were mostly spent attending a meeting Weds-Sat. I matched GeKo's builds for 4.5.2 and 5.0a2, and wrote our status report, and updated the ticket list for June.
18:01:09 <mikeperry> This week, my plan is to help review tickets in https://trac.torproject.org/projects/tor/query?status=!closed&keywords=~TorBrowserTeam201506R, and also follow up on a bunch of company organization things from last week's meeting. I will also be working with Isabela to get deliverables set up for everyone for Q3+Q4 contracts, so keep your eyes peeled for that email.
18:01:18 <mikeperry> I know I really need to devote some more attention to FF38ESR things (prioritize tickets in the ff38-esr tag, do some networking+code review, etc), but I suspect followup from last weeks meeting and this upcoming release will consume a lot of time.. Please be patient, and if anything you're doing is blocking on my input, please bring it to my attention via IRC and/or email.
18:01:30 <mikeperry> that's it for me for now
18:02:12 <mikeperry> who is awake and ready to go next? ;)
18:02:22 * mcs can go next
18:02:27 <mcs> Last week, we helped fix #16014 and #16269.
18:02:32 <mcs> Thanks to GeKo and dcf for merging the patches so they could be included in TB 5.0a2.
18:02:37 <mcs> We created a fix for #16268.
18:02:42 <mcs> We talked to mrphs about fixing #13512.
18:02:49 <mcs> We also did several code reviews (mostly TB 5.0a2 and ESR 38 things).
18:02:52 <mcs> Finally, we made a pass through the Firefox release notes for #16090.
18:02:59 <mcs> This week we work on confirming that various Tor Browser patches are working correctly with a TB built from Arthur's ESR 38 branch, e.g., #12827, #13548, #13900.
18:03:08 <mcs> And we will need push our #15145 changes so they can be picked up in a future 5.0 alpha release.
18:03:14 <mcs> That's all for now.
18:04:03 <mcs> Who else is awake?
18:05:11 <GeKo> ha
18:05:24 <GeKo> here is what I did:
18:06:20 <GeKo> I worked on the two releases, on #16181, #9711 and looked over all the mozillaXX bugs
18:06:31 <GeKo> then I worked on #16011
18:07:25 <GeKo> I looked a bit at #16285, helped with #16253 filed #16300
18:07:51 <GeKo> and thought a bit about Mozilla's tracking protection feature
18:08:16 <GeKo> I think we are good wrt to the mozillaXX bugs, I did not find anything new
18:08:52 <GeKo> althought this whole stuff was kind of depressing given the ton of work we need to to to fix things like the broadcastchannel api
18:08:55 <GeKo> :(
18:09:30 <GeKo> this week I'll work more on the things I tagged with my working tag
18:09:37 <GeKo> that's it for now
18:09:52 <GeKo> oh
18:09:54 <mcs> Yes, there are a lot of bug complex features that are being added without much thought about tracking or fingerprinting.
18:10:31 <GeKo> and I think I'll look over the FirefoxXX tickest as well as we already have some "features" we need to look at
18:10:51 <GeKo> whcih are in this category
18:10:54 <GeKo> *which
18:11:09 <GeKo> ok, that's it now for realz for me
18:12:11 * arthuredelstein can go
18:12:45 <arthuredelstein> This past week I worked on https://trac.torproject.org/projects/tor/ticket/16200, and wrote
18:12:49 <arthuredelstein> a patch to fix cache clearing on New Identity.
18:13:00 <arthuredelstein> I finally landed https://bugzilla.mozilla.org/show_bug.cgi?id=418986 (at least to mozilla-inbound).
18:13:06 <GeKo> \o/
18:13:07 <arthuredelstein> I tried keep the ESR38 branch up to date with TBB5.0a/ESR31
18:13:18 <arthuredelstein> And I started doing some investigation into #13313.
18:13:28 <arthuredelstein> This week I'll keep working on the font problem, and look for more things that need to be fixed in torbutton.
18:13:56 <GeKo> arthuredelstein: you should definitely talk to dcf as he has already invested a lot of work into #13313
18:14:12 <arthuredelstein> And probably look for more fingerprinting patches to upstream to mozilla.
18:14:24 <arthuredelstein> GeKo: Yes, I'll definitely do that.
18:15:09 <arthuredelstein> I'm still trying to figure out whether I should focus on that approach, or just try to write stopgap patch that limits fonts as before
18:15:52 <GeKo> I am in favor of the latter given all the other things that we really need to take care of
18:16:15 <GeKo> and #13313 is nothing we can fix in a week or so
18:16:16 <arthuredelstein> That's good to hear. I'm leaning that way as well.
18:16:33 <mikeperry> or, simply table any font work for now, and just focus on other ff38 issues
18:16:49 <GeKo> yes, I think that might be even better
18:16:51 <mikeperry> there are plenty to choose from :)
18:17:21 <mcs> Do all of the ESR 38 things have tickets at this point or do more need to be created based on #16090?
18:17:32 <arthuredelstein> OK, I'll open a ticket to remind us that the font issue needs to be solved for TBB/ESR38.
18:17:41 <mcs> (it would be good to have tickets for each item so we can divide them up / pick things to work on)
18:18:05 <GeKo> mcs: I've noted on that bug if there are tickets for things mentioned there
18:18:25 <GeKo> thus, we still need to file a couple
18:18:43 <mcs> GeKo: OK.  Maybe you or mikeperry can do that?
18:19:20 <mcs> Or maybe some items need more research too.
18:19:33 <GeKo> I can do that tomorrow
18:19:52 <mikeperry> I will do some today if I can, and update #16090 with any that I do file
18:19:53 <mcs> OK; thx
18:19:56 <GeKo> might be a good time after I finished looking at FirefoxXX tickets
18:20:01 <GeKo> ok
18:20:19 <mikeperry> and GeKo can handle the remainder I guess?
18:20:31 <mikeperry> what are these FirefoxXX tickets? you mean in Mozilla's bugtracker?
18:20:38 <mikeperry> (and also the MozillaXX tickets)
18:21:07 <GeKo> milestone=mozilla32-38
18:21:14 <GeKo> milestone=Firefox32-38
18:21:45 <mikeperry> ok, in mozilla's bugtracker that you linked in #16090. got it
18:22:44 <GeKo> yes
18:23:40 <mikeperry> ok. do we have a boklm?
18:23:49 <boklm> yes
18:23:58 <boklm> Last week I splitted the patches from tb_GECKO380esr_2015050513_RELBRANCH+1 into different topic branches: https://lists.torproject.org/pipermail/tbb-dev/2015-June/000282.html
18:24:05 <boklm> I submitted all branches to Mozilla Try (and added a command in the git helper script to do that)
18:24:14 <boklm> I added a command to fetch build/test results from mozilla treeherder's json for all branches
18:24:22 <boklm> And I started adding patches to disable tests that are expected to fail
18:24:32 <boklm> This week I'm planning to review more test results to find which tests we should disable or fix
18:24:47 <boklm> that's all for me
18:25:23 <GeKo> is there anything we can do against the false positives we often get when runing our tests suite?
18:25:40 <GeKo> see the 5.0a2 results as example
18:27:13 <boklm> It seems we have a problem with the acid3 test page sometimes unreachable with Tor. I will look at hosting these files ourself to avoid that.
18:27:58 <GeKo> ok
18:28:48 <boklm> for the dom-objects-enumeration false positive, I will try to find what is happenning
18:28:51 <mikeperry> I am also wary of the known issues/warnings wrt hardening. those are mostly due to the go binaries, yes? if suddenly we have regressions on additional binaries, do those known issue/warnings get elevated to full failures?
18:29:03 <mikeperry> (for PIE, RELRO, stack canary)
18:32:41 <boklm> for PIE we have a whitelist of the go files. For RELRO, stack_canary, no_rpath I should change it to a whitelist of files too (it's not the case currently).
18:33:35 <boklm> I will look at this this week
18:33:37 <mikeperry> ok. please send the whitelist to tbb-dev or somewhere, esp if it is a larger set than just the go binaries already
18:33:40 <mikeperry> ok
18:33:57 <mikeperry> thanks
18:35:00 <mikeperry> GeKo: so wrt 4.5.2/5.0a2, what were you thinking in terms of release date? I saw the changelogs initially said this Friday. was that just a precaution in case we didn't match, or do you have other reasons to wait?
18:35:40 <GeKo> oh, I seemed to have messed up the dates
18:36:11 <GeKo> what I *wanted* to write was "June 9"
18:37:23 <mikeperry> ok, so I should add blog post writing and file relocating to my TODO list for today, and get everything in place for your tomorrow?
18:37:38 <mikeperry> did you upload signed exes for both releases already?
18:38:03 <GeKo> you can do all this stuff, yes.
18:38:14 <GeKo> if you don't get to it, I'll take care of it tomorrow
18:38:29 <GeKo> all the signatures for 4.5.2 are uploaded, 5.0a2 is following shortly
18:39:21 <mikeperry> ok. I will at least do the blog posts then. worst case all you should need to do is copy some files into place
18:39:34 <GeKo> ok
18:42:12 <mikeperry> ok, anything else then? it sounds like my top priority for moving ff38-esr forward is ensuring that tickets are file for #16090 things, and prioritizing ff38-esr tickets. am I blocking anything else for anyone, or is anyone otherwise stuck/unsure what to do in the meantime?
18:45:22 <mikeperry> ok, I am optimistically taking silence as a "no, all good"
18:46:16 <mikeperry> we may actually be done in under an hour today. hurray! thanks for pre-writing your updates and otherwise making everything go smoothly, everyone!
18:47:04 * mikeperry winds up for the *baf*
18:47:09 <mikeperry> #endmeeting *baf*