18:01:47 <GeKo> #startmeeting tor browser
18:01:47 <MeetBot> Meeting started Tue Jul 12 18:01:47 2016 UTC.  The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:01:47 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:01:53 <GeKo> hi!
18:02:15 <GeKo> welcome to another tor browser meeting on a somewhat unusal date
18:02:22 <GeKo> *unusual
18:02:39 <mcs> hi
18:02:40 <boklm> hi
18:02:53 <GeKo> arthur said he is probably late today. let's get started anyway.
18:03:18 <GeKo> last week was short for me as i took some days off
18:03:47 <GeKo> but i finished #16652 and wrote a small patch for #19585
18:04:35 <GeKo> moreover i looked at #19200 + #18923 and thought a bit about #18762
18:05:01 <GeKo> + i was busy with my backlog today
18:05:47 <GeKo> this (short) week i plan to work on #19200, #19417 and the design documentation
18:05:52 <GeKo> and doing reviews
18:05:59 <GeKo> that's it for me
18:06:39 * mcs will give a report
18:06:45 <mcs> Last week, Kathy and I worked on #19532 (not resolved yet).
18:06:51 <mcs> We also investigated some blog comments and ended up creating #19646, which we did some work on.
18:06:57 <mcs> We also responded to arthuredelstein in #19273 (reminder: the patch for that ticket needs a second review) and we reviewed the fix for #19585.
18:07:11 <mcs> Finally, we commented here: https://bugzilla.mozilla.org/show_bug.cgi?id=1173199#c58 (Mozilla bug for upstreaming the mathml.disabled patch).
18:07:17 <mcs> This week we plan to work more on #19532, #19646, and #19269.
18:07:29 <mcs> Time permitting, we will investigate some TB 6.x updater issues on Windows (reportedly, the MAR file is left on disk and the update history is not shown in the browser).
18:07:38 <mcs> That’s all for now.
18:08:03 <GeKo> mcs: i'll look at your patch this week.
18:08:22 <mcs> GeKo: I thought you might when you mentioned reviews. Thanks!
18:11:17 * boklm will go next
18:11:55 <boklm> This past week I tested a build with the #19528 changes. I made a new version of #18923 running the tests added or modified by one of our patches.
18:12:07 <boklm> On #19410 I added a script to convert the signed dmg files to mar files.
18:12:24 <boklm> This week I'm planning to update #19528 to address the comments from mcs. On #19410 I will add the removal of incremental OSX mar files to regenerate them.
18:12:41 <boklm> That's it for me.
18:12:50 <GeKo> good stuff
18:13:19 <GeKo> fwiw: i am currently testing #18923 which is one of the things on my review list for this week as well
18:13:51 <boklm> ok
18:15:21 <GeKo> i wonder what we should do in case of test starting to fail in general
18:15:26 <GeKo> *tests
18:15:47 <GeKo> the obvious thing often done is just disabling that one.
18:16:09 <GeKo> should we have a procedure in place dealing with it differently?
18:16:28 <GeKo> the one thing i fixed was trivial but what about the two other ones?
18:16:54 <GeKo> how should we integrate that into our dev process?
18:18:00 <boklm> yes, I think it might be good to disable them, unless we can fix them soon
18:18:29 <mcs> Has anyone looked at why tests are failing? (I have not)
18:18:45 <mcs> If the tests are valuable, we should fix them :)
18:19:12 <GeKo> boklm had an idea about why one was failing but no one looked at the other one yet afaict
18:19:50 <GeKo> ok. i think step one should be: as soon as tests are failing someone for the team should look at the reason for it
18:20:21 <GeKo> then we can discuss the benefits/risk of just disabling/fixing it
18:20:52 <GeKo> i think i can put that on my plate for the two things boklm found
18:21:40 <GeKo> in general i am a fan of fixing those issues quickly as the failing tests might hi deeper/other issues
18:21:59 <mcs> Agreed
18:22:04 <boklm> yes
18:22:27 <arthuredelstein> I'm here now -- I can also look at these tests if it would be useful.
18:22:42 <GeKo> hi arthur
18:23:49 <GeKo> ok, thanks. let me put it on my plate this time
18:24:06 <GeKo> i guess there will be other ocurrances in the future :)
18:24:07 <boklm> after all tests are passing and #15994 is done, I am thinking about monitoring the tests logs to find test failures sooner
18:25:17 <GeKo> boklm: oh, speaking of monitoring: do the os x tests not run on our nightlies?
18:25:29 <boklm> ah, I will look at that
18:25:59 <GeKo> ok. i was just wondering as i only see cygwin and fedora results
18:26:45 <GeKo> anyone else here for the status update?
18:26:56 <arthuredelstein> I can give one now
18:27:11 <arthuredelstein> Over the past week, I worked on a hopefully final version for
18:27:15 <arthuredelstein> https://bugzilla.mozilla.org/show_bug.cgi?id=1121643 (not done yet)
18:27:20 <arthuredelstein> I'm hopeful that patch will also fix #18860.
18:27:32 <arthuredelstein> I also worked on https://bugzilla.mozilla.org/show_bug.cgi?id=1235520, which is about rebasing brade and mcs's patch from #6786.
18:27:40 <arthuredelstein> I also did some investigation of #10281 — I was able to build Tor Browser using jemalloc enabled with no problem.
18:27:51 <arthuredelstein> The idea of isolating strings and other buffers seems nontrivial, however.
18:28:06 <arthuredelstein> It's something I will try to discuss with the Mozilla folks and see what their plan is.
18:28:15 <GeKo> cool. yeah that is kind of expected ;)
18:28:31 <GeKo> good idea. can you cc me?
18:28:32 <arthuredelstein> This week I will finish up the bugzilla patches I mentioned, help with reviews needed and have a look at how we could simplify #14429.
18:28:42 <arthuredelstein> GeKo: Yes, will do
18:28:59 <arthuredelstein> That's it for me
18:30:16 <GeKo> if you could give #19484 a quick look that would be neat
18:30:33 <arthuredelstein> Sure
18:31:06 <arthuredelstein> Speaking of the regression tests, the #17809 test is perhaps not so useful
18:31:24 <arthuredelstein> I guess it seemed to me like a good idea at the time, but I wonder if it's more trouble than it's worth.
18:32:30 <arthuredelstein> Here's the source: https://gitweb.torproject.org/tor-browser.git/patch/?id=d9de1a2
18:33:28 <boklm> yes, I think it's difficult to maintain
18:33:29 <GeKo> i actually think this is worth it in general
18:33:55 <GeKo> i could update it and clean it up while updating the design doc
18:34:21 <GeKo> taking the important prefs into account
18:34:29 <boklm> ok
18:34:44 <GeKo> i think testing for a particular noscript version e.g. is not needed
18:35:19 <GeKo> so, i agree, testing all modifications is probably not going to fly
18:36:49 <GeKo> arthuredelstein: if you could run that poc for #8725 that would be neat as well
18:37:19 <GeKo> i am inclined to wait with dealing with this bug for that
18:37:36 <GeKo> not sure whether the result whould change things but it might
18:37:50 <arthuredelstein> Sorry, what poc do you mean?
18:38:10 <GeKo> the one from the sekrit mozilla bug you are added to
18:38:41 <arthuredelstein> Ah, OK. Will do
18:38:49 <GeKo> 1120398
18:38:57 <GeKo> okay, thanks
18:39:18 <GeKo> do we have anything else to discuss?
18:39:31 <GeKo> i have only one thing: the next meeting(s)
18:39:53 <GeKo> i am at PETS next week and chances are high that i will miss the meeting
18:40:15 <GeKo> then i am more or less on vacation the upcoming two weeks
18:41:05 <GeKo> i guess having a meeting next monday would be good though. i can read the backlog in case i don't make it
18:41:35 <mcs> Would a different day/time be better? I assume not.
18:41:40 <mcs> (next week)
18:41:55 <GeKo> probably not.
18:42:04 <boklm> I will be offline next week, but I can read the backlog later
18:42:17 <GeKo> ok
18:43:02 <GeKo> just having mcs and arthuredelstein around might make not much sense
18:43:19 <GeKo> i guess we cancel the one next week
18:43:42 <GeKo> but having one the week thereafter seems good to me due to the firefox release
18:43:47 <GeKo> and mike helping with it
18:43:56 <boklm> yes
18:43:59 <mcs> AGreed
18:44:11 <GeKo> i try to prepare everything but in case there are questions talking about those might be good
18:44:42 <GeKo> okay. i'll send a mail tomorrow and announce both changes leaving the other week open
18:44:47 <arthuredelstein> Sounds good.
18:44:58 <arthuredelstein> That will be the Monday?
18:45:08 <GeKo> yes, as usual
18:45:58 <GeKo> anything else for today?
18:46:33 <GeKo> thanks for the meeting then *baf*
18:46:38 <mcs> thank you
18:46:41 <GeKo> #endmeeting