19:00:41 <GeKo> #startmeeting tbb-dev
19:00:41 <MeetBot> Meeting started Mon Dec 21 19:00:41 2015 UTC.  The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:41 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:00:47 <GeKo> hi everybody!
19:01:06 <mcs> Hello
19:01:30 <GeKo> this week i was mostly busy with the nightmare release. the positive thing is it seems we finally got it out working
19:01:44 <GeKo> and the user impact was not that big
19:01:51 <GeKo> thanks for everybody helping here
19:02:15 <GeKo> i did some hackerone stuff and team planning as well
19:03:04 <GeKo> this week i hope i can get back to #15988, #17858 and #17870
19:03:31 <GeKo> not sure what else will fit into it
19:04:13 <mcs> Is there anything we as a team can do to improve communication with Mozilla to avoid a rerun of last week?  Or maybe it was unavoidable?
19:04:14 <GeKo> there are some things we can discuss later today; one thing i'd like to put on the agende too is thinking about the scope for our 5.5 release
19:04:30 <GeKo> what should get into it and when do we want to get out
19:04:43 <mcs> Discussing 5.5 is a good idea.
19:04:43 <GeKo> (just for you to think about while we are doing status reports)
19:04:53 <GeKo> mcs: i think there is not much we can do
19:05:09 <mcs> GeKo: That's what I thought :(
19:05:09 <GeKo> i mean we have always been gambling as the builds we use are candidate builds
19:05:20 <GeKo> this time we almost lost
19:06:07 <GeKo> i think the best we can do is updating the release process doc to make sure to check immediately before releasing whether there is not another new tag mozilla used
19:06:41 <GeKo> at least that's a thing i want to add althought i'll never forget this again i guess
19:06:48 <GeKo> that's it for me
19:09:42 <GeKo> who's next?
19:09:59 <isis> sorry, what happened was that another tag was pushed with the upstream fixes, and so 5.0.6 was released in rapid succession after 5.0.5… is that right?
19:10:12 <isis> (just out of curiosity)
19:10:17 <GeKo> yes
19:10:23 <isis> oh bummer
19:10:29 <isis> i'm sorry guys
19:10:48 <GeKo> they basically released the second candidate build the same say they shipped the official release
19:11:13 <GeKo> *day
19:11:42 * mcs will go next
19:11:51 <mcs> Last week, Kathy and I posted screenshots of a series of proposed Tor Launcher UI tweaks in #11773. Comments are welcome.
19:12:02 <mcs> We did some research and experimentation for #13252. Comments are welcome there too.
19:12:08 <mcs> We also did a little testing of 5.0.5 and 5.5a5 and triaged assorted bugs.
19:12:13 <mcs> This week we will follow up on #11773 and #13252.
19:12:18 <mcs> Also, we will be taking a few days off this week and next to celebrate Christmas and the New Year.
19:12:23 <mcs> We should be reachable via email.
19:12:29 <mcs> That's all for us.
19:12:37 <GeKo> do that! (taking off)
19:12:56 <mcs> GeKo: Thanks. You too!
19:13:22 <GeKo> re 13252: i think i am fine with the cheating approach to get the signing going asap on the alphas to look for other issues
19:13:53 <GeKo> but i am actually worried about apply restricting their policy further with an update breaking that short-cut
19:14:07 <GeKo> leaving users with a scary warning again
19:14:21 <GeKo> which is especially bad as it worked until then
19:15:42 <GeKo> s/apply/apple/
19:15:54 <mcs> GeKo: Agreed. That is definitely a concern. We could just bite the bullet and switch to a side-by-side approach but it might be messy (we would want only data that changes in the side folder).
19:16:43 <GeKo> yeah
19:17:30 <mcs> I guess we can talk more during the TB 5.5 scope discussion, but it would be nice to have signed Mac OS app bundles for 5.5.
19:18:06 <GeKo> yes, true
19:20:35 * arthuredelstein can go
19:20:44 <arthuredelstein> Last week I wrote a revised patch for bugzil.la/876501.
19:20:52 <arthuredelstein> And then I worked on unit tests for #17785 and #17790.
19:21:00 <arthuredelstein> This week I want to finish those unit tests, and have another look at #14429.
19:21:10 <arthuredelstein> I also want to think about when it makes sense to start rebasing to FF45.
19:21:18 <arthuredelstein> 45 alpha is out now, so we might be able to start soon.
19:21:25 <arthuredelstein> I don't know how much things change between the alpha and the beta/release, though.
19:21:39 <arthuredelstein> And I will also be taking off some days this week and next for the holidays.
19:21:46 <arthuredelstein> That's it for me
19:22:04 <GeKo> arthuredelstein: could you cc me to that bugzilla bug?
19:22:15 <arthuredelstein> GeKo: sure
19:22:34 <arthuredelstein> oh wait, maybe that's the wrong number. Why is it restricted?
19:22:47 <GeKo> dunno, you posted the link :)
19:23:04 <arthuredelstein> Oops, tranposed digits
19:23:13 <arthuredelstein> bugzil.la/867501
19:23:45 <GeKo> aha, ok i cced myself in this case
19:24:51 <GeKo> arthuredelstein: i think starting a first pass early 2016 might be smart (re rebasing)
19:25:06 <GeKo> this leaves us some time to handle trickier cases if there are any
19:25:16 <arthuredelstein> Yeah, I tend to think it will be to our advantage.
19:25:23 <arthuredelstein> Especially since we now have to deal with e10s
19:25:45 <GeKo> well, that's not sure yet actually
19:25:59 <GeKo> oh and we won't have to deal with service workers
19:26:22 <GeKo> which is good news as patching this would probably have been quite time-consuming
19:26:31 <arthuredelstein> Oh, I just realized your right, that it isn't certain we will have e10s
19:26:46 <arthuredelstein> What happened with service workers?
19:26:59 <GeKo> but we might actually want to which makes this even more fun in this case
19:27:01 <GeKo> ;)
19:27:08 <GeKo> (e10s)
19:27:59 <GeKo> arthuredelstein: https://bugzilla.mozilla.org/show_bug.cgi?id=1232029#c0
19:29:18 <arthuredelstein> That would be good.
19:30:52 * boklm can go next
19:31:03 <boklm> This past week I reproduced builds of the releases
19:31:04 <boklm> I made some Windows changes on the testsuite to use a non-cygwin python (marionette doesn't work with the cygwin python as it is mixing some Windows and Cygwin/Unix paths)
19:31:09 <boklm> I started looking at automating the setup of the testsuite dependencies on Windows to make it easier to deploy
19:31:18 <boklm> This week I'm planning to continue on those Windows changes, take some days off then go to Hamburg for CCC
19:31:26 <boklm> That's it for me
19:32:52 <GeKo> thanks. anybody else who wants to give a status update?
19:34:31 <GeKo> alright. let's move to the discussion part then. first the easiest point: next meeting day.
19:34:46 <GeKo> i guess it might not make much sense to have a meeting next monday
19:35:07 <GeKo> so i think jan 4 2016 might make sense then?
19:35:22 <mcs> 4 Jan works for Kathy and me.
19:35:33 <boklm> works for me too.
19:36:18 <arthuredelstein> and me
19:36:35 <GeKo> okay, then be it so. i'll write a mail to tbb-dev later
19:37:28 <GeKo> next: role descriptions. do we want to have more official roles for subsystems of Tor Browser?
19:37:37 <GeKo> would that make sense?
19:38:34 <GeKo> if we want to: what would/should change compared to the status quo?
19:38:54 <GeKo> i mean there are already role indications on the core people page e.g.
19:39:18 <GeKo> and inofficially there are such roles within the team already
19:40:03 <GeKo> (even if not every "sub"system has one that takes especially care of it)
19:40:42 <mcs> If having written role descriptions helps people outside the team, then probably we should do it (but I think we need to remain flexible within the team).
19:40:53 <arthuredelstein> I agree with mcs
19:40:56 <mcs> If it is just for us, I do not need it.
19:41:13 <arthuredelstein> Maybe we could call it a "contact person" for each subsystem, rather than an official role
19:41:37 <arthuredelstein> Although maybe it's better for external people just to write to the tbb-dev list.
19:42:58 <isabela> hi there!
19:43:10 <arthuredelstein> hi!
19:43:15 <isabela> just want to quick say that this was done for the network team to help the team internally
19:43:54 <isabela> is really up to the applications team if you want to use it or not,from what I heard it seems the team has a good structure going on already
19:44:10 <isabela> so dont feel you have to do it
19:44:14 <GeKo> yeah, i think so.
19:44:35 <arthuredelstein> how many people are on the network team, officialy?
19:44:43 <isabela> 5
19:45:19 <isabela> and we are starting with subsystems maintainers
19:46:10 <GeKo> having tbb-dev for external contact is good, i think and having the inofficial way we have now wrt roles seems to be enough for me, too, atm
19:46:34 <isabela> I guess what we want to avoid is to list things that defines a developer at Tor without sharing it with others and getting their opinion
19:46:54 <isabela> is more of an internal thing
19:47:33 <GeKo> ok
19:47:59 <GeKo> so far the current model works out for us, thus, let's stick to it
19:48:30 <boklm> the current model sounds good to me too
19:49:00 <GeKo> ok
19:49:01 <mcs> Sticking with the current model seems fine to me.
19:49:02 <isabela> sounds good :)
19:49:17 <GeKo> last item on my list: tbb 5.5
19:49:34 <GeKo> we have to switch to esr 45 end of may
19:49:53 <GeKo> which is four regular releases away if i counted right
19:50:20 <GeKo> should that be the 5.5?
19:50:50 <GeKo> i guess we want to have the osx signing for 5.5
19:51:20 <GeKo> which means it would be earliest the one at march 7
19:52:10 <GeKo> which would leave just two releases for a 6.0 if we want to have that version for the Tor Browser switched to ESR 45
19:53:07 <GeKo> another option would be to skip 5.5 and just release 6.0 with ESR 45 +  all the changes accumulated so far
19:53:13 <mcs> Is March 7th the next ESR release?
19:53:19 <GeKo> yes
19:53:36 <GeKo> actually March 8th
19:53:56 <arthuredelstein> Officially I think Mozilla only drops ESR38 on May 31.
19:54:09 <GeKo> yes
19:54:14 <arthuredelstein> (looking at https://wiki.mozilla.org/RapidRelease/Calendar#Future_branch_dates)
19:54:23 <GeKo> that is our hard deadline for switching to esr45
19:54:30 <arthuredelstein> Aha, that's what you said :)
19:54:57 <GeKo> i hope so :)
19:55:41 <arthuredelstein> Maybe when we create the first FF45 branch, we call that 6.0 alpha?
19:56:11 <GeKo> yes, sounds good to me.
19:57:01 <arthuredelstein> TB = (FF - 3) / 7
19:57:45 <arthuredelstein> or something
19:58:19 <mcs> I would need to look at what is already in 5.5 before I would say "Wait and fold everything into 6.0"
19:59:05 <GeKo> i am fine with that. i want to do the same over the next days and we can make a final decision during the next meeting
19:59:17 <mcs> Skimming our release notes, it seems ike there are a lot of good things in 5.5.
19:59:49 <mcs> OK. Also, there should be a release near the end of January to correspond with FF 44 / ESR 38.6 (I think)
20:00:11 <GeKo> that's true
20:01:23 <mcs> I do not know if it is realistic to get 5.5 ready by end of January. OS X signing would need at least one test release, so it could not be declared stable that quickly (which is maybe why you mentioned the March date).
20:01:54 <GeKo> yes
20:02:24 <GeKo> looking at our switch to 4.5 and 5.0 there were 4 months between that which makes sense
20:02:45 <GeKo> that would mean having a 5.5 with the next release
20:03:45 <mcs> My instinct is to cut things to fit the January release date and shove TB 5.5 out the door (but others may disagree).
20:04:10 <GeKo> i think either that or no 5.5 at all
20:04:15 <mcs> Agreed.
20:04:37 <arthuredelstein> I think that makes sense.
20:05:18 <GeKo> so let's mull over that in the coming days and make a decision in our meeting at jan 4
20:05:26 <mcs> Sounds good to me.
20:05:48 <GeKo> anything else we should discuss?
20:08:30 <GeKo> okay. thanks for the meeting then! get some rest everybody before we start into an exciting 2016 :) *baf*
20:08:33 <GeKo> #endmeeting