18:00:06 <GeKo> #startmeeting tor browser
18:00:06 <MeetBot> Meeting started Tue May 29 18:00:06 2018 UTC.  The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:06 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:00:10 <GeKo> o/
18:00:31 <GeKo> a bit unusal to have a tor browser team meeting on tuesday but here we go
18:00:38 <boklm> hi!
18:00:48 <GeKo> please update your items on https://storm.torproject.org/shared/tHoN4Ii7rLSjPE0OP4gydX4cMGadsXmRQNc-6lwru0N
18:00:59 <GeKo> and marked the ones you want to talk about bold
18:02:51 <sysrqb> .
18:03:48 <GeKo> alright i don't see anyone having anything marked in bold for the group
18:04:15 <sukhe> GeKo: just a second, reloading
18:04:25 <GeKo> sure
18:04:45 <GeKo> boom
18:04:48 <sukhe> done thanks
18:07:01 <GeKo> sukhe: i think getting the windows nightly builds running is the highest prio right now
18:07:22 <sukhe> ok great, then that's resolved
18:07:58 <GeKo> because time is running short. we need to start building the *alpha* in about three weeks
18:08:24 <GeKo> okay, the user testing for antonela
18:08:43 <GeKo> what we did was compiling custom bundles with #24309 and #23247 applied
18:09:21 <GeKo> as we did not get he linux nightlies going last week we resorted to the esr52-based browser
18:09:57 <GeKo> i heard it was finally working and isabela had some feedback for #23247 which she wanted to post to the ticket
18:10:10 <GeKo> but i don't know whether the whole thing was a success or not
18:10:27 <GeKo> arthuredelstein: yes, both worked
18:10:37 <arthuredelstein> thanks
18:10:48 <GeKo> the problem isa and anto actually had was not the one you and mcs/brade were seeing
18:11:24 <GeKo> but they just overwrite the bundle on macOS with the new one and the stuff in the old profile interfered with the new code
18:11:37 <GeKo> so that they did not see the new display
18:11:44 <GeKo> but it was still at the old place
18:12:10 <GeKo> that's kind of weird but was the issue. not sure how we can prevent that from happening when we actually switch
18:12:15 <mcs> Maybe they had an old Torbutton in their profile.
18:12:19 <GeKo> yes
18:12:38 <GeKo> i guess the regular updater is taking care of that
18:12:46 <mcs> Was the browser version “nightly” or something else?
18:12:51 <GeKo> yes
18:13:04 <GeKo> (i.e. "tbb-nightly")
18:13:26 <mcs> The code that copies over new extensions might be confused when comparing “tbb-nightly” to something like 7.5.x or 8.x
18:13:49 <mcs> (it only copies in a new extension if the version is higher)
18:14:03 <GeKo> yeah, that might be it
18:14:27 <mcs> “Always create a folder and run from there on macOS when testing new stuff”
18:14:31 <mcs> :)
18:14:41 <GeKo> yep, i worked them through that :)
18:14:45 <mcs> thanks
18:14:47 <GeKo> (at least isa)
18:15:28 <GeKo> tjr: any other tickets we should try to uplift for the mingw-w64 build?
18:15:45 <GeKo> i did not try to be that systematic but realized some were missing
18:16:02 <tjr> youll need https://bugzilla.mozilla.org/show_bug.cgi?id=1411401
18:16:19 <tjr> and the two nss patches: https://bugzilla.mozilla.org/show_bug.cgi?id=1448748
18:16:19 <tjr> https://bugzilla.mozilla.org/show_bug.cgi?id=1389967
18:16:37 <tjr> but id like to keep those out of the -esr60 until they're ready
18:17:19 <GeKo> yeah, i was just talking about stuff that already landed
18:18:03 <GeKo> igt0: what's the status with the orfox crashing?
18:18:04 <tjr> ah ok; thats all then
18:21:26 <GeKo> arthuredelstein: could you just do the tor browser patch rebase for what TBA needs?
18:21:56 <sysrqb> (i think so, i'll let arthuredelstein take care of it if he's already planning on it thi week)
18:21:57 <GeKo> for now at least so that sysrqb gets unblocked for TBA
18:22:09 <arthuredelstein> GeKo: Sure -- does TBA need mozilla-central or mozilla-beta?
18:22:17 <GeKo> mozilla-beta
18:22:28 <arthuredelstein> Yup, will do.
18:22:29 <GeKo> we plan to use firefox 61 iirc
18:22:35 <sysrqb> yes
18:22:38 <sysrqb> great, thanks!
18:22:42 <GeKo> i think nightly can wait this week at least
18:23:01 <GeKo> arthuredelstein: we should try to get some hard patches into the alphas as early as possible
18:23:13 <GeKo> so if you could look at e.g. HTTP/2 that would be good
18:23:31 <arthuredelstein> yes, I thought that one and the optimistic socks might be good things to look at
18:23:35 <GeKo> there might not be much code that we need here
18:23:39 <GeKo> yes
18:24:13 <sysrqb> arthuredelstein: i also opened #26233, if it helps for tracking
18:24:23 <GeKo> but i'd like to see some tests that a) HTTP/2 url bar domain isolation is actually working as we want it
18:24:28 <sysrqb> (feel free to rename it, if you want)
18:24:46 <GeKo> arthuredelstein: and there might be more we need to take care of like server push and stuff
18:24:54 <arthuredelstein> yes
18:25:06 <GeKo> okay, good, those two tickets sound like a good plan
18:25:22 <arthuredelstein> Just to understand better, what's the release schedule for TBA going to look like?
18:25:23 <igt0> GeKo, about the crash, I asked for help in the jsapi channel and they said it doesn't make sense :), so my guess is a toolchain issue. For now I am trying to update the latest esr android toolchain.
18:25:55 <GeKo> arthuredelstein: july
18:26:14 <GeKo> see the tbb-dev thread starting on march 13 about remaining things for the first release
18:26:23 <GeKo> there are blockers listed and things we should look into
18:26:42 <GeKo> igt0: ok, thanks
18:26:46 <arthuredelstein> right. So this will be an "alpha" release, correct?
18:26:55 <GeKo> the first one, yes
18:27:17 <arthuredelstein> I'm wondering if we should use ff62 then, because it will be in beta at that point, and then stable on September 4.
18:27:21 <sysrqb> (our alpha, based on FF61 release)
18:27:34 <GeKo> we need some UX help as well. i'll ping anto/isa about that
18:27:35 <sysrqb> hrm, we could i guess
18:27:55 <GeKo> arthuredelstein: we'll i'd like to keep the moving parts as low as we can
18:28:07 <arthuredelstein> right, that's why I'm suggesting ff62 :)
18:28:43 <arthuredelstein> so that TBA alpha and TBA stable are the same codebase for the first releases
18:28:54 <GeKo> well, we already have enough moving parts without instability coming from a beta version for our first alpha
18:29:15 <GeKo> so, there will be other alphas before ff62 will get released
18:29:39 <GeKo> i think we should use the alpha cycle to get used to the rebase provess using mozilla-release
18:29:51 <GeKo> *process
18:30:19 <arthuredelstein> OK! I guess at some point in the future we'll just need to catch up so that our stable aligns with mozilla-release.
18:30:19 <GeKo> it seems to me there is enough complexity in our first alpha that we can switch to the beta in a later one
18:30:30 <GeKo> yes
18:30:36 <arthuredelstein> sounds good to me
18:31:00 <GeKo> we are doing that while the alpha series is under development
18:31:15 <arthuredelstein> makes sense
18:31:23 <igt0> sysrqb, iirc, you said there are few changes we would need to make in the extensions for FF 61, right? Do you know how big are they?
18:31:28 <GeKo> meaning that in september when ff62 gets out we'll still have the alpha series (and no stable one yet)
18:31:59 <GeKo> igt0: they'll be bigger ones as a bunch of interfaces etc. went away
18:32:12 <GeKo> like XUL overlays are gone
18:32:18 <GeKo> (iirc)
18:32:24 <igt0> GeKo, hrm
18:32:37 <GeKo> so, that alone will be cause quite some work
18:32:48 <sysrqb> yeah
18:32:59 <sysrqb> i don't remember which other parts were removed
18:33:25 <sysrqb> but we'll need either webextensions or intengration directly in tor-brower by FF61 and FF62, i think
18:33:30 <sysrqb> *integration
18:33:34 <GeKo> yes
18:34:38 <GeKo> so, getting a branch ready with tor browser + orfox patches for ff61 and then banging at the extension issue (and others) :)
18:35:16 <sysrqb> yep, agreed
18:35:50 <GeKo> okay, while we are at the discussion
18:35:56 <igt0> Are we planning to have the onboarding screen in the alpha release? (even without Tor integrated within TBA) ?
18:36:10 <GeKo> something like that, yes
18:36:25 <GeKo> i plan to ping anto and isa to get that prioritized
18:36:28 <sysrqb> i thought we agreed we must have tor integrated in the first alpha, because we don't have fine-grained ontrol over Orbot
18:36:39 <sysrqb> *control
18:37:10 <GeKo> no, i dont't think so
18:37:32 <GeKo> there are three must haves n the mail thread
18:37:39 <GeKo> 1) not based on esr
18:37:46 <GeKo> 2) no proxy bypass
18:38:13 <GeKo> 3) good hints for users that this is an alpha/possibly showing the missing features
18:38:27 <GeKo> should haves:
18:38:47 <GeKo> 1) what's the updater/updating story (investigate that)
18:39:18 <GeKo> 2) are fpi and fingerprinting patches working if the prefs are flipped
18:39:40 <GeKo> (ideally we would make this work)
18:39:43 <GeKo> that's it
18:39:51 <GeKo> all the other things come in later alphas
18:40:17 <sysrqb> okay i see that mail now
18:40:45 <GeKo> 4) in must haves is branding update
18:41:41 <sysrqb> yes
18:41:44 <GeKo> is there anything more we should talk about regarding preparations for the alpha?
18:42:10 <GeKo> mcs: we should chat about the watershed thing which is scary once you are back
18:42:37 <GeKo> it seems as well the new certutil thing is creating keys in a new database format
18:42:48 <mcs> GeKo: OK. I have not yet read what boklm wrote in the related ticket.
18:42:52 <sysrqb> i don't have anything else right now for the alpha-preparations
18:42:58 <GeKo> (i realized that quite a while after creating the new keys)
18:43:14 <GeKo> so, we might need at least update our signing scripts taking this into account
18:43:35 <mcs> I forgot about the new NSS DB format. Will that cause us a lot of pain?
18:43:48 <GeKo> ha, i don't know
18:44:03 <mcs> Hopefully not then ;)
18:44:12 <GeKo> i guess as long as we start using the new mar-tools from now on for signing we should be fine
18:44:18 <igt0> About the alpha, how are we going to release? (e.g. we need keys for the google play store. Who is going to sign it?)
18:44:40 <GeKo> yes, we need to figure that out
18:45:09 <sysrqb> i can open a ticket for tracking that
18:45:17 <GeKo> that would be a good first start
18:45:46 <GeKo> i can see that it could be helpful taking should have a) into account here too
18:46:04 <GeKo> (what's the updater/update F-Droid/own hosting story)
18:46:15 <GeKo> to have a bigger picture
18:46:47 <GeKo> sysrqb: igt0: it might be smart to start thinking about it and once you have a plan to send an email to tbb-dev to get feedback
18:47:06 <sysrqb> GeKo: sure, will do
18:47:15 <GeKo> if you could put that on your plate for this week that would be smart i think
18:47:20 <igt0> and my last question, are we going to allow users using a deprecated TBA version? (few apps normally open a modal not allowing the user to use the app)
18:47:28 <GeKo> (at least getting started)
18:48:20 <GeKo> good question, i have not thought about that yet
18:49:18 <GeKo> igt0: could you open a trac ticket and start with some pro and con arguments?
18:49:25 <igt0> GeKo, yep
18:49:38 <GeKo> and maybe some links with further reasoning behind the pros and cons?
18:49:41 <GeKo> thanks
18:51:08 <GeKo> oaky, for the nightlies: i'll merge sukhe's patch later today and hopefully the one for #16472
18:51:22 <GeKo> so we'll have nightlies for linux from tomorrow on
18:51:39 <GeKo> please dog-food if you can and start reporting issues
18:52:21 <GeKo> if you find any the "ff60-esr" keyword is your friend. there is already a bunch of tickets tagged with it
18:53:53 <GeKo> right now the two non-toolchain code tickets we should fix for the first alpha are
18:54:07 <GeKo> #26128
18:54:16 <GeKo> and #26039
18:54:36 <GeKo> arthuredelstein: could you pick one of them
18:54:44 <GeKo> pospeselr: could you pick the other one?
18:55:07 <pospeselr> sure enough
18:55:34 <GeKo> i'll triage the tickets as they come in, so that we have an idea what is prio for the alpha
18:55:40 <pospeselr> i'll take 26039, unless arthur objects :)
18:55:56 <arthuredelstein> pospeselr: works for me!
18:56:26 <GeKo> right now all tickets with prio "Very High" and keyword "TorBrowserTeam201805" are on my radar for the nightlies/alpha
18:56:34 <GeKo> thanks to both of you
18:57:12 <GeKo> oh, and one final request: could anyone of you look over #26127?
18:57:29 <GeKo> i want to have that in the first nightlies as well
18:57:50 <GeKo> it's just adding two prefs to make sure our own extensions are treated properly
18:57:59 <GeKo> (as they are legacy ones)
18:58:32 <GeKo> that's all i have for today i think
18:58:38 <GeKo> do we have anything else?
18:59:31 <GeKo> okay, thanks all then *baf*
18:59:34 <GeKo> #endmeeting