18:29:47 <sysrqb> #startmeeting Tor Browser Team Meeting, 2 December 2019
18:29:47 <MeetBot> Meeting started Mon Dec  2 18:29:47 2019 UTC.  The chair is sysrqb. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:29:47 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:29:54 <sysrqb> Hello, hello
18:30:01 <antonela> o/
18:30:02 <boklm> hi
18:30:02 <Jeremy_Rand_Talos> Hello!
18:30:08 <pospeselr> \o
18:30:13 <sysrqb> welcome to December 2019, everyone
18:30:38 <brade> hi
18:31:47 <sisbell> hi
18:31:48 <mcs> hi
18:32:11 <pospeselr> hi
18:33:27 <GeKo> o/
18:34:06 <pili> hi
18:34:17 <sysrqb> pospeselr: hopefully you finished the flu last week, and  it won't carry over into this week ;)
18:34:33 * pospeselr coughs
18:34:44 <pospeselr> I'm mostly fine now :p
18:34:51 <pospeselr> who needs working lungs
18:35:07 <Jeremy_Rand_Talos> pospeselr, (mostly) good to hear
18:35:43 <sysrqb> Jeremy_Rand_Talos: regarding you question for pili
18:36:38 <sysrqb> i would say it won't be very relevent for you because we're not looking at implementing something as large and involved as namecoin for this sponsor
18:36:56 <sysrqb> you can join and follow along if you want
18:37:19 <pili> sysrqb: +1
18:37:25 <sysrqb> but our current plan doesn't involve something at that scale
18:37:51 <Jeremy_Rand_Talos> sysrqb, ok, sounds good.  pili, where would I find the meeting schedule if I decide I have time to follow along?
18:38:08 <pili> Jeremy_Rand_Talos: I normally announce it on tor-project mailing list
18:38:28 <pili> but it's every other Tuesday at 15UTC
18:38:39 <pili> next meeting is on the 10th December
18:38:56 <pili> and our notes are here: https://pad.riseup.net/p/s27-meeting-keep
18:39:19 <Jeremy_Rand_Talos> pili, ah cool, the one Tor mailing list I've still forgotten to sign up for.  I'll make sure to subscribe there.  I might be able to join you on the 10th, I think my schedule is open
18:39:28 <pili> sounds good :)
18:39:53 <pili> that will be the last meeting of the year
18:40:00 <pili> and we'll be back on the 7th probably
18:40:03 <pili> january
18:40:09 <Jeremy_Rand_Talos> cool
18:40:15 <sysrqb> okay, pili, do you want to make your announcements/reminders?
18:40:50 <sysrqb> or, i can tell everyone to look at your items :)
18:41:04 <pili> both :D
18:41:17 <sysrqb> wfm :)
18:41:42 <pili> 1st one: We've started thinking about how to spend the Bug Smash fund and the network team has already tagged a number of tickets they would like the money to be earmarked for :)
18:41:49 <pili> so it would be good if we did the same
18:42:10 <pili> also, in order to know how much money will be needed for each bug we will need a points estimate please
18:42:19 <pili> any questions on the Bug Smash fund?
18:42:34 <antonela> what is the criteria to pick tickets?
18:42:59 <pili> it should be bugs that we can't really get funded any other way and that are important and we want to work on
18:42:59 <antonela> *for picking tickets? (maybe)
18:43:20 <pili> so it wouldn't be a large project, but maybe niggling bugs that have been open for a while and we want to fix
18:43:47 <pili> see https://blog.torproject.org/tors-bug-smash-fund-help-tor-smash-all-bugs for more background
18:44:10 <pili> "This work includes responding quickly to security bugs, improving test coverage, and keeping up with Mozilla’s ESR"
18:44:13 <pili> for example
18:44:50 <antonela> ok
18:44:53 <pili> we could even just tag some of the ESR transition work actually
18:45:16 <mcs> +1
18:45:20 <pili> ok, let's do this, I'll give about a week for people to tag bugs
18:45:40 <pili> and then I'll start filling up with ESR tickets if we don't have any/enough
18:46:07 <brade> +1
18:46:19 <sysrqb> do you know when we'll begin working on these tickets?
18:46:34 <sysrqb> would we start assigning them in January? or later in the year?
18:47:23 <sysrqb> by this i mean, should we consider the money available "now" or should we think about this as another sponsor that starts in $month
18:47:26 <sysrqb> ?
18:47:55 <sysrqb> if you don't know yet, then that's totally fine, too
18:48:16 <pili> I think this is money available now
18:48:38 <pili> as in, we have it already and we need to show how we want to spend it
18:48:45 <sysrqb> okay
18:48:45 <brade> so it covers work we did in October and November?
18:48:52 <pili> this money was raised in augus
18:48:55 <pili> august
18:49:06 <pili> so we could tag work that we did from september in theory
18:50:35 <pili> anyway, I don't want to take up too much more time with this :)
18:50:39 <sysrqb> okay, that sounds good. i guess if we choose any tickets from september then we should add some points (if not already there)
18:50:46 <pili> sounds good
18:51:37 <pili> the other is a friendly reminder for people to update actual points when they are done with a ticket, we're doing pretty good so far on this though
18:52:08 <pili> and finally, a reminder that we're closing down storm in January so please move any documents you want to keep from there to nextcloud
18:52:11 <pili> I'm done :)
18:52:27 <sysrqb> thanks
18:53:19 <sysrqb> okay, i don't see any other bold items from the weekly updates
18:53:55 <sysrqb> and everyone's planned work for this week looks good
18:54:06 <sysrqb> for discussions, the first is the upcoming release
18:54:49 <sysrqb> tomorrow we should be releasing 9.2.0
18:55:10 <sysrqb> (assuming my memory is correct), and 9.5a3
18:55:31 <sysrqb> so far we haven't hit any blockers
18:55:43 <GeKo> *9.0.2
18:56:10 <sysrqb> err, oh, yes, that. thanks
18:56:41 <sysrqb> this will (hopefully) be the last release of 2019
18:57:49 <sysrqb> the next release is on 7 January, so we don't have much time between now and the holidays
18:58:34 <sysrqb> we should try getting as much merged over the next two weeks as we can
18:58:56 <sysrqb> (earlier is better)
18:59:31 <GeKo> earlier and as much :)
18:59:36 <GeKo> as possible
18:59:43 <sysrqb> haha, yes
19:00:14 <sysrqb> okay, any questions about this?
19:01:13 <sysrqb> okay, today we began triaging the tickets we should be working on this month
19:01:26 <sysrqb> and we began assigning reviewers for tickets currently in needs_review
19:01:55 <sysrqb> if you are now a reviewer for a ticket, please review the patch :)
19:02:40 <sysrqb> if you are now a review but you don't think you should be or you'd like someone else to review it, then please let me/GeKo/pili know
19:02:45 <sysrqb> or find a replacement, if you can
19:03:19 <sysrqb> we're in the middle of triaging open tickets remaining tagged for November and December
19:03:38 <sysrqb> and we'll finish those this week and assign owners
19:04:21 <sysrqb> related to the previous topic, if you are assigned one of those tickets then we are hoping a patch for it will be included in the next release
19:04:37 <sysrqb> so please prioritise accordingly :)
19:04:50 <sysrqb> you will probably end up with too many tickets for the month, as usual, though
19:04:57 <sysrqb> so do what you can
19:05:16 <sysrqb> any questions/comments about this?
19:06:05 <sysrqb> okay. #32116
19:06:40 <pospeselr> I would like to not have to edit my .mozconfig each time i start a new branch
19:06:45 <sysrqb> what is the expectatoin of the mozconfig files in the tor-browser repo?
19:06:57 <sysrqb> so, pospeselr voiced his opinion :)
19:07:14 <pospeselr> :3
19:07:17 <sysrqb> (i happen to agree)
19:07:35 <mcs> if the mozconfig files in the tor-browser repo do not match what we use for the builds we ship we should make sure that is very clear to people.
19:07:43 <mcs> maybe comments would do (in the files?)
19:08:40 <mcs> I think it is true that most of us probably maintain their own .mozconfig file (or a set of changes needed) in order to build outside of rbm
19:09:01 <sysrqb> there is another ticket about replacing the current tor-browser mozconfig files with the ones shipped in tor-browser-build
19:09:17 <GeKo> yeah
19:09:40 <GeKo> let me find it
19:10:09 <GeKo> #23656
19:10:26 <GeKo> it's the other way round
19:10:33 <sysrqb> but providing mozconfigs for each platform which are intended for compiling on a developer's computer and not within tor-browser-build
19:10:39 <sysrqb> does have its benefits, too
19:10:44 <sysrqb> ah, thanks
19:10:56 <GeKo> but the idea was to not have to maintain two sets of mozconfig files
19:11:00 <sysrqb> (each or some subset of platforms)
19:11:03 <GeKo> and keep them somewhat in sync
19:11:09 <sysrqb> right
19:11:18 <sysrqb> that is a disadvantage of this plan
19:11:38 <GeKo> because we actually need to update both sets sometimes
19:11:50 <sysrqb> now maintaining multiple platform configs as well as compilation environments configs
19:12:05 <GeKo> (even if we got with the current plan to not have them fully in sync but as two different sets)
19:12:34 <GeKo> s/got/go/
19:13:38 <boklm> if we decide we don't want to do #23656, one possible benefit is we could then use one .mozconfig for all platforms, with some templating for the differences between each platforms
19:13:40 <sysrqb> maybe we can accomplish both tickets by created shared/comment files
19:13:53 <sysrqb> and then overridding any values needed for the different environments?
19:14:09 <sysrqb> similar ot how mozilla current do it?
19:14:23 <sysrqb> ah, interesting idea boklm
19:14:39 <sysrqb> i was also wondering if we could have tor-browser-build generate the configs for us
19:14:49 <sysrqb> but that is...even more complicated than these other ideas
19:17:10 <sysrqb> so.
19:17:47 <sysrqb> maybe for now, we can start with pospeselr plan in #32116, and create a better plan in #23656?
19:18:08 <sysrqb> maybe that results in multiple levels of inheriting comment settings
19:18:28 <sysrqb> or generating the configs in tor-browser-build
19:18:59 <sysrqb> but making the mozconfig files in tor-browser usable seems like a win
19:19:19 <sysrqb> and we can add a note in them saying the official files are found in tor-browser-build
19:19:27 <sysrqb> does that sound reasonable?
19:19:37 <pospeselr> sounds good to me :)
19:19:40 * Jeremy_Rand_Talos notes that the WIP cross-compiled Linux targets that I'm working on might benefit from mozconfig being generated from a common source
19:19:41 <boklm> yes
19:19:58 <Jeremy_Rand_Talos> Because linux-arm and linux-ppc64le are likely to be very similar mozconfigs
19:20:05 <GeKo> works for me.
19:20:06 <sysrqb> s/inheriting comment/inheriting common/
19:20:37 <sysrqb> okay, cool
19:20:54 <sysrqb> #32602
19:21:00 <GeKo> pospeselr: so would you revise that patch accordingly and then i can review it and we get it off our plate?
19:21:07 <pili> this was one that emmapeel wanted to bring to the table :)
19:21:20 <pili> maybe we can discuss it during the last few minutes
19:21:21 <emmapeel> yes! thanks you honor
19:21:55 <emmapeel> i was thinking we should help users take an informed decision when choosing which locale to use on the browser
19:22:09 <emmapeel> but i dont really know where
19:22:40 <emmapeel> if on the browser or where. i have been looking at the 'update ping' stats and it made me feel a bit weary that some locales are used by very few people
19:23:17 <pospeselr> GeKo: which patch?
19:23:37 <GeKo> the one for #32116
19:24:00 <Jeremy_Rand_Talos> emmapeel, apologies for the potentially dumb question, but do we know for sure why certain locales aren't getting many users?
19:24:13 <sysrqb> (pospeselr: adding the comment pointing to the official mozconfigs)
19:24:18 <pospeselr> ahhh ok
19:24:19 <pospeselr> YES
19:24:21 <GeKo> (thanks)
19:25:02 <emmapeel> Jeremy_Rand_Talos: well, i think some may be simply spurious installs. others, simply not enough users. i feel also the website is better now. speaking as a native speaker user, also, that uses the english locale, because of risk of locale fingerprinting :D
19:25:24 <Jeremy_Rand_Talos> Like, there are a range of possibilities, e.g. "those locales are associated with regions that either don't have much censorship/surveillance or don't have much awareness of those issues"
19:25:41 <Jeremy_Rand_Talos> Which is easy to disprove given what locales those are
19:25:45 <Jeremy_Rand_Talos> or not
19:25:48 <emmapeel> yeah for example we have locales like macedonian, with around 50 daily update pings
19:25:51 * Samdney is german but uses the english localization :)
19:26:15 <Jeremy_Rand_Talos> hence why I'm curious if we actually have a good idea of why users aren't choosing those locales
19:27:31 <emmapeel> Jeremy_Rand_Talos: i think the reasons may be not enough users in some locales
19:28:00 <Jeremy_Rand_Talos> emmapeel, hmm, it seems like it might be interesting for Tor Browser to notify the user if they install a different locale TB than their OS locale if their OS locale is available for TB
19:28:03 <sysrqb> the website download page seems like a good place for a warning
19:28:24 <emmapeel> and in other weird locales, i feel it can be forgotten locales and outdates browsers, but i am not sure if that is part of the scope of the ticket
19:29:06 <sysrqb> but this is a difficult problem, because the only default option is en-US, and the requires the person understands US English
19:29:12 <sysrqb> which is a serious usability issue
19:29:22 <sysrqb> *that requires
19:29:23 <boklm> the multi-locales bundles would be an other way to fix this
19:29:40 <sysrqb> yeah, it should
19:29:55 <sysrqb> then there is the download-size tradeoff
19:29:58 <Jeremy_Rand_Talos> the aforementioned install warning could be displayed in the language of both the TB locale and the OS locale, maybe?
19:30:04 <emmapeel> for one side, i want people to use the locales.. i mean it would be sad that after some people translates a locale we publish it saying 'dont use this locale'....
19:30:06 <sysrqb> but it is a good alternative
19:30:39 <sysrqb> emmapeel: for sure, that would be disappointing
19:30:57 <emmapeel> in the browser you actually can have a list of preferred locales, you dont have only one. you say, if this one is not available, then i want this other locale
19:31:38 <sysrqb> hrm. actually, i wonder if the multi-locale bundle would solve this
19:31:39 <emmapeel> so there is some people that maybe is fingerprinting themselves a lot....
19:31:50 <emmapeel> on this long tail languages
19:31:54 <sysrqb> yeah
19:32:17 <sysrqb> and i think the "selected" locale would be identifiable by the leak
19:32:18 <emmapeel> anyway, i saw a problem but i dont have a solution, i just wanted to see what you think about that etc
19:32:27 <sysrqb> even if the browser supports multiple locales
19:32:52 <sysrqb> the locale-specific update pings simply wouldn't be as useful
19:33:23 <sysrqb> emmapeel: it's a good question
19:33:27 <sysrqb> thanks
19:34:10 <sysrqb> okay, any other comments or topics before we end the meeting?
19:34:43 * sysrqb hears nothing
19:34:56 <sysrqb> #endmeeting