15:03:03 <morganava> #startmeeting Tor Browser Release Meeting 2024-08-21
15:03:03 <MeetBot> Meeting started Wed Aug 21 15:03:03 2024 UTC.  The chair is morganava. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:03:03 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:04:12 <morganava> the pad: https://pad.riseup.net/p/tor-browser-release-meeting-keep
15:04:36 <PieroV> o/
15:04:40 <ma1> o/
15:05:09 <morganava> as PieroV pointed out a moment ago, we are 1 month from the *scheduled* 14.0 release
15:05:49 <morganava> boklm, brizental, clairehurst, dan_b, donuts, henry-x, jwilde, ma1, PieroV
15:05:53 <ggus> hello
15:05:54 <morganava> i *think* that's everyonee
15:05:58 <PieroV> Bea is AFK this week
15:06:03 <PieroV> Maybe also next week
15:06:04 <morganava> yes and I think next
15:06:18 <morganava> jwilde may be as well
15:06:36 <PieroV> Next week we're preparing 13.5.3, which might be the last one for the 13.5 series
15:06:39 <donuts> is the target still w/c Sept 16th?
15:06:48 <PieroV> Unless Moz extends Windows 7
15:07:08 <PieroV> And if Moz extends 115, we have some additional time for 14.0
15:07:24 <PieroV> At least, not to have to do backports
15:07:32 <donuts> right, thanks
15:07:38 <PieroV> But it'd be ideal to get everything more on time
15:07:52 <PieroV> Because the ESR after 128 is 140, so we'll have one less release
15:08:32 <donuts> is there any update about extending 115 from Moz?
15:08:48 <PieroV> I think they're working about this right now
15:08:55 <PieroV> Moz had all hands very recently
15:09:02 <PieroV> And they're back from it in these days
15:09:15 <morganava> donuts: according to tjr the plans should be finalised soonish
15:09:26 <donuts> cool ty
15:09:36 <morganava> but as of like yesterday none of the release calendars have been updated to reflect the lengthened ESR115
15:10:04 <donuts> yeah, we will need to do some comms/user-support work around this when we know for sure too
15:10:24 <morganava> yes
15:10:29 <ma1> tails? :P
15:10:42 <morganava> we'll let you know as soon as there is a concrete plan upstream
15:11:12 <morganava> in the meantime we're goingn to assume they do do it and at the very least plan to spilt off the legacy windows deploys to a separate release channel
15:11:32 <PieroV> (I've started working on this ~40 minutes ago)
15:11:42 <donuts> okaydokes :)
15:12:15 <morganava> as I see it we currently have (at least) 2 major problems to fix in the next month, one of which is a blocker
15:13:02 <morganava> 1st being Android apk-size which it looks like we have various fixes in the pipeline
15:13:15 <morganava> which is blocking x86 and x86_64 releases on the play store
15:14:11 <morganava> this isn't a *very* big deal given our isntall numbers for those architectures so I don't think it should necessarily block a 14.0 release
15:14:15 <morganava> but obviously it's not ideal
15:16:11 <morganava> the other one is broken webrtc on Windows; jwilde's been intermittently unavailable and so i'm not sure how much progress has been made there (I'll be catching up with her tomorrow and may need to shuffle that work elsewhere to get the fixes in for 14.0)
15:16:59 <morganava> apart from that, the remainder of the dev work should (mostly) be any patches that come out of our mozilla reviews
15:17:11 <morganava> of whcih there are currently ~190 open as of yesterday out of an initail 256
15:17:15 <PieroV> Today I fixed the problem with onion services
15:17:26 <PieroV> Which wasn't found with audits, but it was a big one
15:17:42 <morganava> yes i saw those threads yesterday, thanks for hopping on those
15:19:01 <PieroV> That is a good segue for one of my topic :)
15:19:06 <morganava> I'll be going through the open ~14.0 Stable labeled issues and start re-prioritising (and updating the bug/feature issues linked from the major release tracking issue)
15:19:17 <morganava> yes PieroV: go for it!
15:19:32 <PieroV> Traditionally, at one month from the release, we had an alpha campaign
15:20:17 <PieroV> Do we have a strategy about this? We're still at 14.0a2 and we haven't even finished auditing yet
15:21:18 <PieroV> But I think it'd be important to start collecting users' feedback asap, as audits won't cover all the functionalities
15:22:53 <donuts> I think we agreed to start public testing from 14.0a4 instead, because of the timeline constraints
15:23:10 <PieroV> I see, it makes sense, thanks
15:23:19 <morganava> donuts: well, depending on when we agreed to that
15:23:34 <morganava> 14.0a4 may have been planned for this week
15:23:43 <morganava> we've had to skip a few alphas because things weren't ready
15:23:44 <donuts> No it says late august in the pad
15:23:57 <morganava> it's august 21 :D
15:24:02 <donuts> "Target RC1 (currently 14.0a4 August 27)"
15:24:04 <PieroV> It'll be 14.0a3 instead
15:24:07 <morganava> ack
15:24:14 <morganava> ah ok
15:24:15 <donuts> is that still accurate, sorry?
15:24:16 <morganava> RC1 hasn't mvoed
15:24:21 <donuts> oh cool okay
15:24:50 <PieroV> RC1 released in 6 days seems early to me ^_^;
15:25:15 <donuts> ha yeah august 27th is sooner than I thought, initially
15:25:25 <morganava> yeah, this is the *first* android release in the series yesterday
15:25:47 <morganava> i mean i'm still fine calling it RC1
15:25:52 <PieroV> 14.0a4 seems more realistic for RC1 to me
15:26:12 <PieroV> And the calendar already says RC1 for 3 September
15:26:35 <morganava> oh so it does
15:27:38 <morganava> alright, then should we then just keep the plan of aug 27 as the kickoff for the call for alpha testers?
15:27:50 <donuts> idk I'm confused now
15:27:59 <ma1> 3 September is also stable 13.5.3 release right?
15:28:03 <PieroV> ma1: yes it is
15:28:06 <morganava> now that we're not running into build/signing issues we should be able to maintain a weekly release cadencce until then
15:28:07 <donuts> when's 14.0a4?
15:28:25 <PieroV> donuts: around September 3
15:28:43 <morganava> yeah currently RC1 would be 14.0a4
15:28:45 <PieroV> We could delay it by one or two days if we need to prioritize 13.5.3 in the same period
15:28:47 <morganava> and RC2 14.0a5
15:28:48 <donuts> okay, so shouldn't we move the call for testers to Sep 3rd then?
15:29:00 <PieroV> donuts: I think it'd make sense
15:29:25 <morganava> what's the downside for keeping it at aug 27?
15:30:01 <PieroV> morganava: we might complete more audits by Sep 3
15:31:33 <donuts> my logic is just that we agreed to do the call at the same time as the a4/RC1 release
15:32:02 <donuts> so if the release has moved, the call should too
15:32:04 <morganava> ok wfm
15:32:57 <donuts> okay cool, I'll tell comms :)
15:35:55 <morganava> ok anything else?
15:36:03 <morganava> other problems/things we've forgotten?
15:36:05 <donuts> did we chat about the first thing in the agenda?
15:36:23 <PieroV> I added it as a generic point, since we talked about having checkpoints
15:36:50 <donuts> well remembered :)
15:36:54 <PieroV> So, we mentioned how we're doing with audits and other blockers, and I don't think there's anything else that planned at this point
15:37:07 <PieroV> There would be a thing
15:37:21 <morganava> donuts: probably anything linked from the 'features' issue is highly likely to be kicked to 14.5 at ths point
15:37:22 <PieroV> And it's the new circuit for this site (aka something that replaces the old new identity)
15:37:40 <donuts> right, new circuit is the one I wasn't sure about
15:37:46 <morganava> but i need to go cleanup gitlab and see how we're sitting
15:37:48 <PieroV> And also a second thing: a checkbox to disable the internet test
15:38:13 <PieroV> As a workaround for the lack of a connection assist opt-in
15:38:21 <morganava> hmm
15:38:23 <donuts> HMM
15:38:28 <PieroV> We've received some complaints about that also recently
15:38:43 <PieroV> But I don't think we have the bandwidth for an actual opt-in for 14.0
15:38:50 <morganava> donuts: https://forum.torproject.org/t/lyrebird-can-prevent-tor-browser-from-working/14178/12
15:38:53 <donuts> it's been this way since Tor Browser 10.5 though, what's changed?
15:39:00 <PieroV> 11.0 or 11.5
15:39:05 <donuts> oh sorry, 11.5
15:39:05 <PieroV> But yes, I've been wondering the same
15:39:07 <donuts> you're right
15:39:16 <ma1> that someone noticed :)
15:39:19 <morganava> well this guy discovered it through some 3rd party fireall thing
15:39:25 <PieroV> Actually someone had already noticed
15:39:28 <PieroV> And we added a pref
15:39:33 <PieroV> But never added a UI for that
15:39:54 <morganava> hm
15:39:59 <donuts> Shouldn't the moat ping only happen after the user presses "Try a Bridge"?
15:40:06 <PieroV> No
15:40:12 <PieroV> Also before
15:40:15 <PieroV> For two reasons
15:40:24 <PieroV> 1. get the list of "frequent" region/countries
15:40:35 <PieroV> 2. the Internet test
15:41:05 <donuts> right, 1. seems optional?
15:41:17 <PieroV> We can ship a list of frequent regions offline
15:41:21 <PieroV> And update it later
15:41:40 <PieroV> When the user clicks on "Try a bridge" for the first time
15:41:57 <PieroV> Or after a bootstrap, if we really wish to have an updated list
15:42:05 <donuts> We could also disable the internet test
15:42:34 <donuts> I'm slightly worried about users reflexively checking the "no moat" checkbox and then breaking a bunch of circumvention stuff
15:42:36 <PieroV> Yep. There's already an invisible opt-out mechanism
15:43:01 <PieroV> It can be made opt-in
15:43:12 <PieroV> (but maybe something to discuss offline?)
15:43:24 <donuts> we had a conversation about this with a wider group including the Community and AC teams in Lisbon, actually
15:43:38 <donuts> I'll need to dig out the notes
15:43:57 <PieroV> We could totally disable the test on Android for now
15:44:01 <PieroV> Since there's no about:config
15:44:12 <PieroV> (at least, if we don't for desktop eventually)
15:44:34 <donuts> I don't think we need an internet test for android anyway
15:44:37 <donuts> doesn't android detect when it's offline regardless?
15:45:21 <donuts> but yes, disabling it on android-only seems like a good short term solution – because the longer-term solution needs some thought put towards it
15:45:26 <PieroV> donuts: yeah, that's another conversation
15:45:38 <PieroV> OS provide a way to test for your "online status"
15:45:51 <donuts> right
15:45:54 <PieroV> But we don't know how much stuff we can use (I think we can on Android, as it isn't an additional permission IIRC)
15:46:09 <PieroV> And in general we should check what Firefox offers, or what we should implement on top of it
15:46:13 <donuts> I think we're using it for the VPN
15:46:34 <donuts> but yes that all sounds reasonable
15:47:04 <donuts> if anyone wants to create issues for this stuff, please feel free to assign to felicia and loop me in :)
15:47:24 <donuts> which would mean our only UXy thing remaining is new circuit (pending some gitlab cleanup), correct?
15:47:50 <PieroV> donuts: yes, I think we can try to check if it's feasible now that we have our patchset
15:47:59 <PieroV> I can quickly sync with Dan and Claire later about this
15:48:19 <donuts> thanks! we can also loop felicia in for UX
15:48:28 <PieroV> donuts: there's already an issue, so will do
15:49:01 <PieroV> Done (https://gitlab.torproject.org/tpo/applications/tor-browser/-/issues/42655)
15:49:15 <donuts> awesome, thanks :)
15:49:24 <donuts> looking forward to being able to do multiple assignees in future
15:50:23 <donuts> oh one question
15:50:34 <donuts> what about mullvad?
15:50:39 <donuts> anything else UXy there? letterboxing backgrounds perhaps?
15:50:56 <donuts> or are we bumping mullvad related things too?
15:51:12 <morganava> for 14.0?
15:51:17 <donuts> ya
15:51:28 <morganava> hmm, rui has opened a handful of issues
15:51:30 <morganava> let me go skim
15:51:34 <donuts> I'm just trying to get a handle on what I should be throwing at Felicia atm
15:51:41 <donuts> but it's a little confusing
15:51:42 <morganava> none of them stood out as being immediate needs tho
15:52:04 <catladyfelicia> Just got a notification about the issue :)
15:52:20 <morganava> mullvad-browser#266
15:52:28 <morganava> mullvad-browser#328
15:52:44 <donuts> morganava: ty
15:52:44 <morganava> mullvad-browser#329
15:52:45 <PieroV> Oh!
15:52:48 <PieroV> That reminds me
15:52:51 <PieroV> We have another blocker for 14.0
15:52:58 <PieroV> Search engines
15:53:03 <morganava> ?
15:53:09 <PieroV> So, since the configuration changed, I restored only DuckDuckGo
15:53:25 <PieroV> As we've had the issue about reviewing the list for a long time
15:53:38 <PieroV> 13.5 still includes Yahoo, IIRC, just saying :D
15:54:24 <PieroV> Worst case scenario I'll choose :D
15:54:43 <donuts> morganava: I'm not sure about #329
15:54:54 <donuts> at least it's a little late in the release cycle to respond to that one
15:55:12 <morganava> hm i'm happyto do a little purge
15:55:26 <donuts> woops lol
15:55:32 <PieroV> The search engine issue is https://gitlab.torproject.org/tpo/applications/tor-browser/-/issues/41835
15:55:41 <morganava> if you make a MR and include a list of what we used ot have vs what you've added back that'd be a good start I think
15:55:58 <morganava> and we can agonise over the particulars there
15:56:10 <PieroV> morganava: what if I used donuts's proposal from that issue instead?
15:56:17 <PieroV> DDG, StartPage and Wikipedia
15:56:26 <morganava> perfect
15:56:27 <PieroV> Remove Goog, Yahoo, Twitter, YouTube
15:56:38 <morganava> yes please!
15:56:40 <ma1> Twitter !!!
15:56:56 <PieroV> DDG and StartPage both with their Onion counterpart
15:56:59 <morganava> it's a shame there isn't an onion-service focused search engine
15:57:02 <donuts> I think that's a good start, we can always make more changes in future
15:57:19 <donuts> otherwise I think we'll risk discussing this issue non-stop
15:57:24 <morganava> yes
15:57:26 <morganava> exactly
15:57:29 <ma1> +1
15:57:31 <morganava> enough talk time for action etc
15:57:42 <donuts> haha exactly
15:57:47 <PieroV> (timecheck)
15:57:48 <donuts> okay thank you both for the list of issues
15:57:57 <donuts> I'll make sure felicia is aware those are the UX issues for 14.0
15:57:59 <morganava> donuts: I think the user-configurable background could be a good task for felicia
15:58:03 <donuts> yeah agreed
15:58:16 <donuts> wait
15:58:18 <morganava> i douobt we'll have dev capacity to iplement anything for 14.0 unless henry-x is stuck with nothing to do
15:58:29 <morganava> but early in the 14.0 cycle we could backport
15:58:34 <donuts> yeah I'm thinking MB-themed background for 14.0
15:58:34 <donuts> anything else for 14.5
15:59:01 <donuts> when we get closer to 14.5, we'll need to take a look at the full list of features we're proposing
15:59:04 <donuts> and then sort by priority
15:59:11 <ma1> if it's just the fixed mullvad-theme bg we can do it for 14. Configurable is 14,5
15:59:18 <donuts> because that list has already grown quite long
15:59:25 <donuts> ma1: yep agreed
16:00:17 <morganava> ok
16:00:19 <morganava> that's the hour
16:00:35 <morganava> anything esle that needs to be squeezed in?
16:00:59 <donuts> all good!
16:01:07 <ma1> all hands!
16:01:11 <morganava> ahaahh
16:01:13 <morganava> yes
16:01:15 <morganava> cya folks o/
16:01:16 <morganava> #endmeeting