17:30:53 <GeKo> #startmeeting tor browser 7/29
17:30:53 <MeetBot> Meeting started Mon Jul 29 17:30:53 2019 UTC.  The chair is GeKo. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:30:53 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:30:55 <GeKo> boom
17:31:05 <GeKo> hello everyone!
17:31:10 <sisbell> hi
17:31:13 <sysrqb> o/
17:31:14 <pili> hihi
17:31:17 <mcs> hi
17:31:20 <acat> hi!
17:31:21 <GeKo> we have another exiting sync on all things tor browser this week
17:31:39 <GeKo> the meeting pad is at https://storm.torproject.org/shared/tHoN4Ii7rLSjPE0OP4gydX4cMGadsXmRQNc-6lwru0N as usual
17:31:56 <GeKo> please add your items and mark them bold
17:32:10 <boklm> hi!
17:33:04 <GeKo> pili: did the one for the network settings in general settings one from last week get resolved?
17:33:09 <pili> ah, no
17:33:20 <pili> I forgot to mention during our sync
17:33:27 <GeKo> okay.
17:33:36 <GeKo> i think that's a new ticket i can file later
17:34:17 <boklm> GeKo: #30126 does not seem related to Google 64bit requirement. Is that a typo?
17:34:35 <antonela> o/
17:35:01 <pili> thanks GeKo
17:35:18 <GeKo> boklm: it is, indeed, fixed. thanks
17:36:00 <GeKo> okay
17:36:11 <GeKo> nothing in bold today?
17:36:26 <GeKo> now is the last chance before we move to discussion items :)
17:37:35 <GeKo> alright, discussion time then
17:37:39 <GeKo> pili you are up
17:37:59 <pili> oh, we're not doing the ESR68 switch for nightlies discussion first?
17:38:22 <pili> anyway, I can go first
17:38:33 <GeKo> no, i thought we do it as the last item
17:39:15 <pili> so, I think one of the items from the retrospective was the fact that it would be good to start estimating our work so that we have a better idea when we can get things done by
17:39:39 <pili> this is also related to figuring out what is our team's capacity
17:40:10 <pili> I would like us to start thinking about this during august with a view to start estimating September tickets
17:40:27 <sisbell> I think we would need to start breaking issues into much smaller pieces that are easier to estimate
17:40:28 <pili> and I thought it would be good to discuss how we can start doing this
17:40:33 <pili> sisbell: definitely :)
17:41:19 <pili> so, one easy way to start is to follow some of what the network team is doing, which is equating 1 day to 1 point and adding these to the ticket in trac under the points field
17:41:52 <GeKo> 1 day = 8hours?
17:42:01 <pili> GeKo: thanks for that, yes
17:42:03 <GeKo> (and the fractions of that for smaller items?)
17:42:10 <GeKo> *then
17:42:18 <pili> I don't think we should go below 0.5 points though
17:42:28 <pili> or 0.25 as the smallest
17:42:37 <pili> one step before that would be to think about what the team's capacity is, as in, how many days a week are realistically spent coding
17:42:59 <pili> so for the network team, it's 3 points per week
17:43:00 <pili> per person
17:43:24 <pili> so that means that the other "points" are spent doing things like email, meetings, code reviews, etc...
17:43:27 <sisbell> We will need to take into account build times to verify fixes, as build times can be very long.
17:44:02 <pili> good point, so I would say build times would add up into this other category i.e non-coding time
17:44:28 <pili> and therefore would be deducted from capacity
17:44:31 <pili> any questions so far? :)
17:45:13 <sysrqb> i like this idea
17:45:33 <GeKo> +1
17:46:33 <GeKo> pili: but if the build takes 10hours that does not mean you spent 10hours in the non-coding time
17:47:14 <pili> no, but maybe there is some context switching time
17:47:38 <GeKo> we need to find a better model than that
17:47:45 <GeKo> yes
17:48:03 <pili> or maybe the build fails and you need to restart it or tweak things and that's not strictly coding time, not sure how that works in practice for people :)
17:48:16 <pili> this is where I need everyone's help and input
17:48:58 <pili> so, maybe we can start with 3 points per person per week and tweak after a few months?
17:49:31 <GeKo> wfm
17:49:52 <pili> also people work different numbers of hours
17:50:03 <pili> so we'll have to account for some individual differences
17:50:19 <pili> ok
17:50:58 <pili> so the next step will be to start adding points to tickets when we do the roadmap for september
17:51:10 <pili> we can do this 2 ways
17:51:58 <pili> 1) everyone votes on points for a ticket (not sure how we'll do this in practice), we discard the highest and lowest estimates and take the median
17:52:21 <pili> 2) the person assigned to the ticket decides the points <- this will be a lot easier
17:53:03 <pili> assigned = ticket owner?
17:53:27 <pili> any ideas? :)
17:53:33 <mcs> What does the network team do?
17:53:43 <GeKo> i think we should start with 2)
17:53:46 <mcs> Option 1) could collectively be very time consuming
17:53:51 <pili> mcs: option 2
17:55:01 <sysrqb> i agree with start with option 2
17:55:23 <pili> ok, so then we roadmap and make sure that the work planned for the month does not exceed capacity given the 3 points per person per week ;)
17:55:52 <mcs> How does the network team account for unplanned work? Re-do their roadmap?
17:56:32 <mcs> I guess we will need to adjust continuously :)
17:56:50 <GeKo> pili: the 3 points were for a 5 day week, right?
17:56:59 <pili> mcs: I need to check with gaba but I guess it depends how much notice we have on unplanned work :)
17:57:00 <pili> GeKo: yup
17:57:05 <pili> so we may want to adjust to 2.5
17:57:13 <GeKo> so, this needs to get adjusted as well
17:57:16 <GeKo> yeah
17:57:29 <GeKo> at least for those not having 5 days (anymore)
17:59:32 <pili> anyway, regarding unplanned work, it's an estimation it's not an exact science :) when we have unplanned work we know that our estimations were probably correct but we had some other stuff that came up
18:01:17 <pili> ok, any other questions? comments?
18:01:30 <GeKo> none from me
18:01:45 <sysrqb> sounds like a plan
18:01:51 <pili> ok, let's see how it goes :)
18:02:19 <pili> maybe the last meeting of august we can review points/estimations
18:02:24 <pili> anyway, I think that's all I had on that
18:02:42 <antonela> should we made estimations just for the current month?
18:03:39 <pili> so, we'll start making estimations for september roadmap items
18:03:48 <pili> we can make them earlier if we want
18:04:01 <antonela> i see
18:04:06 <pili> but when we do the monthly roadmap we should make sure we have the points to ensure we are not planning more work than we can handle
18:04:36 <antonela> groot
18:05:39 <GeKo> pili: okay, the frontdesk thing?
18:05:46 <pili> yup
18:06:07 <sysrqb> sounds weird.
18:06:18 <pili> so basically I noticed that we're getting a bunch of empty emails on frontdesk that are clearly sent from mobile phones
18:06:24 <pili> and I was just wondering what that was about
18:06:35 <pili> I can try to reply to some to ask how we got that message
18:06:35 <sysrqb> good question.
18:07:19 <mcs> Do we know where the frontdesk address is advertised? Our website? Orfox?
18:07:53 <sysrqb> i wonder if someone can look at the full message (headers) and see if there's any useful information
18:08:00 <sysrqb> maybe about the user-agent
18:08:07 <pili> sysrqb: I can try to find that out
18:08:18 <sysrqb> are they all from the same address?
18:09:10 <pili> nope
18:09:44 <pili> let me do a more thorough review and I can create a ticket with the findings
18:09:44 <sysrqb> hrm
18:09:55 <sysrqb> mcs: it's at thebottom of https://www.torproject.org/contact/
18:09:57 <mcs> My experience is that users on mobile devices often don’t know how to leave “compose mode” without sending a message,
18:09:58 <mcs> 29197
18:09:59 <mcs> 29197
18:10:08 <mcs> oops; lost my second line
18:10:16 <pili> mcs: ah, yes, that could be it
18:10:19 <mcs> … so you will end up with a lot of empty messages
18:10:33 <sysrqb> yay mobile
18:10:34 <mcs> Not clickable on that contact page I don’t think
18:10:59 <sysrqb> that is true,only text
18:11:18 <mcs> weird
18:11:48 <pili> ok, so I'll reply to some and try to figure out what might be happening
18:12:19 <sysrqb> thanks :)
18:13:29 <GeKo> okay, i guess that means we are at the last dicussion item?
18:14:22 <GeKo> i just saw pili shared the tbb-9.0-must-nightly link last week already
18:14:42 <GeKo> that's not meant the way that we need to solve all those tickets before any nightly shows up
18:14:56 <GeKo> i like to move faster here
18:16:01 <GeKo> linux is pretty close to be working i think
18:16:15 <GeKo> so we should get that out as fast as possible
18:16:36 <GeKo> my current hope is we have something ready by early next week
18:17:39 <GeKo> as i'll be offline for another couple of days later on
18:17:54 <mcs> brade and I will try to produce a fix for #29197 very soon (the desktop browser is less usable without a working Tor Launcher)
18:18:12 <GeKo> great
18:18:55 <GeKo> just the are minimum
18:19:00 <GeKo> *bare
18:19:20 <acat> and i'll update #30429 with a new branch addressing the comments and rebased to esr68 gecko-dev branch
18:19:34 <GeKo> that should be enough here for now
18:19:39 <acat> i'm just finishing porting some search engines to webexts
18:20:01 <acat> i mean xmls -> webexts
18:20:16 <GeKo> great
18:20:29 <GeKo> i'll review stuff tomorrow
18:21:29 <GeKo> any comments? suggestions?
18:23:22 <sysrqb> nope :)
18:23:44 <GeKo> good, good!
18:23:47 <GeKo> let's do it
18:24:04 <GeKo> i hope as well, we have sorted the linux build related pieces out this week
18:24:11 <GeKo> and then we can iterate from that on next week
18:24:22 <GeKo> this will likely happen like last time:
18:24:36 <GeKo> the builds for other platforms will be broken
18:24:46 <GeKo> and will get added piece by piece again
18:25:05 <GeKo> while we find and fix issues with the already existing nightly builds
18:25:20 <GeKo> given that most issues are platform independent anyway
18:25:50 <GeKo> alright
18:25:57 <GeKo> anything else for today?
18:28:40 <GeKo> thanks then and happy week to everyone *baf*
18:28:43 <GeKo> #endmeeting