[04:49] <didrocks> good morning
[05:00] <duflu> Hi didrocks
[05:00] <didrocks> hey duflu
[05:50] <jibel> hi all
[05:53] <didrocks> salut jibel
[05:56] <duflu> Hi jibel
[06:24] <seb128> hey, good morning desktopers
[06:28] <didrocks> salut seb128
[06:29] <seb128> lut didrocks, en forme ?
[06:30] <didrocks> chaudement
[06:30] <didrocks> et toi ?
[06:30] <duflu> Hi seb128
[06:30] <seb128> ça va bien, c'est tombé de 10°C hier, ça fait du bien
[06:30] <seb128> hey duflu, how are you?
[06:30] <duflu> seb128, OK I guess. How are you?
[06:33] <seb128> I'm good
[06:33] <seb128> trying to think over coffee what we can do to make you hate the trello monthly columns less :)
[06:40] <duflu> seb128, I think another part of the problem is that we want to track SRU work. So the board itself probably shouldn't be called 19.10, but something like "April to October 2019"
[06:41] <duflu> Or a continuous rolling multi-cycle board, with a "Done in 19.10" column
[06:42] <duflu> But those are orthogonal issues to the number of columns
[06:43] <duflu> And a "Done in 18.04.3 column", actually
[06:44] <duflu> And a "Done in 18.04.3" column, actually
[06:48] <duflu> Also, the new wallpaper feels unclean, IMHO :)
[06:48] <duflu> We could just drop the actual Ubuntu wallpaper in there
[07:08] <marcustomlinson> morning!
[07:12] <duflu> Morning marcustomlinson
[07:31] <Wimpress> Morning o/
[07:34] <oSoMoN> good morning desktoppers
[07:35] <duflu> Hi Wimpress and oSoMoN
[07:35] <oSoMoN> hi duflu
[07:35] <oSoMoN> hey Wimpress
[07:37] <duflu> tjaalton, https://lists.freedesktop.org/archives/mesa-announce/2019-June/000525.html
[07:38] <tjaalton> duflu Uploading mesa_19.0.8-0ubuntu1.dsc
[07:40] <tjaalton> if it would not time out this time
[07:40] <tjaalton> didn't, so it's done
[08:01] <Laney> morning!
[08:03] <duflu> Hi Laney
[08:03] <duflu> Thanks tjaalton
[08:12] <seb128> lut oSoMoN
[08:12] <seb128> hey marcustomlinson Laney willcooke
[08:12] <seb128> how are you?
[08:12] <oSoMoN> salut seb128
[08:13] <seb128> duflu, now you want more columns? ;)
[08:13] <oSoMoN> I'm good, how are you?
[08:13] <Laney> hi duflu hi seb128 hey oSoMoN
[08:13] <oSoMoN> hey Laney
[08:14] <duflu> seb128, not really related. On the columns question, I like general non-dated ones. Because dates columns really necessitate a new in progress/blocked/in review/done per month
[08:14] <duflu> That's too many
[08:14] <duflu> And we have a hybrid right now, which is itself problematic
[08:15] <seb128> duflu, the fact that we are not able to drive things through/get blocked by outside factor makes it a bit challening yes :/
[08:15] <duflu> If you want to incorporate months then maybe just pick some colourful tags instead?
[08:15] <seb128> we already have more tags than trello has colors...
[08:15] <duflu> Excellent
[08:20] <didrocks> hey Laney, oSoMoN
[08:20] <didrocks> & marcustomlinson, Wimpress
[08:20] <didrocks> & willcooke
[08:20] <oSoMoN> salut didrocks
[08:21] <didrocks> currently rewritting awk in bash… and the bash version is, in the end, shorter…
[08:21] <didrocks> surprising
[08:21]  * didrocks goes back
[08:22] <seb128> didrocks, mawk and gawk doesn't have a common syntax that things can be based on?
[08:22] <duflu> Rewriting awk or rewriting some awk script? :)
[08:23] <didrocks> seb128: mawk doesn't have arrays, which was the reason we used awk (and in fact gawk) instead of dash scripting
[08:23] <didrocks> well, it has arrays, but not multi-dimensional
[08:24] <seb128> duflu, going back to the trello, I think the month "started" column is small enough that it's fine to no split out things blocked or in review, that should just be reflect in the status of the card
[08:24] <didrocks> duflu: would be better :)
[08:24] <seb128> if they are signed up to be shipped during that month we don't get to discard/bail out because things needs review
[08:24] <seb128> didrocks, ah, I see
[08:25] <didrocks> at least, we'll have 396 tests to proove we didn't regress ;)
[08:25] <seb128> :)
[08:26] <duflu> seb128, that implies the card owner has any control on which month it gets approved and shipped in. If you don't have that control then you end up with a growing backlog month-to-month. Sometimes in shrinks but over the long term it may only grow
[08:27] <duflu> That's where I am now. I have a collection of unreleased cards which I can't do anything more with. The pile will only grow each month, on average
[08:27] <seb128> duflu, if you are not able to control the outcome then the work shouldn't be signed on a monthly column, we still have the untargetted ones
[08:27] <duflu> seb128,  that's everything then
[08:27] <seb128> for you maybe
[08:27] <seb128> the goal of the monthly target was to track things we sign off for shipping that month
[08:27] <seb128> there is plenty of things in that category
[08:28] <seb128> GNOME updates, SRUs, our tooling improvements
[08:28] <seb128> blog posts
[08:28] <duflu> In that case I should have been using the untargeted column
[08:28] <seb128> yes, I think that was pretty much the outcome when we discussed it at the start of the month
[08:28] <duflu> seb128, that discussion was about performance work, which is still in untargeted
[08:28] <seb128> your work is ongoing foundation work that relies on upstream reviews
[08:29] <seb128> not really something where you can target specifics for the month
[08:29] <willcooke> morning all
[08:29] <duflu> Morning willcooke
[08:29] <seb128> hey willcooke
[08:29] <seb128> duflu, well some make sense to target for the current iteration, like SRU regressions
[08:29] <seb128> we own the SRU and ought to unblock it one way or another
[08:30] <seb128> we have the control of reverting or distro patching so we shouldn't end up blocked by upstream/review
[08:30] <duflu> Yes I guess the team as a whole always has enough power to unblock those
[08:31] <seb128> duflu, I think that the rule of thumb is that things in the monthly column are things we don't want to see carried over to the next month/consider a failing if we end up doing that
[08:31] <seb128> those we are usually rolling over should be in the untargetted ones
[08:53] <duflu> seb128, I wonder if some custom stickers would be a good way to show "in review" or "blocked" within the monthly column?
[08:54] <duflu> Looks like we'd have to provide our own images
[09:11] <seb128> yeah, some tag/sticker for that would be useful
[09:23] <Trevinho> morning
[09:26] <seb128> hey Trevinho, slightly shifting to night work hours again?  ;)
[09:27] <Trevinho> seb128: you know, I don't give the best if I don't see the darkness :)
[09:27] <seb128> :)
[09:46] <seb128> bah, I can't reproduce that cogl i386 build hang neither on a chroot nor on a porter box, annoying :/
[10:00] <Trevinho> Laney: so for mutter thing let me know how to proceed, I can see in checking a better fix meanwhile, however if we want mute the crash (not super-high prio as it happens on shell reload only, but still... Might lead to close all the session if g-s doesn't restart the shell properly) that is the faster way also to get .2 in disco, if instead we want the proper fix you know might need the whole upstream
[10:00] <Trevinho> review cycle. So I'd go with this while I propose something else.
[10:01] <Trevinho> I've proposed the git snapshot that includes it, let me onw if you prefer me to drop it
[10:02] <Laney> Trevinho: jadahl tried to ping garnacho about some other idea, might be nice to sync about that
[10:03] <Trevinho> Laney: yeah I read it, and I was already looking how to do it, but I'm not sure isn't much different from what in theory we tried to do
[10:03] <Laney> whatever makes them happy :-)
[10:04] <Trevinho> but in case that would be the road, I'm just saying that we can in the mean time go with this. But if you want to wait more or drop the change instead let me know and I'll adapt the MR
[10:04] <Laney> ok thx, will look
[10:04] <Trevinho> as I wouldn't wait other weeks for .2
[10:09] <oSoMoN> buon giorno Trevinho, I think you mentioned you needed input from French natives on bug #1832875 yesterday, is that still the case and if so what do you need exactly?
[10:10] <Trevinho> oSoMoN: buen día oSoMoN
[10:10] <Trevinho> so... the SRU test case should be amplied, not sure how to test the resolution
[10:11] <Trevinho> what's wrong, not knowing well the french layouts, I don't know how they should be
[10:16] <seb128> Trevinho, https://kali-8fa5.kxcdn.com/wp-content/uploads/2013/10/azerty-qwerty.jpg
[10:16] <seb128> basically check that top-left letter is a A and not a Q
[10:18] <Trevinho> ah, ok I was quite sure it should have been azerty, but you guys know it better :)
[10:18] <seb128> :)
[10:26] <oSoMoN> I personally use a different layout (bépo, a French dvorak variant), but I don't know whether the OSK is expected to offer "exotic" layouts OOTB
[10:54] <duflu> oSoMoN, https://wiki.gnome.org/Design/OS/ScreenKeyboard
[10:54] <duflu> And good night
[11:00] <seb128> oSoMoN, I don't think bepo is going to be handled
[11:01] <seb128> they stated on https://gitlab.gnome.org/GNOME/gnome-shell/issues/475 that they don't intend to handle random variants
[11:01] <gitbot> GNOME issue 475 in gnome-shell "OSK letter disposition does not change when different languages are choosen" [Closed]
[11:01] <seb128> their script is https://gitlab.gnome.org/GNOME/gnome-shell/blob/master/data/update-osk-layouts.sh I think
[11:01] <seb128> which uses www.unicode.org/Public/cldr/latest/keyboards.zip as a source
[11:01] <seb128> that has a qwertz variant
[11:01] <seb128> but not a bepo one
[11:21] <oSoMoN> that's fair enough, until I start using a touchscreen for real keyboard input, then I'll complain loudly
[15:26] <Laney> the sky is ... blue
[15:27]  * Laney calls the military, possibly some kind of weird alien thing is going on, it's supposed to be grey
[15:32] <teward> Laney: actually the sky is orange, you just think it's blue :)
[15:33]  * Laney blames the chemtrails
[17:10] <Eickmeyer> @Laney, teward: sky is gray here. Bland and boring. Just like my soul.
[17:10] <Eickmeyer> (no idea why I went @ on that)
[17:59] <willcooke> night all