=== jono is now known as Guest51 [05:31] smspillaz, I'm looking at http://pad.lv/879237 , Specifically I'm trying to debug why the mouse is offset when citrix is fullscreen [05:31] Launchpad bug 879237 in unity (Ubuntu) "Unity does not allow citrix receiver remote desktop to open full screen" [Medium,Confirmed] [05:32] what I'd like to do is prove whether Unity is passing the wrong pointer location to the app or if it's the app interpreting it wrong. [05:33] I'm looking for a tool like xmon. [05:33] but apparently that doesn't exist any more. [05:33] Does anyone know of a replacement for xmon? === chiluk is now known as chiluk_away [08:34] didrocks, hey [08:34] hey jibel! [08:34] I think I found what's going wrong with utah and autopilot [08:35] really? \o/ [08:35] there is a sed command that modifies the preseed and creates a broken rc.local [08:35] packages are installed from this rc.local [08:35] this is the sed command we talked about yesterday? [08:36] which means that when you run the indicator stack packages are not installed correctly [08:36] didrocks, yup, that thing http://paste.ubuntu.com/1623672/ [08:36] jibel: argh, indeed, this is not good :/ [08:36] (I replaced /etc/rc.local by /tmp/test.local for testing) [08:36] any idea why it's passing on other arch than intel? [08:37] didrocks, no idea [08:38] jibel: nice catch anyway! :-) [08:38] jibel: let's try it once the current unity tests are done [08:39] actually I was so pissed that I had to look closer :) [08:41] jibel: ahah, I know now that it's good to pissing you! :) [08:41] jibel: I'll keep you posted once the machines are free [08:49] jibel: so… the tests machines are free! [08:49] jibel: you can relaunch the indicator tests with whole ppa when you need it :) [08:50] (you need the whole ppa for the new window-mocker and autopilot) [08:50] didrocks, we need a fix in UTAH first [08:50] ok, should I join the qa channel? [08:50] and pay beers/praying? ;) [08:51] didrocks, why did you leave it in the first place ;) [08:51] ahah [08:52] didrocks, but you can ask gema to boost the priority of this bug because AP jobs are guaranteed to fail or not testing the right thing if whole_ppa is disabled. [09:21] chiluk_away: unity doesn't "pass" the pointer location to applications [09:46] didrocks: FYI -- fginther implemented the buildless merge for changelog only MPs. It is enabled. [09:46] mmrazik: excellent! on all projects? [09:46] didrocks: yes [09:52] mmrazik: thanks for the notice! [10:10] Trevinho: yom did you read my comments from yesterday regarding the tests? I brought my raring laptop to work today so we can walk over it if you have few minutes. Thanks [10:55] ... or if anyone else can help here i wouldn't mind either :-) [11:02] didrocks: do you know who knows/maintains ccache? I would like to know if ccache supports compiling with coverage flags [11:03] from what I can see I would say it just does nothing but I wouldn't mind if somebody who knows ccache better confirms [11:03] or if there is some workaround [11:05] mmrazik: we are in sync with debian. Joel Rosdahl is the debian maintainer [11:05] didrocks: thx [11:05] mmrazik: http://qa.debian.org/developer.php?login=joel@debian.org [11:05] If you run 'ccache -s' before and after compiling your source (for the second time) the stats will tell you if object files were reused. [11:05] didrocks: it looks like this fact is affecting our jenkins performance more than anything else [11:06] satoris: cool.. Will try that to confirm [11:06] good hint :) [11:06] compiz build time is 4x bigger when compiling with coverage (after the ccache was initialized by the first build) [11:06] Cache hit (direct) should increase. [11:55] satoris: thanks for the hint. Indeed the only number that gets higher with coverage is "unsupported compiler option" === _salem is now known as salem_ [12:46] sil2100: hey, how are you? :) [12:48] didrocks: hello! Just got back from the city, staying and working at my family's home - equipment set up and ready for action [12:48] didrocks: how about you? [12:48] didrocks: I'm testing andyrock's fix for the quicklists now (building the packages) [12:49] sil2100: I'm great, jumping between a lot of different topics and start of being tired to ensure I'll enjoy the week-end :-) [12:49] sil2100: oh before that, you want action! [12:49] so action :-) [12:49] sil2100: https://jenkins.qa.ubuntu.com/job/ps-indicators-autopilot-release-testing/label=autopilot-intel/112/ has 2 failures [12:49] ;) What's up? [12:49] making the landing failing [12:49] do you think there is a quick fix? [12:49] Ok, let's see which ones are failing [12:49] only intel [12:49] oh [12:50] nvidia passed [12:50] ati has just one [12:50] (and a different one of course) [12:51] Interesting, looking at the videos now [12:51] sil2100: stop watching TV! :-) [12:51] * didrocks runs… [12:51] :| [12:51] ;) [12:51] seems the TV shows are not entertaining though :p [12:52] * sil2100 likes watching autopilot click and type in Unity [12:52] ahah, I knew it! :) [12:53] luv: ok, here? [12:54] Checking the emulator code, since it seems like some strange timing problems, i.e. unity not reacting fast enough and input events getting lost, but this shouldn't happen normally - maybe some additional checks in the HUD could help [13:06] Trevinho: yup :-) im back now [13:08] so the two problems with writing the unit tests are - can't use GetMenus() on mock_icon (b/c it's protected) and mock->app has no windows_ attribute [13:09] luv: I'm giving a look to your branch with hands on, give me few minutes === larsu_ is now known as larsu [13:30] Ho, interesting === dandrader is now known as dandrader|afk [13:53] luv, so... feature side we're ok... [13:53] luv: I've cleaned it up a little and added a test here http://pastebin.ubuntu.com/1624683/ [13:56] luv: so, just do a bzr merge lp:~3v1n0/unity/quicklist-windows-improvements on your branch [13:56] and push it again on yours [13:56] looks great, thanks a million :-) [13:57] luv: we'd have to improve a little one more thing though... :/ === dandrader|afk is now known as dandrader [13:58] luv: that's easy to do... but right now if a window has _ in the title, then it gets drawn as an underscore (i.e. just mkae a new folder /tmp/foo_bar and cd on it in a terminal...). you'll notice the problem [13:59] alright [14:01] oh i see, can we just set a property on dbusmenuitem not to interpret "_"? [14:03] luv: we can once we implement it :) [14:03] luv: it's quite easy... [14:03] didrocks: ok, found one reason for the test failures [14:03] didrocks: those don't happen every time, since it's a time-related issue - will be trying to fix that with a merge request in a moment [14:04] luv: it's basically like doing https://code.launchpad.net/~3v1n0/unity/quicklist-max-label-width/+merge/145862 [14:04] luv: where in case that the new property is enabled, you've to skip the pango_layout_set_markup_with_accel(layout, _text.c_str(), -1, '_', nullptr); into QuickListMenuItem::DrawText [14:05] how can one do "git checkout file" in bzr? [14:05] forget local uncommited changes [14:05] bzr revert [14:05] didrocks: the problem is that when changing the keybinding for HUD in the alternative keybinding tests, when the keybindings are 'restored' back to normal, the restore compiz_option method called then on cleanup doesn't wait a short while for compiz to notice and apply the change properly [14:06] just do bzr revert; bzr merge lp:~3v1n0/unity/quicklist-windows-improvements; bzr commit -m "Mergint with lp:~3v1n0/unity/quicklist-windows-improvements branch" [14:06] luv: ^ [14:06] didrocks: so, sometimes when everything is executed fast enought, Alt is being pressed while compiz still waits for Ctrl+Alt+h as the HUD keybinding, even though it had been changed back to Alt during the last test [14:06] ;) [14:06] * sil2100 fixes that [14:07] yeah i get it now, i have to run commit manually too [14:07] luv: yep [14:08] luv: want to do that final menu thing? [14:10] well, honestly, im not sure if i have time until wednesday. but if it's really as small a change as it looks I should be fine. [14:10] luv: as i said you ony need to follow what done in the branch I linked before, you only have to add a property like MARKUP_ACCEL_ENABLED_PROPERTY [14:11] didrocks: the problem is that the fix will have to go to lp:autopilot sadly :< [14:12] ok, what is confusing me is that it defines a class QuicklistMenuItem but we use DbusMenuitem [14:12] luv: well dbusmenu is rendered by a quicklistmenuitem [14:12] yeah i still would be setting the property on dbusmenuitem though, right? [14:13] luv: yep, however don't worry I'm doing that too... [14:14] yeah, it's better that way. [14:19] sil2100: don't worry, we can rebuild autopilot if needed [14:19] sil2100: nice hunt! [14:24] Trevinho, you made friends by removing the systray whitelist it seems ;-) [14:25] seb128: that was not me, but andy.. :) Anyway we just what what has been decided in bugs ;) [14:25] s/what/did/ [14:25] Trevinho, you are the one that +1ed it :p [14:25] seb128: ah, well :) [14:26] Trevinho, btw congrats on becoming a libwnck maintainer [14:26] seb128: well, no one was taking care of it anymore and we still need it [14:26] seb128: I missed the gnome git push bit, really :) [14:26] Trevinho, the git looks weird [14:28] Trevinho, ignore that, I got confused by the lack of 3.6 releases, the tags were 3.4 in there [14:28] Trevinho, do you plan to roll a tarball or should I have a look at backporting your patches? [14:28] seb128: how you prefer... I'm still newbie as gnome maintainer :) [14:29] Trevinho, tarball makes things easier if there is a stack of patches (which seems to be the case there) [14:29] seb128: well yep... should I announce something or can I just tag the revision, make a tarball and push it in the archives? [14:30] Trevinho, https://live.gnome.org/MaintainersCorner/Releasing [14:31] seb128: yeah, I didn't read all the details since I was not planning to release soon, but at this point it's a good read [14:31] Trevinho, in practice not everyone bother sending an announce for releases so your call if you want to do it or not [14:34] seb128: aok [14:35] luv: yuu can merge again... now looks fine now and we can push approve the branch to merge it in trunk === dandrader is now known as dandrader|afk [14:53] fginther: hello! [14:53] sil2100, hi [14:54] fginther: could you review a quick change for autopilot I proposed? I think it might help in fixing some random failures, 2 of which we encountered during the latest run of the indicator tests [14:54] https://code.launchpad.net/~sil2100/autopilot/set_compiz_option_cleanup_fix/+merge/147379 [14:54] sil2100, yes [14:54] fginther: thank you! [14:57] cyphermox: hey, how are you? [15:01] Trevinho: here we go :-) [15:01] luv: I've found another small issue, pushing in few minutes [15:04] luv: ok, you can pull/merge again [15:05] didrocks: pretty good you? [15:06] cyphermox: I'm great, thanks! [15:06] cyphermox: I was wondering if there was any reason that you didn't publish the misc stack yet (which is in manual publishing mode), do you need any help? [15:06] also, I think mterry (hey!) can publish the unity stack without the indicator one again, can you confirm? [15:07] didrocks, hiyo [15:07] didrocks, indicator tests are actually failing though [15:08] mterry: right, but the unity stack is fine, for the indicator tests failing, sil2100 has an incoming fix for autopilot [15:08] didrocks, do our thresholds mean nothing, sir!? [15:08] (https://code.launchpad.net/~sil2100/autopilot/set_compiz_option_cleanup_fix/+merge/147379) [15:08] mterry: it's a timing issue :) [15:08] didrocks, sure, OK. I can manual [15:08] mterry: so once the misc stack and unity are published, we can have the misc stack rebuilding autopilot [15:09] didrocks: no, I just didn't get to it yet [15:09] the indicator stack using this autopilot retested [15:09] didrocks: waiting for merge approval right now ;) [15:09] and then retired in island :) [15:09] didrocks: cu2d-run -P misc, right? [15:09] cyphermox: right, if the packaging changes putting misc in manual mode suits you :) [15:09] (see the artefacts) [15:10] right [15:11] sure, done [15:12] didrocks: we'll need to deal with autopilot-gtk for jibel, too [15:12] cyphermox: yeah, let's do that now? [15:12] (well, after the publishing) [15:13] cyphermox: as we need to relaunch for the autopilot awesome fix that sil2100 did (and I hope is going to be merged soon) [15:13] fginther: if you have no time to review, I can do it btw (we are a little bit blocking on that fix) [15:14] didrocks: you mean the set_compiz_option_cleanup_fix? [15:14] cyphermox: yeah, so what we are doing is: [15:14] 1. manually publishing the misc task [15:15] 2. mterry is manually publishing the unity stack [15:15] 3. once this autopilot fix is merged and we bootstrapped autopilot-gtk, rebuild the misc stack and publish [15:15] 4. relaunch the indicator tests with the new autopilot, tests should pass [15:15] ok [15:15] 5. profit :) [15:15] luv: thanks... so we can now approve it I think [15:15] well I already triggered the publish for misc [15:16] sweet! [15:16] cyphermox: let's add autopilot-gtk, was there anything missing? [15:16] yeah, both unity and misc are published [15:16] I remember the dep on libindicate-dev was removed [15:16] thanks cyphermox, mterry for the publishing :) [15:16] sil2100's branch is approved too, so it should get merged soon [15:16] yeah [15:16] didrocks: I'm just going to run it through sbuild to be sure [15:16] cyphermox: cool! [15:16] sil2100, reviewed [15:17] cyphermox: tell me how it went [15:17] then we can add it [15:17] sil2100, if you need a fix right now, I can approve as is. [15:17] fginther: it is approved already by cyphermox :) [15:17] but great to have this double checking [15:17] fginther: can you run the merge now? I would hope we don't publish the indicator stack too late on a Friday evening :) [15:18] I trust our tests, but still… ;) [15:18] didrocks, sure [15:18] thanks ;) [15:25] fginther: thanks! [15:25] sil2100, yw [15:25] fginther: I commented on the issue btw. [15:26] luv: it's in merge queue now, prepare for some omgubuntu love! :) [15:26] fginther: could you check the merge proposal? Since I elaborated a bit on that topic [15:26] Trevinho: haha, thanks man! :-) [15:26] fginther: I think we indeed need to find a final good solution and get rid of those waits in overall [15:26] fginther: but I think we'll have to dig a bit into compiz code [15:26] sil2100: can we add a get of the changed value? [15:27] and loop until we have the right value [15:27] ah cool, reading now [15:27] didrocks: yes, that was the idea, but then me and thomi actually thought that a changed value in the gsettings configuration does not give us (probably) guarantee that the setting got applied to the unity shell completely already [15:28] indeed [15:28] valid point [15:28] didrocks: but we might dig into it and see if we can get such certainity from compiz [15:28] Or maybe even add such a thing [15:28] ok ;) I'm afraid we don't have that introspection though [15:29] I tested and anyway compiz takes around 0.3 seconds on my bloated system to apply the configuration change properly [15:29] okay, looking good! [15:29] luv: yw... I wanted that since since the beginning of unity, but unfortunately we took some time to get this... [15:30] cyphermox: \o/ bootstrapped and everything already? (and we have one commit since the boostrap?) [15:30] sil2100, thanks for the additional info in the MP. [15:31] ah, right it's missing bootstrap [15:31] cyphermox: first line: http://bazaar.launchpad.net/~autopilot/autopilot-gtk/trunk/view/16/debian/changelog [15:31] cyphermox: it is bootstrapped :) [15:31] hey, I notice windowmocker isn't in the misc-head.cfg [15:31] what? [15:32] ahh yeah [15:32] I was looking at just the last changelog ;) [15:32] incidently, shouldn't all that get merged somehow? [15:32] cyphermox: it's not? http://bazaar.launchpad.net/~cupstream2distro-maintainers/cupstream2distro/trunk/view/head:/jenkins/etc/misc-head.cfg [15:32] last line :) [15:32] window-mocker [15:32] cyphermox: hum, yes please :) [15:32] cyphermox: just merge them in one [15:37] sorry for dropping, my 1 year old found the power button [15:38] fginther: urgh, this is bad, time change computer? :p [15:39] didrocks, :-) [15:42] cyphermox: still around? ;) [15:42] cyphermox: the autopilot branch is merged, just miss the one changelog before rebuilding the stack [15:43] yeah [15:43] https://code.launchpad.net/~mathieu-tl/autopilot-gtk/changelog/+merge/147398 [15:43] I crashed for some reason [15:43] fginther has his child sabotage his computer, yours is crashing [15:43] crazy place here :) [15:44] cyphermox: https://code.launchpad.net/~mathieu-tl/autopilot-gtk/changelog/+merge/147398 [15:44] cyphermox: empty diff? [15:44] what? [15:44] https://code.launchpad.net/~mathieu-tl/autopilot-gtk/changelog [15:44] heh [15:44] did you commit before pushing? :) [15:44] yeah just a second [15:44] sil2100, https://code.launchpad.net/~sil2100/autopilot/set_compiz_option_cleanup_fix/+merge/147379 merged [15:45] I think that's because of my crash [15:45] probably :) [15:45] there you go [15:46] yeah, it looks like a pretty bad crash :/ [15:46] I'd like to know what happened, but even just syslog is messed up [15:46] fginther: I'll abuse your jenkins power and getting this MP merged quickly if possible (https://code.launchpad.net/~mathieu-tl/autopilot-gtk/changelog/+merge/147398) [15:46] cyphermox: urgh? [15:46] cyphermox: hope you didn't loose any data at least [15:47] didrocks, i'll start it [15:47] heh, I couldn't lose much, it's pretty much all on ubuntu one or on launchpad [15:47] fginther: \o/ [15:47] fginther: thanks a million! [15:47] :) [15:51] didrocks: not looking good: http://people.ubuntu.com/~mathieu-tl/misc/Capture%20du%202013-02-08%2010:43:59.png [15:52] isn't jenkins supposed to review the changelog branch eventually? [15:52] * cyphermox cracks the whip [15:53] :) [15:53] francis started the job === dandrader|afk is now known as dandrader === sil2100_ is now known as sil2100 [16:06] didrocks, https://code.launchpad.net/~mathieu-tl/autopilot-gtk/changelog/+merge/147398 is merged [16:07] fginther: excellent! thanks for tracking :) [16:07] cyphermox: let's add autopilot-gtk to the misc stack? [16:07] and then launch a build? [16:12] fginther: I really wonder if the fasttrack worked again, mmrazik told me this morning, it was working: on https://code.launchpad.net/~ps-jenkins/unity/latestsnapshot/+merge/147388, 54 minutes for landing (and nothing else in the unity queue) [16:15] didrocks, we'll have to restart jenkins for it to pick up these change correctly. Jenkins appears to cache the build triggers that we need to update to turn on fasttrack [16:16] fginther: how do you update your jobs? the manual commands that jibel built worked reliably, maybe we should put that in common in the future [16:16] didrocks, everything is in place otherwise. [16:16] didrocks, I generally use the jenkins web interface [16:16] mterry: cyphermox seems to be on lunch break, as I hope we can publish ASAP the new misc, and then indicator stack, do you have time for adding the components? (I prefer you/cyphermox do it for practicing for the day I'm not around ;)) [16:17] what? [16:17] fginther: waow, and the web interface is caching? weird… I think we'll need to work on the configuration at some point, having one configuration that you can expand/deploy ;) [16:17] didrocks, new component is window-mocker? [16:17] cyphermox, heh [16:17] didrocks, I'm interested if jibel has had success with other methods [16:17] cyphermox: \o/ [16:17] just in time :) [16:17] cyphermox: ping you twice, you probably missed them :) [16:17] yeah, let's add autopilot-gtk [16:17] fginther: yeah, let's discuss that next week? [16:17] didrocks, yes [16:17] cyphermox: ok, tell me if you need my help [16:17] I already have the change if it didn't get crapped by my crash [16:17] heh, let's see :) [16:18] hopefully I'm not missing access for this one [16:18] cyphermox: tell me once I can pull on the archive admin machine [16:18] you should have access, you pushed successfullly yesterday [16:19] so, change, merge, push, publish, notify , correct? [16:19] cyphermox: no merge, just push ;) [16:19] cyphermox: but yeah ;) [16:20] well, yeah [16:20] mterry: window-mocker is already part of the distro, you are soooo old fashion :) [16:21] :) [16:25] ;) [16:25] andyrock: hi! I commented on your merge request for the quicklist fix :) [16:25] Ah! [16:25] I see you fixed it ;) [16:25] cyphermox: so, once you have published the jobs to jenkins, you can run building the misc stack :) [16:26] didrocks: no, still need to notify you no? [16:26] sil2100, not sure what happened here, maybe a cmake issue [16:26] andyrock: maybe during building a package with bzr builddeb adds some flags? [16:26] or is it not order-sensitive? [16:26] andyrock: since I was building a package [16:26] cyphermox: it's not, between the notification and publishing the jobs to jenkins :) [16:26] sil2100, no i get the fail here too now [16:27] cyphermox: I saw you changed the cupstream2distro trunk [16:27] right [16:27] cyphermox: so I pulled on the archive admin machine [16:28] ok [16:28] so the jobs are reconfigured [16:29] cyphermox: sweet, please launch a build :) [16:30] cu2d-run -R misc-head correct? [16:30] why not just rebuild autopilot-gtk? [16:30] cyphermox: the build is per stack, it will see that autopilot and autopilot-gtk are new [16:30] cyphermox: and will only rebuild those [16:31] ok [16:31] so as the two are the ones we are interested in… :) [16:31] right.. [16:31] no need to specify them on the command line [16:31] but good question! :) [16:32] so I'm missing check-whole-ppa then [16:32] cu2d-run complained ;) [16:32] cyphermox: euh non [16:32] check-whole-ppa won't rebuild [16:32] only run the check [16:32] cu2d-run -R misc-head <-- that didn't work [16:32] the command is cu2d-run -R misc [16:32] 2013-02-08 11:31:24,016 ERROR No project or check-with-whole-ppa parameter specified on the command line. This tool is used for those cases. Aborting! [16:32] oh [16:32] so I do need to specify autopilot/-gtk [16:32] right, I did this check :) [16:32] andyrock: anyway, it's fine, I approved - you want someone else from the unity team to take a look before globally approving? [16:33] cyphermox: yeah sorry, I didn't want to enable people to rebuild everything by error :) [16:33] sure, np [16:33] how are projects separated, just space? [16:33] cyphermox: yeah, just spaces [16:34] done [16:34] cyphermox: excellent, its working as you can see [16:35] cyphermox: so then, we'll have a manual publishing to do due to autopilot-gtk with packaging change [16:35] sil2100, it's fine can go in ;) [16:35] cyphermox: and then, we can launch --check-with-whole-ppa for indicators [16:35] with sil2100's autopilot fix [16:35] and cross fingers and everything :) [16:41] didrocks: hey, I was looking at the last changelog published for unity... Since we merged a branch from unity-team on wich I worked I and Brandon, the changelog repets the bug for both (and that's fine), but there's also an entry for both such as [ Marco Trevisan (Treviño) , Brandon Schaefer ] [16:41] didrocks: is that a parsing issue or something wanted? [16:42] Trevinho: yeah, parsing issue, please open a bug with the changelog and link to the commit in lp:unity against cupstream2distro [16:42] Trevinho: I'll provide a testcase that way and fix it :) [16:42] didrocks: nice, thanks [16:44] thanks to you :) [16:47] didrocks: here you are https://bugs.launchpad.net/cupstream2distro/+bug/1119562 [16:47] Launchpad bug 1119562 in Canonical Upstream To Distro "Changelog for shared branch is not generated correctly" [Undecided,Confirmed] [16:49] Trevinho: thanks! === dandrader_ is now known as dandrader|afk [17:57] didrocks: ok, so finally built and all, let's rerun indicators [17:57] cyphermox: \o/ [17:57] cyphermox: --check-whole-with-ppa [17:58] yupyup [18:00] cyphermox: oh, you need manual publication as well for misc [18:00] cyphermox: as there is a new packaging, autopilot-gtk [18:00] jibel: just for your pleasure ^ :) [18:01] right [18:02] didrocks: does that need to be finished before I do the indicator rerun or are we good because it's really about publishing to archive rather than to the ppa [18:02] ? [18:02] cyphermox: no, the publication is just copying to the archive [18:02] cyphermox: so whenever you want (preferably now ;)) [18:03] I hope this time it will be 0 failures [18:03] Since with autopilot and unity, you never know! [18:04] sil2100: I like 0! :) [18:05] aye [18:05] didrocks: done [18:05] brb [18:06] cyphermox: \o/ [18:07] cyphermox: ah, also, the indicators will probably go on manual publishing mode as we have libappindicator and other "new" packages for the system === dandrader|afk is now known as dandrader [18:15] yeah [18:25] cyphermox: and UTAH failed again [18:26] cyphermox: can you restart the job once it finishes? we need to wait for ati to finish [18:27] * sil2100 sighs [18:27] What's up with UTAH? [18:27] again, can't install a machine :/ [18:27] and ati failed as well [18:28] cyphermox: trying to relaunch it [18:28] so, again, have to wait for 1 hour from now to eventually get results [18:30] huh [18:31] :/ [18:49] cyphermox: ok, failing again, can you please go on the qa channel and tell them that it's failing reliably? [18:52] It makes me angry [18:53] same here TBH :/ === salem_ is now known as _salem === _salem is now known as salem_ === salem_ is now known as _salem