[07:44] Saviq: did MacSlow address your needs fixing at https://code.launchpad.net/~macslow/unity8/modal-snap-decisions/+merge/210988 ? i think yes but want to be sure [07:45] tsdgeos, I believe so [07:45] tsdgeos, yeah, looks fine [07:45] tsdgeos, Saviq: I think I've addressed all pending issues with the three related branches... [07:46] tsdgeos, Saviq: ... but let me re-check to be sure. [07:46] MacSlow, I'm doing [07:47] MacSlow, "9 + libunity-api-dev >= 7.80.7," [07:47] MacSlow, missing parentheses [07:47] one sec [07:47] MacSlow, in https://code.launchpad.net/~macslow/unity-notifications/modal-snap-decisions/+merge/212483 [07:47] MacSlow, you missed this comment https://code.launchpad.net/~macslow/unity-notifications/modal-snap-decisions/+merge/212483/comments/518638 [07:48] Saviq, ah ok [07:48] MacSlow, and this one https://code.launchpad.net/~macslow/unity-api/version-bump-to-0.1.3/+merge/217882/comments/518637 [07:48] Saviq: when are we getting the CI builders switched over to utopic? [07:48] tsdgeos, fginther is working on that currently [07:48] oki [07:52] damnit [07:52] dpkg-buildpakcage doesn't build anymore! [07:52] debian/rules:10: *** mixed implicit and normal rules: deprecated syntax [07:52] make: *** No rule to make target 'clean'. Stop. [07:52] dpkg-buildpackage: error: fakeroot debian/rules clean gave error exit status 2 [07:53] any help? [07:53] also [07:54] why has the tabbar stuff been merged without the new tabbar in the sdk? [07:55] tsdgeos, both have been merged [07:55] but sdk not released? [07:56] tsdgeos, https://launchpad.net/ubuntu/+source/ubuntu-ui-toolkit [07:56] hmm [07:57] hmmmm [07:59] ok, i think i may have been trying and un-merged to master branch [07:59] my bad [08:04] sil2100: didrocks: any help with the unity8 package not suceeding with dpkg-buildpackage ? [08:09] tsdgeos: do you have a link? [08:10] didrocks: no, i simply run dpkg-buildpackage and it fails [08:10] didrocks: i can give you a paste if you want [08:10] Saviq, tsdgeos: so... all needed version-bump-related tweaks to lp:~macslow/unity-notifications/modal-snap-decisions lp:~macslow/unity-api/version-bump-to-0.1.3 are done now [08:11] Saviq, I skipped the tap-vs-spaces changes from runnuing "wrap-and-sort -atv" [08:11] Saviq, these always look odd in diffs [08:12] didrocks: http://paste.ubuntu.com/7397410/ [08:12] tsdgeos: why are you not running bzr bd? [08:12] because no reason [08:13] dpkg-buildpackage has always worked, always used that [08:13] you should rather use bzr bd though :p [08:13] but let me look at what's wrong [08:13] why should i use that? [08:13] it's downloading stuff [08:13] i don't want it to download unity8 [08:13] i want to build what i have in there [08:14] tsdgeos: bzr bd downloads, but then incorporate your diff [08:14] exclude .bzr as well and other random files [08:14] doesn't seem very much bandwidth efficient :D [08:14] tsdgeos: well, you just do that once [08:14] when i all want to build is there already [08:14] or you are on a 64kb/s line… [08:15] to care about bandwitdh [08:15] netsplit bad [08:15] anyway, I'm telling you what's supported :p [08:15] you take it or not [08:15] I don't really care ; [08:15] ;) [08:15] tsdgeos, what would be the symptom? [08:15] tsdgeos, tabs in dash work fine for me I think [08:15] tsdgeos, and debian/rules were not touched for ages... [08:15] must be dpkg deprecated something [08:15] sil2100, hey, do you know how to fix: [08:15] on your issue [08:15] debian/rules:10: *** mixed implicit and normal rules: deprecated syntax [08:15] besides bzr bd doesn't work either [08:15] make: *** No rule to make target 'clean'. Stop. [08:15] sil2100, here's our rules: http://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/debian/rules [08:15] it's because you mix: [08:15] so you may as well want to fix that [08:15] hmm I think I know [08:15] build %: [08:15] hmpf [08:15] tsdgeos, http://paste.ubuntu.com/7397402/ should fix the buildpackage issue [08:15] and overrides_* [08:15] so build %: [08:15] and for instance: override_dh_auto_build: [08:16] doesn't mix up well [08:16] didrocks, ah, so we should override auto_build and it'll be fine again? [08:16] Saviq: yeah, just use the override :) === davidcalle_ is now known as davidcalle [08:16] (it's already there) [08:16] indeed [08:17] I guess dh is pickier now [08:18] indeed [08:18] tsdgeos, what would be the symptom of uitk not releasing the new tabbar and us - yes? [08:18] Saviq: tabbar being empty [08:18] tsdgeos, works here (can't swipe with mouse, though) [08:19] Saviq: yes, i think i was using a not properly merged branch. can't swipe where? [08:19] tsdgeos, the tab bar [08:19] didrocks: Saviq: i think i don't understand what fix is needed in the debian/ :D [08:19] incoming [08:20] Saviq: indeed, no swiping, i wonder if that's the new tab bar behaviour [08:20] tsdgeos, http://paste.ubuntu.com/7397428/ [08:23] man [08:23] restart scope-registry killed my session [08:24] Saviq: that patch still doesn't make it work for me, does it for you? [08:25] tsdgeos, indeed, doesn't [08:26] tsdgeos, this makes it work http://paste.ubuntu.com/7397455/ [08:26] tsdgeos, not sure that's all correct though [08:26] he he [08:26] ok [08:26] didrocks, ↑↑? [08:26] Saviq: it's the one that should work [08:26] i'm not the one to comment on it :D [08:26] Saviq: and remove the PHONY build target [08:27] not sure who put that there though (seems quite old) [08:27] didrocks, so that's the question - we had the PHONY 'cause we have a "build" script [08:27] ah ok [08:27] didrocks, mterry put it there 'cause it conflicted [08:27] hum [08:27] didrocks, shall I rename? [08:27] seems a poorly named script :) [08:27] yeah [08:27] sounds saner [08:27] kk [08:35] Saviq: you work on a MR then? [08:41] tsdgeos, yes [08:56] tsdgeos, didrocks, https://code.launchpad.net/~saviq/unity8/fix-rules/+merge/218258 [09:13] MacSlow, there's a conflict in the unity-api MP [09:13] MacSlow, you need to merge trunk [09:13] * MacSlow takes a look [09:14] Saviq: nooooo :D [09:14] Saviq: 77 === renamed file 'build' => 'build.sh' [09:15] mzanetti, sorry, conflicts with the "build" target in dpkg :| [09:15] no I need to type a . + Tab after build... I can't teach my brain to do that after doing just ./build for a year now [09:15] mzanetti, I know ;) [09:15] mzanetti, we should rename the build dir, would solve that ;D [09:15] yeah, that'd work for me I guess [09:15] Saviq, fixed [09:16] MacSlow, kk [09:18] Mirv, hey, if you're around, could you please have a look at https://code.launchpad.net/~saviq/unity8/qt-renames-no-dee/+merge/217444 [09:19] MacSlow, FWIW, there was no need to resubmit the unity-api branch, it's fine to have a branch depend on one that's already merged === dpm_ is now known as dpm === vrruiz_ is now known as rvr [09:23] Saviq: done, looks good now [09:23] Saviq, I wasn't sure about that, thus did it anyway [09:23] MacSlow, k [09:24] let's land this $#!t [09:24] Saviq, +1 :) [09:24] guys, make sure you don't have still any pckage from the demostuff era [09:25] it was playing tricks on me [09:25] dpkg -l | grep demostuff [09:26] right, those had epochs... [09:31] Saviq: approved [09:32] didrocks, thanks [09:32] yw [09:37] Saviq: https://code.launchpad.net/~saviq/unity8/fix-rules/+merge/218258/comments/519833 [09:37] tsdgeos, hmm missed those [09:38] tsdgeos, done [09:39] tsdgeos, https://code.launchpad.net/~saviq/unity8/qt-renames-no-dee/+merge/217444 is fixed, too, and Timo-approved [09:39] * Saviq adds to landing [09:40] ok, i'll top approve === davidcalle_ is now known as davidcalle [10:04] mzanetti: please, see your email. [10:07] Wellark: replied [10:10] mzanetti: thanks! === MacSlow is now known as MacSlow|lunch === _salem is now known as salem_ === davidcalle_ is now known as davidcalle [13:05] garg [13:05] we really need to fix unity7 restarting on unity8 autopilot tests === MacSlow|lunch is now known as MacSlow [13:08] tsdgeos, no, we just need to run unity8 tests under xvfb ;) [13:08] Saviq: i'm usnsure that's the culprit [13:08] it seems to me as if something is really killing my current session [13:09] no idea how that could happen though [13:09] but it makes finishing the tests for the dynamic card stuff hard [13:09] i think i broke something in the autopilot side [13:09] but if i get restarts... [13:15] tsdgeos, should I try? :) [13:16] tsdgeos, I don't think there's anything explicit that would cause that [13:16] i know:/ [13:16] give it a try [13:20] interestingly the thing reboots when "15:17:11.693 INFO process_helpers:146 - unity8 start/running, process 7128" [13:20] http://paste.ubuntu.com/7398652/ [13:22] tsdgeos, so it happens in trunk for you, too? [13:22] yep [13:22] let's see if using kwin instead of unity helps [13:22] * Saviq will try in a sec [13:23] tsdgeos, any test in particular? [13:29] Saviq: unity8.shell.tests.test_emulators.DashEmulatorTestCase.test_open_unexisting_scope seems to be pretty offending [13:29] but it's the first unity8.shell.tests.test_emulators.DashEmulatorTestCase i run separately [13:29] it may be all of them [13:29] let me try some others [13:30] Saviq: hey man. so, this time I think I have a legit issue (don't lol at me ;)). I have an api endpoint that returns results with images like: http://resizing.service.com/foo/// - when I set that as String on the result, the source field of the result Image widget holds a Url, which is url-decoded automatically, which breaks the url. That is, the result is ... [13:30] ... http://resizing.service.com/foo///, which is broken. This may be related to: http://www.macieira.org/blog/2011/09/qurl-in-qt-5-encoding/ [13:33] other unity8.shell.tests.test_emulators.DashEmulatorTestCase made the thing reboot too [13:39] errrr [13:39] it's upstart asserting :S [13:39] in nih_unref [13:39] the gdb backtrace is awfully full of ??? [13:40] tsdgeos, https://bugs.launchpad.net/upstart/+bug/1222705 [13:40] Ubuntu bug 1222705 in upstart (Ubuntu) "init assert failure: alloc.c:633: Assertion failed in nih_unref: ref != NULL" [Medium,Confirmed] [13:40] looks like that [13:40] tsdgeos, it's fix committed, not released yet unfortunately, I wanted to ping jodh about it, but he's off today [13:40] let me install more symbols [13:55] so it's funny [13:55] now that i've installed the debug symbols [13:55] it doesn't crash anymore ^_^ [13:55] was bound to happen [13:57] tsdgeos, you fixed it! [13:58] yeah \o/ [14:17] Saviq, would we have a means to test sound-playback? I'm wondering how to test the sound-hint support in notifications. [14:18] MacSlow, we have a dummy implementation of the Audio component in the PreviewAudio test or so [14:18] MacSlow, we could instrument that a little to what you'd need to do the testing [14:19] MacSlow, tests/mocks/QtMultimedia [14:19] Saviq, when you get a few minutes, I'd like to go over the config for silo 002 and restart builds if necessary. I've trimmed things a bit [14:19] Saviq, ok... I'll look into that and see to add a test before doing a MP for the sound-hint related branches [14:19] mterry, sure, just after standup should be ok [14:19] Saviq, cool thanks [14:19] tsdgeos, Saviq, the builds should all be on utopic as soon as the job updates are deployed [14:20] mterry, btw, what's actually *missing* for us to land that silo? other than a review on the unity8 MP? [14:21] Saviq, I still have a problem with the qmluitests. I don't have the VMs working yet to run these. Is it ok to run them with trusty for another day or two, or should we just disable them until it's ready? [14:21] fginther, it's fine [14:21] Saviq, thx [14:22] fginther, TBH I've almost got autopkgtests for that ready... [14:22] Saviq, as of this morning, just the testing framework changes to use the unlock-device script that unity8 provides (instead of doing it themselves) -- because in split, unlock method changes so we want to let unity8 control that logic [14:22] fginther, not entirely sure it's useful to spend any more time on the qmluitests job [14:22] fginther, is there anyone else using it? [14:22] Saviq, I'm working with doanac and others to get that all tested, but it's been slow going [14:23] mterry, mhm, got it [14:23] Saviq, there are a few other tests that use the VMs, so we'll still need them for a little while at least [14:23] Saviq, do those autopkgtests need to run in an X environment? [14:23] fginther, ok got it, so it's not just about the job but the VMs themselves [14:23] fginther, nope, all xvfb [14:24] Saviq, very cool [14:25] fginther, yeah, with recordmydesktop, too! ;) [14:25] fginther, of any failed tests [14:25] i am suddenly without sound [14:25] Saviq, and yes, it's mostly a matter of setting up the VMs. change the job to work with utopic is mostly a search and replace [14:25] back in a minute === dandrader is now known as dandrader|bbl [14:38] Is there a way to stop nautilus from popping up every time my mako reboots? [14:41] mzanetti: https://code.launchpad.net/~aacid/unity8/dynamic_specialized_cards/+merge/218089 [14:42] tsdgeos, plugins/Dash/CardCreatorCache.qml UNKNOWN *No copyright* [14:43] * tsdgeos fixes [14:43] Saviq, can you pastebin me the current silo 002 config? I'll pastebin back the now-correct one [14:53] mterry, http://pastebin.ubuntu.com/7399057/ [14:57] Saviq, http://paste.ubuntu.com/7399084/ should be sufficient [14:58] Saviq, and everything in there (except for lightdm) should be triggered to build. Order shouldn't matter [14:59] mterry, kicking the whole silo [14:59] Saviq, awesome, thanks [14:59] mterry, except we need landing team ACK for the telephony-service MP which wasn't there before [15:04] Saviq: FYI https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1316183 [15:04] Ubuntu bug 1316183 in qtbase-opensource-src (Ubuntu) "QML Image shouldn't decode it's source field" [Undecided,New] [15:06] karni, every time you write "it's", expand it to "it is" and see if the sentence still works, otherwise you want "its" (like his, hers) ;D [15:09] Saviq: right, that's why wrote it's, because I wanted "yes, it is improperly decoded, ..." :) [15:09] Saviq: its quality is improperly decided, thus, it's improperly decoded :) [15:10] *decoded hehe [15:10] karni, I meant in title [15:10] karni, see ↑↑↑ [15:10] d'UH [15:10] Saviq: thanks pal :D [15:10] karni, I'm just allergic to it, sorry ;) [15:10] mainly because I had to make myself allergic to it to stop making the same mistake [15:11] Saviq: no, man, I appreciate. I'm allergic to that as well, must have been multitasking, I'm embarassed to have done that mistake! :) [15:11] double *r ;) [15:11] :D [15:13] mterry, silo 002 is building [15:13] mterry, CI-SNCF will tell you when done [15:13] Saviq, thanks! [15:14] Saviq, there was also supposed to be a silo last week for ~mterry/ubuntu-touch-session/move-lightdm and another u-t-s branch. Do you happen to know if that just got lost or delayed? [15:14] mterry, let me see [15:15] mterry, looks like it's https://launchpad.net/~ci-train-ppa-service/+archive/landing-013 [15:15] mterry, can you give me a description to put in the train for this? [15:15] Saviq, cool. I assume it's in the queue then. Just didn't want to lose it [15:16] mterry, since this was in the middle of broken spreadsheet, it doesn't have any description [15:16] Saviq, sure. "Allow /custom override of the Touch lightdm config by moving it to /usr/share instead of /etc. And clean up the packaging a bit while there" [15:20] mterry, kk, will try and land it today, too === gatox is now known as gatox_lunch [15:48] Saviq: damn we need to get your dpkg-buildpackage fixes in asap, otherwise CI is useless :D [15:48] tsdgeos, testing right now, will push the button in 5 or so, assuming everything's ok [15:49] \o/ [15:52] tsdgeos, is there anywhere I could test zoomable image on the phone? [15:52] Saviq: changing the code is what i did [15:52] tsdgeos, ok, so no ;) [15:53] to make all the zoomable images actually zoomable [15:53] tsdgeos, well, they're not meant to be zoomable inherently are they? don't we have a bool there? /me looks [15:53] Saviq: we do [15:53] Saviq: but i tested it by making them all zoomable [15:54] tsdgeos, got it [15:54] so i had something to actually test [16:01] tedg, hey, I'm trying to verify the launcher dbus bits working, can you point me to the correct gdbus call, I tried: [16:01] gdbus introspect --session -d org.freedesktop.DBus -o /com/canonical/Unity/Launcher/com.ubuntu.developer.mzanetti.xbmcremote_xbmcremote_1.4.1.0 [16:01] but I'm getting "is not a valid object path" [16:02] Saviq, Yeah, you need to encode the "_" [16:02] tedg, url-encode? [16:02] Saviq, No, "dbus encode" [16:02] ... [16:02] Just a sec [16:03] gdbus introspect --session -d org.freedesktop.DBus -o /com/canonical/Unity/Launcher/com.ubuntu.developer.mzanetti.xbmcremote_95xbmcremote_951.4.1.0 [16:03] Nope [16:03] gdbus introspect --session -d org.freedesktop.DBus -o /com/canonical/Unity/Launcher/com.ubuntu.developer.mzanetti.xbmcremote_5Fxbmcremote_5F1.4.1.0 [16:04] tedg, thanks [16:04] Needs to be in hex [16:04] tedg, same error [16:04] Saviq, Get the list from the launcher: gdbus introspect --session -d org.freedesktop.DBus -o /com/canonical/Unity/Launcher/ [16:05] not a valid object path :| [16:05] Hmm... [16:06] Saviq, try this real quick: gdbus introspect --session -d org.freedesktop.DBus -o /com/canonical/Unity/Launcher [16:06] If that doesn't work I'll get a test system setup and figure out what's up. [16:06] tedg, yeah, that's working [16:06] Saviq: qdbus com.canonical.Unity.Launcher /com/canonical/Unity/Launcher/gallery_2Dapppp org.freedesktop.DBus.Properties.Set com.canonical.Unity.Launcher.Item count 8 [16:06] less -- [16:06] :) [16:07] fix the appId tho [16:07] Guessing the "." are encoded as well. [16:07] Didn't want to do that by hand :-) [16:10] mzanetti, yeah, that didn't complain, nothing on the gallery launcher, though, is that for the deb-installed gallery or should work for click, too? [16:11] mzanetti, and not too many ps? [16:11] and not a valid name any more :| [16:11] ah wrong [16:12] Saviq: yeah... that was in the history because I intentionally tried wrong appids [16:12] Saviq: you need to set visible to true, otehrwise it doesn't display anything [16:12] mzanetti, did that ;) [16:13] hmm... was working for me when I did the review... [16:13] hmm... not that I look at it, I might only have tried with preinstalled apps :/ [16:13] naughty me... [16:14] mzanetti, gallery is preinstalled, but still click [16:14] ok it's not working for me, but I expect this to be because of wrong qdbus calls... [16:16] can't get anything to show up :| [16:18] Saviq: this worked for me just now: http://paste.ubuntu.com/7399555 [16:18] mzanetti, so that should work on the click gallery, you didn't install the .deb one? [16:18] I probably did [16:18] ah countVisible [16:19] that's better [16:19] ok :) [16:19] mzanetti, so what's required after /Launcher/? [16:19] gallery_2Dapp == gallery_app? [16:20] well, the appid. whatever that is :D [16:20] yes. [16:20] the above statement is true [16:20] so what would that be for xbmcremote? [16:20] maybe try replacing the . with _2X (don't know the code for . by memory) [16:21] Saviq: _2E [16:21] * Saviq no gets how is gallery different [16:22] aanyway [16:22] Saviq: this is my launcher config: http://paste.ubuntu.com/7399589 [16:23] seems gallery_2Dapp is gallery-app, not gallery_app [16:23] you need whatever is in the "id" field. [16:23] urlencoded [16:25] YES [16:25] com_2Eubuntu_2Edeveloper_2Emzanetti_2Exbmcremote_5Fxbmcremote_5F1_2E4_2E1_2E0 [16:26] _check_ [16:26] ok, ACK [16:26] gtg === gatox_lunch is now known as gatox === jhodapp is now known as jhodapp|brb [17:35] sufficiently wide image extends to right of device screen in scope result preview, instead of matching available width. known issue, or should I report? === jhodapp|brb is now known as jhodapp [17:50] mterry: hey, is it intentional that you can flick the greeter away the "wrong" direction? [17:51] as in: start dragging from the right edge, move mostly across the screen and flick it back to the right quickly [17:51] mzanetti, yes [17:51] mzanetti, wait, maybe [17:51] it will move over the screen and go away to the right [17:51] :D [17:52] mzanetti, the unlock to the right is intentional. I'm less clear on what design intended for a rapid cancel like that [17:52] I don't think it should unlock in this case [17:52] seems goofy [17:53] mzanetti, I'm not in a position to test it for myself right now (running some tests on my mako), but sounds a little unexpected yeah [17:53] not critical, you can test it later [17:55] mzanetti, maybe file a bug and assign to me to look at [17:56] Else I'll get to it later [17:56] mterry: ack [17:56] mterry: that drop shadow makes such a huge difference :) [17:56] lookin awesome [17:57] :) [18:10] Are annotations supposed to work already? There's related code in Ubuntu SDK scope template, but it does nothing. [18:11] I can't get Annotation to work either. Perhaps it's not hooked up in the shell yet? [18:43] Saviq, hmm, can you add https://code.launchpad.net/~saviq/unity8/fix-rules/+merge/218258 to silo 002 too? Silo build failed without it [19:00] mterry, it's already being landed [19:00] mterry, will kick it again when it's in trunk [19:02] Saviq, ok, thanks [19:03] Saviq, I didn't notice a message in choo-choo channel and the bot there thought the silo's status was still 'preparing packages' -- is that normal for this failure case? [19:04] mterry, the same is in the spreadsheet [19:04] mterry, I think yeah, it's missing a "failed to prepare packages" state [19:05] Saviq, understood. Just means they aren't quite fire and forget [19:05] mterry, this looks like a bug, actually [19:05] mterry, file one against cupstream2distro please [19:05] Saviq, k [19:05] mterry, it definitely should be, that's the whole idea behind the bot [19:07] mterry, well *actually* "mktemp: failed to create directory via template '/tmp/debsign.XXXXXXXX': No space left on device" [19:08] Saviq, where do you see that? [19:08] mterry, wrong job [19:08] mterry, http://people.canonical.com/~rbpark/citrain/ links to that job for some reason [19:09] Saviq, bug 1316276 filed [19:09] bug 1316276 in Canonical Upstream To Distro "Need better feedback when a silo fails to prepare packages" [Undecided,New] https://launchpad.net/bugs/1316276 [19:38] mterry, I wanted to test the u-t-s landing, any pointers? other than ogra's "see if it boots'? ;) [19:39] Saviq, see if it boots, confirm that /etc/lightdm/lightdm.conf.d/52-ubuntu-touch.conf is gone [19:39] Saviq, and confirm that /usr/share/lightdm/lightdm.conf.d/52-ubuntu-touch.conf exists [19:39] mterry, kk [20:03] Saviq: still awake? [20:04] Anyone here know why my unity 7 session is killed when I try and run the unity8 autopilot tests on my utopic desktop? [20:05] thomi, upstart crashes probably [20:06] Saviq: who's looking into the fix? [20:06] thomi, it's fix committed [20:06] https://bugs.launchpad.net/upstart/+bug/1222705 [20:06] Ubuntu bug 1222705 in upstart (Ubuntu) "init assert failure: alloc.c:633: Assertion failed in nih_unref: ref != NULL" [Medium,Confirmed] [20:06] thomi, will talk tomorrow to James to get it released [20:07] Saviq: ahh - FWIW, I don't see anyting from upstart in /var/crash [20:07] Saviq: awesome, thanks man :) [20:07] then I can make sure that unity8 will continue to work under autopilot 1.5 :) [20:07] thomi, it actually asserts, might not leave anything in .crash maybe [20:07] interesting === salem_ is now known as _salem [20:48] mterry, silo 013 ACKed, should land soonish [20:48] Saviq, awesome thanks [21:08] aaargh the tags... [21:15] mterry, I'll kick silo 002 again as soon as I strip tags from trunk... they got in with the latest merge again :[ [21:16] Saviq, someone still has tags on their system? :( [21:16]