[00:31] morning huwshimi [11:41] benji, you reviewing frankban's branch? [12:07] ahoy sinzui [12:07] jujugui two reviews of https://codereview.appspot.com/11804043/ please? [12:08] gary_poster: o/ (which translates to i'll do one) [12:08] thanks bac. we need that landed, and tehn a release today [12:08] benji, hi, you reviewing frankban's branch? [12:09] gary_poster: hi [12:09] :-P [12:09] I can review a branch; once I wake up a little, apparently [12:11] benji :-) sorry, cool about frankban's branch. I misremembered. I need a review of https://codereview.appspot.com/11804043/ so I can start a release [12:11] gary_poster: I'll review https://codereview.appspot.com/11804043/ [12:13] thank you [12:14] bac or benji, I'm sorry, but we need one more for frankban's giganto branch. could one of you volunteer for that as well? https://codereview.appspot.com/11725044/ [12:15] gary_poster: I'm up for a giganto review [12:15] thanks benji [12:16] https://www.youtube.com/watch?v=7AtSHp35xUg [12:18] thanks benji and gary_poster, and sorry for the giganto giant [12:18] * frankban lunches [12:19] we all create monstrosities every know and then [12:22] gary_poster: how do i qa your branch? [12:24] bac, thanks sorry for not clarifying in cover. qa for primary issue is that you have to start a juju core environment and deploy puppet [12:24] bac, I did that with Francesco and Mark Mims [12:25] bac, for other issues: [12:25] gary_poster: are you saying the subordinate issue is sufficiently qa'd? [12:25] bac, yeah, I'm comfortable with it. [12:26] - create a service in sandbox. from environment, click on menu and then destroy. verify that styling is same as before (jujucharms.com). verify that charm browser looks ok. verify that zoomer looks ok. [12:26] ok, and the sam-skin issue is fixed [12:26] i see that [12:27] - you could try subordinates with the inspector. they are...better than they were before. :-/ :-D [12:27] gary_poster: lgmt + qa-ok [12:27] bac, cool, thank you [12:27] gary_poster: i'm having a brain block on subordinates. what is the canonical one to try? i tried rsyslog but it won't build relations to anyone [12:27] bac, puppet is a good one [12:27] ok [12:28] gary_poster: fwiw, testing destroy, i killed a service that had an inspector open. the service went away but the inspector stayed. oopsie. [12:28] bac, yeah, have seen :-/ [12:29] rt [12:29] will make card and/or bug [12:31] actually gary_poster doing qa on your branch i see some weird styling issues on the inspector. the X for closing has dropped from the upper portion of the section to the bottom of it [12:32] bac, huh. maybe it needs sam too? will investigate. ty [12:33] gary_poster: reverting your branch puts it back to the right place [12:33] bac, cool. I suspect I need to put that sam thing on something. [12:45] if anyone (like me) missed the GUI part of the keynote yesterday, you can enjoy it at https://www.youtube.com/watch?v=ZqmFXalqKic&t=714 [12:45] hey hatch_, when you are around lemme know. fun new bug to address asap :-P [12:54] thanks benji, i forgot to watch [12:54] hey, we made a whiteboard! [12:57] new catchphrase: Juju GUI Export -- throw it up on metal! [12:58] lol [13:04] huh, 2 of 3 comments on the video are about the codec support OSCON chose. [13:05] hi sinzui -- seems my branch fixed 2 of the reports but the third is still broken. http://staging.jujucharms.com/reports/charm-contributors.json [13:05] sinzui: can you get a traceback? [13:06] teknico: would it be ok if you and i switched our 1:1 call times with gary as i have another call. it would mean your call with him would be in 24 minutes. [13:07] bac: that's ok for me [13:07] bac: https://bugs.launchpad.net/charmworld/+bug/1204899 [13:07] <_mup_> Bug #1204899: ValueError in report json [13:07] thanks teknico [13:08] bac, teknico thanks. bac, one sec [13:08] thanks sinzui [13:08] bac: I was just extracting the TB from my email when you pinged. === teknico1 is now known as teknico [13:40] guys, I deployed juju-gui with the config " juju-gui-source: trunk", and I still get the old look [13:40] and some sort of "IO error" in the alert box [13:40] "Could not retrieve charms: Error: IO data failure" [13:41] gary_poster: I'm in the 1:1 hangout: https://plus.google.com/hangouts/_/9042d8b55d7f3f374fcd41e731d334ff529957f1?authuser=1 [13:41] in /var/lib/juju/agents/unit-juju-gui-0/charm/release I have "juju-gui-0.3.1+build.595" [13:42] teknico, finishing up, there soon [13:42] javascript console has: [13:42] Failed to load resource: the server responded with a status of 500 (Internal Server Error) [13:43] https://manage.jujucharms.com/search/json?search_text=series%3Aprecise%20owner%3Acharmers <-- "Internal Server Error" [13:44] ahasenack, looking one sec [13:45] ahasenack, was this an upgrade? [13:45] ahasenack, I suggest trying [13:45] gary_poster: no, fresh juju gui depl [13:45] gary_poster: still, the server shouldn't 500 [13:45] really! [13:46] I don't know that this means too, it's in the console [13:46] XMLHttpRequest cannot load https://manage.jujucharms.com/search/json?search_text=series%3Aprecise%20owner%3Acharmers. Origin https://10.0.3.108 is not allowed by Access-Control-Allow-Origin. [13:46] 10.0.3.108 is the juju-gui's unit ip [13:46] ahasenack, agreed, but not the problem that worries me most atm. that's an old API. the question is, why are you getting it. [13:46] ahasenack, oh I know [13:46] ahasenack, [13:47] juju set juju-gui juju-gui-source=stable [13:47] or juju set juju-gui juju-gui-source=0.8.0 [13:47] gary_poster: ok, so trunk doesn't work with manage.jujucharms.com, is that it? [13:48] ahasenack, trunk is a bad name I'm afraid. it is the last release made on the "trunk" series [13:48] ahasenack, we don't release those except for tests [13:48] gary_poster: what about lp:juju-gui? [13:48] ahasenack, if you actually want to use trunk, yes: [13:48] juju set juju-gui juju-gui-source=lp:juju-gui [13:49] ahasenack, sorry for confusion. We'll clarify in docs [13:49] ok, np, I'll try lp:juju-gui, and then stable [13:50] cool thanks ahasenack. stable is what is always supposed to work fwiw. lp:juju-gui is about to have a nice fix for juju core subordinates [13:50] gary_poster: what I'm trying to debug is why my juju-gui deployment ("stable" last time I tried, will verify now) does not show that extra icon for subordinate charms, the one that looks like an electrocardiogram [13:50] oh, I wonder if that's what is happening [13:50] ahasenack, ok, that's a bug I have fixed in a branch that will land within an hour and be released today [13:50] ah, good [13:51] sorry, late for call, running :-) [13:55] gary_poster, Hi, someone just pointed out there seems to be an issue viewing the readme's of some charms. Is this a known issue? [13:56] gary_poster: hey I'm here === hatch_ is now known as hatch [14:00] antdillon, on call, but if it is that sometimes the markdown is not processed properly, yes, that typically means the markdown is broken. is that what you mean, or something else? [14:00] one other issue is that the markdown processor does not honor html comments [14:00] but that's all I know about [14:01] gary_poster, When you are in browse mode and checking out the details of an app and click readme it loads then jumps back to the search results [14:01] hatch, hey. calls. take a look at http://paste.ubuntu.com/5911233/ and see if you get the gist :-) [14:02] gary_poster: i see you're on another call. let me know if you want to reschedule [14:02] bac, no you are next call. be there very soon. [14:02] ok [14:03] antdillon, remember seeing something like that but can't dupe. can you give step by step dupe instructions starting from jujucharms.com? [14:03] gary_poster, Sure, go to https://jujucharms.com/ [14:04] gary_poster, Search "cinder" [14:04] hatch, if you have test ideas lemme know. that was from another report of app spinning forever. they had turned off cookies. could dupe in FF [14:04] gary_poster, Click the cinder app for more details [14:04] gary_poster, Then click read me and it fades back to the search results [14:05] antdillon, got it! thank you, that's new. we'll get on it. thanks a lot! [14:05] gary_poster: ahh ok looking [14:05] gary_poster, No probs [14:05] jcsackett, sinzui ^^^ see what ant reported. any chance of a quick fix in next hour? [14:06] gary_poster, doubtful. I know that bug [14:08] gary_poster, that is the bug I showed you when we QAed 0.8.0 for release [14:08] oh, another variant? [14:08] yep [14:09] sinzui, ok thanks. will investigate asap [14:09] jcsackett, do you know how to reproduce the bug where you choose a charm's readme and instead are returned to search results? [14:10] gary_poster, jcsackett: I search for apache, choose apache, choose readme, then see search results again, and all navigation is broken. I use my back button to get back to the charm [14:11] sinzui: gary_poster it looks like it's due to the interaction with the tab widget #bws-readme getting appended to the url which causes the url to 'appear' changed [14:11] gary_poster: sinzui and then we're smack back into the bug I hope to fix tomorrow #1200743 [14:11] <_mup_> Bug #1200743: complex search/view interactions in fullscreen fail to work

[14:12] rick_h, were did you get the wifi from? [14:12] sinzui: the dealership has wifi :) [14:12] benji: I am glad you survived the monster, thanks a lot for your review! [14:12] I hope you are working on bookie then [14:12] just hopped on after looking at pretty airstreams for a bit [14:12] frankban: my pleasure [14:12] sinzui: :) sure thing [14:13] gary_poster: we can normalize the set/get credentials methods to both return a value then we can clobber the window.seessionStorage property and test that they both return the desired values [14:16] jujugui: can I get two reviews on https://codereview.appspot.com/11831043/ ? [14:17] jcsackett, sure. [14:17] thanks, Makyo. [14:17] no it [14:17] on even [14:18] frankban: I'll review your branch too, if only because your card had just one name in the tags ;-) [14:18] heh [14:19] hatch, window.sessionStorage does not allow clobbering on FF at least [14:20] rick_h, sinzui ack thanks [14:22] gary_poster: ok then we need to change how we assign sessionStorage here so that we can assign it to something else during testing `var sessionStorage = window.testSessionStorage || window.sessionStorage` [14:23] hatch, have a plan. day is horribly busy though. one sec, will get diff in 3 min [14:30] hatch, http://paste.ubuntu.com/5911344/ [14:30] looking [14:31] hatch, with that you can test behavior of stub, test whether the verifySessionStorage is called by the module, test verifySessionStorage itself, and even test whether module.sessionStorage is actually used by getCredentials/setCredentials [14:31] hatch, maybe we don't need all of that :-P but we could [14:32] haha well this is more in depth than I was thinking [14:32] but it does open up more possibilities for testing [14:32] sinzui: you free to chat? [14:32] :) [14:42] abentley: here's my MP: https://code.launchpad.net/~benji/charmworld/bundle-indexing/+merge/176956 [14:42] benji: looking... [14:43] rick_h: I found two incorrectly changed classes from the view-container switchover.....such a minor bug - not bad all things considered :D [14:47] jcsackett: I found svg social icons http://zurb.com/playground/social-webicons maybe we want to use those? [14:49] benji: You've got some conflicts. [14:49] abentley: doh! I'll resolve and re-push. [14:52] abentley: fixed and pushed [14:52] benji: Thanks! [14:54] hatch: they look cool, but given we only need three of the services, and only one resolution, I don't know that it's worth switching to those. [14:54] yeah maybe not :) [14:54] maybe we can use them as charms? [14:54] deploy facebook [14:54] but I've bookmarked them for the future. [14:55] * jcsackett laughs [14:55] oh it's PHP, nm [14:55] :-p [14:55] haha [14:59] http://www.engadget.com/2013/07/25/dell-officially-unveils-precision-m3800-workstation-at-siggraph/?utm_medium=feed&utm_source=feedly [14:59] finally a MBP competitor....maybe [15:01] I just hope it doesn't only have HDMI out [15:12] gary_poster, arosales: We are testing an alternate redirects to manage.jujucharms.com. I think we will have this sorted in production in a few hours. [15:13] gary_poster, I have a cunning idea [15:13] sinzui, ah great to hear :-) [15:13] sinzui, k [15:14] gary_poster, We rejected the redirect rule for /~user and /charmers. I think I can write a rule that appends -0 to the url, so that the gui loads the charm [15:18] bac, do you want to talk about search and api3? [15:19] sinzui: yes [15:19] sinzui: orange hangout? [15:19] hatch: i don't mind the HDMI out, but wish there was a 13in option for that. [15:19] yes, the standup one [15:20] * bac gets tea. one sec [15:28] bac fwiw, I didn't have time to resolve CSS issue so Jeff will tackle it as part of his current inspector styling branch. benji, bac, thank you for reviews [15:29] any time [15:30] oooo only one conflict [15:30] sometimes bzr impresses me [15:32] hatch: other times it represses you ;-) [15:34] gary_poster: looks like my branch preemptively fixed the x bug [15:34] hatch lol cool [15:45] bcsaller: I already have two reviews for my branch, I'll be happy to wait for yours if you want to [15:45] frankban: no need, it looked good, the test were just having strange issues for me [15:46] bcsaller: cool, thanks [15:46] jcsackett: I hope it's a 15" and not a 17" :) but I need something better than HDMI because my monitor is 2560x1600 [15:49] Makyo: you didn't put the time your landing in LHR on the spreadsheet :) [15:52] jujugui call in 8; kanban now [15:54] hatch, alright, will do. Was just going to take the train, figured it didn't matter much. [15:54] hatch: look at the keyboard, no numeric keypad, can't be a 17" [15:55] teknico: oh is that the sign? [15:55] hatch: I mean, the keyboard would be larger and include a keypad if it were a 17" [15:55] yeah I suppose [15:56] unless they went for giant keys :-) [15:56] lol [15:57] hatch: in the article linked within the engadget article, they say it's a 15. [15:57] ohh ok [15:58] http://www.dell.com/learn/us/en/04/campaigns/precision-m3800-workstation?c=us&l=en&s=bsd [15:58] it shows the sides of the machine but I can't tell what the ports are haha [15:58] jujugui call in 2 [15:59] :-) thx [16:01] bcsaller call [16:01] just finishing up coffee [16:02] freaking google plus ... [16:02] yeah free stuff.....damn them!! [16:07] abentley, I have a zany idea to support old charm urls pointing to jujucharms.com [16:08] hatch: it's not free. i'm giving them all my data. :-P [16:08] lol [16:08] sinzui: Oh? [16:09] I can add a rewrite rule that matches the /~ or /charms and append -0 or -HEAD to it so that juju-gui sees a URL it can load [16:09] benji: can you paste your review url so i can ensure we're not overlapping? [16:09] abentley, do you have any idea regarding how to ask for the tip of a charm? [16:10] bac: https://code.launchpad.net/~benji/charmworld/bundle-indexing/+merge/176956 [16:10] thx [16:10] sinzui: In code, or via URLs? [16:11] benji: yep, that's some of what i was going to do [16:11] bac: heh, I'm glad we didn't duplicate work [16:11] we were trying to decide on what to call superset(charms, bundles). i see you picked items [16:11] abentley, in URLs. gui only knows how to handle versioned charms, and passes that to charmworld to get information. [16:12] I can't say I put a lot of thought into "items". [16:14] abentley, A rewrite rule that adds a version makes the old URLs work (me just tested this). I don't think we want to pass a real version. [16:14] sinzui: once we support versioning, I don't think we have URLs for tip. We could add them. bzr uses -1 for tip, but that would give /api3/charm/~foo/bar/baz--1 [16:14] yep [16:14] I imagined that [16:14] gary_poster: holiday filed [16:15] benji, i'm going to lunch nowish. when do you get back from yours? [16:15] bac: lunch now sounds good; I'll do the same [16:15] fried chicken? [16:15] abentley, this is just to support historic URLs. I don't think we want to suggest that users always make versionless links to charms [16:16] frankban, thanks, approved [16:16] * sinzui ponders though that the charm details view needs to state that there are newer revisions [16:16] thanks [16:17] sinzui: We could provide a route that redirects to tip-- /tips/~foo/bar/baz or something could redirect to /~foo/bar/baz-$LATEST [16:17] abentley, this is hacked URL allows askubuntu users to see the charm: https://jujucharms.com/~christophe.sauthier/precise/zentyal-HEAD [16:18] abentley, to be clear, this is about juju gui showing the charm. It needs a version component, and this hack works because charmworld currently shows tip for ever version [16:18] every [16:18] sinzui: Right. [16:19] abentley, this does work: https://jujucharms.com/~christophe.sauthier/precise/zentyal-$latest [16:19] sinzui: so does this: https://jujucharms.com/~christophe.sauthier/precise/zentyal-moo/ [16:19] yep [16:21] If we guarantee that a token in the revision permission always returns tip, I can fix all the cahrm redirects today [16:21] ^ abentley do you have a preferred token to mean tip? [16:22] sinzui: HEAD or TIP are my faves. [16:23] I will use HEAD since it came up in another conversation today [16:36] hatch: btw, that notebook's display size is noted in the specs below the text, and it's 15.6" [16:37] looking for reviews https://codereview.appspot.com/11843043/ [16:37] teknico: ahh ok [16:42] sinzui, you saw email from jcastro? [16:43] check your email for my reply [16:44] gary_poster, sorry had messed up config on nginx for comingsoon, it should be working now [16:44] thanks hazmat! looks good on quick fly by [16:49] bcsaller: review done [16:54] jujugui lf reviews https://codereview.appspot.com/11843043/ [16:55] * gary_poster would but trying to get test passing for deployment. will look soon if no one else has [17:02] howdy [17:02] the "view" page for already deployed, running charms is not nice :(! [17:06] gotwig, I think we know and are working on it, but just to be clear, could you give us details? screenshot might help. To give you an idea of what we will release soon, we will make an inspector that you can see from the environment view itself [17:06] I can give you preview if you'd like [17:07] yeah [17:07] please [17:08] I like this widget like view [17:08] for getting a quick overview of charms that you wanna add [17:08] or the initial charm setup widget thing at the right, is superb [17:08] but after its deployed, setting it up or configuring it is such a huge mess [17:09] gotwig, understanding that this has lots of work needed still, incuding styling, which is why we are not releasing it and working on it, go to http://comingsoon.jujucharms.com/:flags:/serviceInspector/ [17:09] howdy.. sry, one notebook batery died [17:09] jujugui: can i get reviews on https://codereview.appspot.com/11845043 please? [17:09] so much custom styles :/ [17:09] gotwig, deploy something (clunky) and then click on service again (less clunky) [17:10] gary_poster: yip [17:10] the clunky ness of the deploy something (clunky) will be better soon too :) [17:10] :-) [17:10] hatch: dont use soo much custom styles ;P [17:10] jcsackett: I'll swap you a review for my review? [17:10] gotwig, what do you mean? [17:11] hatch: if you mean the one you posted above, i've already looked at yours. :-) [17:11] right now the styling is different all the time [17:11] there should be some sort of consistecy [17:11] like in the overview page for exploring charms [17:11] jcsackett: darn....lol ok I'll do yours [17:12] hatch: thanks. :-) [17:12] gotwig_: yes it is moving towards a consistent style across the whole application [17:12] but still, juju is now nice enough for me to dig some time into ;X [17:12] gotwig, cool :-) [17:12] :) [17:12] I think about doing a meteorJS charm [17:13] gotwig_: i would like to see that. :-) [17:13] realtime javascript framework, very similar to nodeJS [17:13] it uses a lot of handlebar design patters, more info on #meteorJS [17:13] gotwig_: did you see about the charm competition? [17:13] *patterns [17:13] hatch: haha, I am a student xD [17:13] I am not a pro in all this, this is very complex to me [17:14] https://github.com/juju/charm-championship [17:14] hey everyone has to start somewhere :) [17:14] gotwig_, as to styles, ack. we are releasing incrementally, which is part of what you are seeing. we think it is better to release good stuff with some style issues gradually. also inspector should make what you describe better--we will get rid of all internal pages then. [17:14] gotwig_, charm sounds great. ought to try championship anyway. :-) [17:15] IMHO this is better suited for some ultra smart cloud developers ;X [17:15] thats too much money ;D [17:15] heh [17:15] ok [17:16] that could be your too much money :) [17:16] I still am going to work on JuJu, it looks like fun [17:16] well, meteorJS sounds like a great charm anyway [17:16] great [17:16] I already worked with lenses and scopes in the past, trough that stuff I got my Ubuntu Membership in one week :> [17:16] I know this is a whole other category here ;D [17:16] bac: I'm ready to coordinate when you are. [17:17] gotwig_: that's great :) [17:17] you guys are so nice :D [17:17] we like making users happy, if we can :-) [17:17] you just caught us between sprints [17:17] lol [17:17] :P [17:20] bcsaller, "databings"? :-) [17:20] that's not detailed :-) , but we are releasing incrementally. when inspector is ready we will get rid of all of internal bits [17:20] last sentence should have been... [17:20] Databings clean ups + wildcard support [17:20] This branch adds support for wildcards to databings, it [17:20] hey, it was late [17:20] lol [17:20] do you think about responsive webdesign? [17:20] does ubuntu.com have resposive webdesign? [17:21] jcsackett: review done [17:21] gotwig_, yes we think about it, but we don't have it; IIRC yes, ubuntu.com has it or will soon [17:21] gary_poster: are you part of Canonical? [17:21] gotwig_, yes [17:21] gary_poster: :O [17:22] nowadays you guys are everywhere [17:22] heh [17:22] even following me on Google + ;P [17:22] lol [17:22] mwa ha ha ha? [17:23] I sort of know michael hall good.. he woks on Unity [17:23] cool, yeah [17:23] *works [17:23] sorry, my notebook keyboard is sort of lame [17:23] Its a shame my Zenbook from ASUS broke, due to UEFI [17:24] the UEFI destroyed the BIOS chip, and with that the whole motherboard :/ [17:24] I just bought it 3 months ago :X [17:24] yuck :-/ [17:24] jujugui https://codereview.appspot.com/11849043 [17:24] UEFI is such a mess [17:24] On it. [17:24] gary_poster, ^^^ [17:24] hatch I'll look at yours now unless you already have two? [17:24] thanks Makyo ! [17:26] gary_poster: nope only have one [17:26] hatch ok on it [17:26] thanks [17:29] we really need to relook at the CSS rule generation flags in the makefile :-/ [17:29] make clean && make devel again [17:29] hatch: when a test case finishes everything declared inside it should be destroyed right? assuming they weren't globally declared. [17:29] jcsackett: well a view renders a container and attaches events [17:30] if the destructor is never called those will linger [17:30] hatch: what i'm asking is isn't the desctructor called as view falls out of scope then the 'it' block finishes? [17:31] gary_poster: that session storage pretends to work, but doesn't in the case we use it for, a reload, right? Our code only expects to find the value in SS so we force it down that path then? [17:32] hatch I cleaned build and cache and I'm getting something that doesn't look like what I expect you wanted :-) . I merged your branch into trunk fwiw. http://ubuntuone.com/34o3fKV6ESZt2KorbYbdoD [17:32] jcsackett: nope the destructor is purely a framework construct not part of the language [17:32] is that right? [17:32] woah [17:32] no thats definitely wrong [17:32] I'll merge again see what's up [17:33] bcsaller, right, if they have cookies and session storage turned off, they will have to log in again [17:33] yeah, the larger flaw is that SS isn't optional, but this quickly masks it [17:34] gary_poster: can you clear your cache - I can't repro [17:34] hatch, I did for the day. doing again for the week... [17:35] hatch I get the same [17:35] bcsaller, yeah. I figured it was ok, since I figure these users will be small percentage of base, and it only really bites them in the Landscape scenario [17:36] hatch, wanna hangout so you can show me the error of my ways? :-) [17:36] oh meh [17:37] did it fix? [17:37] hatch no, but I need to get this landed, have a call with ben, have lunch, and start the release process all in the past seven minutes ;-) [17:37] this meaning my cookie branch [17:38] ok I can repro in ubuntu chrome [17:38] hmm [17:38] good catch! [17:38] :-) sorry there's more work but glad you could repro [17:39] gary_poster: bug about juju-core subordinate charms in juju-gui confirmed fixed, I see the extra subordinate icon in the UI now === andreas__ is now known as ahasenack [17:40] oh, sorry benji, didn't see your message [17:48] there is a massive font-size difference between osx and ubuntu chrome [17:48] in the inspector [17:50] bac: let me know when you want to chat [17:50] how long do I have to wait for the eye candy :>? [17:50] (to be ready) [17:51] gotwig_: the GUI is usable now, if that's what you're asking [17:51] benji: the new gui :> [17:52] gotwig_: we did a release last week; is that new enough for you? ;) [17:53] benji: you know, the charm view after you already deployed the charm, to configuer stuff, sucks :> IMHO. I cant save changes? [17:53] benji: I go into the deployed charm click on settings or something like this, and e.g try to change the port. how can I save these changes? [17:54] gotwig_: are you using any flags? [17:54]