[00:40] Makyo I will make a release this evening or get someone else to do it tomorrow morning then. Looking at it now... [00:42] gary_poster: are you able to edit benji 's post from the 18th? there is a few spelling mistakes and it doesn't look like I can edit. [00:42] trying [00:42] yes [00:43] hatch, beginning, what else? [00:43] documentation [00:43] at the bottom [00:44] got it thanks [00:44] I think that's all - my computer highlights more because not English (missing u's) ;) [00:44] :-) [00:44] cool [00:44] thanks, fixed [00:44] thanks [00:45] I'm just writing a blog post about the release [00:45] great! put yourself on the spreadsheet, hatch (second page of https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0At5cjYKYHu9odDFwb0gxQTROSzZOX0ZWUEMzZWtJTGc#gid=2) [00:49] will do, should be up shortly, just building trunk so I can take a screenshot :) [00:59] jujugui http://fromanegg.com/post/53479366574/juju-gui-0-6-0-released [00:59] woops didn't mean to ping everyone :) [01:00] hatch :-) 0.6.1 will be out soon hopefully... [01:01] well then I guess I'll have to make another post :) [01:01] :-) [01:15] Makyo, hatch, 0.6.1 released [01:15] aww poop [01:15] when you said soon I thought like in a couple days [01:15] lol [01:15] :-) [01:42] gary_poster, thanks, almost forgot I owed James a dinner [01:42] :-) cool [10:44] gary_poster: did you get my spelling errors in the release posing fixed? If not I can do that now. [10:44] * benji remembers what time it is. [11:29] benji, yeah. [12:09] hey gary_poster, i need to be out for an 11am appt so i'll miss the meeting. sorry for the late notice. [12:09] bac ok thanks for heads up === rogpeppe2 is now known as rogpeppe [12:22] gary_poster: y'day joey was asking me about non-canonical folks being in ~juju-gui-charmers. it seems there are several via ~inactive-charmers -> ~charmers -> ~juju-gui-charmers. i sent him to mimms. seems like a non-issue to me... [12:31] bac, thanks for forwarding to mims. AIUI, charmers are supposed to include non-Canonical folks [12:32] my gut feeling as well. i think joey has a mandate to at least document these things to show they didn't just happen [12:47] I see [13:09] gary_poster: 'bzr tags' for juju-gui doesn't show a 0.5.0 [13:10] bac :-( k looking [13:10] and many have ? instead of a revno. [13:12] I'm sure I tagged it. [13:13] bac r677 [13:13] oh, 0.*5*.0 I did 0.6.0 [13:14] I also did 0.5.0 and aparently didn't tag it. I don't know how that happened. [13:15] maybe you added it to checklist after that benji? I only did it since you added it to checklist [13:15] bac: revision 677 is 0.5.0 [13:15] gary_poster: perhaps [13:22] hatch: when you are around I'd like to get some ideas on the right place to wire up some code (http://paste.ubuntu.com/5786687/) [13:24] benji: I'd just stick the method in the widget with tests and call it manually from the views/editorial.js when the containers are created. Instead of Y.all, it'd be container.get('container').all(). The zoomplane bit shuold be an event that's setup when the .zoom-plane div is rendered. [13:25] rick_h: thanks, I'll try that [13:25] it's the charm-containers job to be optionally draggable, the sidebar's job to tell the container to enable the feature, and the zoom-plane to handle being dropped on. [13:25] gary_poster: i've reverted to 677, the browser is gone...but there is no charm search on the right either, so there is no way to add a charm [13:25] benji: ^ [13:26] bac, looking. [13:26] i think i need to manually turn sandbox on [13:26] bac, never connects for me...yeah I think so [13:28] bac, add user and password also [13:29] gary_poster: it is a mess. i've reverted to latest and will test locally to get it working [13:29] admin:admin should be fine [13:29] ok bac [13:29] so for right now we have charm browser back temporarily [13:30] bac you still should set user, password, and sandbox: true in config-prod.js [13:31] abentley: r=me on the short-url branch. [13:31] rick_h: Thanks! [13:42] gary_poster: uistage should be happily 0.5.0 now [13:48] bac, perfect, thanks [13:50] gary_poster: i didn't take into account local modifications before reverting to 677. i've copied the local changes of app/config-prod.js to app/config-prod-mod.js [13:50] they are trivial, of course [13:50] cool thanks bac [13:53] bac, how much work would it be to set up the old approach on another port, do you think? [13:54] gary_poster: prob not hard, just some poking at the web server config [13:54] gary_poster: i'll be glad to do it after lunch [13:54] bac, ok. thank you, that would be great [14:02] morning [14:21] benji: did you have an updated version of that code? (after ricks comment) [14:22] hatch: it's not done yet [14:22] okee np [14:24] hatch: agree/disagree? /me is curious [14:25] agree [14:25] luca__, thanks for wireframes. you available for guichat to review? [14:26] gary_poster: not at this very moment, I think I'm going to be pulled into an impromptu meeting hehe [14:26] heh [14:27] luca__, ok. I have call in 34 minutes for an hour. can talk before or after. woudl very much like to talk before your EoD today [14:27] gary_poster: Ok, I'll definitely find time :) [14:27] thanks :-) [14:39] gary_poster: I'm free now, if you are [14:39] cool, come on by guichat luca__ [14:41] I didn't know one could do such a thing so easily in Javascript, nice: this['handle' + data.Type + data.Request]() [14:54] jujugui call in 6 [14:54] jujugui call in 6, kanban now [14:54] kanban now [14:54] Oops. [14:54] :-) thanks [14:54] on call [14:59] jujugui call in 2 [14:59] ...1 [15:00] 0! [16:02] hey luca__ , guichat? [16:03] gary_poster: coming! [16:03] cool :-) [16:04] oh and fyi - dart is pretty cool [16:04] hatch, cool :-) [16:04] it's being sold as the 'next javascript' and at least on the very limited hacking i've done with it, I agree [16:05] it's not even v1 though so it's still under heavy development [16:10] rick_h: you have your monitors mounted on an adjustable pole right? [16:13] node 0.10.12 is out [16:28] hatch: yes [16:29] are they adjustable? I'm thinking of grabbing some but not sure what kind to get [16:29] hatch: http://goo.gl/NA7F3 [16:29] hatch: they're semi-adjustable. Cheaper than the fully adjustable and enough movement for my needs [16:29] ahh - don't think that one will work for my primary monitor unfortunately [16:29] darn [17:02] Makyo: reviewing [17:03] Makyo: should the QA not be done on prod? [17:03] or both for that matter... [17:04] hatch, I suppose both, but the goal was to make this a quick run-through, which the sandbox gives us, right? [17:05] I remember issues which only happened in prod [17:10] abentley, rick_h , I see a couple of oddities in search right now. (1) if I search for "ceph," on uistage (old interface) I get an IO error, and on trunk (new interface) I never get results. [17:10] (2) if I search for apache2, I get no icon for the charm [17:10] Is this known? [17:11] ah wait I just got ceph from the new interface... [17:11] gary_poster: looking. It's two diff api endpoints [17:11] well, it was hanging before [17:11] gary_poster: yea, it's throwing a 500 :/ [17:11] maybe I just didn't wait long enough, but apache2 results were very fast [17:12] rick_h: I searched for ceph in my local branch in the charmbrowser, now I can't get the real results, even after refresh and clearing the input box - known bug? [17:12] s/real/original [17:12] gary_poster: oh, it's actually pulling the old site jujucharms.com? [17:12] i just get the ceph result list [17:13] hatch, okay, that's fair, just lets make sure to find a balance. The impression I got from the call was a default-usage script with the 'actively breaking things' thing done by a dedicated QA person. [17:13] Makyo: definitely [17:13] rick_h, sorry, don't understand your question [17:13] abentley: can you peek at staging? http://staging.jujucharms.com/search/json?search_text=ceph [17:14] hatch: what's the url from the network debug tools for your failed request with charmbrowser? [17:14] rick_h, fwiw, on uistage: [17:14] charm_store_url: 'http://jujucharms.com/', [17:14] charmworldURL: 'https://manage.jujucharms.com/', [17:15] gary_poster: yea, I guess they must be pointed at the same place. Hadn't really thought about them being different now. [17:15] rick_h: it's not failing, it's persisting the search results cross reload, I can't get back to the original results with the featured charms and whatnot [17:16] hatch: right, there's no "go to home" button in the UX and we've talked with the UX folks several times about it. I *think* they've got a new home button in the updated UI that huw is starting work on, but it's not there now [17:16] well why does the X when I clear the search box not do that? [17:17] rick_h, oh! I should point config charm_store_url to manage? still doesn't work there but easy to do [17:17] and why does it persist cross reloads [17:17] hatch: it does an empty search and you end up querying all charms [17:17] http://manage.jujucharms.com/search/json?search_text=ceph [17:17] gary_poster: I think it's ok to do that now. I've not verified, but assuming they're the same machine/dns now. [17:17] ok will check locally...later. need some lunch. :-) [17:17] thanks [17:17] gary_poster: right, I ping'd abentley to see if he can check the error log. sinzui gets the error emails but isn't around. [17:18] cool thanks [17:18] gary_poster: so probably some error/issues around the ceph charm causing it to puke [17:18] rick_h: just got back. Looking. [17:18] makes sense [17:21] rick_h: It appears our code isn't forward-compatible with raring: http://pastebin.ubuntu.com/5787391/ [17:21] rick_h: is there a bug number? [17:22] abentley: no, gary_poster just hit it and was asking me if it was a known issue. I'll file a bug now with that traceback. [17:22] rick_h: Cool. I'll look at it soon. [17:22] abentley: ty much [17:25] rick_h: https://bugs.launchpad.net/juju-gui/+bug/1193437 this is pretty critical imho but I left the importance open [17:25] <_mup_> Bug #1193437: Cannot get back to 'featured' charm page after search

[17:25] abentley: bug/card files [17:27] I'd like someone else to confirm that bug but then they will be stuck without a featured page [17:27] :) [17:27] hatch: I understand your point, it's design and it's been brought up. [17:27] hatch: it's getting handled in the updated design work [17:27] no persistence across reloads is a code bug not a design issue [17:28] hatch: I"m confused then. You mean actual browser reloads? [17:28] hatch: the urls udpates, if you reload it should drop you back where you are. It's how it's supposed to work. [17:28] yeah when I hit refresh on the browser, the charmbrowser shows it's spinner, then shows me the full list of charms [17:29] http://localhost:8888/sidebar/search/:flags:/serviceInspector/ [17:29] hatch: right, because you just did an empty search [17:29] that's the url [17:29] hatch: right [17:29] it's how it's supposed to work :) [17:29] so if I make a search then it will never show featured charms again for every following reload? [17:29] "hey hatch, go check out the list of charms available: http://xxx/sidebar/search" [17:29] hatch, Going to remove the second section then, since prod uses a backend by default, rather than sandbox. [17:30] hatch: yes, and that is a known issue and will be addressed by adding a home button into the sidebar heading [17:30] hatch: so once you do a search, there is no way to get the default 'helpful' content back to you right now without url mangling. [17:30] ohh ok now I see what you're saying [17:31] * hatch still thinks an empty search should drop you at the 'home' page [17:31] hatch: then there will be no way to get a list of all charms in the store on one screen. :) [17:31] good? [17:32] why for you hate my use case I might want :P (and my easy hack to find the total number of charms in the store) [17:33] if you go to google and search for blank do you get every site on the internet? :) [17:33] fyi - I have this argument on pretty much every project I've ever been on lol [17:33] hatch: the url says search, I search :P [17:34] hatch: sure, we can rework the patterns, provide a reall "all charms" endpoint and find a way to expose it in the UX. ...but we don't have it right now. [17:35] ok I'll cede my argument.....for now *shifty eyes* [17:36] hatch: wait 3wks and bring it up with design if it's not there. I can't find the doc with the new UX now but I swear I had it and it had a real home house icon to help with things [17:36] can do! [17:43] ahhh ux debates, gota love em ;) [17:48] rick_h: Since this bug affects only the original API, I'd love to just abandon the original API. [17:53] rick_h: how does your truck tow your trailer? pretty stable? it looks way outclassed in that pic :) [18:01] abentley: yea, I was hoping to get to that next week to audit the old calls. sicne they're leaving uistage up on old revs for now we can't remove the old api until that goes away now [18:01] hatch: it's within spec :) 6k lbs max trailer weight and 7.7k max on the car [18:02] wow that thing has a 6k capacity? crazy! [18:02] rick_h: Ack. I'll fix it. It's not a big change. [18:02] hatch: well the trailer weighs 4800 empty I think. Add batteries/junk and you can take it up to 5900 [18:02] haha that's funny [18:03] and it tows well? [18:03] doesn't wag the truck? [18:03] :) [18:03] hatch: so far, only taken it home from the dealer. I've got to load it up in a bit for this weekend's test trial. [18:04] hatch: the big test will be in the fall driving it down to VA from MI for annual 'visit the family' trip [18:06] bac: excellent choice of port number [18:06] :) [18:06] i figured it would be memorable [18:07] rick_h: ahh well make sure to balance the weight out properly [18:07] hatch: got a tongue scale in the garage ready to go :) [18:09] haha well done [18:09] always remember kids, "1/8th the weight or your trailer you will hate" [18:09] your more prepared than most lemme-tell-ya [18:09] benji: HAHAHA [18:09] that's awesome [18:10] hatch: yea, I've been doing research for a while now. I'm kind of an over-research guy. [18:10] yeah same here [18:10] benji: hah [18:20] benji: do you have any objection to me adding typescript and dart to your coffeescript card? as they are sort of in the same 'ballpark' [18:20] hatch: nope, that would be great [18:21] although, I suspect it decreases the chance of choosing one ;) [18:21] haha - maybe that's a good thing lol [18:28] so samsung came out with their new 9 series ultra book with a crazy high res screen [18:28] then they gave it an hdmi out :/ [18:28] so the 13" screen has higher resolution than anything you could hook up to it, what a waste [18:33] rick_h: Could you please review https://code.launchpad.net/~abentley/charmworld/fix-old-search-sort/+merge/170894 ? [18:33] abentley: looking [18:35] abentley: should we add a test/check in import for a series we don't recognize from that list and toss an error to help catch new releases as they come about? [18:35] rick_h: If we aren't done using the old API by the time the next series is announced, I will go on a killing spree. [18:36] abentley: ah, gotcha. Since this was hit with our api2 seearch I figured it'd still be an issue. [18:36] abentley: r=me [18:37] rick_h: No, gary_poster retracted that: ah wait I just got ceph from the new interface... [18:38] rick_h: I've checked myself, and the new API isn't affected. [18:38] abentley: cool, thanks. [18:39] abentley: ah, my fault. I tried to manually build a search call and did /charms/search?text=ceph. [18:48] rick_h: I've filed bug #1193463 about the brittleness of result_sorter, but I don't anticipate we'll have to resolve it, since we'll remove the old API before Ubuntu T is named. [18:48] <_mup_> Bug #1193463: Old JSON search API explodes when new series are introduced [18:48] abentley: cool [19:04] gary_poster: We have a fix for the Ceph search issue. Is Monday soon enough to deploy it? [19:08] * gary_poster on call. abentley yes ok [19:21] gary_poster: heya to tldr on the gui export json thing landing? is that "OSCON"? [19:22] jcastro, gui export might be July 1 or thereabouts. I will get back to you/Antonio about it this afternoon. gui import json is "OSCON" or thereabouts [19:22] ok perfect [19:23] just making sure I was at least -/+ the correct month. :) [19:23] :-) [19:29] wow this new ghost config is a lot less code hah [19:29] like 1/4 [19:29] * hatch thinks he must be missing something [19:36] bcsaller: are you around? [19:36] hatch: yes, on a call [19:36] alrighty ping me when free [19:51] hi Makyo, you i ask you about some stuff in test_sandbox? [19:51] bac, sure [19:52] Makyo: i intended to add expose and unexpose. in doing so, i think for testing it is easiest to do add_units at the same time. [19:53] Makyo: so i'm trying to port the generateServices helper for Go. but, i'm missing something and client.onmessage is not getting called [19:54] bac, oh, hmm... I had something similar happen, but it turned out to be a typo of Receive :/ [19:59] hatch: off call, whats up? [20:01] In order to do this ghost config properly I need to be able to bind events to functions in the viewlets [20:01] was wondering if you wanted to chat about that [20:02] hatch: sure [20:02] guichat is open [20:05] hatch bcsaller sent you some wireframes that might or might not help with clarifying ghost inspector goals [20:05] gary_poster: thanks [20:07] arosales, jcastro: tldr, we can have a GUI-based export for you by Wednesday morning of next week. I'll coordinate with Kapil to connect some of the technical dots, but should work fine. GUI import will still be tracking for OSCON. [20:08] ok so we can do everything we need [20:08] and just slot in the workflow on wednesday afternoon [20:08] toooooo easy [20:08] gary_poster, woot thank you sir [20:09] jcastro, arosales :-) welcome [20:09] gary_poster, and to confirm this export should work with juju deployer? [20:09] arosales, yes [20:09] gary_poster, perfect thank you [20:10] :-) happy to be a part of what you guys are doing [20:10] thanks for including us [20:10] rick_h: Could you please review https://code.launchpad.net/~abentley/charmworld/remove-bzr-ingest-job/+merge/170899 ? [20:11] gary_poster: thanks [20:13] gary_poster, np more folks need to see wow'ed by the GUI :-) [20:13] s/see/be/ [20:13] a-greed [20:15] gary_poster: after looking at the wireframe, it looks like everything here will be pretty straight forward, nothing really needs to change at the architectural level of how Ben and I did the various components [20:16] hatch, excellent [20:19] hatch, fwiw, separate from the architecture, one potentially tricky UX bit is the scrolling behavior he wants for the inspectors. you know how the contact app on iphone works, when you scroll up through the letters, and each letter is a header as you scroll, but then when you get to the next letter it rolls that previous letter away? that's how he wants to handle scrolling in the inspectors when the box size is b [20:19] igger than the content. [20:19] I figure we add that in gradually after we build the main pages :-) [20:20] OOoo pick me pick me, I wana write that [20:20] hatch, consider yourself picked :-) [20:20] Yussssss [20:20] lol [20:21] but yeah...later :) [20:33] bac thanks for sending out email about uistage changes, and for doing the work. [20:34] was trivial [20:37] lol 8080 and 8086 we r such geeks [20:38] hatch, please keep conversation going with Ant. Seemed like a reasonable response from him, but it seems like we have an obvious reply (that he seems like he's even setting us up for) that we are building an app, not sites. I imagine you can suggest a setup that works for both him and us? [20:39] definitely [20:39] will reply shortly [20:41] whenever I see 'Ant' I always think of the build system [20:49] :-) [20:49] thanks [20:50] bac, do you mind if I move our weekly Thursday call to my 9:30 AM, 1330 UTC [20:51] gary_poster: that's fine. will be 830am for me 1/2 year [20:51] not too early [20:51] ok thanks bac [20:53] benji, do you mind if I move our Thursday call one hour earlier to my 2:30, 1830 UTC? [20:53] gary_poster: nope, that's fine [20:53] thanks