=== alexpilotti_ is now known as alexpilotti [10:00] hi luca: how do you want the "Your service has been added" to be dismissed? for the current ghost inspector? once for all future added services (onboarding style)? for a specific service? [10:17] frankban: are you talking about the black pop-up? [10:18] luca: I am talking about https://bugs.launchpad.net/juju-gui/+bug/1379655 [10:18] Bug #1379655: The "this service has been added" notification can't be dismissed permanently [10:18] frankban: oh! right! [11:50] morning [13:43] hatch: ! hey, you disappeared for a week :P [13:43] *yawn* [13:44] hey rick_h_ :) I did! [13:44] as ordered :P [13:44] hah [13:44] welcome back to one week before launch. Get to work! :P [13:44] tbh I spent 1/2 of it in bed with a cold :/ [13:44] lol [13:44] hatch: :( [13:45] hatch: when you get caught up let me know and we can chat. We're in crazy pre-launch mode [13:45] you bet, I'm going to 'start' in 15m, just catching up on the remaining emails [13:45] hatch: rgr [13:45] * rick_h_ goes to make some coffee [14:01] rick_h_: alll ready [14:04] hatch: k, standup roomn [14:27] rick_h_: https://github.com/hatched/juju-gui/tree/react-spike there is only a single commit so you can get the entire diff by going to https://github.com/hatched/juju-gui/commit/7db9fa3d3f976a1d052d9320dd82140c21262c6d [14:27] hatch: ty [14:27] just of course keep in mind that this is not the 'proper' way :) [14:28] oh course [14:55] jujugui call in 10 [14:55] errr 6 [14:55] :p [14:56] was just about to say the timestamp disagrees. [14:56] 4 here [14:56] :P [14:56] rick_h_: ntptrace plz. [14:56] but my clock is off and I can't change it for some reason lol [15:02] rick_h_: standup. :p [15:18] mannnnn I'm 1500 points away from the next Altitude level with Air Canada :/ rick_h_ can we shoehorn a sprint in before the end of the year? :) [15:21] hatch: I'm taking a weekend trip out to Seattle for just that purpose :-) [15:21] lol [15:23] Makyo: right now one of the added services bugs is that dragging the canvas unfades any faded services. Any pointers on where to look? [15:26] kadams54: look in topology.js (I think) [15:27] I'm pretty sure that's where it sets the translate [15:27] and then calls the d3 methods [15:27] going from memory here :) [15:28] I'm pretty bummed that apple didn't release a 4k cinema display but instead a cobbled together 5k in an imac :/ [15:29] hatch: It's a hard-knock life. [15:29] kadams54, grep for 'show' in app/views/topology [15:29] I'm actually very surprised they did the 5k display [15:30] they had to overdrive the thunderbolt to be able to do it [15:30] seems like an odd choice [15:30] kadams54, I'm hunting too [15:31] Makyo, hatch: yeah, I did look around for "show" but didn't find anything. [15:32] kadams54, Try breaking service.js:1508 [15:33] I'm wondering if it's registering a click along with the drop handler, and clearing the state(which shows all services) [15:33] kadams54, that occurs in relation.js:722 [15:34] rick_h_: are you ready for some awesomely amusing news? [15:35] lazyPower: maybe [15:35] rick_h_: friend reaches out over email - "The juju gui just ate all my units, what happened? I just see one service and i cant move it on the canvas" -- i reply "Juju upgrade juju-gui" - response: "Awesome! that worked. ta" [15:35] i have no idea which version they were running, but thats third party verification you guys fixed it [15:36] lol [15:37] we fixed a bug we may or may not have known existed [15:37] awww yeah *puts on sunglasses* [15:37] actually, come to think of it, i deployed that environment around the same time i upgraded my DO env to machine view.... [15:37] hah i may be the only one breaking the gui with whatever i was doing on manual envs. [15:37] hatch: TP-LINK TL-WDR4300 rocks [15:38] hatch: new tagline - "juju gui team, fixing lazypowers broken deploys since 2014" [15:38] lazyPower: haha [15:39] jrwren: I was thinking of getting the new monster linksys [15:39] lazyPower: yea, manual provider and such needs more love but getting better [15:39] jrwren: but I like the price of this one better :) [15:40] rick_h_: its definately getting better. This cycle with juju-alpha builds, and the new gui have been an exceptional experience. [15:40] lazyPower: <3 [15:41] hatch: remember, linksys:cisco::toyota:lexus :p [15:42] hatch: the tp-link also supposedly runs openwrt very nicely. I just haven't found the need to change from default firmware yet. [15:42] jrwren: since I own two Toyota's that means I should buy the linksys right? [15:42] mount it on the back as a spoiler [15:44] jrwren: however unlike Toyota's I've had to replace every one of my old linksys routers after a year because of failures. :) [15:45] hatch: terrible analogy. I fail. [15:45] lol [15:45] It also fails in that the linksys router has a TON of features whereas Toyota's ususally fall short on the feature category :P [15:46] So maybe in my experience if linksys routers they are the Chevy of the router world lol :P [15:55] kadams54: there appears to be some added services bar tests that block the test runner - are there any running settimeouts? [15:55] yeah, always failing. [16:12] I have finally found a use for generators in javascript [16:12] wow that took too many years [16:12] :P [16:22] note to self: when spamming 'enter' make sure you don'thave any pending updates which may ask to install === fabrice is now known as fabrice|family [16:44] nope, still don't like the white sidebar :P [16:44] lol [16:48] rick_h_: crit bug before release https://bugs.launchpad.net/juju-gui/+bug/1383381 [16:48] Bug #1383381: autoplaced units don't show up in machine list [16:50] hatch: on the tests, no, no setTimeouts, just ones that are waiting for an event to fire before calling done(). [16:51] kadams54: yeah ok thanks I saw that after (sorry I forgot to postback) it looks like some of the events were taking forever to propagate in the branch I was working in [16:51] :-() [16:51] Smiley fail [16:51] :-( [16:51] lol [16:51] it's fixed now - not sure why the branch I was working on caused those events to hang [16:51] not even related haha [16:51] oh our test suite.... [17:03] rogpeppe: do we want GetArchive to return a complete id or to modify in place the given one? [17:03] frankban: let's return a new id [17:04] rogpeppe: +1 [17:05] frankban: rogpeppe other chan [18:43] rick_h_: feature request to re-write juju-gui in go using https://github.com/robertkrimen/otto [18:43] :X [18:43] j/k btw [18:44] ummm, wow [18:44] lol crazy right [18:44] well, better option that dart :P [18:44] * rick_h_ pokes hatch in the ribs [18:44] ouch [18:44] lol [18:45] dart syntax is nicer than go :P [18:45] hah [18:45] any syntax > go [18:45] :X [18:45] lol!! [18:45] lies! perl < go < * [18:46] :P [18:46] haha [18:46] haha - perl CAN be written nicely [18:46] go can't :P [18:46] i wrote juju bindings in perl :( [18:46] im so ashamed [18:46] did it look like #$^&*(*&^%$%^&*&^%$%^&* [18:46] :D [18:46] haha [18:46] no, that's how you make an ajax request over cgi [18:46] rofl [18:47] not a juju websocket connection :P [18:47] my eyes only bled a little [18:50] just remember, go syntax is ugly like that because the C compiler is too slow, so we have to help the compiler be fast. Its just a slight improvement over nasm. [18:52] haha [19:36] test_fakebackend is over 3k lines :/ [19:46] and now to make it bigger! :P [19:47] bwuahahahaha [19:47] good news is that it now supports bundles with multiple relations per endpoint [19:47] from a bundle yaml file [19:47] :P [19:48] and all the current tests pass [19:48] so yay [20:08] kadams54: hey does added services still need to be under a flag? [20:08] it's close enough that it'll definitely be in the next release right? [20:08] * hatch just keeps forgetting to use the flag ;) [20:12] I bet it was quiet in here last week [20:12] aren't you all happy I'm back?? [20:12] :P [20:12] hatch: yes, still behind a flag until it's ready for release :) [20:12] hopefully EOW [20:12] ahh that's how we're doing it? cool [20:12] hatch: well there's still a few cards of work [20:13] ohh [20:13] looked done to me lol [20:13] we had UX updates last week and there's the 'remember me' stuff still [20:13] ohh gotcha [20:14] don't suppose those UX updates change the background colour? ;) [20:15] :P [20:16] in lucas blog post there were some awesome mockups [20:16] after seeing them I wonder why we didn't use those hah [20:17] I should probably keep quiet or we'll get a new UI this week [20:17] lol [20:19] the new images are in the email? Or card? [20:19] https://drive.google.com/drive/u/1/#folders/0BwDPGKe0SiMbU0RfOXZIXzJlODg/0B7XG_QBXNwY1V3B3dDNvYXJGRE0/0B7XG_QBXNwY1NEtGaHJYZGM4enM/0B7XG_QBXNwY1N3Rld2RvdkFUMWs [20:20] oo new google drive ui [20:22] added services w/machine view is sure nice [20:22] ppl will like that [21:14] jujugui lf a review and qa https://gist.github.com/bcsaller/131454a8fabb9e0f8154 [21:14] oops [21:14] https://github.com/juju/juju-gui/pull/623 [21:14] ^ that one [21:18] hatch: will try to look tonight after the boy goes to bed [21:18] sure np, it's not blocking anything else [21:19] thanks [22:05] jujugui anyone around? for a loop at a wip? [22:05] Luca: hey [22:05] hatch: heya [22:06] hatch: how's it going? [22:06] I read your blog post - looks good [22:06] some of the prototypes looked awesome :) [22:06] hatch: thanks [22:06] definitely want some of that stuff! lol [22:06] Haha [22:08] yeah some of those look really good [22:08] are some of those ideas going to come back? [22:08] morning huwshimi [22:08] Morning [22:09] Which ideas are you talking about? :O [22:09] Morning huwshimi [22:09] Luca: well they go by so fast....but they all have the black inspector open [22:10] Luca: Morning [22:10] Luca: I tweaked your bugs today, let me know if we need to chat on any of them (and that can be tomorrow) [22:10] rick_h_: https://github.com/hatched/juju-gui/compare/juju:develop...hatched:deploy-target?expand=1 initial go at the state stuff - thoughts? [22:11] hatch: oh! The gif. I'm not site if any of them will come back, I would love to see analytics though. [22:11] aww darn - I just like the styling of some of the tokens - they look 'tighter' [22:11] rick_h_: OK, did you tweak them in GH or Kanban? [22:11] Luca: GH and turned them into kanban [22:12] Luca: per the email I sent out to you and the guys working on the items [22:12] huwshimi: isn't it very early there? [22:12] rick_h_: cool, I'll check it out [22:12] hatch: 9:12am [22:12] oh did the tz's change? [22:13] hatch: Yeah, we changed to dst [22:14] ohh ok [22:14] rick_h_: I just saw your email. Thanks, the more eyes the better :) if we add some new ones do I just leave them unsigned or assign them to you do you know that I've added them? [22:20] Luca: definitely, I'll track the emails as they come in and triage them [22:20] Luca: but I had to cut down a bit. We had 36 high priority bugs for 3 peopla and 5 days of work [22:20] rick_h_: cool, I'll do that [22:20] Luca: so I'm trying to help get it down to the ones required for the guys to do over the next 4 days to get to release [22:21] rick_h_: yeah, I can't see many being added but the guys are some design bits which should be wrapped up tomorrow [22:21] Luca: rgr [22:22] the question will be are they release blockers or not and how they line up with current release blocking tasks [22:23] Well, its the Get Started page and 404 page and some content updates [22:23] We could launch without Get Started and add it for ODS [22:26] Luca: ok, I think those are on the current list. [22:28] rick_h_: cool [22:31] rick_h_: is it safe to assume people would specify the 'bundle:' prefix when using deploy-target? [22:32] so it would be a trivial flag to determine if it's a bundle or not [22:32] hatch: no, the urls that will come in will have ~user/bundle I think for now, but I think that'll be going away [22:33] I'm trying to figure out a way to determine simply from the id string whether it's a bundle or not [22:33] unfortunately we have too many optional fields lol [22:34] hatch: so figure out what we want it to be and we can update the front end to send it [22:35] well if it can send the fully qualified id bundle:mediawiki/7/single and cs:precise/mysql-48 etc [22:35] then we'd be golden [22:35] easy to indexOf on the prefix [22:35] can we do that? [22:36] yes, just make sure we document it so we can make sure we do that on the other end [22:36] hatch: at least for initial release [22:36] sounds good [22:44] Night all [22:51] bundle deploys - aww yeah [22:56] Super interesting, if depressing: http://www.npr.org/blogs/money/2014/10/17/356944145/episode-576-when-women-stopped-coding [22:57] Makyo: tldl ? [22:57] hatch, still listening :P [22:58] oh lol [22:58] that topic is in-vogue so I try to avoid it at all costs [22:58] Got hooked by the graph. I always get hooked by graphs :/ [22:58] so many idiots [22:59] like why can't we all just be friends? :) [22:59] because there are canadians in the world? :P [22:59] bwu-ha-ha-ha! [23:00] hahaha [23:00] oh u suck [23:04] Well I'm taking off - see you all in the am