| huwshimi | hatch: Any ideas how to not actually include that file in our build step? | 00:33 |
|---|---|---|
| huwshimi | (the yui3-skin-sam file) | 00:33 |
| hatch | umm | 00:34 |
| hatch | what's it called? | 00:34 |
| huwshimi | hatch: I dunno. yui3-skin-sam.css? | 00:35 |
| hatch | oh crap, I was going to grep for it but I'm trying to fix my vm right now so I cant :) | 00:35 |
| hatch | so anyways, check in the make file, and the merge-files merge-files.js files | 00:35 |
| huwshimi | hatch: Oh actually could it just be including the skin file for each widget? | 00:40 |
| hatch | thats possible too - or it was just copying the css over | 00:41 |
| huwshimi | hatch: There's a this call which might be doing that: merge.getYUIFiles() | 00:43 |
| hatch | sorry I gota run for a couple hours | 00:43 |
| hatch | when I get back I can take a closer look | 00:43 |
| huwshimi | hatch: It's all good, thanks :) | 00:44 |
| gary_poster | huwshimi, hi! I kicked comingsoon. You should have regenerated images and css now. Please report on the mailing list if you see some good changes. | 01:25 |
| gary_poster | hatch, gave you LGTM and QAOK. | 01:25 |
| huwshimi | gary_poster: Thanks a lot! | 01:25 |
| * gary_poster runs away again. | 01:25 | |
| gary_poster | welcome :-) | 01:25 |
| hatch | huwshimi: back | 03:34 |
| hatch | annnd my vm is bricked....awesomer! | 03:34 |
| huwshimi | hatch: Oh dear | 03:37 |
| hatch | I'm mostly pissed that people in the real world are probably having the same issues | 03:37 |
| huwshimi | hatch: I found this, it looks like this person is trying to do the same thing: http://yuilibrary.com/projects/yui3/ticket/2530057 | 03:38 |
| hatch | yeah - so were you able to find all instances of it in our codebase? | 03:39 |
| huwshimi | hatch: Yeah, the only references were in our CSS or HTMl | 03:42 |
| hatch | oh ok awesome | 03:42 |
| huwshimi | (Which are all now gone) | 03:42 |
| huwshimi | hatch: I would have thought it was a common thing to not include any of the base CSS. | 03:44 |
| hatch | yeah the GUI on this vm is totally borked | 03:58 |
| hatch | basically it looks like vm's do not love 13.10 | 04:02 |
| huwshimi | ouch | 04:02 |
| hatch | looks like people are having similar issues on virtual box and parallels | 04:03 |
| hatch | I guess when they made the changes to the display stuff noone checked to make sure it works in vm's :D | 04:03 |
| hatch | huwshimi: did you happen to get your branches landed? | 04:05 |
| hatch | er branch? | 04:05 |
| huwshimi | hatch: the yui skin one? | 04:05 |
| hatch | the bundle labels one | 04:05 |
| hatch | er | 04:05 |
| hatch | search labels | 04:05 |
| huwshimi | hatch: Oh, no I didn't. I'm happy to just change the labels or leave them with the flag. I don't think it matters that much, but if you or someone else has strong feelings I can fix it up... | 04:07 |
| hatch | not strong feelings, I was just commenting because I don't really see what bennefit they give us :) | 04:08 |
| hatch | ""it WILL be 'featured' next week, but today its 'featured charms'"" | 04:08 |
| hatch | :D | 04:08 |
| huwshimi | hatch: I'm happy to change it if you like :) | 04:10 |
| hatch | well don't if there was a reason for the flag | 04:11 |
| huwshimi | hatch: Nope, only cause Gary mentioned to do it that way :) | 04:13 |
| hatch | :) | 04:32 |
| gary_poster | hatch, not quite sure what you and huwshimi were referring to, but I'm happy to listen. I probably can be convinced by whatever you prefer | 12:21 |
| gary_poster | rick_h__, do you want a re-review or are you good (deploy-quicksearch)? | 12:21 |
| rick_h__ | gary_poster: I'm good, just wanted to publish the notes and take another look at the updated diff with the s/deploy/deployService | 12:21 |
| rick_h__ | gary_poster: tryingto get lbox to go through before I leave for the dr appt | 12:22 |
| gary_poster | cool, good luck. :-) any progress qith git/travis, or is that in the back seat for now? | 12:22 |
| rick_h__ | and of course, some test pyjuju test is trying to make me cranky this morning | 12:22 |
| gary_poster | :-) | 12:23 |
| rick_h__ | gary_poster: back seat, was going to play with it some at CHC tonight. | 12:23 |
| gary_poster | cool | 12:23 |
| gary_poster | whatever CHC is :-) | 12:23 |
| rick_h__ | gary_poster: but to do the github 'hook' you have to have a url you get POST'd to with a json payload | 12:23 |
| rick_h__ | gary_poster: so was looking at what it would take to do that | 12:23 |
| rick_h__ | gary_poster: sorry, coffeehousecoders http://royaloak.coffeehousecoders.com/ | 12:23 |
| rick_h__ | my weekly promised 'coding/social' time from my wife :) | 12:23 |
| gary_poster | oh, fun | 12:23 |
| gary_poster | :-) | 12:24 |
| rick_h__ | ok, off to doc. Hopefully this submit will go through. /me shakes fist at lbox! | 12:25 |
| gary_poster | :-) bye | 12:25 |
| gary_poster | frankban, welcome back. :-) approved expenses | 12:26 |
| benji | bac: can I get a quick review of https://codereview.appspot.com/19430043/ | 12:32 |
| bac | benji: sure | 12:32 |
| benji | thanks | 12:32 |
| bac | benji: our path parsing is suboptimal. :( | 12:33 |
| benji | heh, yeah | 12:33 |
| bac | benji: but this branch makes it less so. thanks. LBBGTM | 12:37 |
| gary_poster | Fruitastic! | 12:42 |
| bac | benji: is your card ready to move? | 12:51 |
| benji | bac: yep (to landing); I'll move it. | 12:51 |
| bac | yay, space | 12:51 |
| bac | gary_poster: if we're swarming on bundles is 5 the right lane limit? | 12:52 |
| gary_poster | probly not, thanks. thinking. 8 of us...I don't do much, staring out the window all day long...Francesco is only occasioanly working on this but is now...ok, I'll push it up to seven for now and will decrease to six soon after | 12:54 |
| gary_poster | Done | 12:54 |
| bac | ty | 12:55 |
| frankban | gary_poster: thanks! :-) could you please ping me when you have a minute for a call? | 13:00 |
| gary_poster | ack, will do | 13:03 |
| frankban | thanks | 13:03 |
| benji | gary_poster: my timing is poor: rick_h__ is AFK and I'm ready to talk to him about the proof stuff. Do you know enough to get me started? Does he have WIP? | 13:03 |
| gary_poster | benji, I think he's almost ready to start it. I suggest looking at the normal priority ones | 13:04 |
| benji | k | 13:06 |
| gary_poster | frankban, I am now having fun filing an RT. Let me know when you are ready, and if you have a hangout or you want me to make one | 13:10 |
| rick_h__ | benji: so I was going to start that card now that my current branch is landing | 13:19 |
| rick_h__ | benji: can you request the deploy if you and bac are ok with it? I meant to get back with you guys after the call yesterday | 13:19 |
| benji | rick_h__: I don't see any reason we can't deploy. This will be the first deploy I have requested (I think) so I'll look up the how-to. | 13:20 |
| rick_h__ | benji: sec, I've got a doc I think to look at | 13:20 |
| rick_h__ | benji: shared a gdoc with you. I think bac has a more official one, but this was my notes from my time requesting it | 13:21 |
| frankban | gary_poster: ready when you are, https://plus.google.com/hangouts/_/7acpjjctqbq3ojqji2r6llsln8?hl=en | 13:21 |
| benji | thanks | 13:21 |
| rick_h__ | bah, gary_poster beat me to landing so I got the conflict :P | 13:22 |
| benji | I hate google docs so much. | 13:27 |
| rick_h__ | lol | 13:27 |
| rick_h__ | brb, moving to coffee shop | 13:28 |
| bac | helpful tip, when paychex.com displays a message saying "System down for maintenance." it really means they don't support your browser. | 13:31 |
| benji | rick_h__: we don't want to update the charm, right? | 13:32 |
| benji | (It seems to me we should always update the charm, even if it doesn't need it, just for hygene.) | 13:33 |
| rick_h__ | benji: oh hmm, I don't think we have any charm updates but not sure. | 13:38 |
| rick_h__ | bac: lol big business translations | 13:38 |
| bac | benji, rick_h__: are you looking at this doc https://docs.google.com/a/canonical.com/document/d/1G6IoLyDz3VSw7lMRL7QUbaKKXropxyVm3uOFYsqKSeU/edit#heading=h.rrofd67agcbz | 13:41 |
| benji | bac: nope, but Rick's is essentially the same. | 13:42 |
| rick_h__ | bac: no, but that's the 'better doc' I mentioned. Just didn't have a link for it handy | 13:42 |
| bac | rick_h__: perhaps we should move to OneTrueDoc | 13:42 |
| rick_h__ | bac: definitely, where is that one located in? | 13:43 |
| rick_h__ | bac: I'm all for using it, just didn't have it handy and had a personal 'notes' doc that's mostly just copied irc conversations lines :) | 13:43 |
| bac | rick_h__: the one i just pasted is curtis' and it seems up-to-date | 13:43 |
| rick_h__ | bac: ah, I was lookin in the charmworld folder for a doc | 13:44 |
| bac | rick_h__: if you have more info in yours perhaps you could add it to curtis' | 13:44 |
| rick_h__ | bac: I foud it now in the juju-gui doc | 13:44 |
| rick_h__ | errr folder | 13:44 |
| bac | benji: i see you landed r432 of charmworld but it isn't deployed on staging. did you use 'lbox submit'? | 13:44 |
| benji | bac: I did. Darn! | 13:45 |
| benji | My next branch will be to disable lbox submit. | 13:45 |
| benji | (without disabling lbox propose) | 13:45 |
| bac | benji: i think we can just change branch ownership. i'll do that now. | 13:45 |
| benji | bac: k | 13:46 |
| bac | benji: first, let me bump staging to r432. | 13:46 |
| benji | k | 13:46 |
| gary_poster | bac or benji, could you give me an example of the charmworld deployer file URL for bundles? That is, the URL that the deployer can use? | 13:46 |
| benji | gary_poster: http://manage.jujucharms.com/~benji/bundle/wiki/5/wiki/json | 13:47 |
| gary_poster | perfect thanks benji | 13:47 |
| rick_h__ | antdillon: if you get a sec, the '+' to deploy from the quicksearch has hit comingsoon. http://comingsoon.jujucharms.com/:flags:/charmworldv3/searchDeploy/ | 13:52 |
| rick_h__ | antdillon: that icon looks a bit tiny though. Can you double check/grab me an updated png if that's not correct please? | 13:52 |
| benji | rick_h__: deploy done | 13:52 |
| rick_h__ | benji: awesome, thanks | 13:52 |
| rick_h__ | http://manage.jujucharms.com/api/3/bundle/proof woot | 13:52 |
| rick_h__ | marcoceppi: so the first stage of the proof url is on manage.jujucharms.com ^^ | 13:53 |
| marcoceppi | rick_h__: spectacular | 13:53 |
| antdillon | rick_h__, Sure | 13:53 |
| rick_h__ | marcoceppi: please let me know when proof is updated and we'll get that updated in charmworld and proofing bundles on ingest | 13:53 |
| marcoceppi | rick_h__: I'll run my tests and get 1.1 ready for release | 13:53 |
| rick_h__ | benji: ^^ can do run in parallel once marcoceppi's proof update hits if you want. Just a heads up. I'm working on the relation proof bits now, but it won't change the api or how proof needs to get added to bundle ingest | 13:54 |
| rick_h__ | bah, "can be" man wtf is today monday? | 13:54 |
| gary_poster | benji, so the best way to generate that URL from https://manage.jujucharms.com/api/3/bundle/~benji/wiki/wiki/ is to use the given id, insert "bundle" after the name, prefix the charmworld root, and suffix with "json"? | 13:54 |
| rick_h__ | gary_poster: I think the *version* is required? /me double checks | 13:55 |
| rick_h__ | gary_poster: nvm, ignore me. Guess not | 13:56 |
| gary_poster | :-) | 13:56 |
| gary_poster | benji, would it be both trivial and better to add a "deployer_file_url" key or similar? | 13:56 |
| gary_poster | bac or rick_h__ opinions also welcome | 13:56 |
| rick_h__ | gary_poster: well I would prefer it if it was an api url myself, but that was a point of discussion among us previously | 13:57 |
| gary_poster | oh | 13:57 |
| rick_h__ | https://manage.jujucharms.com/api/3/bundle/~benji/wiki/wiki/data or something | 13:57 |
| * gary_poster steps on landmine BOOM | 13:57 | |
| hatch | morning | 13:57 |
| benji | gary_poster: yep, a pointer to the deployer file from the JSON would be good | 13:57 |
| rick_h__ | we can support adding a trailing /someting to pull the deployer file straight from the api url we already have | 13:57 |
| gary_poster | yeah I wondered about that. I'd be OK with either | 13:58 |
| rick_h__ | or we can add a new key to the deployer dict of data, so yes, there's a couple trivial ways to do it | 13:58 |
| gary_poster | but we already have the other support | 13:58 |
| gary_poster | right | 13:58 |
| gary_poster | seems like adding a key to the API output would be fastest and reasonable | 13:58 |
| bac | benji: staging is now on 432 | 13:59 |
| benji | bac: cool | 13:59 |
| benji | we should add the current revno to the heartbeat page | 13:59 |
| rick_h__ | benji: +1 it was always talked about but never done | 13:59 |
| gary_poster | bac, benji, rick_h__ I will make a high priority card for this, because it will really make the deploy tab make more sense than it does now. volunteers welcome. I think proof is more important than this, though, fwiw. I welcome disagreement on my relative prioritization of this | 14:01 |
| benji | +1 | 14:01 |
| antdillon | rick_h__, I think they also look a little like close icons because of the colour but I'm grabbing a designer/UX :P | 14:01 |
| rick_h__ | antdillon: thanks, it's behind a feature flag so not exposed to users yet. Appreciate you peeking at it and agree it needs a little more <3 | 14:02 |
| antdillon | rick_h__, Luca is not about as he has been recalled to Oakland ... poor guy ... spoke to spencer who checked the design and said it matches so I guess it's good | 14:06 |
| rick_h__ | antdillon: ok, yea I mean I landed it since it does match the design. Just think it's not really looking as good as hoped in real life. | 14:06 |
| rick_h__ | antdillon: appreciate it if we could put something on the back burner maybe to see what luca thinks when he's back as I know search is his baby | 14:07 |
| antdillon | rick_h__, I will suggest we work out a hover and possibly a click state for the button. We have designs for them | 14:07 |
| rick_h__ | antdillon: hmmm, hover for just that icon? | 14:07 |
| antdillon | rick_h__, Also could it have a few px padding to increase the hit area so you dont have to be a sniper to get it | 14:07 |
| rick_h__ | antdillon: we have a hover effect/color change for the item as a whole | 14:07 |
| rick_h__ | antdillon: yea, if we were to do a bigger icon I figured that would help so didn't do the padding for now | 14:08 |
| rick_h__ | antdillon: but can look at doing that as a quick drive-by 'tweak' to make it a bit nicer | 14:08 |
| antdillon | rick_h__, Let me get the design for the tick box on inspector its quite close | 14:10 |
| hatch | gary_poster: any card in mind that you'd like done? | 14:10 |
| antdillon | hatch, Morning, my branch is good for review if you have a min# | 14:10 |
| hatch | sure thing | 14:11 |
| antdillon | hatch, Thanks! | 14:11 |
| rick_h__ | adeuring: ping | 14:12 |
| rick_h__ | adeuring: can you add the bzr revno in that heartbeat branch you've put up real quick as a drive-by please? | 14:12 |
| hatch | antdillon: is there a bundle/charm that shows it working? | 14:12 |
| adeuring | rick_h__: nice idea, but let me add that to a follow-up branch; I want to add more heartbeat data anyway | 14:12 |
| rick_h__ | adeuring: ok, thanks. benji bac ^^ any other requests while adeuring is tweaking heartbeat? | 14:13 |
| antdillon | hatch, Yeah try "workitems-tracker" | 14:13 |
| gary_poster | hatch, yeah, helping ant sounds good. after that, any card in bundles is good, preferring high and then normal. non-blocked JS-only ones include "bundle details opens with service selected..." (which I *think* has some tricky svg work because we have to make our own squircle, but I could be wrong; or import control; or you can take over on an import bug diagnosis I'm doing for Maarten Ectors for a yaml file he gav | 14:13 |
| gary_poster | e me. | 14:13 |
| benji | if it is (very) easy, uptime (since we only have one backend app server it will be an easy way to verify deployments and restarts) | 14:14 |
| rick_h__ | adeuring: ^ ? | 14:14 |
| antdillon | rick_h__, https://drive.google.com/a/canonical.com/file/d/0B2sqJRO7vy3ZZ2N0XzlTUTZSYnM/edit?usp=sharing | 14:14 |
| adeuring | rick_h__, benji; nice idea; I'll add it | 14:15 |
| rick_h__ | antdillon: that doesn't have the deploy button though in the search results from what I can see? | 14:15 |
| bac | rick_h__, adeuring: revno and uptime would be swell | 14:15 |
| rick_h__ | antdillon: or were you linking me that for another reason? | 14:15 |
| rick_h__ | thanks a ton adeuring! | 14:16 |
| antdillon | rick_h__, No but has a the notifications list checkbox, my guess is Luca will want states for the icon | 14:16 |
| hatch | gary_poster: ok sounds good | 14:16 |
| rick_h__ | antdillon: I'm confused. So the + should be a checkbox? Or you just mean you guys want to generate a hover state for hte + icon and get that to us to update? | 14:17 |
| hatch | antdillon: so if the url in the description is a link then it's working? | 14:17 |
| antdillon | rick_h__, Yes they will need to review and will get hover and maybe click states once Luca lands | 14:18 |
| rick_h__ | antdillon: ok, well will put it on hold and leave it as is for now then. Thanks for checking up. | 14:18 |
| antdillon | hatch, Yes and also the links the commits, check mySQL or wordpress for that one | 14:18 |
| antdillon | rick_h__, Np | 14:18 |
| hatch | cool looks to be working good here | 14:19 |
| hatch | antdillon: it's being proposed now | 14:22 |
| antdillon | hatch, Thanks a lot | 14:22 |
| hatch | gary_poster: so the service selected thing....that means just pick any service on the canvas and select it and add in a 'clicking a service adds a ring around it' ? | 14:24 |
| gary_poster | hatch, yes | 14:24 |
| gary_poster | hatch, squircle in particular, per design | 14:25 |
| gary_poster | hatch, squircle selection thing should always be in play | 14:25 |
| hatch | hmmm | 14:25 |
| gary_poster | that is, if someone clicks on a service, squircle should show up there | 14:25 |
| hatch | well what the heck it's about time I learn svg anyways :) | 14:26 |
| gary_poster | I started trying to make my own svg squircle at the time. there are a few generators and then I put it in inkscape to shrink. I would do a pre-imp with Makyo first though | 14:26 |
| gary_poster | I was just playing one night at the sprint | 14:26 |
| hatch | ohh - I was thiking of generating it on the fly | 14:27 |
| hatch | can d3 not do that for us? | 14:27 |
| jcastro | gary_poster, for these bundles I am working on | 14:27 |
| gary_poster | squircle is not a common-enough shape | 14:27 |
| jcastro | you don't support showing colocation on the gui do you | 14:27 |
| hatch | ahh | 14:27 |
| jcastro | or can I do like "all in one wordpress" | 14:27 |
| hatch | antdillon: https://codereview.appspot.com/19790044/ submitting now | 14:28 |
| gary_poster | jcastro, nope. :-/ and colocation in gui is being pushed back to post 14.04 unless I can find a way to sneak it in :-/ | 14:28 |
| jcastro | ugh, what? | 14:28 |
| jcastro | didn't know that | 14:28 |
| gary_poster | yup | 14:28 |
| gary_poster | other company-internal priorities took over | 14:28 |
| jcastro | so ... will the bundle work or just look wrong? | 14:28 |
| hatch | and of course there are conflicts haha | 14:29 |
| gary_poster | jcastro, I think it will work. | 14:29 |
| jcastro | ok so I should give you split up bundles, not colocated ones? | 14:29 |
| gary_poster | jcastro, I *think* the GUI will just not show you that you are on the same machine, unless you look really close | 14:29 |
| antdillon | hatch, Awesome, thanks | 14:29 |
| gary_poster | jcastro, yeah, I suggest we do not colocated for now. We can strategize about colocated later | 14:29 |
| benji | heh, the charmworld addition of assertNoException to TestCase is funny | 14:29 |
| jcastro | gary_poster, also, when bundles v1 lands ... what does the CLI of deploying it from the charm store look like? | 14:30 |
| jcastro | like can I do `juju-deployer <flags> http://jujucharms.com/blah`? | 14:30 |
| benji | evil, in fact; it will hide assertion errors in tests | 14:31 |
| gary_poster | jcastro, yes. We will be filling in the exact files to this page, but as preview: http://comingsoon.jujucharms.com/sidebar/search/bundle/~hatch/wiki/7/TestBundle/:flags:/charmworldv3/?text=hatch#bws-deploy | 14:31 |
| jcastro | juju quickstart <path to deployer file> <--- awesome! | 14:32 |
| hatch | antdillon: another thing to add to your list of things to do....merge in trunk pre-push ;) | 14:32 |
| rick_h__ | hatch: it's been a busy trunk today for what it's worth | 14:32 |
| hatch | oh that's fine | 14:33 |
| antdillon | hatch, Ah lol it was :( | 14:33 |
| hatch | the conflict was easy to resolve :) | 14:33 |
| jcastro | gary_poster, one more question! The deployer files can be deployed from the store though right? | 14:33 |
| jcastro | like if I wanted to tell someone "here's the URL for the best wordpress bundle" and then they don't have to like download the deployer file each time, etc. | 14:34 |
| hatch | antdillon: except now qa doesn't pass | 14:34 |
| gary_poster | jcastro, not sure what you mean, but...ah, yes. Correct. | 14:34 |
| jcastro | right, | 14:34 |
| jcastro | juju quickstart http://blah | 14:34 |
| antdillon | hatch, Damn ... so close | 14:35 |
| gary_poster | that wil work, and also "juju quickstart bundle:~hatch/wiki/7/TestBundle" will work, jcastro | 14:35 |
| hatch | antdillon: do you have time right now to debug it? | 14:35 |
| gary_poster | (but deployer will only support the http ones) | 14:35 |
| jcastro | got it | 14:35 |
| antdillon | hatch, I make time | 14:35 |
| benji | ah, it won't hide AssertionErrors... but it's still silly | 14:35 |
| antdillon | hatch, Not all of it, just for this | 14:36 |
| jcastro | and as far as workflow etc, as far as people submitting bundles, I know we agreed to make them the same as charms | 14:36 |
| jcastro | but like will this be a separate page in the docs or? | 14:36 |
| hatch | antdillon: :) sounds good - so the conflict is because tmplData was renamed to templateData | 14:36 |
| antdillon | Ah ok, easy enough | 14:36 |
| hatch | but the error is visible when you try to view the mysql charm, it throws an error that recent_commits is undefined | 14:36 |
| rick_h__ | jcastro: Makyo submitted a doc update to evilnickveitch, I believe, for bundles. | 14:37 |
| jcastro | ah, perfect, thanks | 14:37 |
| antdillon | hatch, So there need error checking there | 14:38 |
| hatch | guess it's a good thing I merged and re-qa'd lol | 14:39 |
| antdillon | hatch, Good catch! | 14:39 |
| bac | jujugui: i'm moving the trunk for charmworld on launchpad right now. please don't land anything until i'm done. | 14:41 |
| rick_h__ | bac: rgr | 14:41 |
| jcastro | rick_h__, where'd you guys hide the ability to show all charms again? | 14:44 |
| rick_h__ | jcastro: heh, manage.jujucharms.com? http://manage.jujucharms.com/charms ? | 14:45 |
| rick_h__ | jcastro: otherwise wait for this to load https://jujucharms.com/fullscreen/search/?text= | 14:45 |
| rick_h__ | (empty search) | 14:45 |
| jcastro | huh weird, empty search wasn't working for me | 14:46 |
| jcastro | oh, you url hacked it? | 14:46 |
| rick_h__ | jcastro: yea, you should be able to replicate just by hitting <enter> in the search field | 14:46 |
| rick_h__ | jcastro: yea, <enter> in the jujucharms.com search field works for me :/ | 14:47 |
| rick_h__ | it just takes many seconds to load due to pulling the giant json api call and rendering all the tokens | 14:49 |
| jcastro | ok I have 5 bundles | 14:49 |
| jcastro | with the discourse one I gave you that gives us 6 | 14:50 |
| rick_h__ | jcastro: coolio, enough to flesh out the featured section | 14:50 |
| jcastro | ugh | 14:50 |
| jcastro | dude you are killing me | 14:50 |
| rick_h__ | jcastro: how so? | 14:50 |
| jcastro | the GUI exports "cs:precise/jenkins-8" and so on | 14:50 |
| jcastro | mind if I cut out the -8 and all the version URLs? | 14:50 |
| rick_h__ | jcastro: yea, I mean if you export a bundle and want to load it in another env it should be exactly the same | 14:50 |
| rick_h__ | jcastro: pulling out the version is ok/safe | 14:51 |
| jcastro | fair enough | 14:51 |
| rick_h__ | jcastro: but I think that's not the best behaviour for users in real scenarios | 14:51 |
| Makyo | jujugui call in 9 | 14:51 |
| rick_h__ | ah crap, /me runs back to house | 14:51 |
| bac | benji: join call now so i can ask you some questions? | 14:52 |
| benji | bac: sure | 14:52 |
| jcastro | rick_h__, sure, but for people who want a wordpress blog I want them getting the latest one | 14:52 |
| jcastro | hatch, did you ever get your ghost stuff in launchpad? That'd be a cool bundle | 14:53 |
| jcastro | rick_h__, ok, should I send these to you? | 14:53 |
| hatch | jcastro: nope my night was busy, and then my VM decided to explode | 14:53 |
| hatch | I seem to have a difficult time keeping Ubuntu running | 14:54 |
| gary_poster | jcastro, sometimes we want exact bundles, sometimes we want most recent bundles. medium term--soon enough--I think it will be good if we provide versions in our default descriptions, as we do now, but we communicate that a new version is available. should be easy enough in gui. less easy elsewhere, perhaps. but being precise is a good place to start IMO | 14:55 |
| * jcastro nods | 14:55 | |
| jcastro | ok so these are simple ones I got right now | 14:55 |
| jcastro | I'll make some with crazy monitoring, etc. too | 14:55 |
| jcastro | but those will require testing | 14:55 |
| jcastro | these should be enough to unblock you though | 14:55 |
| gary_poster | awesome thanks jcastro. are these approved/recommended also? | 14:56 |
| jcastro | I stuck to all charms in the store already | 14:56 |
| gary_poster | good | 14:56 |
| jcastro | so these should be safe/recommended | 14:56 |
| gary_poster | jcastro, I mean, are the bundles promulgated | 14:56 |
| jcastro | no, they're on my disk | 14:56 |
| gary_poster | oh! duh | 14:56 |
| gary_poster | sorry | 14:56 |
| jcastro | I was like man, we can do that? | 14:57 |
| gary_poster | heh, yeah | 14:57 |
| jcastro | that would be cool! | 14:57 |
| gary_poster | that's what we want | 14:57 |
| jcastro | got it | 14:57 |
| gary_poster | cool thanks | 14:57 |
| jcastro | ok so what steps do I follow for that? the doc that bzc gave nick? | 14:57 |
| jcastro | the one with the "it must be in this crazy long url" :p | 14:57 |
| gary_poster | heh | 14:57 |
| jcastro | lp:~<your launchpad id>/charms/bundles/<your bundle name>/bundle | 14:58 |
| jcastro | do those promulgate? or do I need to propose it to something after that? | 14:58 |
| gary_poster | rick_h__, can you help jcastro with getting bundle in store? jcastro, https://docs.google.com/a/canonical.com/document/d/1paE70HEVgKA6LIjngzMa83OYr5jkZrOJ0hVNrYUuva0/edit# should start you maybe? | 14:58 |
| Makyo | jujugui call in 1 | 14:59 |
| gary_poster | thanks Makyo | 14:59 |
| jcastro | I'll follow that and catch you guys on the flipside of the meeting | 14:59 |
| gary_poster | thanks | 15:00 |
| hatch | rick_h__: what was the blocker? | 15:20 |
| rick_h__ | hatch: running CI before auto landing a branch once it got a LGTM | 15:20 |
| benji | bac: you were going to stay | 15:21 |
| hatch | ohh I thought it had a onComment hook | 15:21 |
| antdillon | hatch, Fixed up the linkify branch when you have time | 15:21 |
| hatch | antdillon: cool thanks | 15:21 |
| gary_poster | he *said* he was going to stay | 15:21 |
| rick_h__ | hatch: yes, and you can force a re-run of a travis-ci build, however you can't feed any info to that build to say "This time, in your after_success hook, go ahead and merge with trunk and close this pull request" | 15:21 |
| rick_h__ | hatch: so the api needs updating to allow data passed to the 're-build this CI build with this extra 'approved' flag' | 15:22 |
| rick_h__ | hatch: which isn't happening | 15:22 |
| hatch | ahhh | 15:23 |
| hatch | hmm | 15:23 |
| rick_h__ | jcastro: cool so you want to push this up? So from that google doc you need to create a bzr repo with them in it, push them to that url, and then promulgate them like charms. | 15:23 |
| hatch | well...right now even after lgtm we have ot manually trigger it | 15:23 |
| hatch | so I suppose that's not really an issue | 15:23 |
| hatch | right now being, with lbox | 15:23 |
| rick_h__ | hatch: so the issue is that you manually trigger and it runs for 7min. Then you have to MAKE SURE you come back in that 7min and land it | 15:23 |
| * benji waits for bac to come back. | 15:24 | |
| rick_h__ | hatch: right, the other thing to do is to recreate lbox submit and try to work the flow into a manual step there | 15:24 |
| hatch | doesn't lbox do that right now? | 15:24 |
| bac | jujugui: charmworld trunk is now owned by ~juju-gui-bot. rejoice | 15:24 |
| benji | yay | 15:24 |
| bac | benji: hangout? | 15:24 |
| gary_poster | yay | 15:24 |
| * Makyo commences rejoicing. | 15:24 | |
| benji | bac: https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.j0rk5d371ph8331ijtf48t2uj0?authuser=1 | 15:24 |
| rick_h__ | bac: woot | 15:24 |
| hatch | rick_h__: can the local script get a notification that it passed successfully? | 15:24 |
| rick_h__ | hatch: it'd have to poll/ping | 15:24 |
| rick_h__ | hatch: but if it lives as a service somewhere, the after_success hook could ping something | 15:25 |
| rick_h__ | hatch: I'm looking at stridercd and wonder if we could self-host something that could do it, but don't know enough about it yet | 15:25 |
| hatch | ahh....hmm how unfortunate | 15:25 |
| rick_h__ | hatch: otherwise it's basically setup jenkins + a bunch of scripts | 15:25 |
| hatch | yah...which kind of sucks | 15:26 |
| rick_h__ | hatch: but that'll take time to try out/setup get right | 15:26 |
| rick_h__ | hatch: and I'm doing this on my own time right now | 15:26 |
| gary_poster | I forgot to say that our bundle deploy developer story is the broke, because errors don't give helpful info and I can't get breakpoints to stop where I expect them to in Chrome. :-( Maybe the second issue is PEBKAC, but the first is real. I'm going to investigate and make a card once I know more | 15:26 |
| rick_h__ | gary_poster: the quickstart deploy? or from the gui itself? | 15:27 |
| gary_poster | gui | 15:27 |
| rick_h__ | gary_poster: hmm, well there's not much to break into right? I mean it loads the file and calls out to the deployer and just gets back whatever comes back from the delta? /me should look at the code before speaking...but wth :) | 15:27 |
| gary_poster | rick_h__, :-) this is with the sandbox | 15:28 |
| hatch | rick_h__: so does this mean that travis doesn't really give us anything if we have to switch to jenkins anyways? | 15:28 |
| gary_poster | rick_h__, is there a way to run a bundle through the charmworld proofer right now? I'd like to see if it can tell me what is wrong with Maarten Ector's files | 15:29 |
| rick_h__ | hatch: correct, travis is cool, but the lack of the ability to trigger a new build via the api or to pass data to builds we re-start means we cannot run tests before LGTM lands the branch | 15:29 |
| rick_h__ | gary_poster: yes, sec | 15:29 |
| gary_poster | thanks | 15:29 |
| hatch | rick_h__: yeah that kind of sucks...I wonder if the pro version adds these things | 15:30 |
| hatch | if not, it sounds like it's a pretty basic tool | 15:30 |
| rick_h__ | hatch: no, it's relying on users to do the merge or for something else to pick it up and do it. It's a gap in the feature set imo, but from irc they're not sold it's their job | 15:30 |
| rick_h__ | gary_poster: http://paste.mitechie.com/raw/1058/ | 15:31 |
| rick_h__ | gary_poster: now that's using the tools I used to dev/check things | 15:31 |
| gary_poster | rick_h__, awesome thank you | 15:31 |
| rick_h__ | gary_poster: feel free to swap httpie/curl or whatever if you prefer | 15:31 |
| gary_poster | ack | 15:31 |
| rick_h__ | bac: if you get a second, see jcastro's request for some more cli examples in the bundles doc https://docs.google.com/a/canonical.com/document/d/1paE70HEVgKA6LIjngzMa83OYr5jkZrOJ0hVNrYUuva0/edit | 15:32 |
| bac | rick_h__: ok, thanks | 15:32 |
| jcastro | bac, here's what we do for charms | 15:33 |
| jcastro | https://juju.ubuntu.com/docs/authors-charm-store.html#submitting | 15:33 |
| hatch | rick_h__: ahh gotcha | 15:33 |
| jcastro | we'd basically need those kind of steps, but for bundles | 15:33 |
| * bac looks | 15:39 | |
| hatch | I'm not sure which is worse, the wiki search or the charm search :P | 15:39 |
| rick_h__ | lol | 15:39 |
| rick_h__ | hatch: how is charm search letting you down? | 15:39 |
| hatch | rick_h__: haha no the wiki one is | 15:39 |
| hatch | its index must be....basic :) | 15:40 |
| rick_h__ | hatch: yea, I think it's simple wordpress mysql fulltext | 15:40 |
| bac | jcastro: the google doc we wrote was for nick use to craft docs for juju.ubuntu.com not as a final product. i've not gotten any feedback from him. but i'm happy to add whatever you want to that google doc. | 15:41 |
| jcastro | oh ok | 15:41 |
| gary_poster | rick_h__, proofer found a bug in our bundle export (we should honor option type) so that's good. :-) thank you | 15:43 |
| rick_h__ | gary_poster: awesomeness | 15:43 |
| rick_h__ | I love it when a plan comes together | 15:43 |
| gary_poster | Hey hatch, do you have a minute to help check something out? I feel like I'm encountering a Chrome debugging problem, but I want another pair of eyes | 15:49 |
| hatch | sure | 15:49 |
| rick_h__ | antdillon: nice blog post! | 15:49 |
| gary_poster | where blog? | 15:49 |
| hatch | rick_h__: link? | 15:49 |
| rick_h__ | http://design.canonical.com/2013/10/cloud-sprint-what-ive-learned-about-testing-and-code-reviews/ | 15:49 |
| hatch | cool thx | 15:49 |
| hatch | gary_poster: have a hangout url? | 15:50 |
| gary_poster | hatch https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.j0rk5d371ph8331ijtf48t2uj0 | 15:50 |
| gary_poster | sorry was looking at blog :-) nice post antdillon | 15:50 |
| hatch | I now have the new hangouts | 15:50 |
| antdillon | gary_poster, rick_h__ Thanks, you guys were the stars! | 15:52 |
| Makyo | Timeboxing the autoposition stuff, saving it at lp:~makyo/juju-gui/investigate-autopos couldn't find a clean way to get the centering stuff to work with it without being hacky, so I'll come back to it later. | 16:04 |
| Makyo | gary_poster, next step? Just pick a bundle card? | 16:04 |
| hatch | Makyo: lets pair on my branch | 16:04 |
| hatch | just in a call right now | 16:04 |
| Makyo | hatch, okay | 16:05 |
| hatch | Makyo: https://plus.google.com/hangouts/_/calendar/Z2FyeS5wb3N0ZXJAY2Fub25pY2FsLmNvbQ.j0rk5d371ph8331ijtf48t2uj0?authuser=1 | 16:08 |
| hatch | so I'm back to the old hangouts....lol | 16:08 |
| hatch | so confusing | 16:08 |
| rick_h__ | gary_poster: watching the charmers hangout and having them fight with the name to use for the 'deployer file' that contains multiple bundles I realized we've got a disconnect between quickstart and the gui bundle support. | 16:16 |
| rick_h__ | gary_poster: in the gui, we push the whole deployer file through | 16:16 |
| rick_h__ | gary_poster: but in the quickstart we ask you to select the bundle in that file to actually deploy | 16:16 |
| gary_poster | rick_h__, if there is only one it is automatic | 16:16 |
| gary_poster | I think that helps? | 16:16 |
| rick_h__ | gary_poster: right, but from the gui, if we pull marcoceppi's "wordpress" deployer file that contains bundles for 'basic', 'w/memcache', 'ha' versions | 16:17 |
| rick_h__ | the gui will try to deploy all three right now | 16:17 |
| rick_h__ | gary_poster: where the quickstart will ask you to select from basic, ha, etc | 16:17 |
| gary_poster | rick_h__, why would you do that in the GUI? We separate things in charmworld. confused | 16:17 |
| rick_h__ | gary_poster: oh, /me goes to check ingest. Missed/forgot that it split a deployer up. | 16:18 |
| gary_poster | rick_h__, that's the second name in the url | 16:19 |
| rick_h__ | gary_poster: ah right, the basket/bundle bit | 16:19 |
| rick_h__ | gary_poster: ok, yea. Just forgot that as the examples we've done have been all simple one-one | 16:19 |
| * rick_h__ is happy to have been mistaken | 16:19 | |
| gary_poster | rick_h__, right. So, bundle:~marcoceppi/wordpress/6/ha | 16:20 |
| rick_h__ | gary_poster: yea, will be interseting to see how well that reads/searches in the gui then. | 16:20 |
| gary_poster | rick_h__, they are treated as completely separate in the gui | 16:20 |
| gary_poster | rick_h__, we want related bundles in the future | 16:21 |
| gary_poster | should be trivial | 16:21 |
| gary_poster | they point to the other bundles in the basket | 16:21 |
| rick_h__ | gary_poster: right, I'm looking to see if we're displaying 'wordpress' or 'ha' has the name then in the gui? and if it should be somethig like 'wordpress/ha' | 16:21 |
| gary_poster | rick_h__, for instance, concrete example: http://comingsoon.jujucharms.com/sidebar/search/bundle/~hatch/wiki/6/envExport/:flags:/charmworldv3/?text=hatch | 16:22 |
| === rogpeppe2 is now known as rogpeppe | ||
| gary_poster | wiki == wordpress, envExport == ha | 16:22 |
| rick_h__ | gary_poster: right, and it won't come up for a search for 'wiki'. So if I was looking for 'wiki bundles' I wouldn't see this one which is why we're searching for 'hatch' | 16:23 |
| rick_h__ | gary_poster: or am I just confused on the user experience on this? | 16:23 |
| rick_h__ | e.g. http://comingsoon.jujucharms.com/sidebar/search/bundle/~makyo/openstack/1/envExport/:flags:/charmworldv3/?text=envExport | 16:24 |
| gary_poster | rick_h__, going to be late for lunch, have to run | 16:24 |
| rick_h__ | two very different bundles, found via a search for 'envExport' | 16:24 |
| rick_h__ | gary_poster: rgr, have fun! | 16:24 |
| gary_poster | thx | 16:24 |
| hazmat | jujugui, just a heads up and search for feedback re unit placement in deployer.. rough capabilities outlined here http://pastebin.ubuntu.com/6330874/ .. tl/dr ditched machine placement in favor of deploy-with emulation, ie. to: service | 16:36 |
| rick_h__ | hatch: got a sec to chat? | 17:16 |
| hatch | rick_h__: 2 mins? | 17:16 |
| rick_h__ | hatch: rgr | 17:16 |
| hatch | noice done | 17:19 |
| hatch | rick_h__: link? | 17:19 |
| rick_h__ | hatch: https://plus.google.com/hangouts/_/76cpip79jiokcbcjn1g80l5po4?hl=en | 17:19 |
| hatch | Makyo: so I have the highlight working and whatnot now | 17:31 |
| hatch | but it looks like garbage | 17:31 |
| rick_h__ | hazmat: is there really the idea of an relation without an endpoint? /me is looking at http://bazaar.launchpad.net/~juju-deployers/juju-deployer/trunk/view/head:/deployer/relation.py | 17:32 |
| hatch | have a quick second to take a peek? | 17:32 |
| Makyo | hatch, sure | 17:32 |
| hatch | join in the link that rick posted above :) | 17:32 |
| hazmat | rick_h__, well a relation with one endpoint is a peer relation | 17:32 |
| rick_h__ | hazmat: so if the endpoint is mysql vs mysql:db it's the peer relation? and it just *knows* as there can be only one? | 17:33 |
| hazmat | rick_h__, most of the peer disambiguation comes from charm examination where the rel type is specified | 17:34 |
| hazmat | mysql vs mysql:db is just qualified vs unqualified endpoint name | 17:35 |
| hazmat | ie juju add-relation mysql wordpress and juju add-relation mysql:db wordpress are equivalent | 17:35 |
| hazmat | qualified is needed in the case of ambiguity like adding a relation between mysql and mediawiki (which has both master/main and slave db rels to mysql) | 17:35 |
| rick_h__ | hazmat: ok, so if there's no qualifier on the endpoint it's assumed to be the first one in the list? Or some sort of set(mysql) intersect set(wordpress) | 17:36 |
| hazmat | rick_h__, i'm not really understanding the question.. a relation with a single endpoint must be a peer relation if its valid, qualified vs unqualified is only relevant/required for cases of ambiguity. | 17:40 |
| rick_h__ | hazmat: well your mysql to wordpress example isn't peer? | 17:40 |
| rick_h__ | hazmat: mysql has a peer relation of 'cluster' from what I can tell and not sure how that would work to wordpress | 17:40 |
| hazmat | rick_h__, no.. a peer relation is only a service's relation to itself.. the mysql/wordpress and mediawiki examples are all client/server | 17:40 |
| hazmat | rick_h__, peer relations can't be made between services, they are intra service relation, vs client/server as inter-service relations | 17:41 |
| rick_h__ | hazmat: ok, so juju add-relation mysql wordpress does not specify which endpoint, just services. (forget peers for the moment) so it'd be valid to have in the deployer relations: [ [ "mysql", "wordpress" ] ] ? | 17:42 |
| hazmat | rick_h__, the service is the endpoint | 17:42 |
| hazmat | ie juju add-relation <service_name:rel_name> <service_name:rel_name> | 17:43 |
| hazmat | <> == endpoint | 17:43 |
| hazmat | and really its <service_name[:rel_name> | 17:43 |
| hazmat | oops | 17:43 |
| hazmat | <service_name[:rel_name]> | 17:43 |
| rick_h__ | hazmat: sorry, I'm doing poorly at picking names to use. I want to call the key name (website) the endpoint and wordpress the service | 17:43 |
| hazmat | yes re your deployer relations, that's valid | 17:44 |
| rick_h__ | I'm trying to figure out how to validate [ [ "mysql", "wordpress" ] ] as a relation from a deployer file matching up https://manage.jujucharms.com/api/3/charm/precise/wordpress and https://manage.jujucharms.com/api/3/charm/precise/mysql | 17:44 |
| rick_h__ | hazmat: so ok, that's valid and the way it's valid is that juju does some work to match up the two services and finds a common relation to build? Just first available? | 17:45 |
| hatch | rick_h__: hazmat hey were you guys inviting me to a call? | 17:45 |
| hazmat | ah.. cool. so let's explore the use case | 17:45 |
| hatch | can you link it here? | 17:45 |
| hazmat | argh.. pythonhosted doc site is down.. or i could link to deployer docs | 17:45 |
| rick_h__ | hatch: we're just chatting in irc atm. We can move to a call if it'll help | 17:45 |
| hazmat | rick_h__, let's move to g+ | 17:46 |
| hatch | nope I just had a ton of hangout requests on my phone for some reason | 17:46 |
| rick_h__ | hatch: yea, I'm looking at http://bazaar.launchpad.net/~juju-deployers/juju-deployer/trunk/view/head:/deployer/relation.py | 17:46 |
| hatch | said it was you guys inviting me haha | 17:46 |
| rick_h__ | hatch: hazmat https://plus.google.com/hangouts/_/7ecpjqd9br9gv2ar0e17d70tkk?hl=en | 17:46 |
| hazmat | rick_h__, http://pythonhosted.org/juju-deployer/config.html#relations | 17:46 |
| rick_h__ | hazmat: ah, that makes it clear | 17:47 |
| hazmat | well, not sure it actually answers your question | 17:47 |
| benji | bac: [I was at lunch.] My phone reminds me, but your concern is noted and appreciated. | 18:07 |
| bac | benji: quick doc-only review: https://codereview.appspot.com/19890043 | 18:08 |
| * benji looks | 18:08 | |
| bac | benji: you can get even with the nagging tomorrow! | 18:08 |
| benji | bac: I'll need a reminder for that. | 18:09 |
| hatch | Makyo: so apparently a squircle is a real thing - not just rounded corners so I'll create one of those and then place it behind the service icon so that should hopefully get rid of the ugly gaps | 18:27 |
| Makyo | hatch, yeah, it's just a matter of finding the right angles. | 18:27 |
| Makyo | If not, punt to design; they'll already have that info. | 18:28 |
| hatch | yeah it doesn't look like it's quite a squircle either after fudging with it :) | 18:36 |
| hatch | it's a touch taller than wider | 18:36 |
| hatch | at least the 'generic' icon is | 18:36 |
| hatch | so I'll leave as is and get a proper one from design | 18:36 |
| rick_h__ | lol @ " so apparently a squircle is a real thing" | 18:36 |
| hatch | haha - I just assumed it was a square with rounded cornders | 18:37 |
| hatch | blah | 18:37 |
| hatch | corners | 18:37 |
| bac | benji: do you know LGTM reviews on rietveld are moved to LP merge proposals? is that a side-effect of 'lbox submit'? if so, that's a problem. | 18:39 |
| bac | s/know LGTM/know when LGTM/ | 18:39 |
| benji | bac: I'm pretty sure it is because LP is subscribed to the RV emails | 18:39 |
| bac | benji: your review is not shown on LP so 'bzr rv-submit' fails unless i use --force | 18:40 |
| benji | hrm | 18:40 |
| bac | benji: which is ironic given the content of this branch | 18:40 |
| rick_h__ | bac: is that the email from sinzui? | 18:40 |
| bac | rick_h__: no, those are separate. he was just pointing out you can do it all on LP without using lbox. | 18:41 |
| rick_h__ | bac: ah ok. Wondered if it was an issue with the owner movement and watching emails | 18:41 |
| benji | bac: heh, indeed | 18:41 |
| bac | benji: as to your nit, there is an irc client called 'rooms' which i blame for my confusion. i'll try to remember IRC = modern CB | 18:42 |
| benji | heh | 18:43 |
| bac | benji: your branch from earlier today did not get the rv comments moved to the LP mp: https://code.launchpad.net/~benji/charmworld/bug-1229179/+merge/193131 | 18:51 |
| bac | so that branch is merged with no approval votes | 18:52 |
| benji | :( | 18:52 |
| benji | are we the first people to use rv-submit? | 18:52 |
| bac | benji: compare to https://code.launchpad.net/~bac/juju-gui/bundle-change-log/+merge/193134 | 18:54 |
| bac | benji: i used it week before sprint | 18:54 |
| bac | i guess i may try with --force to see what it does | 18:54 |
| benji | bac: it is odd that Jeff's comment is in LP but mine is not | 18:55 |
| bac | benji: i just replied to your review on my reitveld. i noted mp+193298@code.launchpad.net was a recipient. the comment got posted at LP. i wonder if the email from google just got dropped. | 18:59 |
| benji | or maybe there is something about my emails that it doesn't like | 19:00 |
| bac | benji: try another LGTM msg | 19:00 |
| benji | bac: I am known to RV as benji.york@gmail.com, but LP knows me as benji@benjiyork.com and benji.york@canonical.com | 19:01 |
| bac | benji: have your reviews not made it through in the past? | 19:01 |
| benji | <shrug> | 19:01 |
| bac | benji: OH, on yesterday's review of my branch jeff's made it but yours did not. that's what you meant. hmm, that does make you look like a culprit. | 19:02 |
| bac | benji: would it kill you to add that email address to LP? | 19:02 |
| benji | bac: done | 19:02 |
| hatch | jujugui lf a reivew/qa for https://codereview.appspot.com/19920043/ | 19:26 |
| hatch | just updating some small cleanups for that branch right now | 19:29 |
| * benji heads out for a bit | 19:41 | |
| sinzui | bac, I added ~juju-gui-developer to ~ce-orange-squad to ensure everyone has maintainer privs in charmworld. gary_poster needs to accept the team invite | 19:42 |
| gary_poster | ok on it | 19:42 |
| gary_poster | uh, I don't see the email. Trying LP itself... | 19:42 |
| sinzui | gary_poster, bac, just lost network. he doesn't know | 19:42 |
| gary_poster | sinzui, https://launchpad.net/~juju-gui-developers is the wrong team | 19:45 |
| gary_poster | that's someone from the old juju-gui project | 19:45 |
| gary_poster | We are https://launchpad.net/~juju-gui | 19:46 |
| bac | gary_poster: sinzui has left the building | 19:46 |
| bac | gary_poster: can you reject the bad and add the right one? | 19:46 |
| gary_poster | bac, no I'm not a member. :-) um | 19:46 |
| * benji realizes he's too early and says put for a while longer. | 19:47 | |
| bac | gary_poster: well, you're a member but not an admin | 19:47 |
| bac | hazmat, can you add ~juju-gui to ~charming-devs on LP? | 19:47 |
| gary_poster | bac, LP doesn't think so. For example, https://launchpad.net/~ce-orange-squad/+mugshots | 19:47 |
| bac | https://launchpad.net/~charming-devs/+members shows gary_poster | 19:48 |
| sinzui | gary_poster, which team needs power | 19:48 |
| bac | gary_poster: oh, curtis wanted to add us to orange? | 19:48 |
| sinzui | yes I have left | 19:48 |
| gary_poster | bac, yes | 19:48 |
| bac | so. damned. confused. | 19:48 |
| gary_poster | sinzui, https://launchpad.net/~juju-gui-developers is not us. https://launchpad.net/~juju-gui is us | 19:48 |
| bac | but that would work... | 19:48 |
| sinzui | gary_poster, I was taking a circuitous route to ensure teams juju-qa and juju-gui can land branches in charmworld. orange is a shell of a team | 19:49 |
| hazmat | bac, sure in a meeting atm though eta 20m | 19:49 |
| bac | sinzui: that'll work great if you add ~juju-gui | 19:49 |
| sinzui | I don't see that team? | 19:50 |
| bac | hazmat: ack | 19:50 |
| gary_poster | sinzui, you don't see https://launchpad.net/~juju-gui ? Public, and Orange squad is member | 19:51 |
| bac | sinzui: orange is a member of that team. you don't see it b/c it would cause a cycle and salgado is too smart for that | 19:51 |
| gary_poster | heh | 19:51 |
| sinzui | gary_poster, something about team types prevents that team being a member. | 19:52 |
| sinzui | I will sort this out after children are moved | 19:52 |
| gary_poster | sinzui, I suspect it is the cycle that bac mentioned | 19:52 |
| bac | To the Bac Cycle | 19:53 |
| bac | so i think we need to patiently await kapil to fix it. perhaps he can make one of us an admin too. | 19:54 |
| benji | I can't believe I have never made that pun. | 19:56 |
| jcastro | heya bac | 20:04 |
| jcastro | https://code.launchpad.net/~jorge/charms/bundles/wordpress/bundle | 20:04 |
| jcastro | ok so does the store automatically index that or ... ? | 20:04 |
| rick_h__ | jcastro: yes, it'll ingest it like charms in about 15-20min | 20:04 |
| bac | jcastro: what he said | 20:05 |
| jcastro | ok | 20:05 |
| jcastro | now ... how do we gate what shows up in the gui? | 20:05 |
| bac | jcastro: ummm | 20:05 |
| rick_h__ | jcastro: right now there's a feature flag. :flags:/charmworldv3 | 20:05 |
| jcastro | got it | 20:05 |
| bac | jcastro: but no gating | 20:05 |
| jcastro | so like, I'd go into manage and check the box like a charm? | 20:05 |
| bac | anything pushed to LP that smells like a bundle will show up | 20:05 |
| jcastro | got it | 20:05 |
| jcastro | anything with charms/bundles in the namespace? | 20:06 |
| bac | no, it has to match charms/bundles/<name>/bundle | 20:06 |
| jcastro | got it | 20:06 |
| bac | so they won't just be there by accident | 20:06 |
| jcastro | and as far as featuring it, I assume I can check a box in the manage.jujucharms UI like I can a charm? | 20:07 |
| bac | jcastro: yes. | 20:07 |
| bac | jcastro: but it has to be promulgated first, right benji? | 20:07 |
| sinzui | bac, you are right. | 20:07 |
| jcastro | ok so how do we promulgate? | 20:08 |
| bac | sinzui to allow me to complete my experiment would you add me to ~ce-orange-squad ? | 20:08 |
| sinzui | I will | 20:08 |
| sinzui | bac, done | 20:09 |
| bac | sinzui: ta | 20:09 |
| gary_poster | hey hatch, do you have a reviewer? | 20:09 |
| hatch | umm | 20:09 |
| hatch | I don't think so | 20:09 |
| bac | jcastro: i think you have to use the charm tool and i'm unsure if marco has updated it tbh | 20:09 |
| gary_poster | hatch, ok I'll take it | 20:09 |
| jcastro | ah got it | 20:10 |
| jcastro | that's an important bit to note! | 20:10 |
| gary_poster | hatch, here's a sketch of a fix for Maarten's bug: http://paste.ubuntu.com/6331944/ | 20:10 |
| jcastro | but for these first few ones you guys are going to manually promulgate them? | 20:10 |
| gary_poster | hatch, you up for making a high card to address that and the "let's show a reasonable error to the user and maybe even put something nice in the console" issue? | 20:10 |
| hatch | sure thing, can you email me the yaml file so I can test? | 20:11 |
| jcastro | bac, I can confirm it's got promulgate support, perfect! | 20:11 |
| bac | good to know. | 20:11 |
| gary_poster | hatch sent | 20:11 |
| bac | benji: bzr rv-submit now works for me. it'll work for us all when we get ~charming-devs updated. lordy that was convoluted. | 20:12 |
| gary_poster | hatch that svg looks like a rounded rectangle, not a squircle | 20:14 |
| hatch | cool thanks | 20:14 |
| gary_poster | am I right/wrong? | 20:14 |
| hatch | gary_poster: it is a rounded rectangle | 20:14 |
| hatch | :) | 20:14 |
| hatch | the squircle didn't fit | 20:14 |
| gary_poster | heh | 20:15 |
| hatch | it actually looked worse | 20:15 |
| gary_poster | heh | 20:15 |
| gary_poster | ok | 20:15 |
| hatch | my guess is they have a different squircle calculation | 20:15 |
| gary_poster | hatch one other idea...one sec | 20:15 |
| hatch | sure | 20:15 |
| gary_poster | hatch dumb idea: maybe try putting the official charm template icon in inkscape, https://juju.ubuntu.com/docs/media/icon.svg ? | 20:17 |
| gary_poster | In retrospect that seems like the obvious thing to do | 20:18 |
| hatch | right it does doesn't it... | 20:18 |
| * hatch fail | 20:18 | |
| * gary_poster fail :-) | 20:18 | |
| hatch | well at least I caught an egg that rolled off the counter with my foot just now | 20:18 |
| hatch | so that's one win | 20:19 |
| hatch | lol | 20:19 |
| gary_poster | impressive! the circus is always another option for you. | 20:19 |
| hatch | yeah I'd make a good carny | 20:19 |
| hatch | actually no, i'd make a horrible one | 20:20 |
| gary_poster | :-) | 20:20 |
| hatch | haha | 20:20 |
| gary_poster | heh | 20:20 |
| bac | rick_h__, benji: i thought abel was going to show the revno on http://staging.jujucharms.com/heartbeat | 20:21 |
| bac | has the mongo test ever passed? | 20:22 |
| rick_h__ | bac: he was going to do it as a follow up | 20:22 |
| bac | oh | 20:22 |
| rick_h__ | bac: his branch was up for review already and he wants to add some more things | 20:22 |
| rick_h__ | bac: so he's agreed to do that and uptime in the follow up branch | 20:22 |
| bac | rick_h__: ok. | 20:23 |
| hatch | does anyone know if x11 applications run on mavericks? | 20:26 |
| hatch | the internet seems to have mixed reviews | 20:26 |
| gary_poster | hatch they used to, but yeah, I've heard of maverick problems | 20:29 |
| gary_poster | hatch, give it a try? :-) | 20:29 |
| gary_poster | or enjoy your vm :-)\ | 20:29 |
| hatch | haha I'm tryign to avoid more issues LIKE my vm :D | 20:29 |
| hazmat | bac done re ~gui->~cdevs | 20:29 |
| gary_poster | hatch, did you check with Makyo--should we declare the highlight image and then make repeated references to it, rather than making an image for each one? | 20:30 |
| marcoceppi | rick_h__: some weird things happening | 20:30 |
| rick_h__ | marcoceppi: ok, what's up? | 20:30 |
| marcoceppi | rick_h__: http://paste.ubuntu.com/6332077/ | 20:30 |
| marcoceppi | says can not find precise/wordpress and mysql, but it should | 20:30 |
| hatch | gary_poster: so on select move the highlighted image instead of having multiple? | 20:30 |
| gary_poster | hatch, no, svg has some spelling where you declare an image once and then reference it repeatedly. We use it for something or other. :-P If Makyo says we don't need it, that's fine. | 20:31 |
| rick_h__ | marcoceppi: hmm, yea. | 20:31 |
| bac | hazmat: thanks. sinzui may want to add ~juju-qa. | 20:31 |
| Makyo | gary_poster, hatch - We could stick it in a ref, but I don't think that buys us too much. It will still be cached, just like the service blocks in the real env. view. | 20:32 |
| gary_poster | Makyo, hatch, cool thanks | 20:33 |
| Makyo | If it's small enough, we could stick it in a ref and it would be one less round-trip, but we'll have to see that down the road. | 20:33 |
| Makyo | Not hard to do, just an investigation task for later. | 20:33 |
| gary_poster | cool | 20:33 |
| marcoceppi | rick_h__: I'll add a prettier --debug output for proofing | 20:35 |
| gary_poster | hatch, """Code LGTM with ultra trivial. I asked you to investigate using the official Juju squircle SVG (https://juju.ubuntu.com/docs/media/icon.svg). I will QA after you let me know to do so.""" | 20:36 |
| hatch | yeah I'm attempting it now | 20:37 |
| hatch | :) | 20:37 |
| gary_poster | cool | 20:37 |
| hatch | note to self....don't double click the header of inkscape | 20:39 |
| hatch | it....closes | 20:39 |
| hatch | if I have to do more of this stuff I think I'll just buy Sketch hah | 20:40 |
| rick_h__ | marcoceppi: I'll have to get back to you. It's the one freaking test I don't have is store_url: cs:... | 20:40 |
| rick_h__ | well at least the test fails. yay bugs | 20:43 |
| marcoceppi | rick_h__: cool, thanks | 20:43 |
| rick_h__ | oh wait, nvm. This is charm:, not store_url | 20:44 |
| rick_h__ | I do have a passing test for that, wtf | 20:45 |
| rick_h__ | ah hell, they're versionless | 20:46 |
| * rick_h__ grumbles in a corner. | 20:46 | |
| hatch | gary_poster: yeah...this looks 100x better :) | 20:46 |
| gary_poster | awesome, hatch! | 20:46 |
| sinzui | gary_poster, bac, Lp Ui doesn't let you uninvite, and the API only supports membership changes from the invitee perspective. This is an example of how to bend Lp API to do the right thing: http://pastebin.ubuntu.com/6332145/ | 20:47 |
| gary_poster | sinzui, lol, impressive | 20:47 |
| hatch | the zooming still appears to add some aliasing issues but we can deal with that later | 20:48 |
| hatch | it's reproposing now for qa | 20:48 |
| gary_poster | cool thanks hatch, lemme know and I'll merge | 20:48 |
| hatch | gary_poster: all done | 20:52 |
| gary_poster | ack, on it | 20:52 |
| gary_poster | hatch QA good! Thanks! | 20:55 |
| hatch | excellent thanks | 20:56 |
| hatch | gary_poster: so does the browser not cache svg hrefs? | 20:59 |
| hatch | re your ref comment | 20:59 |
| gary_poster | hatch it does, AIUI, but <Gary starts waving has hands, looking in a somewhat panicked way towards Makyo> I think that using the ref approach means that the file is only *parsed* once for that svg. The way it is now, the file is cached, but the file is parsed every time you add the image. Maybe. :-P | 21:01 |
| hatch | I'll accept your reasoning and substitute it for my own | 21:01 |
| gary_poster | lol | 21:01 |
| gary_poster | sounds good | 21:02 |
| Makyo | gary_poster, hatch the more I think of it, the less ref makes sense, because refs are NOT cached. | 21:02 |
| Makyo | It made sense considering the service blocks in the environment view, because we only loaded that once, but we would load the bundle view once for each bundle viewed. | 21:03 |
| gary_poster | why are they not cached? That seems silly | 21:03 |
| gary_poster | to not honor the usual HTTP headers | 21:03 |
| Makyo | Sorry, disconnect here. Refs are blocks of SVG that are basically stored in variables that can be used anywhere, such as for the service blocks in the env view. They are part of the DOM, not linked externally. | 21:04 |
| Makyo | We were considering moving the service blocks to a ref at the beginning of the <svg> element so that it would be loaded with the page rather than externally. This doesn't work quite as well for something that will be loaded over and over again. | 21:05 |
| Makyo | We could ref just the image tag, but that doesn't buy us a whole lot, because it's still loaded once on the first view and then cached, just like if we use image tags directly. | 21:06 |
| rick_h__ | bac: benji can I get a quick review of https://codereview.appspot.com/19990043/ please? | 21:06 |
| gary_poster | Oh, so there's no cached calculation of some sort? | 21:06 |
| Makyo | I think the solution is as we have it; the browser will take care of the caching. | 21:06 |
| rick_h__ | short branch, bugfix for marcoceppi's proofing | 21:06 |
| benji | rick_h__: sure | 21:06 |
| rick_h__ | thanks benji | 21:06 |
| Makyo | gary_poster, ref would be the solution if the calculation were simple. if we can get the squircle outline down to a few lines of SVG, then maybe that will be the way to go, but if it includes gradients or anything, it's a net loss. | 21:07 |
| gary_poster | huh | 21:07 |
| Makyo | (we couldn't really tell if the select was a border with an interior drop-shadow, though we think not, so as an increment, we made it just a border) | 21:08 |
| Makyo | (If it's just a border, yes to refs; if more, yes to images) | 21:08 |
| gary_poster | interesting. I don't get all of it, but that means I need to go read up on SVGs at some point. In THE FUTURE. :-) Thanks Makyo. | 21:08 |
| bac | rick_h__: how does that guarantee you get the latest revision? | 21:09 |
| rick_h__ | bac: the resolve_charm_from_description method adds a sort and pulls the latest | 21:10 |
| * bac looks | 21:10 | |
| rick_h__ | bac: I just had to change the 'query' part that this builds | 21:10 |
| bac | rick_h__: ok, so that already existed. gotcha. | 21:10 |
| rick_h__ | bac: rgr | 21:11 |
| bac | rick_h__: remember you'll need to land with 'bzr rv-submit'. | 21:12 |
| rick_h__ | bac: yea, I was giong to ask how I'm supposed to do this now | 21:13 |
| * rick_h__ goes to get rv-submit | 21:13 | |
| bac | benji: any idea why the charm-queue collection doesn't exist on staging? is it ephemeral, getting torn down after ingest is run? | 21:13 |
| bac | benji: this is shown on staging.jujucharms.com/heartbeat | 21:14 |
| * bac bbiab | 21:14 | |
| hatch | gary_poster: so re your fix for maarten the issue was the subordinates were being given a num_units of 1 when it should be undefined? -and- services that don't have a bum_units, but aren't subordinates should default to 1? | 21:20 |
| hatch | just confirming | 21:20 |
| hatch | well...I mean, it makes sense...just making sure that was the goal :) | 21:21 |
| gary_poster | hatch, my fix was #1. I moved behavior #2 down the stack from where it was (see the last bit of the patch), which seems reasonable and was necessary for #1. | 21:22 |
| hatch | ok great - just wanted to clarify so I can write a proper test | 21:22 |
| Makyo | Man, I hope this works... | 21:23 |
| Makyo | (╯°□°）╯︵ ʎʇᴉlᴉqᴉʇɐdɯoɔ ɹǝsʍoɹq | 21:23 |
| hatch | rofl | 21:23 |
| hatch | best! | 21:23 |
| gary_poster | lol | 21:24 |
| Makyo | Funny thing is that'd probably be just a bunch of boxes in IE | 21:25 |
| hatch | haha, sad but true | 21:26 |
| gary_poster | jcastro, just deployed a bundle with 8 haproxy units using official charm and all failed in "config-changed". Known issue? | 21:26 |
| jcastro | I don't know every known issue on all 130 charms off the top of my head unfortunately, heh | 21:26 |
| hatch | amature | 21:27 |
| hatch | ... | 21:27 |
| * hatch runs | 21:27 | |
| hatch | lol | 21:27 |
| gary_poster | jcastro, :-) sure, just figured that one was important | 21:27 |
| gary_poster | ok, shutting it down | 21:27 |
| jcastro | https://bugs.launchpad.net/charms/+source/haproxy/+bugs | 21:27 |
| jcastro | did you relate it to anything? | 21:27 |
| * Makyo dogwalks, then quick errand to ship something. Back in a few. | 21:29 | |
| rick_h__ | benji: thanks for the review. The urls should be url-safe, so thought they'd be regex safe. However, I guess there could be a [] or () and still be url safe | 21:29 |
| gary_poster | jcastro, was connected to daisy | 21:29 |
| rick_h__ | benji: I'll update it to at least ^ and check for \d at the end and do a better match | 21:30 |
| rick_h__ | and document the rest | 21:30 |
| jcastro | hey rick_h__ | 21:30 |
| jcastro | still no bundles showing up for me in manage. | 21:30 |
| jcastro | debug now or wait until tomorrow? | 21:30 |
| rick_h__ | jcastro: yea, notice. I'm not sure what's up. I'll have to check staging to see if there was some error pulling it in | 21:30 |
| rick_h__ | jcastro: tomorrow, I'm about to run out the door and past EOD | 21:31 |
| jcastro | I agree, beer time. | 21:31 |
| rick_h__ | CHC time | 21:31 |
| jcastro | hey so for now | 21:31 |
| jcastro | I'll keep the one-shot all-on-one-unit ones out of there | 21:31 |
| jcastro | and I'm going to do the winning bundle connected to the rails charm as our "crazy one" | 21:31 |
| rick_h__ | jcastro: k | 21:31 |
| gary_poster | hatch, should I maje a card? Do you want one for the patch, and one for the better error reporting, or combine them? | 21:33 |
| hatch | I reviewed a branch from frankban this morning that added the better error reporting | 21:34 |
| hatch | which....there isn't a card for | 21:35 |
| hatch | hmm | 21:35 |
| gary_poster | hatch, "review deployment notifications..." | 21:35 |
| gary_poster | exists | 21:35 |
| gary_poster | in review lane | 21:35 |
| gary_poster | but that was about gui side | 21:35 |
| gary_poster | not fake backend | 21:35 |
| gary_poster | IIUC | 21:35 |
| hatch | ohh right that was | 21:35 |
| hatch | I guess they could be combined | 21:38 |
| gary_poster | hatch, I made two cards. if you combine them then so be it :-) | 21:45 |
| * gary_poster runs away | 21:45 | |
| gary_poster | night all! | 21:45 |
| hatch | night, thanks | 21:46 |
| huwshimi | Morning | 21:59 |
| hatch | morning sashimi | 21:59 |
| hatch | jujugui lf a review https://codereview.appspot.com/20040043/ | 22:05 |
| huwshimi | hatch: So it turns out most YUI widgets include their base CSS and yui3-skin-sam CSS in the skins/sam css | 22:12 |
| hatch | right but we don't include that file though do we? | 22:13 |
| huwshimi | hatch: Yep. | 22:13 |
| huwshimi | hatch: I figured out a way to not include the sam skin files, but now everything breaks because it doesn't include the base css that we need :( | 22:14 |
| hatch | oh intersting - I thought we removed that long time ago | 22:14 |
| hatch | ohh | 22:14 |
| huwshimi | hatch: Well, it's per widget | 22:15 |
| huwshimi | hatch: So ./node_modules/yui/tabview/assets/skins/sam/tabview.css contains styles that are prefixed with .yui3-skin-sam, but also includes the base CSS for that widget | 22:15 |
| hatch | ohhh I see where they are coming in from | 22:16 |
| hatch | that's....unfortunate | 22:16 |
| huwshimi | hatch: Ah, but those skin files look like they're including the *-core.css file | 22:16 |
| huwshimi | hatch: So if we include those instead we should be fine. | 22:16 |
| hatch | assuming that they are split up correctly hehe | 22:17 |
| huwshimi | yeah... | 22:17 |
| huwshimi | hatch: I'm guessing if we want the core files we'll need to include them manually | 22:18 |
| huwshimi | hatch: Unless you have any ideas on how to get the loader to do it... | 22:18 |
| hatch | hmm | 22:18 |
| hatch | I thought there was a way tos pecify the skin for the css in the loader config | 22:18 |
| hatch | so maye specifying an invalid one would just load the core | 22:19 |
| hatch | but it's been so long | 22:19 |
| hatch | I couldn't say for sure without some hacking on it | 22:19 |
| * huwshimi wonders if rick_h__ is around and has any ideas... | 22:19 | |
| huwshimi | hatch: I tried to do that, but I don't think I was doing it correctly as it didn't change anything. Maybe I'll give it another go... | 22:20 |
| gary_poster | hatch, LGTM. /me collects more trash from house for trash day... | 22:20 |
| gary_poster | hi huwshimi | 22:20 |
| huwshimi | gary_poster: Hey! | 22:20 |
| hatch | thanks | 22:24 |
| hatch | huwshimi: maybe we create a yui-module-css.css file and copy the core stuff in there | 22:25 |
| hatch | if you know which ones we need it'll be pretty easy to write a small script to create that file | 22:25 |
| huwshimi | hatch: I can actually just use the current CSS builder to do that, I was just hoping we could use the Loader to find all the neccesary css so we don't miss things in the future. | 22:26 |
| huwshimi | I can't even get it to use the night skin | 22:27 |
| huwshimi | I mustn't be doing this right | 22:27 |
| hatch | ohh - well they are working on moving yui to use pure | 22:27 |
| hatch | so when that change comes I rekin we'll have some issues :) | 22:27 |
| hatch | can you gist your code so I can take a look? | 22:27 |
| huwshimi | hatch: here are all the places I've tried changin the skin :) | 22:29 |
| huwshimi | hatch: http://paste.ubuntu.com/6332657/ | 22:29 |
| hatch | looking | 22:29 |
| hatch | http://yuilibrary.com/yui/docs/slider/#skins | 22:30 |
| huwshimi | hatch: Right, but that's per widget | 22:31 |
| hatch | :/ man there is really 0 documentation on this | 22:32 |
| huwshimi | hatch: http://yuilibrary.com/yui/docs/tutorials/skins/ | 22:32 |
| hatch | so when you set skin to night it doesn't load the night files? | 22:33 |
| hatch | I think you'll need to run make again | 22:33 |
| huwshimi | hatch: yeah, loads the same set of files | 22:33 |
| hatch | :/ | 22:34 |
| huwshimi | hatch: in merge-files.js in the combineCSS function you can console.log(files) and it will output the list of files it's include every time you run make devel | 22:35 |
| hatch | and that list never includes the night css? | 22:35 |
| huwshimi | that's correct | 22:35 |
| hatch | looks like fetchCSS only has three options | 22:35 |
| hatch | neither is what we want | 22:36 |
| hatch | unfortunately | 22:36 |
| hatch | http://yuilibrary.com/yui/docs/api/classes/config.html#property_fetchCSS | 22:36 |
| hatch | we could create a group for the css core files | 22:36 |
| hatch | so we set fetchCSS to false | 22:37 |
| hatch | then manually include the css core files | 22:38 |
| hatch | manually being in the loader | 22:38 |
| hatch | the last time I did that was probably 1.5 years ago though so I'd have to find the syntax again | 22:38 |
| huwshimi | hatch: We already set fetchCSS to false, but that only sets it use the local CSS, not use CSS from the YUI CDN | 22:39 |
| huwshimi | hatch: It's ok, I can just manually include the CSS and we can try and remember to add css for each widget when required in the future | 22:40 |
| huwshimi | hatch: I would assume that somehow we should be able to change the skin though :) | 22:41 |
| hatch | yeah - I would too... | 22:41 |
| hatch | in my personal projects the first thing I do is fetchCSS: false | 22:41 |
| hatch | haha | 22:41 |
| huwshimi | :) | 22:41 |
| hatch | the skins are just too ugly | 22:42 |
| hatch | they remind me of the 90s | 22:42 |
| hatch | lol | 22:42 |
| huwshimi | hehe | 22:44 |
| hatch | there are WAY too many places to add notifications | 22:52 |
| huwshimi | hatch: do you think it would be too hacky to modify the list of paths to grab the -core.css file in stead of the ./skin/sam file? | 22:55 |
| rick_h__ | huwshimi: what's up? see I was ping'd earlier | 22:55 |
| hatch | where is this list of paths? | 22:55 |
| huwshimi | hatch: It comes from loader | 22:55 |
| hatch | oh....then yeah :) | 22:56 |
| huwshimi | rick_h__: We're just trying to figure out how to get the YUI Loader to pick up the -core.css files for the widgets instead of the skin/sam files | 22:56 |
| huwshimi | rick_h__: Any ideas? | 22:56 |
| hatch | but maybe if you document it enough :) | 22:56 |
| huwshimi | hatch: As in yes, it would be too hacky? | 22:56 |
| hatch | yeah it's pretty hacky, but if there is enough documentation then maybe not :) | 22:57 |
| rick_h__ | huwshimi: we don't use the YUI loader? | 22:57 |
| rick_h__ | huwshimi: yea, fetchCSS false in the GlobalConfig (and each section of it) should turn off any css loading and thought we already did that. | 22:58 |
| hatch | that's what I thought too, but I guess it's not working for whatever reason | 22:58 |
| huwshimi | rick_h__: That makes it use our local files, but it's still using the sam skin files, we want the core ones | 22:59 |
| rick_h__ | huwshimi: right, but don't we auto combine the sam ones? We don't load them from external sources do we? | 22:59 |
| huwshimi | So we get paths like node_modules/yui/widget-base/assets/skins/sam/widget-base.css and we want node_modules/yui/widget-base/assets/widget-base-core.css | 22:59 |
| rick_h__ | bah, wish we just used a damn combo loader and could do it right | 22:59 |
| rick_h__ | huwshimi: oh, well did we take off skin-sam from the <body> tag? | 22:59 |
| rick_h__ | or all containers? | 23:00 |
| huwshimi | rick_h__: Yep, I've done that and now want to get rid of the CSS :) | 23:00 |
| * rick_h__ tries to think back to how we got this working in the first place | 23:00 | |
| huwshimi | rick_h__: I've fixed up any issues from dropping the yui3-skin-sam classes. I just want to drop the CSS as well. | 23:00 |
| rick_h__ | huwshimi: k, looking | 23:01 |
| hatch | rick_h__: +1 on using the loader properly :) | 23:01 |
| rick_h__ | huwshimi: did you change the config in the lib/merge_files.js? | 23:02 |
| rick_h__ | huwshimi: that's not looking to ignore css and that's what builds things from YUI | 23:02 |
| rick_h__ | huwshimi: you'd have to debug what merge-files.js is doing and try to config it to skip the files even though combo-loading rules would say to get them | 23:02 |
| rick_h__ | huwshimi: I would think it'd be two passes, one for css and one for js but not 100% since merge-files is part magic trying to 'pretend' to combo load | 23:03 |
| huwshimi | rick_h__: But I don't want it to ignore the CSS, I just want it to pull the core files instead of the skin files | 23:03 |
| rick_h__ | huwshimi: right, but it must be deciding what to pull for some reason there | 23:03 |
| huwshimi | rick_h__: I figured it would be using the YUI() config | 23:05 |
| rick_h__ | huwshimi: so combineCSS is getting sent a list of files but not seeing who's building the list atm and calling combineCSS | 23:05 |
| huwshimi | rick_h__: It comes from getYUIFiles().css and a couple of extras in 'merge-files' | 23:06 |
| rick_h__ | huwshimi: ok, so that loader then. Something is telling it to get sam skins. /me looks some more | 23:08 |
| huwshimi | rick_h__: Yeah, I can't figure that one out | 23:09 |
| rick_h__ | huwshimi: http://yuilibrary.com/yui/docs/tutorials/skins/ | 23:09 |
| rick_h__ | huwshimi: try with skin: null in the loader config? | 23:09 |
| rick_h__ | See the section "Option B" | 23:09 |
| rick_h__ | I'm just guessing here | 23:09 |
| rick_h__ | if all the skin-sam is pulled from the html, and there's noting in the deps for it, figure should be ok | 23:10 |
| huwshimi | rick_h__: I've been trying | 23:10 |
| huwshimi | rick_h__: Not even changing it to 'night' changes anything | 23:11 |
| rick_h__ | huwshimi: push your branch up? | 23:11 |
| huwshimi | one sec... | 23:12 |
| hatch | sorry guys I gota run for a few hours but I definitely want to know the outcome of this :) | 23:12 |
| huwshimi | hatch: Thanks! | 23:13 |
| huwshimi | I've actually gotta head to the doctor in 10 minutes | 23:13 |
| huwshimi | rick_h__: I'm going to push this with all my changes it. I'll be pretty dirty. | 23:14 |
| rick_h__ | huwshimi: that's dine | 23:14 |
| rick_h__ | fine | 23:14 |
| huwshimi | rick_h__: I can fix it by doing "cssFiles.push(path.replace('skins/sam/', '').replace('.css', '-core.css'));" but that's a hack :) | 23:20 |
| rick_h__ | huwshimi: lol | 23:21 |
| huwshimi | rick_h__: I'm sorry I have to head for a bit. I'll be back shortly. If you do discover anything let me know, otherwise we can take a look another time. Thanks for your help! | 23:26 |
| rick_h__ | huwshimi: http://paste.mitechie.com/show/1059/ works to ignore the skin sam, it doesn't load -core. You have to pull those. | 23:49 |
| rick_h__ | -core is meant to be shared across skins so we've basically turned off skins here and so you'll have to pull what you want manually | 23:49 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
