|
[13:04] <rick_h_> this is interesting reading. I think this somewhat fits a model I like of one smart bit of code doing the work and logic using a bunch of dump components http://r.bmark.us/u/c65d3981a861f0 |
|
[13:06] <redir> rick_h_: interesting, does jujugiu map to that model? |
|
[13:07] <rick_h_> redir: in parts |
|
[13:07] <rick_h_> redir: so the State system helping to managing dispatch fits that a bit. Components request a change of state and the state system is a dispatcher from there |
|
[13:08] <redir> the store in the flux model == statesys |
|
[13:08] <rick_h_> redir: and there's the code that handles the deltas from juju-core. It's something we probably don't do well enough, reusing that dispatcher that we use to talk over the websocket in the direct user interactions |
|
[13:08] <redir> and components == view? |
|
[13:08] <rick_h_> redir: yea, somewhat |
|
[13:08] <rick_h_> the dispatcher is the sys which stores the current state |
|
[13:09] <rick_h_> and then the views are triggered and told "Hey, this is the state, update your stuff man" |
|
[13:09] <redir> I can't quite put my finger on it but is seems to echo the ORM vs. anti-ORM model proponents converstations as well |
|
[13:10] <redir> "Hey view, you better get your state together!" |
|
[13:10] <redir> I like it |
|
[13:11] <rick_h_> heh, well I always wrap my orm around a layer of code |
|
[13:11] <rick_h_> that's more app-specficic and provides a cleaner/simpler api |
|
[13:13] <redir> I should look at bookie and see how you did it there. |
|
[13:15] <rick_h_> redir: did your branches land? |
|
[13:15] <rick_h_> redir: or are they in limbo landing at all? |
|
[13:15] <redir> I think they both landed |
|
[13:15] <rick_h_> ok cool, will move them over to daily call |
|
[13:16] <redir> "bam!" to quote emeril |
|
[13:16] <redir> thanks for the nudges |
|
[13:16] <redir> WRT store stuff |
|
[13:16] <rick_h_> all good, it's my job :) |
|
[13:18] <redir> http://bit.ly/1qFzThG I can't reproduce that. |
|
[13:18] <redir> the code appears to have changed significantly since then |
|
[13:19] <redir> June 2012 |
|
[13:19] <rick_h_> redir: cool, then mark the bug as invalid and note you can't reproduce it and ask them to reopen if they find a method of reproducing |
|
[13:19] <redir> should I go back and see if I can see it in the past and note where it may have been fixed? Ask someone? comment that I can't reproduce? |
|
[13:19] <redir> perfect |
|
[13:20] <jcsackett> huh, wasn't actually on IRC... |
|
[13:20] <jcsackett> morninng all. :) |
|
[13:20] <rick_h_> party |
|
[13:28] <jcsackett> i love "party" as a morning greeting. :p |
|
[13:28] <redir> especially at the end of the night |
|
[13:29] <jcsackett> ...fair. :p |
|
[13:29] <redir> the charm browser == CW? |
|
[13:30] <rick_h_> heh not really |
|
[13:30] <rick_h_> the charm browser is the UI components in the GUI that allow for browsing charms |
|
[13:30] <rick_h_> it loads it's api data from charmworld |
|
[13:30] <rick_h_> which loads it's metadata from the juju store and other parts of the world |
|
[13:30] <redir> looking through the other store bugs and they seem to need some clarity |
|
[13:30] <rick_h_> and it all gets called 'the store' at times confusing the $#@$#@ out of people |
|
[13:31] <rick_h_> redir: ok, feel free to add some notes/etc |
|
[13:31] <rick_h_> redir: a little triage/etc. Mark bugs as incomplete if they really do need more info to move forward and the like |
|
[13:31] <rick_h_> redir: maybe even put a card up for "store bug triage" and move it to done when you've gone through |
|
[13:31] <redir> k |
|
[13:31] <rick_h_> hit me up if you've got any questions |
|
[13:32] <redir> lots |
|
[13:32] <redir> :) |
|
[13:32] <redir> can prolly run through them in 10 minutes though |
|
[13:49] <redir> gah. |
|
[13:53] <BradCrittenden> jcsackett: could you review https://codereview.appspot.com/98280043 at your convenience? |
|
=== BradCrittenden is now known as bac |
|
[13:53] <jcsackett> bac: sure. |
|
[13:53] <jcsackett> bac: i'll be reviewing it soon. soritng out a bit of branch goofiness at the moment. |
|
[13:54] <bac> jcsackett: ty |
|
[13:55] <jcastro> hey rick_h_ |
|
[13:55] <jcastro> http://dash.juju.solutions/solutions |
|
[13:56] <jcastro> https://github.com/chuckbutler/juju-is-dashing |
|
[13:56] <rick_h_> jcastro: purdy |
|
[14:01] <redir> very metro or aol depending on how old you are |
|
[14:11] <bac> hi abentley. we're having login issues with charmworld on canonistack and the new staging version on prodstack. both are using new URLs. is it required to register those URLs with SSO for things to work? |
|
[14:11] <bac> jcastro: i'll bet that is real pretty on your windows phone |
|
[14:12] <jcastro> hah |
|
[14:12] <abentley> bac: No, registering the URLs is only needed for SSO to "trust" the site. Without it, you'll get a page asking you what you want to share. |
|
[14:12] <abentley> bac: But if you share everything, it should have the same effect. |
|
[14:12] <abentley> bac: Note that everything defaults to /not/ shared. |
|
[14:12] <bac> abentley: ah, ok. there is some other problem then as when we return to charmworld we're not logged in |
|
[14:13] <bac> abentley: still i'll ask webops to register us |
|
[14:13] <abentley> bac: At login time, charmworld tells SSO where to send users once they've authenticated. |
|
[14:17] <rick_h_> hatch: http://r.bmark.us/u/c65d3981a861f0 from one of the tech mailing lists this morning. |
|
[14:18] <rick_h_> hatch: seems to match some of the 'make one part smart, all the rest stupid components' stuff we've been talking about lately |
|
[14:19] <hatch> checking |
|
[14:24] <hatch> I got a sun burn yesterday |
|
[14:24] <hatch> lol |
|
[14:24] <rick_h_> lol |
|
[14:36] <jcsackett> bac: looking at your MP now. |
|
[14:37] <hatch> rick_h_ While I agree with the outcome of that article, they are doing MVC wrong, no wonder it wasn't scalable lol |
|
[14:38] <rick_h_> heh |
|
[14:38] <rick_h_> hatch: they're also on a different scale than most |
|
[14:39] <hatch> "A Flux TodoMVC Tutorial ..." |
|
[14:39] <hatch> haha so ironic |
|
[14:39] <hatch> :P |
|
[14:40] <hatch> yeah, very scale, much codebase |
|
[14:50] <rick_h_> jujugui call in 10 |
|
[14:51] <jcsackett> jujugui: can someone look at https://github.com/juju/juju-gui/pull/322, please? |
|
[14:52] <hatch> on it |
|
[14:59] <rick_h_> jujugui call in 1 |
|
[15:01] <rick_h_> jcsackett: ^ |
|
[15:01] <rick_h_> kadams54: ^ |
|
[15:02] <jcsackett> rick_h_: yeah, fighting with G+ again. |
|
[15:03] <bac> redir: http://qa.manage.jujucharms.com |
|
[15:03] <bac> on canonistack |
|
[15:19] <hatch> good afternoon antdillon |
|
[15:22] <antdillon> hatch, hi there |
|
[15:40] <redir> bac tx |
|
[15:43] <hatch> Makyo you asked if the simulate false should have been committed......after working with this machine view stuff - I'm thinking it should be off by default because we have no visual indication that it's on, thoughts? |
|
[15:44] <rick_h_> hatch: make it obvious by dumping to the console some info? |
|
[15:45] <hatch> well the simulator is very broken for the machine view - so it needs to be disabled every load if you're going to be using the mv anyways |
|
[15:45] <hatch> very is probably an exaggeration |
|
[15:46] <rick_h_> well I think that's what I like about this. It points out issues easy to sweep under the rug for now. |
|
[15:46] <rick_h_> I don't think turning it off by default because it's annoying right now is the answer tbh |
|
[15:46] <rick_h_> you can flip it for your dev cycle and get the work done |
|
[15:47] <jcsackett> rick_h_: re: your comment about tests for the generateDescription stuff--i'm game for that, but will get to it in a follow up as that functionality already exists untested in the code. |
|
[15:47] <jcsackett> i've planned the whole rest of my day to be "add tests" cards, so it should be handled soon. |
|
[15:47] <rick_h_> jcsackett: right, but now you made a bunch of changes to it. While you're there, it makes sense that your changes turn into "here's what it should be now" |
|
[15:48] <rick_h_> I'm just saying one test that's got an array of "change object => html string" mapping to loop and verify |
|
[15:48] <hatch> ohhhhhh kay |
|
[15:48] <jcsackett> rick_h_: fair enough. |
|
[16:02] <hatch> jujugui are there no placeUnit() tests? |
|
[16:03] <hatch> rick_h_ I created a card for ^ |
|
[16:03] <rick_h_> hatch: hmm, I thought there were some that frankban and Makyo did |
|
[16:03] <rick_h_> hatch: but yea, ok, card is good |
|
[16:04] <hatch> I thought so too but a grep in the test folder turned up 0 results |
|
[16:04] <rick_h_> it might be indirect I guess. So yea, direct tests of it welcome |
|
[16:50] <hatch> ugh test suite |
|
[16:59] <jcsackett> ^ + 1000 |
|
[17:01] <Makyo> jujugui quick review/qa https://github.com/juju/juju-gui/pull/323 |
|
[17:06] <jcsackett> hatch: i've pushed up changes to my PR, not sure if you want to take a look again or not. |
|
[17:06] <hatch> should I trust you? :P |
|
[17:10] <jcsackett> hatch: i'm generally trustworthy. |
|
[17:10] <jcsackett> hatch: however, "trust, but verify" is a saying for a reason. :p |
|
[17:10] <hatch> haha |
|
[17:10] <hatch> I just looked, all looks good :) |
|
[17:11] <jcsackett> hatch: cool, thanks. |
|
[17:42] <hatch> jujugui looking for a review/qa on https://github.com/juju/juju-gui/pull/321 |
|
[17:42] <Makyo> On it |
|
[17:43] <hatch> thx sir |
|
[17:47] <hatch> ahh it's raining out.....awesome |
|
[18:00] <hatch> rick_h_ I think the kanban on deck for project 1 is out of date |
|
[18:00] <rick_h_> hatch: looking |
|
[18:00] <rick_h_> hatch: I gardened a bunch of it today but not done yet |
|
[18:01] <rick_h_> hatch: let's chat paticulars after my calls |
|
[18:01] <hatch> cool np - I'm looking for my next card - I was thinking I might try and tackle the 'ghost machine' in the ecs addMachine call bit.... |
|
[18:01] <hatch> that might be something to pair up with Makyo on |
|
[18:02] <rick_h_> hatch: for now let's grab one of the cards in the ready to code and do some cleanup? |
|
[18:02] <hatch> can do! |
|
[18:02] <rick_h_> hatch: or I could really give you the hell task of trying to get that search/state card going again which will probably have the rebase from hell now |
|
[18:02] <rick_h_> kadams54: 1-1? |
|
[18:02] <kadams54> Ah yes |
|
[18:02] <hatch> lol....hm |
|
[18:03] <rick_h_> hatch: I'm trying to be nice and find time to get back to that one myself :P |
|
[18:03] * hatch picks something from ready to code really fast |
|
[18:03] <hatch> sorry busy, got another card |
|
[18:03] <hatch> ...lol |
|
[18:04] <redir_> anyone have any pointers to fixing dns slownes swith dns-masq |
|
[18:04] <redir_> ? |
|
[18:05] <redir_> under 14.04 |
|
[18:08] <bac> redir, no, its built into my wap |
|
[18:10] <redir_> bac wap? |
|
[18:10] <bac> wireless access point |
|
[18:10] <rick_h_> hatch: did you have time to chat now? |
|
[18:11] <hatch> otp 5 mins? |
|
[18:11] <rick_h_> hatch: :P fine be that way |
|
[18:12] <redir_> bac me too. I'll just deal with it |
|
[18:12] <redir_> for another day |
|
[18:13] <hatch> rick_h_ ok rdy |
|
[18:13] <rick_h_> hatch: stand up hangout |
|
[18:13] <bac> redir: sorry, i may have been too cryptic. my router runs dns_masq, so i don't have to on my machine. works well. |
|
[18:14] <hatch> im there |
|
[18:14] <bac> redir: but consequently i don't know anything about setting it up |
|
[18:22] <redir_> oic |
|
[18:22] <redir_> found it here |
|
[18:22] <redir_> I think' |
|
[18:29] <rick_h_> bac: there was an email about this bug https://bugs.launchpad.net/juju-core/+bug/1318711 that mentioned In the |
|
[18:29] <rick_h_> mean time, you can do "juju scp -- -r $source $target" |
|
[18:29] <_mup_> Bug #1318711: juju scp reports -r as an invalid option <ssh> <juju-core:Triaged> <https://launchpad.net/bugs/1318711> |
|
[18:30] <rick_h_> bac: not sure if you saw the email on that but seemed good to know |
|
[18:30] <bac> rick_h_: that is good to know |
|
[18:31] <bac> i shall add that to the bug report |
|
[18:31] <rick_h_> bac: yea, just looking noticed that his reply didn't get into there thanks |
|
[18:33] <rick_h_> jcsackett: call? |
|
[18:33] <hatch> Makyo hey how goes the review? |
|
[18:33] <Makyo> code's okay, just QAing |
|
[18:34] <jcsackett> rick_h_: oh shoot, didn't realize it was 2:30. |
|
[18:40] <lazyPower> Question, aside from doing charm create and making a bunch of dummy charms, is there a mode in the gui that i can use to mock a deployment with charms that dont exist yet? I'm trying to explain and diagram something out, and i'm looking for a better technique than clipping screenshots to build this in inkscape. |
|
[18:41] <rick_h_> lazyPower: so, funny you should ask that |
|
[18:41] <rick_h_> lazyPower: with the work that bac is just landing, you could run a charmworld instance, feed it local 'fake charms' and then point a gui at it |
|
[18:41] <lazyPower> oh thats awesome! |
|
[18:42] <rick_h_> lazyPower: but no, there's nothing that just let's you pretent eh GUI is a wireframe tool at the moment |
|
[18:42] <rick_h_> so you still have to create the fake charms, but you can feed them into a fake charmworld/GUI in sandbox demo mode to get the UI you're looking for to show off |
|
[18:42] <bac> lazyPower: grab lp:charmworld and look at docs/index.rst for --local-repo |
|
[18:43] <lazyPower> i think i can do that with local charms at jujugui.com for the time being can't i? |
|
[18:43] <rick_h_> lazyPower: ah, good point. yes |
|
[18:43] <rick_h_> we can do local charms in sandbox mode |
|
[18:43] <lazyPower> hmm.. |
|
[18:44] <lazyPower> i really want a juju gui wireframing tool. I'll pin this and come back to it. TA gentlemen |
|
[19:10] * rick_h_ escapes from today. Have a good day all. |
|
[19:34] <hatch> so hows everyones afternoon going? |
|
[19:35] <bac> hatch: mine is going well. a bit warm but i'm fighting the urge to turn on a/c. |
|
[19:49] <hatch> bac yeah? What does the temp need to be to turn that on? |
|
[19:49] <hatch> I'm usually turning it on at around 18c |
|
[19:49] <redir> later rick_h_ |
|
[19:50] <bac> hatch: well it is 28c now... so, 30c i'd say |
|
[19:51] <redir> 30c == ac to me if humid |
|
[19:52] <hatch> jeebus |
|
[19:52] <hatch> 30c is me curling up in the basement |
|
[19:53] <hatch> lol |
|
[19:53] <redir> hatch: also depends if the windows are open, breeze etc... |
|
[19:53] <hatch> granted it does hit 30+ during the summer but those days are rare |
|
[19:53] <hatch> redir let the smog roll right through? |
|
[19:53] <hatch> :P |
|
[19:53] <redir> smog? |
|
[19:53] <redir> this ain't LA |
|
[19:54] <hatch> lol, with all those ppl there isn't smog? |
|
[19:54] <hatch> I figured just that many people exhaling would create smog lol |
|
[19:57] <redir> there's definitly some grime, but that is mostly just cars. |
|
[19:57] <redir> but no smoggier than DC |
|
[19:57] <redir> maybe less |
|
[20:00] * Makyo takes dogs for a walk while fighting CI |
|
[20:33] <hatch> Chrome is 50% slower running our test suite than safari https://twitter.com/FromAnEgg/status/467039921501655040 |
|
[20:38] <bac> redir: the ngram branch did not deploy cleanly to the qa machine. i think it requires the ES index be blown away before the mapping will work. |
|
[20:41] <bac> redir: that means before we can deploy it to production we'll need to write a migration to remove the index. i'll make a card. |
|
[20:42] <redir> i see |
|
[20:42] <redir> I think we (you) had mumbled something about that in vegas |
|
[20:43] <hatch> jujugui lf a quick review/qa https://github.com/juju/juju-gui/pull/324 |
|
[20:43] <redir> bac how do migration scripts work? |
|
[20:43] <bac> redir: i was probably mumbling more than usual in vegas |
|
[20:43] <bac> redir: precariously |
|
[20:44] <bac> redir: we should probably chat tomorrow for a more thorough explanation |
|
[20:45] <redir> bac: works for me |
|
[20:46] <redir> if that other script you had worked... |
|
[20:46] <bac> redir: ngrams now seem happyish: http://qa.manage.jujucharms.com/search?search_text=erp&op= |
|
[20:46] <bac> redir: but only on that machine |
|
[20:48] <redir> bac and that search box doesnt' use ngrams I don't think |
|
[20:49] <bac> redir: ah, yes, you're correct. |
|
[20:49] <redir> cuz we only fixed up api search to use it |
|
[20:49] <bac> redir: so, an api test will be required before proclaiming victory |
|
[20:51] <redir> http://bit.ly/1jjJUNp |
|
[20:51] <redir> bac ^ |
|
[20:51] <bac> redir: that looks like good news |
|
[20:52] <bac> redir: too bad there isn't a couscous project that we could've matched |
|
[20:52] <bac> or countchocula |
|
[20:52] * redir gets hungry at mention of couscous |
|
[20:52] <redir> going to danrea tonight |
|
[20:53] <redir> da andrea |
|
[20:53] <bac> i hope that is better than it sounds |
|
[20:53] <redir> http://bit.ly/1jjKqet |
|
[20:54] <bac> redir: so, in summary, the qa machine is happy and the jenkins lander seems happy again. though QA for your branch failed since it didn't install cleanly. |
|
[20:54] <redir> which is because the index fails to update without wiping it? |
|
[20:55] <redir> perhaps related to jcsackett's issue qa-ing, maybe. |
|
[20:55] <bac> redir: yes. the webops will get unhappy if we tried to deploy like that on production |
|
[20:56] <redir> heheh |
|
[20:56] <bac> redir: you should offer to help that restaurant pick a new palette for their web site in exchange for a free dinner |
|
[20:56] <redir> rightly so |
|
[20:56] <redir> yeah hopefully they spend more time in the kitchen than building www sites |
|
[20:58] <redir> google maps now has a plane as a transportation option |
|
[21:00] <hatch> oh cool |
|
[21:21] <hatch> jujugui anyone available for a review? https://github.com/juju/juju-gui/pull/324 |
|
[21:21] <Makyo> hatch, on it |
|
[21:22] <hatch> thx |
|
[21:22] <hatch> it' a quick one |
|
[21:23] <Makyo> hatch, can you look at #323? I'm not sure what's going on with CI. |
|
[21:23] <hatch> sure |
|
[21:23] * rick_h_ looks at CI |
|
[21:24] <rick_h_> Makyo: saucelabs fail |
|
[21:24] <rick_h_> Makyo: just keep pushing it tbh |
|
[21:24] <Makyo> alright, will do |
|
[21:25] <rick_h_> Makyo: well that error usually is a Sauce issue. Might have hit it at a bad time. If it fails like that again, after someone else got a success I'd think maybe there's an issue in there in the tests |
|
[21:26] <hatch> Makyo looks like the first one was a real failure, the second was a sauce labs failure |
|
[21:26] <hatch> I also added one comment to it if you like |
|
[21:26] <Makyo> Alright, but I'll deal in a bit. One thing at a time. |
|
[21:35] <Makyo> hatch, was there no test impact for your branch? |
|
[21:36] <hatch> Makyo there were test already - they just continue to pass :) |
|
[21:36] <Makyo> hatch, just making sure. |
|
[21:36] <Makyo> Another case of not-really-unit-tests |
|
[21:42] <hatch> yup |
|
[21:43] <hatch> I've talked with a few people who run other large scale js apps and how they do their tests |
|
[21:43] <hatch> maybe in 3 years when we get some downtime we can refactor our test suite |
|
[21:43] <hatch> :) |
|
[21:47] <rick_h_> hatch: :P or in the next 6mo we'll work on it. |
|
[21:47] <hatch> that's a lie and you know it!! |
|
[21:47] <rick_h_> hatch: not fair considering you and I just talked about that work starting after the next 2wk cycle |
|
[21:47] <hatch> lol |
|
[21:47] <rick_h_> ok, well yea lies all the way down |
|
[21:48] <hatch> rick_h_ did you see my G+/tweet about safari vs chrome? |
|
[21:49] <hatch> crā crā! |
|
[21:49] <rick_h_> hatch: yea, I want to try that out. I'm curious if that's a chrome dev version thing, a chrome on osx thing, etc |
|
[21:49] <hatch> I'm on the GA release |
|
[22:37] <hatch> Makyo are you around still? |
|
[22:37] <Makyo> hatch, yeah |
|
[22:37] <hatch> so I'm looking into this issue with placeUnit not placing in containers |
|
[22:37] <hatch> *existing containers |
|
[22:38] <hatch> actually......hangout? |
|
[22:38] <hatch> https://plus.google.com/hangouts/_/gzx4obd5lqizoj2vofrv5zs4waa?hl=en |
|
[23:00] <hatch> I came across a cool song "Will Chaplin - Eye of the Pyramid" he was on The Voice last year |
|
|