[11:34] <rogpeppe> if anyone wants to take a look at the latest allWatcher CL, it's here: https://codereview.appspot.com/7594048/
[13:51] <rick_h_> gary_poster: the hacking doc doesn't mention submitting changes to the charm. I'm looking to just self-review/land the changes https://code.launchpad.net/~rharding/charms/precise/juju-gui/subapp_config/+merge/153557 is there a process I need to make sure I follow? e.g. is there a tarmac/test runner for the charm?
[13:55] <benji> rick_h_: Gary is out of the office today.  I'm not completely sure what the charm landing procedure is, but I'm sure there are tests and reviews.  Let me look real quick and see if anything jumps out at me.
[13:55] <rick_h_> doh gary_poster is out today isn't he
[13:55] <rick_h_> benji: thanks, yea there are tests, and instructinos for running locally. 
[13:55] <rick_h_> benji: but curious for this small change if it's safe to merge with trunk and push directly or if something mechanical should be looking/doing that?
[13:56] <rick_h_> benji: we don't run tarmac on our charm, but it doesn't have the nice tests you guys have so figured it might be here.
[13:56] <benji> nope, there is not automation as of yet
[13:56] <rick_h_> benji: ah ok cool then. 
[14:27] <benji> jujugui: call in 3
[14:28] <rick_h_> hatch: https://code.launchpad.net/~rharding/juju-gui/subapp_updaet is my changes applied to your subapp branch. You can pull them in from there if you'd like. Sorry for the typo in the name. Can't seem to type today.
[14:30] <hatch> ahh very cool thanks
[14:42] <rick_h_> goodspud: so on icon related notes. Is there someone we can ping to make charmworld/juju-gui leaders by examples and getting something fancy in .svg for our charms?
[14:46] <goodspud> rick_h_  I'd love to say yes but our designer hasn't adopted IRC as an acceptable form of communication
[14:46] <goodspud> BUT he has just recently discovered his work email
[14:46] <rick_h_> goodspud: is there someone I should send an email to?
[14:46] <goodspud> greg.viercant@canonical.com
[14:46] <rick_h_> lol
[14:46] <hatch> lol!!
[14:46] <rick_h_> goodspud: thanks, I shall begin ancient ineffecient communication patterns with said person :)
[14:47] <goodspud> He's got a bit of work to do today but he's going to produce some documentation re icons on Monday
[14:47] <hatch> no irc means they can claim the 'email got lost'
[14:47] <hatch> :)
[14:47] <goodspud> I sit beside him... cc me into it
[14:48] <goodspud> rick_h_, you could describe IRC as ancient as well couldn't you? Even older than email?
[14:48] <hatch> ooo he's got you there
[14:49] <rick_h_> is it? I figured email came first. After all I'm not counting finger on an old unix system as irc 
[14:49]  * rick_h_ heads to wikipedia
[14:49] <hatch> 1982 SMTP
[14:49] <hatch> 1988 IRC
[14:49] <hatch> OOO
[14:49] <hatch> soo close
[14:50] <rick_h_> email 1982 and irc 1988
[14:50] <rick_h_> I win!
[14:50] <goodspud> Damn it!
[14:50] <hatch> yeah except I totally wiki faster than you
[14:50] <hatch> :P
[14:51] <rick_h_> must be that fancy canadian internet
[14:51] <goodspud> I accept defeat
[14:51] <hatch> it's ok as time goes on, the difference becomes negligable
[14:52] <hatch> negligible even
[14:52] <hatch> people who research dino's even thing 1M years is no big deal
[14:52] <hatch> :P
[14:55] <arosales> retrospective this week?
[14:56] <hatch> sure thing, when are you thinking?
[14:56] <rick_h_> I thought there was none due to the people out.
[14:57] <hatch> well there wasn't - but we could if people had things they wanted to complain about ;)
[14:57] <goodspud> arosales, Gary said yesterday that we'd skip the weekly meeting because half of the people are away
[14:57] <arosales> goodspud: ah, ok 
[14:57] <goodspud> But, up to those who are still around really
[14:57]  * arosales wasn't aware of the yesterday's comments
[14:57] <goodspud> I can certainly tell you all where we are at in the design deparment
[14:58] <arosales> no need to meet on my account or anything. I wasn't aware of the logistics.
[14:58]  * arosales puts headset up
[14:59] <goodspud> arolsaes, no worries. Just know that everything is alright.
[14:59] <goodspud> (hopefully)
[15:00] <hatch> rick_h_ I saw you merged in that charm code - I THINK that we are supposed to tag things like that as trivial before merging...I don't know where or how, but I saw Ben do that during the sprint
[15:01] <rick_h_> hatch: ah, sorry. 
[15:01] <hatch> oh it's no problem, I don't even know if that's true - I just saw him do it
[15:01] <hatch> haha
[15:01] <hatch> :)
[15:07] <goodspud> rick_h_, regarding your email....
[15:07] <goodspud> You're asking for an icon for the Juju GUI Charm
[15:07] <goodspud> AND
[15:08] <goodspud> One for the charm store/charm world?
[15:08] <goodspud> (assuming there is not Charm for deploying the Charm Store?)
[15:09] <rick_h_> goodspud: so I think for now the primary one is the Gui since it'll be front/center
[15:10] <rick_h_> there's talk of people doing their own charm stores at some point, and I'm working on both projects atm so I figured I'd ask for both :)
[15:10] <rick_h_> goodspud: but both are charms from canonical and I assume at some point we'd want some sort of icons for things we put out there. 
[15:11] <goodspud> rick_h_. OK. We have our branding icon for the Canonical Charm Store (it's the 4 j's in a rosetta pattern). Mark liked it so we've stuck with it
[15:11] <rick_h_> goodspud: cool
[15:12] <goodspud> rick_h_, but yes we could definitely work on an icon for the Juju GUI charm itself
[15:12] <rick_h_> goodspud: awesome! appreciate it. 
[15:12] <goodspud> rick_h_, I'll beat our designer into submission and chain him to his desk until he does it
[15:12] <goodspud> PS: I'm only saying this because he's not on IRC
[15:13] <hatch> lol
[15:13] <rick_h_> goodspud: make sure he gets occassional crackers and cheese. We're not cruel people :)
[15:13] <goodspud> rick_h_, He's from "Up Norf" so I'l lthrow in some breaded scampi and chips
[15:13] <rick_h_> ah, yea I don't speak 'over the pond'. I'll try to get some reading time in this weekend.
[15:16] <hatch> it would be awesome if we could trigger our charm tests via the gui
[15:16] <goodspud> rick_h_, I've been here 6 years and I'm still getting used to it
[15:31] <hatch> goodspud: where are you originally from?
[15:31] <goodspud> hatch, a small country I like to call "Godzone" but the rest of the world calls "New Zealand"
[15:32] <hatch> lol
[15:32] <goodspud> hatch, the locals call it "Aotearoa"  -  the and of the long white cloud
[15:32] <goodspud> er, *land
[15:32] <hatch> so it's always cloudy there?
[15:33] <goodspud> Yep
[15:33] <goodspud> Except at the moment... it's suffering a drought
[15:33] <hatch> our license plates say 'land of the living sky' - what it should have said was 'damn cold, stay away'
[15:34] <goodspud> hatch, Bwa ha ha ha ha
[15:34] <goodspud> Only in winter surely
[15:34] <hatch> yeah only in winter :) In the summer it's actually really nice here
[15:35] <goodspud> I've only been to Canada in Winter - bloody freezing
[15:35] <goodspud> And then it was only Toronto/London (ontario)
[15:35] <hatch> oh pfft that's probably the warmest place :P
[15:36] <hatch> imho if you realy want to see Canada you need to get out of the big cities
[15:36] <hatch> much the same as anywhere I suppose
[15:38] <goodspud> hatch, that's my opinion of most places I go... which is why I like to go by motorbike when I travel
[15:38] <goodspud> But, I went to Toronto before I had a motorbike
[15:39] <goodspud> it did only get down to -17
[15:40] <hatch> haha -17 is cold enough though
[15:40] <hatch> :)
[15:40] <hatch> buy a motorbike...that's a great idea!
[15:41] <goodspud> For a moderate climate boy who had only been in the snow skiing once, it was definitely cold
[15:41] <goodspud> Yeah mang, motorbikes are awesome
[15:41] <goodspud> Just don't crash
[15:41] <goodspud> But then don't crash in any vehicle
[15:41] <hatch> :) I had 2 gsx-r 1000's but sold the last one a few years ago
[15:41] <hatch> I'm scared of the other drivers haha
[15:42] <hatch> one almost took me out on the freeway - that's when it went up for sale
[15:43] <hatch> fyi all it looks like this node issue was caused by the update
[16:04] <benji> man, I am not feeling 100% today; I hope I'm not coming down with something
[16:04] <hatch> uh oh
[16:04] <hatch> time go to pick up some OJ
[16:37] <hatch> arg
[16:38] <rogpeppe> gary_poster: did you see my proposed branch, BTW?
[16:40] <benji> rogpeppe: Gary is out of the (virtual) office today.
[16:40] <rogpeppe> benji: ah, thanks
[16:40]  * rogpeppe thinks we should probably use /away more
[16:42] <hatch> rogpeppe: I have a q which you might know....is it documented somewhere about the relation between CS path and LP repo name?
[16:42] <rogpeppe> hatch: CS path?
[16:42] <hatch> charmstore
[16:42] <hatch> so for example...
[16:42] <rogpeppe> hatch: ah! hmm, it might be
[16:43] <hatch> I can't use this charm because it's not called juju-gui  lp:~hatch/charms/precise/juju-gui-conflict/trunk
[16:43] <hatch> I change the path to be juju-gui and it deploys using cs:~hatch/precise/juju-gui
[16:44] <hatch> butI can't use cs:~hatch/precise/juju-gui-conflict
[16:44] <hatch> so instead of me fighting with this I was hoping there were some 'rules' :)
[16:45] <rogpeppe> hatch: i don't know this off the cuff, i'm afraid
[16:45] <rick_h_> hatch: time to push to lp:~hatch/charms/precise/juju-gui/conflict 
[16:45] <hatch> darn - I'm moreso curious why I can't specify an absolute path to begin with but I'm sure there is a logical explanation behind it
[16:45] <hatch> rick_h_ that won't work, it expects it to end in 'trunk'
[16:46] <hatch> at least that's the only way I got it to deploy
[16:46] <rick_h_> hatch: hmm, so you're not pushing a local charm to your env?
[16:47] <rick_h_> guess not if you're referring to the remote cs:...
[16:47] <hatch> yeah I'm actually debugging this CI Stuff so I can't do it locally
[16:48] <hatch> matsubara: it looks like you're having issues staying connected :)
[16:48] <rogpeppe> hatch: i *think* this might be logic in question: http://paste.ubuntu.com/5617026/
[16:49] <hatch> ahah!
[16:49] <hatch> so if I wan'ted to file a bug about that who where is the logical place to do so?
[16:49] <rogpeppe> hatch: that's inside juju-core/store/lpad.go
[16:49] <rogpeppe> hatch: juju-core
[16:49] <hatch> gotcha
[16:49] <hatch> thanks a bunch
[16:49] <rogpeppe> hatch: np
[16:49] <hatch> I knew 'how' to get it to work but wasn't sure why
[16:50] <hatch> rogpeppe: is there an unminified version?
[16:50] <rogpeppe> hatch: ?
[16:51] <hatch> oh I just see 'u', and 'nil' those looked minified to me
[16:52] <hatch> or is nil a go keyword?
[16:52] <rogpeppe> hatch: it is
[16:52] <hatch> oh heh
[16:52] <hatch> :D
[16:52] <hatch> go looks pretty easy to follow
[16:53] <hatch> I'm going to guess those are 'famous last words' :D
[16:53] <rogpeppe> hatch: it's designed to be easy to follow. that doesn't mean it always is though :-)
[16:53] <rogpeppe> hatch: more accurately, nil isn't a go keyword, it's a predefined identifier. you have have variables and struct fields named "nil" if you really want :-)
[16:54] <hatch> ohh
[16:54] <hatch> well I'll be sure to add go to my list of 'languages to learn'
[16:55] <hatch> well at least learn well enough to know what I'm looking at :D
[16:57] <hatch> wow it's snowing here, can't even open the backdoor haha
[16:59] <hatch> ok here is to hoping that I  have solved the node issues
[16:59] <hatch> (by reverting to an old version) :-(
[17:17] <hatch> alright so it looks like that the PPA that we use for node no longer includes the working version of node
[17:18] <hatch> https://launchpad.net/~chris-lea/+archive/node.js/
[17:18] <hatch> so we can manually build from source
[17:18] <hatch> OR wait
[17:18] <hatch> anyone have any input?
[17:19] <hatch> https://github.com/isaacs/npm/issues/2907 here is the issue - looks like it's a regression
[17:19] <benji> hatch: I haven't tried to build the gui in a while; are you blocked?
[17:19] <hatch> well it's blocking the CI
[17:20] <hatch> issue is that I don't know how long until a fix...could be days...could be weeks
[17:20] <hatch> but our system isn't really setup to build node right now
[17:20] <benji> that issue say that it it fixed in 0.8.14
[17:20] <hatch> scroll down ;)
[17:20] <hatch> wayyyyyy down
[17:21] <benji> heh
[17:21] <benji> does it happen on trunk?
[17:22] <hatch> yep
[17:22] <hatch> it's an issue with that version of npm
[17:22] <rick_h_> hmm, can't you lock a previous version in a repo?
[17:23] <hatch> rick_h_ I am but the PPA only has node 0.10
[17:23] <hatch> looks like he removed 0.8.14
[17:23] <hatch> https://launchpad.net/~chris-lea/+archive/node.js/
[17:23] <benji> oh, that's evil
[17:24] <benji> actually, that may be the default (or only) PPA behavior
[17:24] <hatch> ahh
[17:24] <rick_h_> benji: yea, now that I think of it I don't often recall tons of back revisions
[17:24] <hatch> https://launchpad.net/~chris-lea/+archive/node.js-legacy
[17:24] <hatch> yay!
[17:25] <rick_h_> hatch: there you go. I'd say go that route vs doing our own compiling of fresh packages until npm fixes itself
[17:25] <hatch> oh yeah for sure - I REALLY didn't want to set up our system to build it hah
[17:26] <benji> this page shows lots of old builds: https://launchpad.net/~chris-lea/+archive/node.js/+builds?build_text=&build_state=built
[17:27] <benji> the old ones are available for download, so they might be available in the archive
[17:28] <rick_h_> just changing the ppa source to the archive should unblock and once the bug is fixed can be switched back.
[17:28] <hatch> yup
[17:32] <hatch> we should make our own repo for things and then deploy from there so we always have working versions haha
[17:34] <rick_h_> one thing we did in charmworld was to create a download cache of all of our node deps and npm install from local only. 
[17:34] <rick_h_> I didn't look at the npm bug close enuogh to see if it was only a problem fetching remote packages or a more general npm issue
[17:36] <hatch> yeah that's something we should also do - would probably speed up the installs too
[17:36] <hatch> put that on the infinitely long todo list
[17:36] <hatch> haha
[17:41] <benji> hatch: how do you provoke the problem?  A make clean-all; make; make test-prod worked for me.
[17:42] <hatch> sudo apt-get update; rm -rf /path/to/trunk/node_modules; make clean-all; make devel
[17:43] <hatch> should do it
[17:43] <hatch> node --version needs to output 0.10.0
[17:52] <hatch> yay install working now - tests running
[17:52] <hatch> and it's not even lunch yet
[17:52] <hatch> w00t
[18:00] <benji> I can't reproduce the problem.
[18:00] <benji> (and by the way, make clean-all should remove the node_modules directory)
[18:07] <rick_h_> benji: but npm is a system wide thing. It's like needing to update apt-get itself. it's not in the node_modules
[18:07] <rick_h_> benji: just like the node binary itself
[18:08] <benji> ah... that's stupid
[18:08]  * benji looks around for where the modules are stored
[18:08] <rick_h_> benji: :) well virtualenv does bring it's own pip to install. but at least it's part of the virtualenv
[18:09] <rick_h_> which npm and which node should show a system-wide install from taht chris lea ppa
[18:10] <hatch> I've solved it
[18:10] <hatch> benji: you need to be running the most recent version of node and npm
[18:10] <hatch> for it to break
[18:10] <hatch> so probably don't
[18:10] <hatch> :D
[18:11] <benji> hatch: cool, what is the solution?
[18:11] <benji> I am runnign the latest from the PPA but I haven't tried the remove-the-node-modules bit yet, so that's prbably why I can't reproduce the problem.
[18:13] <hatch> using node v 0.8.14 and whatever npm comes with it
[18:13] <hatch> I'll monitor the situation and when they release the new version I'll upgrade to 0.10 which requires some other small changes
[18:22] <benji> I'm confused.  I'm running node v0.10.0 and npm 1.2.14 with no problems.  It's not something that we have to nail down now, but I'm concerned that other developers (and users) will see breakage.
[18:24] <rick_h_> benji: yea, in that bug report there's reference to a 'sudo npm cache clear' command. Wonder if there's a cache in place keeping you happy
[18:25] <benji> I wonder where that cache is ~/.npm perhaps
[18:40] <benji> yay!  I broke it!
[18:40] <benji> it is indeed the cache that was saving me (good thing I put it aside rather than killing it outright)
[18:40] <rick_h_> benji: heh, caution is always a good thing
[18:41] <benji> hatch: this seems like something the other developers should be aware of, will you write an email to the juju-gui-dev list (whatever it's called) and tell everyone what to do/not do?
[19:00] <benji> OK guys, I'm not feeling well and I'm not getting anything done so I will take a late lunch and stop a bit early.  See you all on Monday.
[19:07] <hatch> benji: I can do that
[19:07] <hatch> get better - see you Monday
[19:12] <hatch> rick_h_ are you around?
[19:33] <hatch> jujugui anyone around able to give me a quick review on a one line diff https://codereview.appspot.com/7759050/
[19:33] <hatch> ok 2 line diff, 1 line change :)
[19:34] <bac> hatch: yep
[19:34] <hatch> thank yas
[19:35] <bac> hatch: did
[20:31] <rick_h_> hatch: what's up?
[20:32] <hatch> you are no longer needed
[20:32] <hatch> ;)
[20:32] <hatch> unless you want to help me debug my flickr carousel :P
[20:40] <hatch> Good news is that the CI is back up and running albeit one obscure test failure
[20:44] <rick_h_26> hatch I hate obscure 
[20:45] <rick_h_26> cool news that things are trucking at EOW though
[20:45] <hatch> yeah but somehow the unit tests are failing when they pass no problem (obviously because you can't commit if they fail)
=== rick_h_26 is now known as rick_h_droid