[07:17] <urulama> fabrice: can you connect to canonical irc server?
[07:20] <fabrice> urulama: I am in
[07:21] <urulama> fabrice: yeah, seems to be back now
[07:21] <urulama> fabrice: and good morning :)
[07:21] <fabrice> urulama: good morning
[14:46] <hatch> uiteam need reviews and qa on https://github.com/juju/juju-gui/pull/703
[16:04] <hatch> jcsackett: kadams54 heyyyal how goes the reviews? :)
[16:04] <jcsackett> hatch: had to tie up the loose end of what i was working on pre-standup.
[16:04] <kadams54> hatch: Going to start in on them just as soon as I land this switchboard card
[16:05] <hatch> ok I'm working on the last card I have which isn't blocked by this so that's why I have a bit of urgency :)
[16:05] <jcsackett> hatch: i did make a cursory read through and didn't see anything alarming.
[16:05] <hatch> I don't like stacking branches
[16:05] <hatch> when possible
[16:07] <hatch> I'm removing the editorial.js and charmresults.js files right now - nice way to lose about 500loc's of bloat :)
[16:09] <jcsackett> hatch: ok, i've only got one question on your PR.
[16:10] <hatch> thanks jcsackett just going to reply
[16:12] <hatch> jcsackett: replied
[16:12] <hatch> jcsackett: https://github.com/juju/juju-gui/blob/develop/app/store/charmstore-api.js#L114
[16:12] <hatch> is the method in question
[16:15] <jcsackett> kadams54: any chance you can QA hatch's branch? i've done review but desperately need to find food.
[16:18] <hatch> jcsackett: go eat!
[16:18] <hatch> :)
[16:19] <jcsackett> hatch: don't have to tell me twice. :p
[16:21] <hatch> uiteam lf two reviews on a code removal branch (real easy review/qa) https://github.com/juju/juju-gui/pull/704
[16:21] <kadams54> jcsackett: yeah, I can QA.
[16:24] <rogpeppe> frankban: BTW in your disambiguation, how can you tell if there are two bundles in a basket?
[16:24] <rogpeppe> frankban: do you check out the lp branch?
[16:32] <hatch> haha the gopher china conference character is awesome https://plus.google.com/111292884696033638814/posts/331ckwTux34
[16:42] <frankban> rogpeppe: not sure if I understand your question, but no lp is not involved
[16:44] <rogpeppe> frankban: ah, i think i see - you've got an old URL and you want a new one, and all the basket metadata is in charmworld
[16:44] <frankban> rogpeppe: you can pass bundles to quickstart using two different forms: 1) new-style mediawiki-single, or u/foo/bar: those are translated to api4 bundle/mediawiki-single and ~foo/bundle/bar without problems
[16:44] <frankban> rogpeppe: no need to go to charmworld
[16:44] <frankban> rogpeppe: the second syntax is the legacy one, e.g. bundle:mediawiki/single
[16:45] <rogpeppe> frankban: i guess i still don't see how you can do the translation unambiguously
[16:46] <rogpeppe> frankban: without looking at the actual basket
[16:46] <frankban> rogpeppe: in this case check if bundle/mediawiki-single exists in the charm store. if so, then it's done. If not, get the original bundle from bundle/mediawiki, and validate that "single" is in the original content. if so, proceed with bundle/mediawiki
[16:46] <rogpeppe> frankban: how do you know that the original bundle is at bundle/mediawiki?
[16:47] <frankban> rogpeppe: because, starting from a bundle:basket/bname syntax, the resulting bundle in the charm store can be either bundle/basket-bname or bundle/basket
[16:48] <rogpeppe> frankban: but what if the basket name itself contains a hyphen?
[16:49] <frankban> rogpeppe: we start from bundle:basket-with-hyphen/bname -> bundle/basket-with-hyphen-bname or bundle/basket-with-hyphen
[16:49] <rogpeppe> frankban: you try all possible split points?
[16:50] <frankban> rogpeppe: I don't need to, because the legacy syntax separates the basket name from the bundle name, so we know what the precise basket is
[16:50] <rogpeppe> frankban: (because the bundle name in the basket can also contain arbitrary numbers of hyphens of course)
[16:50] <rogpeppe> frankban: ah of course!
[16:51] <rogpeppe> frankban: that's the bit i hadn't appreciated
[16:51] <frankban> cool
[17:24] <ahasenack> hi guys, where are juju-gui's logs?
[17:24] <ahasenack>  6731 ?        Ss     0:01 /usr/bin/python /usr/local/bin/runserver.py --logging=info --guiroot=/var/lib/juju-gui/juju-gui/build-prod --sslpath=/etc/ssl/juju-gui --charmworldurl=https://manage.jujucharms.com/ --apiurl=wss://10.96.4.18:17070 --apiversion=go
[17:24] <ahasenack> when i access it with a browser, it's stuch at the login box. No form yet, just a spinning wheel
[17:24] <ahasenack> hm, javascript console has
[17:24] <ahasenack> Firefox can't establish a connection to the server at wss://10.96.3.86/ws/environment/c703e120-5d19-4a93-8642-8194b330ebb1/api.
[17:25]  * ahasenack checks those ips
[17:25] <ahasenack> 3.86 is the actual juju-gui unit, btw
[17:25] <rick_h_> ahasenack: yep, so it should be talking to the gui unit to establish the websocket
[17:25] <rick_h_> ahasenack: so that it proxies through to juju for you 
[17:26] <rick_h_> ahasenack: so interesting that you can get the JS/html but not the wss
[17:26] <rick_h_> ahasenack: the gui logs are in /var/log/upstart/guiserver.log?
[17:26] <rick_h_> might be something of interest there on the unit
[17:27] <ahasenack> I see a log of GETs there, all seem to have succeeded
[17:27] <ahasenack> lemme restart it
[17:28] <rick_h_> ahasenack: are you referencing the gui via a differnt dns name than the url it's connecting to the wss to?
[17:28] <ahasenack> no, all ips
[17:29] <rick_h_> k
[17:37] <ahasenack> rick_h_: it doesn't load with firefox
[17:38] <ahasenack> rick_h_: loads with chromium
[17:38] <ahasenack> rick_h_: https://bugs.launchpad.net/juju-gui/+bug/1397296
[17:38] <mup> Bug #1397296: GUI inaccessible in Firefox 33.0 <landscape> <juju-gui:New> <https://launchpad.net/bugs/1397296>
[17:38] <ahasenack> (filed by someone else)
[17:39] <rick_h_> ahasenack: ok, so you're in safe mode currently?
[17:39] <ahasenack> no, tried private mode only
[17:39] <rick_h_> ahasenack: with the latest gui 
[17:39] <rick_h_> ahasenack: right, that's what I mean
[17:39] <ahasenack> I just did "juju deploy juju-gui"
[17:39] <rick_h_> you're getting this in FF in private mode?
[17:39] <ahasenack> yes, private mode too
[17:39] <rick_h_> and without private mode?
[17:39] <ahasenack> rick_h_: same
[17:40] <ahasenack> rick_h_: if you have the company vpn, you can try
[17:40] <ahasenack> rick_h_: it's 10.96.3.86. Try with FF and chromium
[17:40] <rick_h_> ahasenack: not set it up yet, but will add a note to QA this again then. originally we could not dupe and we only noted it was report in private mode
[17:40] <rick_h_> ahasenack: rgr
[17:41] <rick_h_> ahasenack: sorry, I'm leanving for a dr apt in 5min but will add the bug to the board to take another look at with your info
[17:41] <ahasenack> rick_h_: and I have a demo in 20 :)
[17:41] <rick_h_> ahasenack: use chrome?
[17:41] <ahasenack> will use chromium
[17:41] <rick_h_> ty
[17:41] <ahasenack> sure
[17:41] <rick_h_> ahasenack: trusty? 
[17:41] <rick_h_> ahasenack: for the gui deploymenht?
[17:41] <rick_h_> ahasenack: or precise just to be 100%
[17:42]  * hatch reads backlog
[17:42] <ahasenack> rick_h_: trusty
[17:42] <hatch> someone else had this exact same problem and we weren't able to reproduce it
[17:42] <ahasenack> trusty on an lxc on a bare metal machine behind the company vpn
[17:42] <ahasenack> hatch: can you try the ip right now?
[17:43] <hatch> hmm I don't have the vpn set up on this machine, I can though
[17:44] <rick_h_> hatch: ok, updated the bug, we should try to dupe locally as well just to see if it's gotten 'worse' 
[17:44] <ahasenack> well, feel free to poke. I can't give you the creds, I'm using this for a demo in a few minutes
[17:44] <rick_h_> hatch: going to run to PT appt, but yea please run with this if we have a major FF issue we need to know asap
[17:44] <hatch> rick_h_: on it
[17:44] <rick_h_> ty
[17:44] <hatch> ahasenack: definitely, demo away, I'll get setup here to test it - if you can keep it up after the demo that would be really helpful :)
[17:45] <ahasenack> sure
[17:46] <ahasenack> is there a setting to stop the gui from trying to move the services around?
[17:47] <ahasenack> I'm trying to rearrange them, but they all snap back to their original bad positions
[17:48] <hatch> ahasenack: there isn't. When you drag them are you dragging for more than 3s? 
[17:48] <ahasenack> perhaps
[17:48] <ahasenack> is that the trick?
[17:48] <hatch> ahasenack: there was an old bug which would cause the update from juju-core to override the drag position
[18:01] <hatch> ok when you're done your demo
[18:31] <hatch> kadams54: you get that review and qa done?
[18:32] <kadams54> hatch: almost
[18:34] <hatch> ahasenack: I'm not able to connect to that ip, ping when avail
[18:36] <ahasenack> hatch: here
[18:36] <kadams54> hatch: done.
[18:36] <ahasenack> hatch: I guess you were not given scapestack access in your vpn then
[18:45] <hatch> kadams54: if you have time to do 704 that would be awesome :)
[18:45] <hatch> it's just unused code removal
[18:46] <kadams54> hatch: yeah, I can take a look
[18:46] <kadams54> hatch: Code removal? +1'd without even looking :-)
[18:46] <hatch> thanks - ci hung on the run so I'm just killing it now
[18:46] <hatch> lol - yeah like 500locs out!
[18:46] <kadams54> Woo hoo!
[18:47] <hatch> that's a considerable amount of code that no longer needs to be sent over the wire