[11:06] <frankban_> dimitern: morning, when you have time, could you please take a look at https://codereview.appspot.com/50090044/ ?
[11:07] <dimitern> frankban_, sure, will look after the standup
[11:07] <dimitern> frankban_, morning :)
[11:11] <frankban_> dimitern: thank you! IIRC, I need two reviews, and when everything is ok I just need to approve on launchpad, correct?
[11:12] <dimitern> frankban_, not anymore, just one LGTM is enough, and then you need to set the commit message on the MP first and then approve it
[11:13] <frankban_> dimitern: cool thanks
[11:20] <dimitern> frankban_, reviewed
[11:21] <dimitern> frankban_, usually, we set the MP commit message to the same as the description (look at the bzr rv-sumit plugin you can use to approve and submit for landing a CL - it's awesome)
[11:21] <dimitern> s/rv-sumit/rv-submit/
[13:38] <benji> jujugui: I need someone with some git fu; I don't have any ideas left on how to get my branch merged.  (other than hand-editing a diff and starting a new branch with it)
[13:44] <gary_poster> benji you tried cherrypicking?
[13:45] <benji> gary_poster: nope; I'll google it.  Unfortunately I asked that question at the wrong time, I need to run Owen to day care :\ now
[13:45] <bac> hi jujugui. I've got a laptop with no display this morning. trying to get my env and vim moved to another. on here via tablet.
[13:45] <gary_poster> benji alternatively, FBOW the git pattern of rebasing doesn't seem to be any better than breating a diff and that applying it as a patch
[13:45] <gary_poster> ok benji
[13:45] <gary_poster> ack bac
[13:45] <bac> s/vim/vm/
[13:46] <bac> I don't use no vim
[13:46] <gary_poster> :-)
[14:03]  * bac_ some progress
[14:18] <bac> nick bac
[14:19] <bac> doh
[14:22] <rick_h_> hello all, thanks for the review/landing gary_poster 
[14:22] <gary_poster> hey rick_h_.  you there safely?
[14:22] <gary_poster> welcome
[14:22] <rick_h_> gary_poster: yep, long long trip but alive and well
[14:22] <gary_poster> cool :-)
[14:23] <rick_h_> surviving first day of meetings. Just one more today I think. I'll try to attend to the standup from here. It'll be 6pm here
[14:23] <gary_poster> ok cool rick_h_
[14:23] <rick_h_> I'll try to look into the qa issue on the bws branch. Thought that worked. Thanks for catching it
[14:23] <gary_poster> np.  very nice otherwise
[14:40] <frankban> guihelp: is anyone available for a quick review + QA of https://codereview.appspot.com/59560044 (GUI charm)? thanks
[14:40] <gary_poster> frankban: on it
[14:40] <frankban> gary_poster: thank you
[15:01] <gary_poster> hey frankban, this isn't a bug in your branch, but potentially in core (and therefore a problem for us).  I tried to use a local environment with the juju I had currently installed (1.16.5-saucy-amd64).  My ~/.juju/environments/local.jenv says 'password: ""'.  An empty password does not work to log into the GUI.  I noticed that the user is "" also in the jenv file.  Perhaps that's the problem, but I tried logging in 
[15:01] <gary_poster> on the console with a an empty password and a user name of "user-" and it did not work.  That doesn't really make sense anyway.
[15:02] <gary_poster> So...
[15:02] <gary_poster> Any ideas on what we should do there?
[15:02] <gary_poster> One option is to see if it works on the most recent released juju
[15:02] <gary_poster> and if it does, pretend everything is OK:-)
[15:03] <gary_poster> I'm fine with that, but wanted to see what you thought
[15:03] <gary_poster> It's certainly not ideal :-/
[15:03] <gary_poster> ah
[15:04] <gary_poster> my environments.yaml does have an admin-secret for my local env
[15:04] <gary_poster> it just wasn't copied over to the jenv
[15:04] <gary_poster> it works fine
[15:04] <frankban> gary_poster: you should have an admin-secret field in your jenv. I also have user and password as empty strings there, and that's confusing
[15:04] <gary_poster> frankban: with recent release? trunk?
[15:05] <frankban> gary_poster: this is 1.16.5-saucy-amd64, trying trunk now
[15:05] <gary_poster> ok
[15:05] <gary_poster> I hope it is fixed in trunk.
[15:06] <gary_poster> If so, the legacy message can say "if there is no password there, try the environments.yaml" file.  That's unpleasant, but accurate, and at least the new version would be OK
[15:06] <frankban> gary_poster: same thing in trunk, user and password empty, the real password is found in the admin-secret field of the jenv
[15:06] <gary_poster> frankban: :-(
[15:06] <gary_poster> um
[15:06] <gary_poster> so, ideally
[15:07] <gary_poster> we'd get Tim or someone to treat this as critical
[15:07] <gary_poster> and fix it
[15:07] <gary_poster> before the next release
[15:07] <frankban> gary_poster: so we always have the admin-secret in the jenv, but we also have a confusing user and password fields
[15:07] <gary_poster> oh
[15:07] <gary_poster> oic
[15:08] <gary_poster> hm
[15:08] <gary_poster> and it is correct for ec2
[15:08] <gary_poster> (that is, you can use password)
[15:08] <gary_poster> ok.  so, right\
[15:09] <gary_poster> IMO ideally we'd have juju core treat this as a critical fix
[15:09] <gary_poster> then the legacy message would be, roughly, "look in the jenv file for password, or if that's empty, look for admin-secret"
[15:09] <gary_poster> and the new message would be "look in this file right here for the password"
[15:10] <gary_poster> I think that would be OK, if we can get juju core to address before next release.  WDYT frankban?
[15:11] <frankban> gary_poster: oh... in trunk, the jenv file includes the password field once you deploy the first charm
[15:11] <gary_poster> frankban: :-/
[15:12] <frankban> gary_poster: I mean, the password field is no longer empty
[15:12] <gary_poster> right
[15:12] <gary_poster> well, I guess that would be OK for the GUI
[15:12] <gary_poster> so we can treat this as not critical for us
[15:12] <gary_poster> and adjust the legacy message.  agree?
[15:13] <gary_poster> getting some water, brb
[15:13] <frankban> gary_poster: yes, when the user see the message the password field should be populated, even in local envs. so I'd need to change the messages to point to the password field (rather than the admin-secret one), and the legacy message should mention the environments.yaml file for juju < 1.16
[15:17] <gary_poster> cool frankban.  proceeding with qa
[15:24] <gary_poster> frankban: this is too long, but it is my first stab at a suggestion for adjusting the legacy help text.
[15:25] <gary_poster> The password for newer Juju clients can be found by locating the Juju environment file placed in ~/.juju/environments/ with the same name as the current environment.  For example, if you have an environment named "production," then the file is named ~/.juju/environments/production.jenv.  Look for the "password" field in the file, or if that is empty, for the "admin-secret".  Remove the quotes from the value, and use 
[15:25] <gary_poster> this password to log in.  The password for older Juju clients (< 1.16) is the admin-secret for the environment you are using in ~/.juju/environments.yaml.  Note that using juju-quickstart (https://launchpad.net/juju-quickstart) can automate logging in, as well as other parts of installing and starting Juju.
[15:25] <gary_poster> Still adjust that
[15:25] <gary_poster> adjusting
[15:26] <gary_poster> The password for newer Juju clients can be found by locating the Juju environment file placed in ~/.juju/environments/ with the same name as the current environment.  For example, if you have an environment named "production," then the file is named ~/.juju/environments/production.jenv.  Look for the "password" field in the file, or if that is empty, for the "admin-secret".  Remove the quotes from the value, and use 
[15:26] <gary_poster> this to log in.  The password for older Juju clients (< 1.16) is in ~/.juju/environments.yaml, and listed as the admin-secret for the environment you are using.  Note that using juju-quickstart (https://launchpad.net/juju-quickstart) can automate logging in, as well as other parts of installing and starting Juju.
[15:27] <gary_poster> Best so far ^^^
[15:27] <frankban> gary_poster: so here is the situation: admin-secret is always populated correctly. the password field is populated in trunk when at least one charm is deployed after bootstrapping. the password field is never populated in 1.16
[15:27] <frankban> gary_poster: that's great! thank you!
[15:28] <gary_poster> cool frankban.  proceeding to Juju trunk noq
[15:28] <gary_poster> now
[15:30] <rick_h_> jujugui anyone got a sec to re-qa https://github.com/juju/juju-gui/pull/98 please? /me looks at it and realizes pushing a branch done while tired on a plane was not a good plan. 
[15:30] <gary_poster> :-)
[15:32] <gary_poster> rick_h_: I'm doing it while waiting for the charm to set up
[15:32] <gary_poster> for a different qa
[15:32] <gary_poster> oh it's up :-P
[15:32] <gary_poster> 1 sec then
[15:40] <gary_poster> frankban: LGTM and QA OK.  Thanks!
[15:41] <frankban> gary_poster: thank you!
[15:41] <gary_poster> :-)
[15:41] <frankban> gary_poster: filed two holidays
[15:42] <gary_poster> frankban: ack, approved
[15:43] <frankban> gary_poster: re paragraphs, yes that's a GUI branch (to make the text not to be escaped) and a subsequent charm branch. the first one could also help with the quickstart link (i.e. a target blank), thnaks
[15:44] <gary_poster> cool
[15:50] <gary_poster> jujugui call in 10
[15:50] <gary_poster> rick_h_: I am slowed down because I'm trying to figure out a nice way to update my git qa-pr version of your PR
[15:51] <gary_poster> with your new changes
[15:51] <gary_poster> simply deleting the branch and then running qa-pr again was insufficient
[15:51] <gary_poster> I assume it has something to do with the git fetch we are doing
[15:52] <gary_poster> but currently flailing
[15:52] <gary_poster> if you have suggestion, it would be welcome :-)
[15:53] <gary_poster> ah-ha
[15:53] <rick_h_> gary_poster: hmm, so if you git checkout develop && git branch -d qa-... and then another git qa-pr it won't work?
[15:53] <rick_h_> gary_poster: or you found a way?
[15:54] <gary_poster> rick_h_: correct, that does not work, and I think it is because remotes/pr/98 already exists
[15:54] <gary_poster> andthere is no logic to update that remote
[15:54] <gary_poster> so I need to update that remote somehow...
[15:55] <gary_poster> nad then the merge will work
[15:55] <gary_poster> trying to delet remote now...
[15:59] <gary_poster> oops
[15:59] <gary_poster> jujugui call in 1
[16:09] <Makyo> Finally!
[16:25] <rick_h_> man, you'd think on a canonical connection this wouldn't be an issue. 
[16:25] <gary_poster> heh, yeah
[16:26] <Guest92118> Grrr.
[16:26] <Guest92118> Services are down, whoops.
[18:32] <bac> so my mbp is back.  i'm beginning to think vmware ate my graphics card but can't explain how it took multiple reboots to clear up.
[19:07] <bac> gary_poster: i reviewed the work i did on friday for juju-core 1.17.2 QA and what i said in the meeting was wrong.  the QA was clean.  no remaining issues.
[19:12] <gary_poster> great bac
[19:27] <hatch> freenode I'm in you!
[20:12] <rick_h_> gary_poster: heads up, quickstart package was pushed upstream yesterday. 
[20:12] <rick_h_> gary_poster: and casey is going to look into how to do a charm store deploy for 1.17.2
[20:12] <gary_poster> rick_h_: awesome!  thanks!
[20:12] <gary_poster> great
[20:13] <gary_poster> rick_h_: if you want some demo cred, you can demo local charm deployment from github.  It's a *great* story
[20:13] <gary_poster> demo cred is good :-)
[20:13] <rick_h_> gary_poster: cool yea I htink there's a demo session. hatch can you shoot me some instrucitons?
[20:13] <gary_poster> I think it will blow people away
[20:14] <gary_poster> be sure to give Dimiter a shout out
[20:14] <rick_h_> sure thing, I'll see if I can get setup to duplicate. You need trunk right?
[20:14] <gary_poster> hatch just gave you +1 with lots of small comments
[20:15] <hatch> gary_poster thanks will get right on the changes
[20:15] <hatch> rick_h_ you bet, can you get juju-core trunk working on your local machine?
[20:16] <rick_h_> hatch: yea send me a link to those build docs and I'll see if I can get it done. It's late here (10pm) but hopefully find time to tru it out
[20:16] <gary_poster> rick_h_: for best experience, yes, use Juju trunk.  Then you start up GUI, go to https://github.com/hatched/ghost-charm and click on the "download zip" button on the right, drag it in to the GUI, configure it, and deploy.
[20:16] <gary_poster> On LXC on my network it shows up in a minute or so
[20:16] <gary_poster> and then you can expose it and actually use ghost
[20:16] <gary_poster> really awesome
[20:17] <hatch> yeah I'll send that in an email too, just in case :)
[20:17] <rick_h_> woot
[20:17] <hatch> rick_h_ https://code.launchpad.net/~go-bot/juju-core/trunk here is juju-core trunk, read the README for detailed install instructions
[20:19] <gary_poster> hatch does teh ghost charm download something over network?
[20:19] <gary_poster> would be more reliable/faster for rick if he had one that did not
[20:20] <hatch> gary_poster yes it requires access to a number of external resources https://github.com/hatched/ghost-charm/blob/master/hooks/install
[20:20] <gary_poster> so obviously this evening you need to adjust ghost so that default install does not need network access, like GUI ;-)
[20:21] <hatch> haha, probably not until ghost stabilizes their build setup :)
[20:21] <gary_poster> hatch, consider packaging zip, or giving rick a branch that includes zip, like we do for gui?
[20:22] <gary_poster> also then emphasizes advantage of being able to work with local charms for demos:
[20:22] <gary_poster> make quick hack, ready to demo
[20:22] <hatch> I'll have to look into the current ghost build process to see if I can package all of it's deps or if it needs some 'installed' 
[20:22] <gary_poster> (as soon as the icon actually shows up!)
[20:22] <hatch> haha
[20:22] <gary_poster> oh hatch right, npm install.
[20:22] <gary_poster> nm then
[20:22] <gary_poster> might not help much
[20:23] <gary_poster> benji gets lucky PR 100
[20:23] <hatch> I think they install sqlite so that probably needs to be 'installed'
[20:23] <hatch> not just exacted but I can look into this 
[20:24] <gary_poster> cool :-)
[20:24] <rick_h_> there's an AP right by the demo space so should be good connection there. Look slike friday is open so I'll get a chance to test it out before hand
[20:24] <gary_poster> cool, perfect
[20:25] <hatch> rick_h_ I'll send the demo instructions to peeps just so everyone has them 
[20:25] <rick_h_> hatch: thanks, appreciate it
[20:25] <rick_h_> ok, bed time, I'm out for real this time
[20:25] <rick_h_> zzzzzzzz
[20:25] <hatch> :) night
[20:25] <gary_poster> :-) night
[20:29] <gary_poster> hatch, when you have a sec (maybe after you have branch in), would like to talk with you about https://bugs.launchpad.net/juju-gui/+bug/1274955 , maybe on a hangout
[20:29] <_mup_> Bug #1274955: Juju GUI slow in disconnected environment due to using Google font APIs <micro-cluster> <juju-gui:Triaged> <https://launchpad.net/bugs/1274955>
[20:30] <hatch> sure, want to do it now?
[20:30] <gary_poster> :-) sure
[20:30] <BradCrittenden> see ya rick_h_.  have a good evening.
[20:31] <hatch> gary_poster https://plus.google.com/hangouts/_/76cpjggu44rijpcdrj53ubb2qs?hl=en
[20:31] <gary_poster> hatch, lol calling you from another, k one sec
[20:31] <hatch> haha sorry
[20:33] <hatch> gary_poster I'm going to restart
[20:33] <hatch> maybe that will help
[20:33] <gary_poster> ok hatch
[21:41] <gary_poster> jujugui, tiny charm branch for review: https://codereview.appspot.com/52230045
[22:06] <huwshimi> Morning
[22:25]  * Makyo -> dogwalk, back in a few.