[00:00] bkerensa: ^^ Well done! :) [00:02] lol [00:02] SpamapS: Does that mean it passes muster? [00:03] bkerensa: yes, it also means it will be regularly tested by our automated testing system [00:04] bkerensa: and when the built in backend charm store is out, it will be available as 'juju deploy subway' (without a local copy) [00:05] SpamapS: Good stuff well I hope to make improvements on it in the future including upstart job handling and eventually a upgrade hook (if the author continues to add features) [00:05] bkerensa: fantastic [00:06] bkerensa: to make sure you have the latest, do 'bzr pull --remember lp:charms/subway' [00:07] SpamapS: :P I dont get to be maintainer of my charm? :P [00:07] bkerensa: after you push your changes into your own branch (lp:~bkerensa/charms/oneiric/subway/trunk) you can propose them for merging into the official charm with 'bzr lp-propose' [00:08] bkerensa: you can be the maintainer if you want. You just have to become a member of charmers... which is really easy.. you just ask, and 2 of us +1 you ;) [00:08] bkerensa: the trade off is, you are expected to also do charm reviews and bug triage. :) [00:09] uhh maybe after 12.04 release :P I am far to busy atm :) [00:09] bkerensa: we do it this way so that we can be sure that the charms are always up to date. [00:10] bkerensa: also your version of the charm will be available in the store at cs:~bkerensa/subway [00:10] nice [00:10] is this store web-based or does it ship in a gui or something? [00:10] bkerensa: we just want to keep the default set extremely high quality.. which means you have to seek our sponsorship for changes, or commit to being a full "charmer" [00:11] bkerensa: the store is just a backend service right now. Its undergoing live deployment right now. [00:11] bkerensa: notice if you type 'juju deploy subway' right now you get a DNS error ... [00:11] SpamapS: ahh yeah [00:12] bkerensa: there's also a readonly charm browser here http://charms.kapilt.com/~bkerensa/oneiric/subway [00:14] SpamapS: pretty neat browser [00:15] hazmat: ^^ another fan of your work. :) [00:31] SpamapS, its also available now at jujucharms.com [00:31] SpamapS, just put that together today after a request jcastro [00:32] sweeeeeet [00:34] hrm.. [00:34] there seems to be a bug in "broken" hooks [00:34] SpamapS, ? [00:34] http://paste.ubuntu.com/897196/ [00:35] thats while inside db-relation-broken [00:35] happens not in debug-hooks as well [00:35] SpamapS, what version? [00:35] we put a broken hook context into place [00:35] ii juju 0.5+bzr494-1juju4~oneir next generation service orchestration system [00:35] ugh [00:35] oh.. that is the departed hook context [00:35] hazmat: breaks the mysql charm in an interesting way :-/ [00:36] SpamapS, so what would you expect it to tell you [00:36] SpamapS, the relation has been broken, its not coming back [00:36] there are effectively no members of the other side [00:36] so relation-list returns what? [00:36] hazmat: I was using it to hack around bug #791042 [00:36] <_mup_> Bug #791042: *-relation-broken has no way to identify which remote service is being broken < https://launchpad.net/bugs/791042 > [00:37] SpamapS, ah so you wanted to get the remote service name.. [00:37] right, I'm sure there are other ways to do what I want, which is to note that it is broken so the next time around certain things happen [00:37] hmm. we're doing some work right now to get relation-ids attached to relations.. and a relation-ids command.. to do relation cli api from any hook.. [00:37] I can probably just use more predictable usernames, and remove the username.. and that will achieve what I want. [00:38] that would still be available in a broken context, but its rather bespoke as a resource identifier [00:38] hazmat: as long as its the same during joined/changed/departed and broken :) [00:38] it would be [00:38] I can use it as the username/pass [00:38] it would look like db:42 [00:38] well not the pass [00:38] but the username [00:39] hazmat: so relation-list is just not allowed in broken? [00:40] hazmat: also I still get '{}' when using juju-log in the local provider. Did we ever figure that out? [00:41] SpamapS, yeah.. i did a have quick look at that one.. but its low on priority list to fix [00:42] SpamapS, i think it should just return an empty list if invoked during broken [00:42] its debatable [00:42] its effectively useless then, and at least the error informs somewhat [00:43] SpamapS, re juju-log its effectively printing its return value.. re the {} [00:43] ah [00:44] anyways.. family time.. ttyl [00:44] hazmat: have a good weekend! [00:44] SpamapS, you too! === jelmer_ is now known as jelmer [01:29] <_mup_> juju/relation-ids-command r501 committed by jim.baker@canonical.com [01:29] <_mup_> Fixed tests [01:35] <_mup_> juju/relation-id-option r499 committed by jim.baker@canonical.com [01:35] <_mup_> Merge upstream === hspencer is now known as hspencer[afk] [02:47] <_mup_> juju/relation-ids-command r502 committed by jim.baker@canonical.com [02:47] <_mup_> JUJU_RELATION default [03:03] <_mup_> juju/relation-id r493 committed by jim.baker@canonical.com [03:03] <_mup_> Rework internal relation id parsing [03:05] <_mup_> juju/relation-hook-context r500 committed by jim.baker@canonical.com [03:05] <_mup_> Docstrings [03:06] <_mup_> juju/relation-id r494 committed by jim.baker@canonical.com [03:06] <_mup_> Merged trunk [03:06] <_mup_> juju/relation-hook-context r501 committed by jim.baker@canonical.com [03:06] <_mup_> Merged upstream [03:07] <_mup_> juju/relation-ids-command r503 committed by jim.baker@canonical.com [03:07] <_mup_> Merged upstream [03:10] <_mup_> juju/relation-ids-command r504 committed by jim.baker@canonical.com [03:10] <_mup_> Docstring [03:23] <_mup_> Bug #963640 was filed: Implement external relation id support < https://launchpad.net/bugs/963640 > [03:24] <_mup_> Bug #963642 was filed: Implement child relation hook context with lookup by relation id < https://launchpad.net/bugs/963642 > [03:26] <_mup_> Bug #963643 was filed: Implement relation-ids hook command < https://launchpad.net/bugs/963643 > [06:19] * SpamapS pokes at the mysql charm === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [09:55] jimbaker, rock the house === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [14:36] <_mup_> Bug #963931 was filed: possibly non-deterministic test juju.unit.tests.test_lifecycle.UnitRelationLifecycleTest.test_join_hook_error < https://launchpad.net/bugs/963931 > === mars is now known as Guest83177 [17:32] I'm trying to set up a local provider, How much time should it spend on 'Creating master container...' on a decent hardware but a slow connection? Is it trying to download something? [18:38] How can I find out whats going in inside lxc when juju debug-log says 'Creating master container...'? [21:16] doh.. I just discovered help2man [21:16] and it works perfectly w/ argparse generated output [21:16] * SpamapS does a little dance [21:16] yay, manpages for juju! [21:20] nice [21:20] \o/ [21:20] * imbrandon is still trying to fight with it on OSX [21:20] ok gas station , brb [21:20] SpamapS: btw mentioned you in my blogpost , didnt think youd mind given the context [21:23] imbrandon: we're all public figures. :) [22:17] SpamapS: haha true === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [23:33] SpamapS: wanna know something else about nginx that makes it even more of a magic wand, and as i thought of it is gonna make a good case for a _only_ nginx charm [23:33] SpamapS: # server { [23:33] # listen localhost:110; [23:33] # protocol pop3; [23:33] # proxy on; [23:34] # }:) [23:34] :) [23:34] it speaks more than just http{,s} heh [23:35] often overlooked , but if you already knew my bad :) [23:45] imbrandon: epic [23:45] holy crap, just thinking about that actually [23:51] marcoceppi: heh [23:51] yea [23:52] its kinda epic and hard to wrap head arround all at once [23:52] takes a bit to sink in [23:52] but its not an after thought either, its fully suported and from the begning was there [23:52] brb afk a min