UbuntuIRC / 2012 /12 /04 /#juju-gui.txt
niansa
Initial commit
4aa5fce
raw
history blame
13 kB
=== Makyo is now known as Makyo|out
=== rogpeppe1 is now known as rogpeppe
=== mattuk1972_ is now known as mattuk1972
[13:14] <gary_poster> teknico, how is your card going?
[13:17] <gary_poster> benji, should I review your branch or wait until you have resolved the issue Makyo raised?
[13:18] <gary_poster> would it be useful to see if I can duplicate his issue?
[13:18] <benji> gary_poster: I don't have a strong opinion either way.
[13:18] <benji> it couldn't hurt :)
[13:18] <gary_poster> benji, ok, I'll see if I dupe. If I do, I'll wait on its resolution before a full review.
[13:19] <benji> k
[13:23] <hazmat> gary_poster, meeting this am is a one time evt
[13:24] <gary_poster> benji, duped. Also, make debug, at least, seems worse in that it will repeatedly try to make files that should already have been made, such as sprite file. trunk is quiet.
[13:24] <benji> hmm
[13:24] <benji> ok, thanks
[13:25] <gary_poster> hazmat ack thanks. 10:30 will normally work fine, just not this week?
[13:27] <hazmat> gary_poster, just not today
[13:27] <gary_poster> cool hazmat thanks. will adjust, and reply to mail
[14:10] <bac> hi mattuk1972 would you mind taking a look at the visuals for the charm store filter widget to ensure it looks like you expect?
[14:26] <mattuk1972> bad, sure in a meeting - will be out in a few
[14:26] <mattuk1972> bac,
[14:50] <bac> benji: second review done.
[14:50] <benji> bac: thanks
[14:58] <mattuk1972> bad, could i see it without the 1px stroke? it seems too much here -also the text and tick need some padding to bring them in a bit. basically the same as the text has before you click in it to open the chooser.
[15:01] <bac> mattuk1972: you can play with the CSS in your browser, as we did before
[15:02] <bac> mattuk1972: i had trouble placing that tick mark. i know it is poorly placed but i cannot figure out how to move it in. the CSS that should do the trick is ineffective.
[15:03] <mattuk1972> bac, yep it doesn't seem to line up with the text either
[15:03] <mattuk1972> should i fond another way to show selected?
[15:03] <bac> nope
[15:10] <bac> mattuk1972: my "nope" reply was to your previous message. as to finding another way of marking the selected item that may be a good idea but i'd wait until i've exhausted all attempts at getting the check mark to place correctly.
[15:14] <goodspud> gary_poster, could you please send me a new invite to our meetings
[15:16] <goodspud> And how do I sign up to our mailing lists (like juju-gui peeps)?
[15:17] <gary_poster> goodspud, mailing list: please join https://lists.ubuntu.com/mailman/listinfo/juju-gui
[15:17] <gary_poster> we are trying to dump peeps
[15:18] <gary_poster> goodspud, reinvited you to mtg
[15:50] <goodspud> Thanks gary_poster
[15:50] <gary_poster> welcome
[15:52] <goodspud> ... although i haven't received the meeting invite email yet....
[15:56] <gary_poster> goodspud, don't know. deleted and added you. what do you need? should be on your calendar
[15:57] <goodspud> Nope, not on my calendar. I figured I'd get an email with the details... stoopid technology
[15:58] <gary_poster> bac bcsaller benji frankban hazmat Makyo mattuk1972 teknico call in 2, unless hazmat and my call goes over...
[15:58] <gary_poster> goodspud, I used [email protected]
[15:58] <gary_poster> another?
[15:59] <goodspud> That should work... perhaps the mail server is super slow. I'll have some patience
[15:59] <gary_poster> ok
[15:59] <gary_poster> looks like call is going over...?
[16:00] <gary_poster> yup
[16:00] <gary_poster> sorry
[16:00] <gary_poster> don't know how much yet
[16:01] <gary_poster> bac are you willing to run the meeting today?
[16:03] <bac> gary_poster: doing so
[16:04] <gary_poster> thank you
[16:11] <bac> Makyo i'll review your endpoints card. i added my tag
[16:11] <Makyo> bac, thanks.
[16:12] <hazmat> gary_poster, reap the whirlwind :-)
[16:12] <gary_poster> bcsaller, ready in 10 if you want to pair
[16:13] <benji> bcsaller: if you want to see what I was talking about in the standup, take a peek at line 50 of lib/server.js in trunk
[16:13] <gary_poster> hazmat it seems like a whirlwind certainly :-)
[16:13] <bcsaller> benji: I'll take a look
[16:14] * gary_poster restarts
=== gary_poster is now known as gary_poster|away
[16:15] <bcsaller> benji: that is insane, its under assets as well which isn't even the right place, so rather than fix a relative path issue there was a dynamic workaround it looks like
[16:18] <goodspud> bcsaller, just as an aside, what part of the world are you in at the moment?
[16:19] <bcsaller> goodspud: refactoring the frontpage view to better support the upcoming features and to allow more of us to work on it at once
[16:21] <bcsaller> goodspud: unless you mean geography, in which case NYC
[16:22] <goodspud> bcsaller, geographically.... but good to know what you are working on anyway
[16:22] <goodspud> :)
[16:22] <goodspud> bscaller, just seemed like you'd moved from San Fran... it was bright and sunny out your window
[16:23] <bcsaller> goodspud: today it is. I'm only here for a while, maybe another couple months
=== gary_poster|away is now known as gary_poster
[16:26] <gary_poster> bcsaller, I lied. Decided to make some gazpacho for a bit. back soon. will ping
[16:26] <bcsaller> gary_poster: np
[16:28] <bac> umm, the cold soup of spain
[16:29] * bac heads to tapas joint
[16:32] <teknico> gary_poster, I'd like some, thanks :-)
[16:51] <gary_poster> :-)
[17:05] * bcsaller grabs some food, back in a few
[17:37] <benji> Makyo: my branch should be much saner now, I would love it if you would try it again
[17:50] <benji> bac: I suspect my branch does most of what you want for your card; I would be happy to discuss it if you like
[17:50] <bac> benji: i figured it might
[17:51] <benji> gary_poster: my branch is ready for your review (https://codereview.appspot.com/6865054/)
[17:51] <gary_poster> benji, don't you already have two reviewers? I'm OK with it if they are, though I'm happy to look if you'd liek me to.
[17:51] <gary_poster> or even like
[17:52] <benji> gary_poster: yep I have two, so unless you are super excited about reviewing it, no need
[17:52] <bac> benji: is your branch merged?
[17:52] <benji> bac: if lbox does it, then yes... I thought it did
[17:53] <benji> "No new revisions or tags to push."
[17:53] <gary_poster> benji, not super excited :-)
[17:53] <benji> :)
[18:00] <Makyo> benji, I'll forever wonder about that regex, always, in the back of my mind..
[18:00] <benji> Inflicting obscure psycological pain is my greatest joy in life.
[18:03] <bac> benji: the rule for $(BUILD_ASSETS_DIR)/images fires everytime 'make build' is used, whether it needs to or not. i can't figure out why
[18:03] * benji looks
[18:04] <benji> bac: did you uncomment the debug helper at the top of the Makefile? It can help a little.
[18:04] <bac> benji: nope.
[18:04] <benji> bac: I don't see that behavior, do you want to pair on it?
[18:05] <bac> sure
[18:05] <benji> bac: you get tb ready and I'll make a hangout
[18:07] <benji> bac: https://plus.google.com/hangouts/_/15f7b6bcb0736024fb1834d49ca7c6acf56a4080?authuser=0&hl=en-US
[18:17] <hazmat> hmm.. the charm readme covers testing, but really it should be talking usage for a user
[18:20] <hazmat> also
[18:20] <hazmat> no mechanism to change config after deploy
[18:27] <Makyo> hazmat, should I create a config-changed hook card?
[18:27] <hazmat> Makyo, sounds good
[18:27] <Makyo> hazmat, can do a README card, too, while I'm at it.
[18:32] <Makyo> hazmat, for the README, should the testing information be moved to a different file, or just kept at the end of the current one?
[18:34] <gary_poster> The config changed is a good idea, but also a nice to have IMO: I would make a crad but put a low priority. README for users is a normal priority one
[18:36] <hazmat> Makyo, separate file.
[18:37] <hazmat> gary_poster, requiring config at deploy time is a bad semantic for a charm, its more than a nice to have imo
[18:37] <hazmat> Makyo, i'd suggest DEVELOPER.txt
[18:38] <hazmat> gary_poster, we have several folks wanting to do a demo of it.. the deploy process is already convulted enough, having to switch out defaults at deploy time is kinda of painful to..
[18:39] <hazmat> this is my current instructions on deploy fwiw http://pastebin.ubuntu.com/1410842/
[18:39] <gary_poster> Technically, it's a nice to have. It will still work fine without it, and that's the way I've worked with other charms--with a config file. But eh.
[18:40] <hazmat> for a staging env its also echo / cat to config.yaml and pass along for deploy.. much nicer if we can juju set staging=true
[18:40] <hazmat> gary_poster, i consider charms requiring a config file to generally be broken
[18:40] <hazmat> er.. badly done let's say
[18:40] * gary_poster not going further with it :-)
[18:41] <hazmat> it means they take a bunch of config options which they don't accept
[18:42] * hazmat takes a shallow breath and moves on
[20:52] <Makyo> Anyone else unable to get to the board?
[20:53] <bac> Makyo: yes, i cannot connect to the server either
[20:54] <bac> Makyo: https://leankitkanban.com is gone too
[20:57] <bac> from twitter: LeanKit ‏@LeanKit
[20:57] <bac> We have an unexpected outage at the moment. The team is diagnosing. Resolution ASAP.
[20:57] <Makyo> bac, ah, thanks! Not urgent, just realized I should move cards.
[20:57] <bac> all of this happened an hour after they announced their new Chief Marketing Officer. coincidence?
[21:00] <Makyo> Maybe they hit the big red Door Open button in the server room?
[21:02] <benji> heh
[21:03] <hazmat> having odd issues with the charm http://pastebin.ubuntu.com/1411132/
[21:05] <benji> hazmat: I touched those rules recently, but I didn't have any problem with them; looking at it again to be sure
[21:08] * bac gives in to frantic dog. bbiab.
[21:11] <benji> hazmat: I can't reproduce the problem
[21:21] <hazmat> benji, try it with jitsu deploy-to 0
[21:21] <benji> hazmat: try what? the charm? I was just running the make file; I haven't deployed the charm before
[21:38] <hazmat> benji, the charm yeah..
[21:38] <Makyo> hazmat, the GUI dir is chowned in hooks/start, but make is run in hooks/install. I've not run into the problem with juju deploy, but maybe jitsu deploy-to is calling it differently?
[21:38] <hazmat> Makyo, the hooks get called the same.. the deploy is the same .. its just different on the front end cli
[21:41] <Makyo> hazmat, ah, alright. Curious that it's a permission denied thing, since it's in the same hook that's creating the directory.
[21:52] <hazmat> hmm.. these hooks are idempotent
[21:53] <hazmat> er. aren't
[21:53] <hazmat> can't rerun it to resolve, as it will barf on the checkout
[21:54] <hazmat> full charm log fwiw.. bottom is a resolved --retry ... http://paste.ubuntu.com/1411248/
[21:57] <hazmat> Makyo, good point.. i think i foobar'd it running by hand.. it looks like its needs root..
[21:57] <hazmat> aha.. the actual error..
[21:57] <hazmat> http://paste.ubuntu.com/1411255/
[21:58] <Makyo> Though the 'ln -sf ....yui' bit succeeded this time.
[22:00] <hazmat> Makyo, yeah.. i was running it as a normal user instead of as root with the first pastebin
[22:00] <Makyo> hazmat, oops, I see.
[22:02] <Makyo> Oh, huh. ln succeeds if the target exists, and creates a bad link.
[22:02] <Makyo> Suppose that...sort of makes sense. Can make the link, create the file later.
=== gary_poster|away is now known as gary_poster
[22:04] <gary_poster> bcsaller, worth checking up now, or do it tomorrow?\
[22:04] <bcsaller> gary_poster: lets do it tomorrow
[22:04] <gary_poster> cool bcsaller. have a good evening
[22:05] <bcsaller> you too
[22:13] <hazmat> benji, i think we need to revert that last commit..
[22:13] <hazmat> gary_poster, it works on 262
[22:15] <gary_poster> hey benji. I asked hazmat to revert. Please make a card for getting that landed again and move it through as "critical"
[22:15] <hazmat> benji, fwiw the error is http://paste.ubuntu.com/1411255/
[22:16] * gary_poster stepping away
[22:16] <gary_poster> have a good night
[22:16] <benji> gary_poster: will do
[22:18] <bac> thanks for the review benji
[22:19] <benji> hazmat: if you feel so inclined, I would appreciate testing the pre-reverted code with just this bit reverted in the Makefile: http://paste.ubuntu.com/1411289/
[22:24] <hazmat> benji trying out
[22:26] <hazmat> unrelated.. the makefile doing a bzr file check outside of a target is actually quite painful when there is no local bzr repo.. the charm uses a lightweight co.. and its a multi-second pause
[22:26] <hazmat> and of course wouldn't play nice with a tarball
[22:44] <hazmat> benji, those diff chunks don't match up
[22:45] <hazmat> hmm... or i already applied..
[22:46] <hazmat> yeah.. no diff to revno 264
[22:49] <hazmat> benji, perhaps you mean the inverse of the diff you pasted
[22:55] <hazmat> no help with the inverse either.. reverting