=== hazmat` is now known as hazmat [11:07] morning [11:15] hi rick_h_ [11:22] howdy frankban, welcome back. [11:22] hopefully had some fun while away? [11:24] rick_h_: thanks, yeah, I was able to relax a bit, and my grandma surgical operation went well [11:25] oh, very good to hear [11:38] rick_h_: quick chat after my lunch? [11:39] frankban: sure thing [11:39] cool thanks [11:39] * frankban lunches [11:56] rick_h_: any chance i can ask you to look at part 1 of 3 of tests for token rendering? https://github.com/juju/juju-gui/pull/328/ [12:03] ahoy monday people [12:03] jcsackett: sure thing, can in a few min [12:03] rick_h_: awesome, thanks. [12:03] morning, bac. [12:36] jcsackett: can you take a peek at huw's branch while I look at yours? [12:37] rick_h_: sure. [12:50] rick_h_: happy to chat, but i'll have to ping you when i'm back from the doc. [12:50] rick_h_: i'll say now though, that doing the "render, test output" approach is exactly what i meant when i said we *weren't* directly testing method, and instead just doing implicit testing. [12:51] if that's what we *want*, ok, but we're not unit testing. [12:51] jcsackett: yea, understand. This one I think hits the nerve a bit because all these tests are "render" tests [12:51] rick_h_: I am available when you want [12:52] I think there's a time/place for both but this one sure feels like the wrong side of it to me [12:52] frankban: cool, hangout url? [13:02] everyone have a good weekend? [13:05] redir: yes thanks. so, how is the store dependencies investigation going? After a quick look, I saw we'd have some indirect dependencies to handle (utils, schema, environs) and maybe some others. Also we'd need the testing package for testing. [13:08] frankban: Ididn't get too far. I only saw a direct dependency on juju-core/charm but didn't look at it's dependencies. [13:09] frankban: I have one last think on CW to finish so it can be migrated with ngrams. [13:10] redir: ack [13:11] also it didn't occur to me until you mentioned it that 'go list' probably wouldn't list test dependencies by default. [13:25] redir: http://pastebin.ubuntu.com/7488514/ [13:33] frankban: right and XTestImports for test deps? [13:35] redir: AFAICT tests depends mostly on the "testing" package. unfortunately it does not seems we have a Deps for tests, so we need to check each individual package. There is stuff like launchpad.net/juju-core/environs/config (for charm tests) but I guess we can fix that. [13:41] redir: http://pastebin.ubuntu.com/7488568/ [13:41] frankban: XTestImports []string // imports from XTestGoFiles lists imports from external packages AFAICT. Albeit not recursively [13:41] frankban: nice [13:55] rick_h_: should I start QA’ing the URL now? or wait for a later build? [13:56] luca: you can qa now. I'll triage stuff that's still todo and such [13:56] rick_h_: ok [13:56] rick_h_: should I just file bugs? [13:56] luca: maybe just notes in a google doc? And I can reply/note and we can create bugs from that [13:56] luca: probably less work on your end than filling out the bug form over and over [13:57] rick_h_: ok, sounds good. [14:55] jujugui call in 6 [15:00] think I have the migration test done. will commit after running fill suite [15:00] jujugui call now [15:00] * rick_h_ gets in the call [15:01] rick_h_ I hadn't realized it was 11, still not back home. [15:01] jcsackett: gotcha [15:01] Makyo: ^ [15:14] rick_h_: paste it here please [15:14] https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0ApyaFXSrLF38dGVCYUhGLTRXMXRaVXBCMUFjQlJMTkE&usp=drive_web#gid=3 [15:49] As long as that was, it's still shy of the marathon 3 hour meetings I've done in the past. [15:52] redir: do you have a branch you want me to look at? [15:52] bac: gonna make lunch then hit you up but there is a branch at ~reedobrien/es-migration [15:53] bac: if you can sanity chech that, I'd appreciate it. And i'll update the docs after lunch [15:53] redir: cool. will lunch now too. [15:53] awesome [15:53] redir: i think you mean sanity czech [15:55] kadams54 3h meetings? lol ouch [15:56] it was kind of funny that there was a 'reed' column and an 'agreed' column :D [15:57] kadams54: yea, still a sucky way to start off the week. "Yay 45min planning poker meeting" [15:57] we have too much stuff 'on deck' I need to do some more cleaning [15:57] The best part was that at the end of the 3 hours we had to schedule another planning meeting because we'd never gotten to the actual planning poker part. [15:58] lmao, that seems like a can of fail [16:00] rick_h_: i am finally back; want to chat? [16:01] jcsackett: sure thing [16:01] hangout url? [16:01] rick_h_: https://plus.google.com/hangouts/_/g3c2nla25fgahq4pm3h7w5ocdua?authuser=2&hl=en [16:32] bac I can be a real slav when it comes to typing [16:37] * rick_h_ goes to make some lunchables [17:09] jujugui: we're still doing 2 space indents, right? i've found some objects (tokens) appear to have 4 space, wondering if those should be cleaned up or not. [17:10] jcsackett: yes, 2 space indent though in line breaks in () I think the linter wants 4 [17:10] jcsackett: we just make the linter happy :P [17:11] rick_h_: yeah, in line breaks i know are 4--i've upset the linter before. :p [17:12] but this is like "var thingy = Y.base.create(..." and events line is 4 space, which i think should be 2. [17:12] i'll clean 'em up as a separate trivial branch while i'm working on them. [17:17] rick_h_: pushed the changes we chatted about in review/chat. i think between that and follow up we're set, if you want to take another look at the PR. [17:18] jcsackett: looking [17:20] jcsackett: lGTM [17:20] rick_h_: thanks. [17:22] is CI still going bananaballs? it sent a failure back "connection reset by peer" against my PR earlier that seemed to involve the last 12 branches. [17:22] looking [17:27] jujugui tiny docs branch: https://github.com/juju/juju-gui/pull/329 Currently, the icon shows unknown, which I'm looking into, but I want opinions on removing -W from SPHINXOPTS [17:31] BradCrittenden: back? [17:35] Makyo: can things sphinx warns us on bork the docs? or just result in them being kinda ugly? [17:35] i'm usually good with warnings *not* being errors, so as long as they don't include this option b/c warns can be particularly bad, i'm +1. [17:35] jcsackett, Worst case would be a broken link, I think. Things that are actual errors will still be errors and cause make docs to fail. [17:36] Makyo: seems fine to me then. [17:37] * jcsackett goes to comment as such on PR. === BradCrittenden is now known as bac [17:38] hey redir [17:38] jcsackett, thanks. Just need to see if saucelabs will ever show anything but 'unknown'. [17:39] rick_h_, do we have saucelabs creds somewhere so I can investigate this real quick? [17:39] Makyo: sec [17:41] bac: thoughts [17:41] ? [17:42] redir: oooking now [17:42] er, looking now [17:59] redir: have you tried locally spinning up and populating a version 26 of the db and then applying this migration? [18:00] bac how do I populate a version 26 of this locally? [18:00] and where are versions 1-25? [18:00] redir: oh, don't ask that. :( [18:01] bac the short answer is no [18:01] bac but if I understand the migration code, it shouldn't touch mongo [18:01] version 27 [18:01] redir: shelve your 027, blow away the ES, then load it back up. that should give you v26. [18:02] ingest some stuff. then run the 027 migration [18:02] I can do that. Are you verifying? [18:02] redir: that's true [18:02] bac& [18:02] redir: i will. [18:02] also shouldn't I then start at version 0 and ensure all migrations work [18:03] ls [18:03] oop [18:03] s [18:03] redir: i'm also confused by your 'exists_in_index'. seems overly complicated. why not just call get(id, kind) and let it handle the charm/bundle differences? [18:04] bac I borrowed that from another test [18:04] redir: it can go from 0 directly to 26. the way migrations work kind of got messed up a while back [18:05] bac understood [18:05] redir: ok. it may be required but seems too complicated to me [18:05] there are gaps in my understanding, too [18:06] bac initially I thought I'd write a few tests, but then realized there wasn't that much to test... [18:07] bac: would you like me to remove the helper? [18:08] not yet [18:13] bac: I dropped everything am re ingesting with code @ revno 511. Then will initialize db versioning, update to tip, and test migration [18:15] redir: i'm doing the same but going back to r508, since that is the version on production, which is what we care about [18:15] jujugui anyone have time off in the next 2 weeks that's not submitted they know about? [18:15] redir: see manage.jujucharms.com/heartbeat to check the revno [18:15] rick_h_: i'm comtemplating either this friday or next tuesday. will decide rsn [18:15] bac: thanks will revert to 508 [18:16] bac: rgr thanks [18:17] rick_h_: pretty sure you saw my moving swap day from may 9th to this coming friday; also memorial day which i'm not sure i filed yet. [18:17] jcsackett: rgr, yea got the holiday on everyone [18:17] rick_h_, Part of Friday the 30th for closing. [18:17] Makyo: cool thx [18:17] Like, two hours in the afternoon, closing is at 4, EOD is at 5. [18:17] Makyo: understood [18:18] Makyo: congrats. hadn't heard it was all going through [18:18] just trying to pick out how many cards to assign over [18:19] bac, thanks. Things got pushed up by the seller. [18:21] redir: i completed a 026 - 027 migration and it all went well. you want to do a hangout to discuss the steps? it's kind of non-intuitive [18:22] I shouldn't follow intuition if it isn't intuitive so yes, I think that sounds like a splendid idea:) [18:22] redir: can you spin one up and invite me? [18:22] bac sure [18:22] rick_h_: you mean in the team calendar or in the canonical admin thing? [18:22] redir: well more that I'm not aware of :) [18:23] redir: which is mostly canonicaladmin [18:23] rick_h_: OK then quite possibly [18:23] but it is in the team calendar [18:23] rick_h_: I'll follow up after chatting with bac [18:28] redir: okie dokie. I'm free whenever you're ready [18:46] so...i should probably have asked this weeks ago...but can anyone explain why shipit is a squirrel in a hat? [18:47] Github ¯\_(ツ)_/¯ [18:47] lol, much better answer than mine [18:47] * jcsackett laughs [18:49] yay kanban board is updated and ready for 2wk cycle again. [18:49] Feels good, like cleaning up that messy garage that's been fugly for too long [18:51] jujugui pausing the saucelabs indicator card, since they don't seem to show anything but unknown for us without further config. Any suggestions on a next card? [18:53] Makyo: the world is our oyster? ghost config, new machine UI controls, the bug around the relation lines in maint. [18:53] Aha, sounds good. Will do. [18:54] oh hmm, should have subtracted some card points for today. Oh well, we're a bit over subscribed [18:54] Makyo: your expenses and done right? Just catching up on the misc end of the board [18:55] Oops, yeah, sorry, those are done. [18:58] rick_h_: prolly after your call at this point [18:59] redir: I'm all good, my call is cancelled [18:59] redir: so whenever you're free [19:00] rick_h_: OK gimme 10 [19:08] rick_h_: ring [19:47] Makyo: did you see the email from Menno regarding unit status? [19:48] Oops, no wonder email was quiet, 2FA time. [19:48] Checking now. [19:48] Makyo: cool, it's from last night. If you get a few min can you read that over and make sure we're cool and maybe reply back. [19:48] Sure thing [19:48] Makyo: and if we need to build up any todo's we get that in [19:48] Makyo: ty much [19:51] * rick_h_ runs away. Have a good day everyone. [19:51] later rick_h_ [19:51] bac yt? [19:51] redir: yes [19:52] thoughts on where to put migration helper files [19:52] ? [19:52] bac: I dumped the pre 027 mapping. so we can load it in the tests [19:53] redir: in a tests/data (or similar) subdir? [19:53] cw.migrations.versions.tests.data ? [19:53] yes. [19:54] k [19:54] tx [20:21] bac, what exception was raised on staging? [20:21] NumberParsingFooThingy [20:22] that was on qa.m.j.c -- haven't rolled it to staging yet. [20:22] from __magic_errors__ import NumberParsingFooThingy [20:22] it is an ES exception [20:23] how do I look at the specifics:) [20:23] not sure where qa.m.j.c is [20:36] redir: actually NumberFormatException [20:36] wha [20:36] qa.manage.jujucharms.com is the system we maintain on canonistack [20:37] bac is there a traceback? [20:37] I have a test and it doesn't reproduce [20:37] redir: no, this is all from memorty from the failed deploy last week [20:37] s/memorty/memory/ [20:38] bac, not sure how to reproduce then. I can put the mapping right over the old one and search and it still works [20:38] i can try to reproduce it by going back to r508, resetting the version to 026, and going forward [20:38] bac that wouldbe cool I think [20:38] redir: it may be data driven [20:38] I don't know what I am trying to reproduce [20:38] redir: i mean, there are charms that cause it to be thrown [20:47] redir: success (in failing) http://paste.ubuntu.com/7490215/ [21:39] jcsackett: did you get anywhere with huw's branch? [21:39] rick_h_: i left comments; did you want me to fixup and land it? [21:40] it only needs one other test, which i can absolutely add and land this evening. [21:40] jcsackett: ah missed the comments in my email today [21:40] jcsackett: did you give it a +1 otherwise so he can land it after the test? [21:40] rick_h_: no, but i will now. [21:40] jcsackett: awesome thanks [21:41] Makyo: the pull request with the icon, should we close then then since that turned into paused slack? [21:59] rick_h_, sorry, was dogwalking. Yes, I say close it. It's going to take more work than is worth for a slack task right now. [23:06] Morning [23:18] morning huwshimi [23:21] rick_h_: The card "Wire existing containers and machines into the unit token." has been done unless I misunderstand the card. [23:22] hi huwshimi [23:22] bac: Hey :) [23:22] * bac not really here, just saying hi [23:22] huwshimi: it needs to be live updating and should be selectable when you click on the placement button [23:23] huwshimi: clicking on the placement icon in an unplaced unit token doesnt' do anything atm [23:23] huwshimi: there's some html there, but it's not in final form [23:23] rick_h_: Oh, I see, it's the placement that needs to be done. Nevermind me :) [23:24] yea, we've concentrated on drag-n-drop, but now we need to start looking at the click/form/drop down UI [23:32] bac: that is totally different than NumberFormatException. [23:33] But I have a test for MapperParsingExceptionError now [23:36] of course I feel like this now: http://bit.ly/TnzYrY