/srv/irclogs.ubuntu.com/2012/07/23/#launchpad-dev.txt

StevenKI thought I could click on 'Private: Some' on a +sharing page to remove access. :-(00:42
wgrantStevenK: You can00:48
StevenKwgrant: Except it doesn't work on qas00:48
wgrantStevenK: Hm, but the FF is enabled...00:50
wgrantOh00:50
wgrantStevenK: You're trying on launchpad?00:50
wgrantflacoste owns that on qas, rather than ~launchpad00:50
wgrantSo you can't edit it00:50
StevenKAh00:50
wgrantwebops: could you please reassign https://launchpad.net/launchpad from ~flacoste to ~launchpad?00:51
wgrantBah00:51
wgranthttps://qastaging.launchpad.net/launchpad00:51
StevenKHahahaha00:51
hloeungwgrant: "No items matched "~launchpad"00:52
wgranthloeung: Drop the ~00:52
hloeungwgrant: done00:53
wgrantIndeed, thanks.00:53
wgrantStevenK: Try again?00:53
StevenKRight, now it has edit icons and such00:53
StevenKBefore, the text were links, which gave the impression00:54
StevenKwallyworld: You haz QA01:00
StevenKNot that it matters01:01
wallyworldyes, will do it soon01:01
StevenKwallyworld, wgrant: Either of you want to review https://code.launchpad.net/~stevenk/launchpad/auditor-for-packageupload/+merge/116180 ?01:54
wgrantStevenK: ಠ_ಠ ಠ_ಠ02:05
wgrantAuditorClient.recieve is awful02:06
wgrantOne doesn't compose query strings using string formatting operations.02:06
wgrantAlso, consider using lazr.uri to compose the rest of the URL02:06
wgrant>>> u = lazr.uri.URI(scheme='http', host='foo', port=1234, path='/')02:07
wgrant>>> u.append('log/').replace(query='somequery')02:07
wgrantURI('http://foo:1234/log/?somequery')02:07
wgrantAlso, datetime.now() is local time, which is wrong02:08
StevenKwgrant: I didn't know about lazr.uri, but it doesn't seem like it will help much, given the URLs for auditor are simple.02:10
StevenKwgrant: If not string formatting operations, then what?02:10
wgrant>>> urllib.urlencode([('correct=encoding', 'is&good?')])02:11
wgrant'correct%3Dencoding=is%26good%3F'02:11
StevenKwgrant: http://pastebin.ubuntu.com/1105753/02:16
wgrantImproved.02:17
StevenKBut still not good? :-P02:17
StevenKwgrant: Diff updated, makes use of urlencode rather than '&'.join() and datetime.utcnow()02:29
wgrantStevenK: Have you discussed all this with lifeless?02:46
StevenKwgrant: He just gave his 2 cents on the MP itself02:57
StevenKSo I'll be refactoring after lunch02:57
StevenKwgrant: But thanks for the hints about making use of urlencode.03:01
wgrantStevenK: In general if you're using string formatting to produce a machine-readable string in a format that you didn't design yourself you probably have a terrible bug03:02
lifelessintel should quit the wifi chip business.03:11
StevenKOr perhaps you should switch to Ethernet03:16
lifelessyeah, no.03:19
lifelessI'd rather whinge than fix the problem :)03:19
lifelessActually, cables around home are problematic when you have a 11month old.03:20
lifeless+2 cats with cable fethishes.03:20
StevenKHaha03:20
StevenKYes, I can remember having a cat with a cable fetish03:20
lifelessAnyhow, I see my name in backscroll.03:20
StevenKPretty much every cable that the TV/DVD player/PS2 used had bite marks/exposed wiring on it03:21
StevenKExcept for power cables, she never bit those for some reason03:21
lifelessso ...03:27
lifelesswhats 'this' under03:27
lifeless14:29 < StevenK> wgrant: Diff updated, makes use of urlencode rather than '&'.join() and datetime.utcnow()03:27
lifeless14:46 < wgrant> StevenK: Have you discussed all this with lifeless?03:27
lifelessStevenK: ?03:38
StevenK[12:57] < StevenK> wgrant: He just gave his 2 cents on the MP itself03:50
StevenK[12:57] < StevenK> So I'll be refactoring after lunch03:50
StevenKlifeless: ^03:50
wgrantwallyworld: What do you feel is the best way to customise vocab content? Sometimes we pass a context into the vocab constructor, but a couple of weeks ago I used a different method (constructing a SimpleVocabulary using a subset of the terms from the original vocab) in BranchEditView.setUpWidgets, since it's more a case of the view knowing which information types to show.03:53
wgrantI'm not sure it's great to push that knowledge down to the vocab itself03:54
wgrantSince InformationTypeVocabulary really shouldn't know about branches or bugs03:54
wgrantBut I want others' opinions.03:54
wallyworldwgrant: hmmm. the way i've used it based purely on the context03:54
wallyworldbut i've also seen a SImpleVocab constructed03:54
wgrantwallyworld: The context being the pillar?03:54
wallyworldyes03:55
wgrantThat works sometimes03:55
wgrantBut then you have artifact-specific restrictions03:55
wgranteg. a branch can't change to Private Security if it's stacked on a Proprietary branch03:55
wgrantA bug can't be Proprietary if it affects 10 projects03:55
wallyworldso we need to do the construction at the point where we have the info to make a decision03:55
wallyworldso maybe a method on the moel03:56
wallyworldwhich the view can call to get the vocabl to use03:56
wallyworldor even a method on the service03:56
wgrantYeah, that's sort of the design I went for with Branch.information_type03:56
wgrantThere's Branch.getAllowedInformationTypes03:56
wallyworldyes, that sounds like something reasonable03:56
wgrantAnd BranchEditView.getInformationTypesToShow reduces that further03:56
wgrantAnd BranchEditView.setUpWidgets turns getInformationTypesToShow into a vocab03:57
wgrantIt seems to work pretty well03:57
wgrantBut I need to do something similar for bugs now03:57
wallyworldso the view is responsible for presentation logic, the model for model logic etc03:57
wgrantYep03:57
StevenKwgrant: FTR, I've never liked the pattern of teaching the vocab stuff03:58
wgrantgetAllowedInformationTypes is used by transitionToInformationType to verify that the type is legal03:58
wallyworldand the model/service apis can also be used to validate for when called over the ws api03:58
wgrantgetInformationTypesToShow filters that list down to remove unlikely types from the UI03:58
wgrantI mostly just need to work out how to integrate this with the JS choicesources, but I think it should be easy enough03:58
wallyworldfor the choicesources, just put stuff in the json cache03:59
wgrantYeah03:59
wallyworldand then js can construct it03:59
wgrantI'm just hoping they don't rely on having a registered vocab03:59
wgrantBut I don't think they do03:59
wallyworldthe choice sources widget?03:59
wallyworldno03:59
wgrantYeah03:59
wallyworldyou just pass an array of items04:00
wgrantAh, InformationTypeVocabulary isn't even registered in the first place, excellent.04:00
StevenKwgrant: If you want to rip out the context stuff out of that vocab you would be my hero04:00
lifelessStevenK: Ah. So I didn't see that - see under iwlwifi fail :)04:01
lifelessStevenK: As for the MP, I didn't review it, just glanced at it and saw the client was inlined which seemed strange to me.04:02
StevenKlifeless: Yeah, and I'm refactoring it out again04:02
StevenKWell, half of it out04:02
StevenKSome of it is LP specific04:02
lifelessthe config bits?04:03
lifelessthey are sure04:03
StevenKlifeless: And object_to_enterpriseid() and object_to_enterpriseid() calls04:04
StevenKSigh, enterpriseid_to_object() even04:04
lifelessright, thats a layer thing though I suspect ;)04:04
wgrantI guess vocabs really want a filter() method that returns a subset.04:06
StevenKTypeError: must be type, not classobj you make me very sad06:24
wgrantYou fail at new-style classes.06:26
StevenKObviously06:28
StevenKwgrant: I thought it was just __metaclass__ = type and move on06:30
wgrantYes.06:30
StevenKHmmm, the egg didn't change.06:31
=== almaisan-away is now known as al-maisan
adeuringgood mornig08:04
czajkowskiStevenK: morning any reason why your branch has never landed? https://bugs.launchpad.net/launchpad/+bug/79053509:17
_mup_Bug #790535: Distribution:+ppas timeouts <timeout> <Launchpad itself:Triaged> < https://launchpad.net/bugs/790535 >09:17
rick_h_ccccccbgjgvchvctjfjjeluutgulchrvkjjhtftctfcc09:55
rick_h_ccccccbgjgvcugtgelehbubljkrlledibbfhgrjulcdd09:55
rick_h_ccccccbgjgvccbtrielunjgkdgvkhldctjkdvbkbfjhl09:55
rick_h_bah, sorry09:56
StevenKczajkowski: Because it's not the right fix.10:43
czajkowskirick_h_: I know the feeling10:43
rick_h_czajkowski: monday monday10:44
StevenKI think the fix is to delete Distribution:+ppas for being a crime against humanity.10:44
czajkowskiStevenK: well possibly10:45
rick_h_adeuring: morning, did you get a chance to catch up on the branch from thurs/friday?10:54
adeuringrick_h_: ummm. somewhat.10:55
rick_h_it's listed as needing qa and deryck rolled it back in 1566110:55
rick_h_I really want to try to get the report a bug link landed, 15657, so wanted to bug you on that to see if we can clear it up to try to get a NDT10:56
adeuringrick_h_: in that case, I'd prefer a rollback. The problem with the buildbot failure: the test call apply_async() -- when celeryd is at least suipposed not to run -- and a subsequent check if there is a message for the task queued fails: No messages at all.10:57
rick_h_adeuring: right, the rollback is in the queue from deryck on friday10:58
adeuringi habe no clue how this can happen...10:58
rick_h_but is it ok then to just qa yours as qa bad for now?10:58
rick_h_and try to get through to the rollback for a NDT?10:58
adeuringrick_h_no, deryck just disabled the test10:58
rick_h_ah, doh ok then10:58
rick_h_sorry, didn't realize that.10:59
=== matsubara is now known as matsubara-afk
wgrantrick_h_, adeuring: Bug #102710311:39
_mup_Bug #1027103: TestRunMissingJobs.test_run_missing_ready_does_not_return_results disabled due to spurious failure <spurious-test-failure> <test-system> <Launchpad itself:Triaged> < https://launchpad.net/bugs/1027103 >11:39
wgrantPlease revert ASAP if it needs it11:39
rick_h_wgrant: just saw the MP email on that one11:39
rick_h_wgrant: so in process11:39
=== bac changed the topic of #launchpad-dev to: http://dev.launchpad.net/‹ | On call reviewer: - | Firefighting: - | Critical bugs: 4.0*10^2
=== benji changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: benji | Firefighting: - | Critical bugs: 4.0*10^2
benjigary_poster: coming12:11
wgrantadeuring: Hm, did the rollback get bounced by PQM?12:39
adeuringwgrant: its running thourgh ec2 right now.12:39
wgrantadeuring: Oh, no point or time for that12:39
wgrantlp-land12:39
wgrantWe can't deploy until it's landed, so even if there was a chance of failure we still lose nothing by landing it now.12:40
adeuringok, i'll try it...12:40
deryckMorning, all.13:01
czajkowskideryck: ello13:06
deryckadeuring, https://plus.google.com/hangouts/_/ceb51247bd8993e84395ae3e71619d1a725b8063?authuser=0&hl=en13:33
=== slank` is now known as slank
=== slank is now known as Guest49605
deryckwgrant, hi there.  Was there something actually qa-bad about adeuring's branch or are you reverting as a matter of conscious?  or to get the deploy going, or something else?13:49
wgrantderyck: Ah, adeuring said here that it should be reverted.13:50
deryckwgrant, ah, ok.  Fair enough.  Thanks!13:50
czajkowskigary_poster: what make/model headset do you have btw ?13:52
gary_posterczajkowski, Sennheiser...um...looking...13:52
gary_posterczajkowski, not sure, but Sennheiser PC 151 looks like newer version (mine does not have noise canceling microphone but looks very similar and the price of about $50 USD seems to match my memory)13:55
czajkowskigary_poster: cheers13:56
czajkowskiadding to my wishlist13:56
gary_poster:-) cool13:56
czajkowskigary_poster: you're very clear on calls13:56
gary_posteryay czajkowski :-)13:57
mgzhe also enunciates well which helps no doubt13:57
gary_posterheh13:58
czajkowskithis is rather true, I however speak way too fast and live over a flight path, not a good combination13:58
gary_posterit's all the many years of singing training13:58
wgrantOver a flight path? Impressive.13:58
gary_posterheh13:58
* czajkowski peers at wgrant 13:58
czajkowskiwgrant: lots of planes fly over the house reguarly, hows that mr. pedant :)13:59
gary_posterlol13:59
wgrantAw13:59
wgrantAnd there I was thinking that it was cool that you lived in the stratosphere.14:00
czajkowskiwgrant: one day we're going to meet and I'm going to be in throwing distance of you!14:01
gary_posterhe moves quickly14:01
gary_posterprobably can dodge well14:01
gary_posterno actual experience from throwing end though14:01
=== al-maisan is now known as almaisan-away
Bert_2Hi, I'm editing all the launchpad templatefiles to comply with the requirements of canonical I change all icons and the name Launchpad, now am I allowed to have launchpad in the css and new imagefilenames, or do I need to change that as well ?14:33
Bert_2en thandige is datk ook ni kan zien of een file edits nodig heeft of ni omda zo goed als elke file i18n:domain="launchpad" bevat14:34
Bert_2sorry, wrong channel for that last (dutch) message14:34
deryckBert_2, I don't honestly know, but we could find out.  I'm sure the intent is to protect branding of the launchpad site.  So I don't think anyone will care if this is css class names and such internal to the site.14:46
czajkowskimrevell: do yo happen to know re Bert_2 request14:47
mrevellI expect deryck is correct.14:48
mrevellBut IANAL. Bert_2 If you email help@launchpad.net we could forward it to our legal team.14:48
Bert_2perfect, I'll email to help@launchpad.net then14:50
Bert_2thx a lot, let's hope the new logo, icons, and name will be enough, cause manually editing each template file is already taking a lot of time :p14:50
Bert_2k, you've got mail ;)15:00
=== cr3_ is now known as cr3
=== cr3 is now known as Guest64961
=== matsubara-afk is now known as matsubara
sinzuibenji: do you have time to review https://code.launchpad.net/~sinzui/launchpad/adapt-subscription-to-pillar/+merge/11609516:18
benjisinzui: sure, if after lunch is fine16:19
sinzuiIt is, thank you16:19
=== Guest64961 is now known as cr3
=== cr3 is now known as Guest9828
=== salgado is now known as salgado-lunch
=== Guest9828 is now known as cr3
=== cr3 is now known as Guest1039
=== deryck is now known as deryck[lunch]
benjisinzui: your branch looks good17:51
sinzuithanks benji17:52
benjimy pleasure17:52
=== almaisan-away is now known as al-maisan
rick_h_benji: couple of reviews your way please. Pre-landing bugfix with JS gardening, no real changes, just cleaning up some LoC to get a credit for the bugfix coming later.18:17
rick_h_https://code.launchpad.net/~rharding/launchpad/tab_client/+merge/116329 and https://code.launchpad.net/~rharding/launchpad/garden_client/+merge/11633018:17
benjirick_h_: cool, I'll take a look18:17
rick_h_ty much!18:18
=== Guest1039 is now known as cr3
=== cr3 is now known as Guest88086
=== salgado-lunch is now known as salgado
=== deryck[lunch] is now known as deryck
=== matsubara is now known as matsubara-afk
=== al-maisan is now known as almaisan-away
=== bac-lunch is now known as back
=== back is now known as Guest15663
=== bac is now known as Guest27230
=== Guest88086 is now known as cr3
=== cr3 is now known as Guest21652
allenapbenji: Thanks for the review :)19:43
benjimy pleasure19:43
lifelesso/19:44
=== bradcrittenden is now known as bac
allenapo/ lifeless19:47
abentleyderyck: even though I have bugs.dynamic_bug_listings.enabled, I still don't see the sorting losenges and config wheel.  Any idea what I need to do?20:05
deryckabentley, hmmm, weird.  thinking....20:06
deryckabentley, I wonder if some other flag is conflicting and needs turning off.  Something off in production by default but on in dev?20:07
abentleyderyck: err, seeing "ReferenceError: YUI is not defined"20:07
deryckabentley, ah, js build issue then.20:07
abentleyderyck, rick_h_: Did make clean; make run, still getting ReferenceError: YUI is not defined.20:11
rick_h_abentley: not sure. You have build/js/yui? combo load is on by default now so should be looking in there20:12
deryckabentley, I would think make run should call  make jsbuild, but might be worth trying a make clean_js && make jsbuild20:12
rick_h_abentley: check the network and see if it has a combo load call and if so what came back from that20:12
abentleyrick_h_: I do have build/js/yui.  It has a yui-3.3.0 symlink that is broken.20:13
rick_h_abentley: ok, so make clean_buildout and make again20:14
* rick_h_ dbl checks clean_buildout20:14
rick_h_abentley: yea, that should hopefully fix your symlink? Should be linked from the dep cache20:14
abentleyrick_h_: The link is still broken.  It points at ../../../../build/js/yui-3.3.0, but that doesn't exist.  ../../../../build doesn't exist.20:23
abentley../../../../ is the directory containing my launchpad branch.20:24
=== salgado is now known as salgado-afk
rick_h_abentley: so you don't have build in the root LP branch dir? along with bin, lib, etc?20:25
abentleyrick_h_: I do.20:25
abentleyrick_h_: It's looking one parent directory above that.20:25
abentleyrick_h_: My working tree is at ~/launchpad/colo, but it's looking for build in ~/launchpad.20:27
rick_h_abentley: looking for what originally creates the build directory itself, sec20:27
rick_h_that's buildout that creates the dir and sets up the YUI 3.3 symlink that the combo-rootdir then links to build/js/yui20:29
rick_h_abentley: check out the [yui] block in buildout.cfg and see which command isn't working right that causes you to not have the right path20:29
rick_h_abentley: you still in dev mode? In production mode it setups up /srv/launchpad.net for the build dir20:29
abentleyrick_h_: still in dev mode AFAIK.20:30
rick_h_abentley: yea, so check out the buildout install yui and see if it's erroring on something in your env and if that's correct then combo-rootdir comes in and symlinks to the build dir specified which is just 'build/dir' from the calling point20:31
abentleyrick_h_: I don't really know what to look at.20:32
rick_h_abentley: so in the buildout [yui] block should be a series of commands that extract yui from the download-cache and setup the point for the symlink into build/js/yui-3.3.020:33
abentleyrick_h_: I have that, and a yui symlink that links to yui-3.3.0.20:35
rick_h_ok, and that's an ok symlink?20:35
abentleyrick_h_: Yes.20:36
rick_h_abentley: then run ./bin/combo-rootdir build/js and it should get you a good symlink into the build/js/ directory20:36
abentleyrick_h_: The broken symlink is Inside yui-3.3.0, and is named "yui-3.3.0".20:36
rick_h_hmmm, wtf20:37
rick_h_your buildout.cfg is still pulling download-cache/dist/yui-${:yui_version}.tar.gz right?20:37
abentleyrick_h_: http://pastebin.ubuntu.com/1107124/20:40
rick_h_abentley: ok, I see something strange20:45
rick_h_so in build/js/yui-3.3.0 I see all of YUI and a symlink that's ../../../../build/js/yui-3.3.0 i'm not sure where it came from20:46
abentleyrick_h_: So if you see it, it's not what's breaking me.20:47
rick_h_right20:47
abentleyrick_h_: In the network view, I see a 404 for https://bugs.launchpad.dev/+combo/?yui/yui/yui-min.js&lp/meta.js&yui/loader/loader-min.js20:48
rick_h_do you have the build/js/lp dir?20:49
abentleyrick_h_: yes.20:49
rick_h_is the meta.js in there?20:50
abentleyrick_h_: yes.20:50
rick_h_if so can you get just https://bugs.launchpad.dev/+combo/?lp/meta.js20:50
rick_h_without a 404?20:50
abentleyrick_h_: No.20:50
rick_h_ok, so you don't have convoy setup in apache then20:51
abentleyrick_h_: http://pastebin.ubuntu.com/1107142/20:51
rick_h_right, the apache config should be updated to not proxy +combo urls and instead serve via convoy20:52
rick_h_abentley: look in configs/development/local-launchpad-apache and diff against your apache config20:53
abentleyrick_h_: I just ran "make install", and that's giving me plausible results.20:54
abentleyrick_h_: Yes, all looks good now.20:54
rick_h_abentley: ok awesome20:55
=== benji changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: - | Firefighting: - | Critical bugs: 4.0*10^2
Ergo^evening21:46
Ergo^hello, ive seen launchpadblog entry about career opportunity pointing here so i can see what you guys do, so i've decided to pop in21:47
Ergo^hi czajkowski21:48
czajkowskiErgo^: hiya, which role?21:51
Ergo^czajkowski, http://blog.launchpad.net/were-hiring/were-hiring-software-engineers-to-work-on-cloud-projects-in-canonical21:52
czajkowskiErgo^: ah that one.21:52
czajkowskiErgo^: well if you have questions many of the team are on here, gary_poster might also be able to help21:52
czajkowskior you can apply via the link in the post21:52
Ergo^ah, ok, i was thinking i would fit the role well, when do you guys plan to hire someone ?21:54
Ergo^my firstborn is supposed to be born this week so im not sure its right time to apply ;-)21:55
Ergo^gary_poster, hello gary, what pythonic web frameworks do you guys mostly use?21:56
czajkowskiErgo^: soon, he may not be around as it's possibly coming to his EOD21:57
czajkowskiErgo^: where are you based?21:58
Ergo^czajkowski, Poland, so its midnight for me21:58
czajkowskiah most of hte EU folks are well past EOD21:59
czajkowskisorry21:59
Ergo^czajkowski, no problem21:59
Ergo^i might try tommorow21:59
czajkowskiearlier in the day might be bette22:00
czajkowskir22:00
czajkowskigood night22:00
Ergo^night!22:00
rick_h_doh, just missed him. I know Ergo^ from the pyramid/pylons irc world22:14

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!