/srv/irclogs.ubuntu.com/2011/01/07/#launchpad-dev.txt

lifelessoff for a bit, back later to talk with bigjools00:26
thumperoh FFS00:50
thumperanyone else getting make build fail?00:52
wgrantthumper: JS issues?00:52
thumperOSError: [Errno 4] Interrupted system call00:52
thumperwriting WADL00:52
wgrantAh, not the same one then.00:52
wgrantI've not seen that.00:52
* thumper is frustrated00:52
* thumper repulls00:53
wgrantIs ec2 land broken for anyone else? NameError: global name 'AuthorizeRequestTokenWithBrowser' is not defined01:02
thumperwgrant: I can't even get make build running on devel01:13
thumpermwhudson: can you do a 'make clean build' on devel for me?01:14
thumpermwhudson: I want to see if it is just me :)01:15
mwhudsonthumper: i'm running make build now as it happens01:15
wgrantAh, updating just pulled in a new launchpadlib.01:15
mwhudsonthumper: i want it to work though, so i'm not going to pull/make clean just yet :)01:15
thumperit seems that the WADL generation is failing01:15
wgrantthumper: WADL build worked on an up-to-date devel.01:16
thumperI'm trying to run a test, and it fails too01:16
thumperand I don't know why01:16
wgrantAlthough I didn't clean.01:16
thumperwgrant: didn't for me :(01:16
* wgrant cleans.01:17
* thumper tries yet again01:17
mwhudsonhm01:17
mwhudsonOSError: [Errno 2] No such file or directory: './lib/canonical/launchpad/icing/yui/dom/dom-style-ie-min.js'01:17
wgrantmwhudson: rm -r lazr-js/build01:17
mwhudsonthe thing i love about launchpad development is the reliable infrastructure01:17
mwhudsonwgrant: thanks01:18
wgrantthumper: A clean devel build worked fine too :(01:22
thumperit worked for me now01:22
thumpertest still failed01:22
wgrantUh.01:23
wgrantlaunchpadlib 1.9.1 is broken.01:23
wgrantlaunchpadlib/launchpad.py:206 references a class that no longer exists...01:23
mwhudsonnice01:25
mwhudsonhttp://www.netsight.co.uk/misc/xkcd-buildout.png/view01:31
lifeless\o/01:32
wgrantYup.01:32
wgrantI wonder if versions.cfg will like me downgrading launchpadlib to 1.6.5.01:33
mwhudsonthat's quite a big downgrade?01:36
lifelessno01:36
lifelessthe version numbers had quite a big upgrade01:36
mwhudsonheh ok01:36
wgrantYay, ec2 works again.01:37
mhall119can someone tell me why this user is only having their identity URL passed back in the OpeniD response? http://ubuntuone.com/p/X2P/01:43
maxbmhall119: LP only gives more than the identity URL to sites it has specifically been told to trust01:57
Ronniemaxb: can you explain the differences between these 2 screenshots:01:59
Ronniehttp://ubuntuone.com/p/X2O/01:59
Ronniehttp://ubuntuone.com/p/X2P/01:59
Ronnieboth share the same root site, but got other response from lp01:59
maxbhuh. no, I cannot explain that01:59
StevenKDifferent account02:00
StevenKLook at the top line, "Logged in as"02:00
maxbyes... but... why should that influence what data is passed?02:02
StevenKmaxb: I have no idea, but it's a difference02:04
Ronniemaxb, StevenK ill try if i can get the openid request to lp02:06
Ronniethis is the form HTML that is submitted to LP: http://paste.ubuntu.com/551335/02:15
Ronniefor the user Ronnie02:15
=== Ursinha is now known as Ursinha-zzz
wgrantStevenK, maxb: I wonder if that account might not be linked to a Launchpad one.03:28
wgrantNot sure how to check.03:28
maxbActually.... how does the login service know about LP team participation at all?03:39
maxbIs the crazy cross-replica replication thingy still in place?03:40
wgrantmaxb: Yeah, it's all a bit nauseating.03:41
wgrantWe have triggers copying account, person and teamparticipation into lp_account, lp_person and lp_teamparticipation. Those three are replicated to the SSO DB.03:41
wgrantI hope to eventually coerce people into doing it via the API instead.03:45
wgrantBut everyone seems happy enough with this replication madness...03:45
wgrantHm.03:57
wgrantAlmost bug #70000003:57
* spm starts filing bugs like a crazy bug filer04:11
mwhudsonspm: if they're all about loggerhead, no fair04:15
spmmwhudson: no worries. I've gotten bored with reporting codebounce bugs anyway.04:15
rockstarthumper, keep your own bugs. :) https://bugs.launchpad.net/launchpad/+bug/69156304:18
_mup_Bug #691563: bundle merge plugin needs refactoring to improve how it handles errors and send emails <Launchpad itself:New> < https://launchpad.net/bugs/691563 >04:18
thumperrockstar: it seemed to me it was a tarmac bug04:18
thumperthe code was in tarmc04:18
rockstarthumper, no, launchpad maintains a fork of tarmac, and the bug is in that fork.04:18
thumperah...04:19
thumperwhy does it have a fork?04:19
thumperthat seems like bollocks04:19
rockstarthumper, because that bundle merge plugin is kind of bollocks.04:19
rockstarIt's necessary for launchpad because the test suite takes so long.04:19
* thumper sighs04:20
thumperrockstar: shouldn't the plugin be a separate project04:20
thumper?04:20
thumperor part of the LP tree?04:20
rockstarthumper, no idea.  That's for the maintainer to sort out.04:20
* thumper sighs again04:21
rockstarIt might *kinda* be Tarmac's problem since plugins don't seem to play as well with Tarmac as they do with bzrlib.  I need to investigate that.04:21
mwhudsonwow, 'make build' just takes a ridiculous length of time now04:22
* thumper gets hit by the friday afternoon bug04:22
mwhudsonit could hardly be worse if lp was written in c++ :(04:22
thumpermwhudson: that it does :(04:22
rockstarthumper, yeah, I hit that about noon today.04:22
rockstarmwhudson, are you going to the launchpad epic?04:23
mwhudsonrockstar: no04:23
mwhudsongoing to the rally though04:23
thumpermwhudson: ha... you miss out on the beer I was going to buy you04:23
rockstarmwhudson, ah, I thought that was this week.04:23
thumperrockstar: no, next04:23
mwhudsonif i'm going to spend 2 weeks away from home for work, it's not going to be in january04:23
thumperrockstar: you aren't going either are you?04:23
* thumper wanders out of the office04:24
rockstarthumper, no.  lifeless suggested it, but with U1 web losing a member soon, they won't spare the week for me to go.04:24
rockstarI might have pushed harder if it wasn't in Dallas04:24
* mwhudson imagines that possibly uds-o will be more popular04:25
mwhudsonwgrant: did you unbreak ec2 land yet?04:37
wgrantmwhudson: Probably needs a leonardr.04:37
mwhudsonok04:37
wgrantmwhudson: I just hacked versions.cfg back to launchpadlib 1.6.504:38
wgrantThen reran buildout.04:38
wgrantAll works.04:38
* mwhudson uses ec2 land from db-devel instead04:38
mwhudsontime for beer04:52
mwhudsonsee some of you in dallas, i guess04:52
pcjc2are all the scripts / pre-written sql / .... available in the LP sources?06:24
pcjc2I saw mention of LOSA scripts somewhere, and wondered where they lived. (This was for playing with my local dev instance)06:24
wgrantMost things should be scripts in the tree.06:26
wgrantThe LOSAs don't much like running SQL directly.06:27
pcjc2thanks06:28
lifeless=== Top 10 Time Out Counts by Page ID ===07:14
lifeless    Hard / Soft  Page ID07:14
lifeless      67 /  350  POFile:+translate07:14
lifeless      64 /  234  BugTask:+index07:14
lifeless      46 /  325  Distribution:+bugs07:14
lifeless      20 / 3690  Archive:+index07:14
lifeless      20 /  346  Distribution:+bugtarget-portlet-bugfilters-stats07:14
lifeless      10 /   81  MaloneApplication:+bugs07:14
lifeless       8 /  212  Question:+index07:15
lifeless       7 /  116  ProjectGroupSet:CollectionResource:#project_groups07:15
lifeless       7 /   11  Archive:+copy-packages07:15
lifeless       7 /    4  ProjectGroup:+milestones07:15
=== almaisan-away is now known as al-maisan
mrevellHello09:07
adeuringgood morning09:18
lifelessbigjools: hi09:29
* bigjools wearily waves to lifeless09:29
lifelessyou need more sleep?09:29
bigjoolssleep, drugs, you name it09:30
lifelessugh09:31
lifelesshope you're better for the thunderpic09:31
lifelessso you wanted to talk realtime; I don't know if its urgent to do so or not, but i've split my day today so we can, nowish, if you want to.09:31
bigjoolsshould be ok for the epidome, thanks09:31
bigjoolslifeless: yeah, it would be useful to talk.  How much longer are you around?09:32
lifelessits 10:30 pm now09:32
bigjoolsoh, arse09:32
lifelessso I'm here09:32
lifelessbut not hugely awake, definitely not going to be capable of rational thought for much longer :)09:33
bigjoolsso, let me explain the problem09:33
lifelesswould voice be better?09:33
bigjoolsand then you can decide if you want to leave it until later09:33
bigjoolsyeah, if you can put up with my blocked-sinuses-voice09:33
bigjoolsI'll just fire up the Quattro09:34
lifelessI can09:35
bigjoolsI'm on mumble09:35
lifelessskype would be better mumble is still unbearable for me09:36
bigjoolsok09:36
lifelessis that ok?09:36
bigjoolsyup09:36
bigjoolsnow if skype would just connect09:37
lifeless>-<09:37
bigjoolsit's not connecting :/09:38
lifelessI can ring your phone ?09:39
bigjoolsit's finally in09:39
lifelesshah, that threat made it connect09:39
lifelessbigjools: https://dev.launchpad.net/Foundations/JobSystem10:02
jcsacketthenninge: i see your branch merged, how's it going?10:37
henningejcsackett: going well, we have not encountered any other problems.10:37
jcsacketthenninge: cool. can we qa-ok the blocked revision by 1400 UTC (i believe that's 1pm your time)?10:41
henningejcsackett: I was still waiting fot the build to finish - which I just see it has. ;)10:42
henningeand it's 3pm my time (GMT+1 in the winter)10:43
henninge;)10:43
jcsacketthenninge: time zones are hard. :-P10:43
jtvhenninge: interesting obstacle to Q/A… I don't have permission to set Launchpad usage on a test project in staging!10:45
jtvYet I *can* access the form.  Just can't post.10:45
henningejcsackett: technically I'd have to wait for the revision to be available on staging and verify that all is running well. We'll be doing more of that today.10:45
henningejtv: url?10:45
jtvhenninge: https://translations.staging.launchpad.net/wdiff/+configure-translations10:45
henningeah, danilo just did that. Maybe the permissions issue slipped qa?10:46
jtvjcsackett: you worked on the new model for setting LP usage, right?10:46
jtvhenninge: danilo just did what?10:46
henningejtv: that page10:46
jtvWell grr.10:46
allenapgmb: I got that weird thing with kwallet yesterday. I purged kdebase-runtime to get rid of it.10:47
henningeI reviewed it  ... and did some qa.10:47
jcsackettjtv: yeah, some time ago. if you have .Edit permission on that project you should be able to change it.10:47
gmballenap: Ok, Ta. I'll give that a shot10:47
jtvjcsackett: I don't, but I have Translation admin privs.10:48
henningejtv: I, too, would have assumed I could do it as a rosetta admin.10:48
jtvWe used to be able to do it.10:48
jcsackettjtv: actually, i think that should do it too.10:48
jcsackettjtv: translations were a little odd for the usage stuff (more roles to think about), i believe we went with as permissive a model as possible.10:48
jtvAnd it works partially: I do get to access the form.  Only a POST breaks on me.10:48
jcsackettjtv: that does sound like a bug.10:49
jtvRather.10:49
henningejtv: you can set the translation focus10:49
henningeand the other stuff on that page10:49
jtvEvidently.10:49
jtvJust not usage, then.10:50
henningeThey used to be on different forms.10:50
henningeThis is not a regression, though. We weren't allowed to do that before either. We don't have Edit rights on projects.10:51
jtvI thought we were allowed to set usage though.10:51
henningejtv: it's not "official rosetta" anymore, remember?10:52
henningeofficial_rosetta10:52
jtvSo?10:52
jtvWe could set usage.10:52
jtvWe're being told we can still set usage.10:52
jtvWe can't set usage.10:52
jtv→ Bug.10:52
henningeok10:52
henningebut then the bug is older than the creation of this new dialog?10:53
henningejtv: you should find a prooject owned by registry admins.10:55
jtv10:55
henningejtv: we are part of that team and so have edit rights on the projects10:55
henningeI use gedit10:55
jtvThanks.10:56
henningegedit project.10:56
jtvThat should be in main.  ☺10:56
henningeah yes, in main10:58
=== al-maisan is now known as almaisan-away
=== yofel_ is now known as yofel
=== Ursinha-zzz is now known as Ursinha
bigjoolsIf I type "launchpad" in the product search popup it says "too many matches" ....11:41
=== matsubara-afk is now known as matsubara
jcsacketthenninge: can you qa bug 690196 and bug 694703, or does doing so depend on the fix we're waiting on for staging?11:48
_mup_Bug #690196: POTMsgSet.singular_text may need to be a method <lp-translations> <qa-needstesting> <Launchpad itself:Fix Committed by henninge> < https://launchpad.net/bugs/690196 >11:48
_mup_Bug #694703: Make partial translations exports work for upstream <qa-needstesting> <upstream-translations-sharing> <Launchpad itself:Fix Committed by henninge> < https://launchpad.net/bugs/694703 >11:48
henningejcsackett: looking now.11:55
bigjoolshow the heck can I end up with lib/mailman/Mailman/MemberAdaptor.pyc being owned by root?13:10
jcsacketthenninge: sorry to keep pestering you, but we're about 50 min from the original time for merge--how's it looking?13:12
henningejcsackett: one qa still missing but I just igned off the big one.13:17
jcsacketthenninge: fantastic! :-)13:18
bigjoolsjcsackett: when is PQM closing?13:18
jcsackettbigjools: i think we're still going with 1400 UTC, as henninge is in a position to qa everything blocking db-stable before then, i think.13:19
bigjoolsok.  Bugger.13:19
jcsackettbigjools: what are you trying to land?13:19
bigjoolsit's a one-liner to fix https://bugs.edge.launchpad.net/launchpad/+bug/69982013:20
_mup_Bug #699820: BuildFarmJob.date_finished is set in two places <buildd-manager> <buildfarm> <tech-debt> <Launchpad itself:In Progress by julian-edwards> < https://launchpad.net/bugs/699820 >13:20
bigjoolsI need it so I can do a graph13:20
bigjoolsso not desperately urgent but kinda annoying :)13:20
jcsackettbigjools: dig.13:20
bigjoolsI can always ask for RC forgiveness13:20
* jcsackett nods.13:20
bigjoolsI'll grab you later, thanks13:20
jcsackettnp. :-)13:21
* bigjools goes to eat while tests are running13:21
jcsacketthenninge: if you were able to qa-ok the big one, that implies bug 697845 is good too, right?13:22
_mup_Bug #697845: Translations import script on staging fails <oops> <qa-needstesting> <upstream-translations-sharing> <Launchpad itself:Fix Committed by henninge> < https://launchpad.net/bugs/697845 >13:22
jcsacketti am looking at it now and can mark it as such if you agree.13:23
bigjoolshaving said that, I've got some weird shit13:23
jcsackett?13:23
bigjoolsrunning "make" or "bin/test" is blocking on opening the KDE wallet (like the gnome keyring)13:23
bigjoolsWTF13:23
jcsackettbigjools: the kde wallet is now used by launchpadlib, right? of course, that shouldn't come up in make...13:24
bigjoolsurrrgghhh really?13:24
bigjoolsthis Is Bad (TM).13:24
bigjoolsbin/compile_templates does it too13:25
jcsackettbigjools: make clean it first, maybe? that seems to be the magic fix from time to time.13:26
bigjoolsI did that :/13:26
jcsackettone sec; i'll update my devel and see if i can replicate.13:26
bigjoolsbin/compile_templates13:27
bigjools(32547) KWallet::Wallet::openWallet: Pass a valid window to KWallet::Wallet::openWallet().13:27
bigjoolsit works, except when running the tests as the librarian fails to start13:28
bigjoolsre-created it on 2 desktops now :/13:28
bigjoolsyeah, it seems to be launchpadlib13:29
jcsackettit didn't replicate on mine, but then i realized i am a fool to think it would--i don't have wallet/keyring.13:30
bigjoolsah13:31
bigjoolswhat do you have?13:31
jcsackettsorry, bigjools.13:31
jcsacketti use wmii--so nothing.13:31
* jcsackett likes keyboard controls13:31
jcsackettthis is for my dev environment--i keep my launchpad stuff all set up in a VM.13:32
=== almaisan-away is now known as al-maisan
bigjoolsI sent an email to -dev13:36
bigjoolsit's caused by the log line above and the librarian layer sees "unclean output"13:36
jcsacketthenninge: we need qa for bug 69470313:39
_mup_Bug #694703: Make partial translations exports work for upstream <qa-needstesting> <upstream-translations-sharing> <Launchpad itself:Fix Committed by henninge> < https://launchpad.net/bugs/694703 >13:39
pcjc2hi guys.. have a bit of a problem - trying to setup bugmail to go to our old mailing list (mailman)13:39
pcjc2but bugmail appears to originate from the LP account holder, not LP its-self. Is there any known way to make it work without having to accept non-members postings?13:40
maxbpcjc2: Can Mailman accept based on the presence of a header? If so, match on an X-Launchpad-Message-Rationale: header being present13:43
pcjc2couldn't see it, but will investigate13:43
pcjc2maxb: Testing now - there is a spam filter which lets me match on headers14:00
pcjc2(and choose to accept a message based on a given match) - but I'm not sure if it will get that far for a non-subscribed user... waiting for bugmail to be sent out, and we'll see ;)14:00
pcjc2(Is bugmail a cronjob?)14:01
=== jcsackett changed the topic of #launchpad-dev to: Launchpad Development Channel | 11.01 Release Week 3 | PQM in RC for devel | RM: jcsackett | firefighting: - | Get the code: https:/​/​dev.launchpad.net/​Getting
bigjoolsbenji: pingaling14:22
benjibigjools: heh14:22
bigjoolsbenji: francis said you might be able to help with my launchpadlib problem - did you see my email to -dev?14:22
benjiI don't think so, let me look.14:23
bigjoolsok, thanks14:23
benjibigjools: bummer; let me fire up my KDE VM and see if there's anything I can do about it14:24
bigjoolsbenji: the issue is that the stupid piece of crap PyKDE4 library is logging debug output :/14:24
bigjoolswhich makes the librarianlayer blow chunks14:25
benjiyep14:25
bigjoolsthough why that layer needs to use launchpadlib is interesting14:25
bigjoolsI am looking at the PyKDE4 module to see if it has any "stop logging" parameters14:31
benji+114:31
pcjc2https://launchpad.net/~pcb-bugs14:39
pcjc2Is there anyway to make mail sent to that address come _from_ some fixed address, such as noreply@launchpad.net ?14:39
pcjc2pcb-bugs@lists.sourceforge.net  is going to bounce any non-subscribed users bug comments, which is not what we want14:39
bigjoolsbenji: I can see the problem, it's a bug in the keyring module14:40
bigjoolsit's passing 2 args instead of 314:40
bigjoolswhen calling openWallet14:40
pcjc2And I've checked.. it isn't apparently possible to accept mail based on a header at that stage. Can spam filter it by headers... but mails from non-subscribed addresses are dropped into the moderation queue before the spam filter would get a chance to accept them14:40
benjiif that's the code I think it is, that's intentional (perhaps still a bug, but intentional); let me look at it14:41
benjithe above was directed toward bigjools14:42
bigjoolsbenji: http://api.kde.org/pykde-4.3-api/kdeui/KWallet.Wallet.html#obj18800948414:42
bigjools"You can pass 0 if you don't have a window"14:42
* bigjools haxors14:43
bigjoolsno luck14:50
benjibigjools: right, 0 won't reduce the spurious output: http://pastebin.ubuntu.com/551502/14:53
bigjoolsyeah just saw that14:53
bigjoolshttp://api.kde.org/4.x-api/kdelibs-apidocs/kdeui/html/kwallet_8cpp_source.html#l0025714:53
bigjoolsit's using the kDebug ostream, there must be a way of changing that14:53
bigjoolslog level14:53
bigjools"kdebugdialog"14:55
bigjoolsnow, how do we script that14:55
bigjoolscrap, it doesn't actually help14:58
=== matsubara is now known as matsubara-lunch
bigjoolsbenji: it seems as though it's impossible to turn that output off :/15:13
bigjoolsthere's a bug in the logger somewhere15:13
benjibigjools: I'm pursuing making a do-nothing window to pass it so it won't log the message15:14
bigjoolsthat'll work15:14
benjibigjools: will you try this patch to see if it fixes it for you? http://pastebin.ubuntu.com/551508/15:19
bigjoolsyup, one sec15:19
* benji turns on the coffee pot in the meantime.15:19
bigjoolsbenji: success!  I had to use this diff though: http://pastebin.ubuntu.com/551512/15:25
benjibigjools: correct me if I'm wrong, but that looks the same sans some whitespace15:26
benji(same as my patch)15:26
bigjoolsbenji: you were using "kwallet_module"15:26
bigjoolsthat's not in my egg's code15:27
benjioh, I'm using the trunk which changed that a little, cool15:27
bigjoolsmine was grabbed not so long ago, is it out of date already?15:27
benjithe current python-keyring trunk hasn't been released; I'll make this change and do a release today15:28
bigjoolsah ok15:28
bigjoolsyou probably want to add a comment about that weird code :)15:28
bigjoolsbenji: thanks a million for helping, I am unblocked15:30
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
=== leonardr_ is now known as leonardr
=== matsubara-lunch is now known as matsubara
=== Ursinha is now known as Ursinha-lunch
=== beuno is now known as beuno-lunch
=== al-maisan is now known as almaisan-away
=== almaisan-away is now known as al-maisan
=== al-maisan is now known as almaisan-away
=== almaisan-away is now known as al-maisan
=== al-maisan is now known as almaisan-away
=== gary_poster is now known as gary-lunch
=== beuno-lunch is now known as beuno
=== benji is now known as benji-lunch
=== Ursinha-lunch is now known as Ursinha
=== gary-lunch is now known as gary_poster
=== benji-lunch is now known as benji
=== leonardr is now known as leonardr-afk
=== leonardr-afk is now known as leonardr
=== matsubara is now known as matsubara-afk

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