[00:09] <mtaylor> hey lovely people...
[00:09] <mtaylor> https://bugs.launchpad.net/drizzle/build/+bug/440264
[00:10] <mtaylor> how do I remove it from being targetted to the build series?
[00:10] <poolie> hello
[00:10] <poolie> i think you cannot :-(
[00:10] <poolie> you need to just mark it invalid
[00:10] <mtaylor> that makes bunnies cry
[00:11] <poolie> and me too
[00:11] <poolie> there is a bug for his
[00:13] <poolie> mtaylor: https://bugs.edge.launchpad.net/malone/+bug/253600
[00:13] <poolie> put sad bunny ascii art on it :)
[00:13] <mtaylor> poolie: done
[00:15] <poolie> heh
[00:15] <poolie> you can also vote for it :)
[00:19] <mtaylor> also ... it would be great if there were a way to make a series an invalid target like you can for milestones
[00:20] <poolie> to mark the series as no longer active?
[00:20] <poolie> there is a series status
[00:20] <poolie> it seems like if the series is eg obsolete you should not be able to target bugs to it
[00:21] <mtaylor> poolie: it would seem that, wouldn't it?
[00:21] <mtaylor> poolie: alas, that doesn't have that effect
[00:22] <poolie> https://bugs.edge.launchpad.net/malone/+bug/437297
[01:06] <lifeless> leonardr: hi
[01:06] <lifeless> leonardr: are you up?
[01:08] <leonardr> lifeless: yeah, but not working anymore
[01:08] <leonardr> what's up?
[01:09] <lifeless> just catching up on some bzr bugmail
[01:09] <leonardr> ok
[01:09] <lifeless> jam has tested importing lplib at 500ms or so
[01:09] <poolie> heh, that's nothing :)
[01:09] <lifeless> which is more than bzr takes to start and complete status on many trees :)
[01:10] <poolie> https://bugs.edge.launchpad.net/launchpadlib/+bug/522957 1250ms overhead to start sending a single request
[01:10] <lifeless> I'm wondering if you will consider this a bug {it makes it really hard to consider using lplib for bzr}
[01:10] <poolie> lifeless: it means we should be careful about when it's loaded but we probably want to do that anyhow
[01:11] <lifeless> poolie: I am not convinced its safe to use at all at this point
[01:11] <poolie> cause of load time, or something else?
[01:11] <lifeless> poolie: it can easily swamp our actual pull performance
[01:11] <lifeless> poolie: load + https + round trips + needs a cache to work at all
[01:12] <lifeless> in contrast with e.g. AWS's apis which can be done with a single round trip safely over http (assuming you don't consider the submitted or returned data confidential)
[01:13] <lifeless> we may be much better off extending the existing http XMLRPC server lp has for us
[01:13] <wgrant> +1000000
[01:13] <poolie> ah
[01:13] <poolie> i think using it for lp: urls would be very poor
[01:14] <leonardr> lifeless: take a look at bug 521927
[01:14] <lifeless> otoh if leonardr considers these issues bugs we can work on it
[01:14] <poolie> i think using it for commands that are specific to launchpad, like lp-propose are fine
[01:14] <poolie> i commented on the lp: bug that any rpc separate to starting the ssh connection is really a kind of waste
[01:14] <lifeless> poolie: I'll buy appropriate, I'm not convinced about fine. I've so far seen nothing perform well with lp apis.
[01:15] <poolie> lifeless: aiui amazon's approach is to send application-level signed/encrypted messages over plain http
[01:15] <poolie> which is quite interesting
[01:15] <poolie> well
[01:15] <poolie> it's not ideal
[01:15]  * lifeless is in a 'performance matters' mindset today ;)
[01:15] <poolie> i think it's better than not having the feature at all, or implementing a separate mechanism
[01:15] <poolie> launchpad should have an api mechanism that is fast
[01:16] <leonardr> lifeless: performance is going to be my next project after i get multi-version working
[01:16] <leonardr> i was going to work on server-side performance but it seems like client-side is hurting people more
[01:16] <poolie> lifeless: did you try hydrazine?
[01:16] <poolie> it's interesting
[01:16] <poolie> it is actually useful
[01:16] <lifeless> poolie: I haven't tried it yet.
[01:16] <poolie> but also it shows you the stripped-down rpc performance
[01:16] <lifeless> poolie: I've used lplib interactive quite a bit
[01:17] <poolie> ./bugclient --debug -c 'bug 123456' -c 'triage high confirmed +foo'
[01:17] <poolie> right, it's like that with a bit less typing
[01:17] <lifeless> leonardr: when you start thinking and working on performance, I'd be *delighted* to talk with you about concerns, constraints and ideas.
[01:18] <leonardr> lifeless: sure
[01:18] <lifeless> leonardr: as would other folk in bzr team I'm sure; we've all got various amounts of skin in the performance game ;)
[01:18] <poolie> or other organs
[01:18] <leonardr> i don't want a bunch of bugs saying "it's too slow" but separate bugs for ideas like "keep the http connection open" and "don't round-trip for the wadl every time" are good
[01:18] <lifeless> leonardr: jam was measuring just import time, not time to grab the wadl, FWIW.
[01:19] <poolie> leonardr: +10
[01:19] <poolie> same for us
[01:19] <leonardr> lifeless: i'd consider that a third bug
[01:19] <poolie> specific bugs that can be individually fixed
[01:19] <poolie> i'll file that
[01:19] <lifeless> leonardr: is there one for 'ship the wadl', and 'have it precompiled' ?
[01:19] <lifeless> poolie: I suggested on the bug that jam file that; you might like to put the number in that bug for reference.
[01:20] <leonardr> lifeless: there is not. i'm against both those ideas but i'm open to some modified form of 'have it precompiled' (ie. 'compile it lazily')
[01:21] <lifeless> leonardr: I suspect theres a systemic problem in that lp's interfaces are very non-batchy / job-submission style, but good web performance needs batchy/job-submission style.
[01:21] <poolie> i think if it's actually cached properly it need not be bad
[01:21] <poolie> 1MB the first time you use it is not awful
[01:21] <poolie> bug 522963
[01:21] <lifeless> leonardr: well I'll file them
[01:22] <lifeless> (unless you want to convince me now that they are bad things to do [I don't suggest doing that, its late for you])
[01:22] <leonardr> lifeless: that comes from our decision to save time by publishing the internal interfaces
[01:22] <leonardr> now that we have the basic stuff published, we can take a step back and design something better
[01:23] <poolie> lifeless: maybe that's just subsidiary to bug 521927 or a suggested fix for it
[01:23] <poolie> s/that/shipping or trimming it
[01:24] <poolie> oh and that reminds me of another malone bug
[01:24] <lifeless> poolie: interesting ssh trivia, thanks.
[01:24] <poolie> yes, isn't it
[01:25] <poolie> it's a bit hard to actually know for sure
[01:26]  * leonardr sneaks off while lifeless and poolie are talking
[01:27] <poolie> heh, good night
[01:27] <poolie> and thanks, though apis are slow they're useful
[01:30] <lifeless> right, two bugs filed
[01:30] <lifeless> poolie: I think there is a significant difference between 'ship the wadl' and 'does not cache the latest version'
[01:30] <lifeless> mainly around the use cases of 'give me the latest model' and 'I know that I work with model XYZ'
[01:30] <poolie> the first would invalidate the second
[01:30] <poolie> mm
[01:31] <poolie> fixing either would help
[01:31] <lifeless> so the first doesn't invalidate the second :)
[01:31] <lifeless> though they are similar
[01:38] <poolie> does compiling wadl take a long time?
[01:39] <lifeless> I'm not entirely sure
[01:43] <poolie> as much as $10?
[01:43] <poolie> :)
[01:43] <lifeless> poolie: more if roaming.
[01:43] <lifeless> poolie: I didn't want to exaggerate ;>
[01:43] <poolie> even more if you want to use a satelite phone
[01:43] <lifeless> yup
[01:44] <poolie> even more if you're in 1982
[01:44] <lifeless> stuck back in the wild west with a delorean
[01:44] <lifeless> I need some war of the worlds I think
[07:59] <mrooney> hello! anyone know if anything is wonky with staging.lp.net? I'm getting a 401 Unauthorized via the LP API but can change the service to edge and it works fine.
[08:01] <mrooney> does mars happen to have any ideas? :)
[08:01] <thumper> mrooney: it is possible that staging is being updates
[08:01] <thumper> updated
[08:05] <mrooney> thumper: well I gave it a peek and it looks fine and is functional, I thought when it is being updated you can't view it as a webapp either?
[08:05] <thumper> mrooney: yeah, that is normally the case
[08:05] <mrooney> it is a mystery then
[08:05] <thumper> mrooney: do you have an authorized token for staging?
[08:05] <thumper> mrooney: that is more likely the issue
[08:06] <mrooney> that could be! doesn't it ask you to create one if you don't have one?
[08:06] <mrooney> I did at one point as this used to work, maybe it expired but the edge one didn't?
[08:12] <mrooney> looks like I just had to delete my ~/.launchpadlib and it re-prompted me and all is well
[08:25] <mrooney> thanks!
[08:49] <om26er> how can I make a package backported to karmic ? when I am actually using lucid
[08:49] <om26er> *while uploading to pap
[08:49] <om26er> ppa
[08:51] <noodles775> om26er: can you use pbuilder locally to check the source, and then just upload to your PPA?
[08:51] <noodles775> s/check the source/check the source in karmic
[08:51] <om26er> noodles775, ok
[09:55] <edakiri> I filed Bug  #463144 .  The same issue affects LaunchPad.  Should I file a separate bug for LaunchPad or should it be listed in the same bug?
[10:02] <adeuring> noodles775: could you pelase ui-review this MPhttps://code.edge.launchpad.net/~adeuring/launchpad/bug-283941-show-patch-numbers-on-upstream-report/+merge/19405 ? A screenshot: http://people.canonical.com/~adeuring/upstreamreport.png . the branch adds the right-most column
[12:26] <sebner> I can't access several sites (e.g https://edge.launchpad.net/ubuntu/+source/ufl) when I'm logged in. Not allowed here
[12:26] <sebner> Sorry, you don't have permission to access this page.  ... With another browser (not logged in) it works. Anyone an idea?
[12:28] <wgrant> Bug 514824
[12:28] <wgrant> sebner: ^^
[12:30] <sebner> wgrant: ohhhhhhh, thanks :)
[14:18] <ccm> hey, do you know who is the contact for the @ubuntu.com mail server?
[14:18] <ccm> i am getting terribly spammed
[14:22] <mars> ccm, might be better to ask in #ubuntu-signpost
[14:24] <ccm> mars: okay, thank you
[14:24] <mars> ccm, they should be able to direct you to a channel with sys admins (might be #ubuntu-dev)
[14:24] <ccm> that sounds great. hopping there
[14:42] <lfaraone> mars: he should have gone to #canonical-sysadmin
[14:46] <mars> lfaraone, is that in the ubuntu docs?
[14:48] <lfaraone> mars: probably not, but they're responsible for the mailserver.
[14:48] <mars> lfaraone, it was a question about *.ubuntu.com, so I assume that the ubuntu docs are the place to go
[14:48] <mars> hmm
[14:50] <jpds> mars: https://wiki.ubuntu.com/UbuntuEmail
[14:53] <mars> jpds, ok, thanks
[14:54] <mars> so the question is: did the #ubuntu-signpost people manage to direct him to the right wiki page and/or channel?
[14:54] <jpds> There is no #ubuntu-signpost channel...
[14:55] <mars> ?
[14:56] <edakiri> Uh! mars is a troll!  (just kidding)
[14:56] <mars> looking back through the browser history.... could have sworn I read "....#ubuntu-signpost is staffed by volunteers..."
[14:56] <mars> edakiri, :)
[14:57] <mars> edakiri, I have been helping people to engender feelings of goodwill.  I was going to begin DCCing you all momentarily, but it appears I will have to build yet more goodwill first.
[14:58] <mars> ha!
[14:58] <mars> jpds, https://help.ubuntu.com/community/Signpost/Destinations#irc-ubuntu-signpost
[14:59] <jpds> mars: Interesting.
[14:59] <mars> and I know #ubuntu+1 is real, so why no #ubuntu-signpost? :)
[15:46] <micahg> are the amd64 lucid chroots broke?
[15:47] <micahg> actually, the PPA gets from archive, so I'll go ask in -idevel
[15:47] <bigjools> micahg: yes but someone is on it
[15:47] <micahg> ah, k
[15:47] <micahg> thanks
[15:48] <bigjools> the udev issue?
[16:01] <micahg> bigjools: yeah, I think so
[17:24] <bdmurray> unsubcribing a team from a bug is raising an error message for me "Server error" but seems to work.  is this known?
[17:38] <Ursinha> bdmurray: don't think so, is it consistent?
[17:39] <bdmurray> Ursinha: is 2 out of 2 times consistent?
[17:39] <Ursinha> bdmurray: hmm, yes
[17:40] <Ursinha> bdmurray: you tried two times in a short period?
[17:40] <bdmurray> Ursinha: well in the past hour with 2 separate bug reports
[17:41] <Ursinha> bdmurray: hmm, right, I'll take a look
[17:41] <Ursinha> thanks
[19:14] <adiroiban> hi. Any ideas how can I access entries from Launchpad API using launchpadlib?
[19:14] <adiroiban> I'm trying to get a language https://launchpad.net/+apidoc/index.html#language
[19:17] <tsimpson> adiroiban: what issue are you having?
[19:18] <adiroiban> I'm kind of dumb, and I don't know how to query a language from launchpadlib
[19:19] <adiroiban> I can query persons... but I got stuck at quering a language
[19:19] <tsimpson> you usually get a language from a person, or you can use the .load method of your Launchpad instance
[19:20] <tsimpson> as you can see though, a language object doesn't really contain any data, so there's not much point in loading it
[19:22] <adiroiban> tsimpson: thanks. lp.load did the trick
[19:23] <adiroiban> I guess there is no API query languages in a similar way as persons or projects
[19:24] <tsimpson> it's really just an object representing a string, which happens to represent a language code
[19:27] <adiroiban> tsimpson: true. I was looking after something similar to the LP page for a language https://translations.launchpad.net/+languages/ar
[19:28] <tsimpson> not much (any) of the translations stuff has been exported to the API as far as I can see
[19:30] <tsimpson> it's mostly just people/teams, projects and bugs so far
[19:31] <adiroiban> yep. np. I will read about writing LP APIs and try to submit a patch
[19:31] <adiroiban> is this part of registry?
[19:31] <rhpot1991> I think something might be wrong with the 32bit chroot: http://launchpadlibrarian.net/39317246/buildlog_ubuntu-lucid-i386.mythbuntu-common_0.43-0ubuntu2~ppa1_CHROOTWAIT.txt.gz
[19:31] <adiroiban> the code for languages is in lp/services/worlddata
[19:33] <tsimpson> adeuring: I'm not sure where it is, you'll have to do some digging :)
[19:37] <tsimpson> https://dev.launchpad.net/Hacking has some help on finding stuff
[19:40] <adiroiban> thanks
[19:46] <blueyed> again, I cannot (finish) report(ing) a bug: OOPS-1509B1894
[19:49] <mars> blueyed, just looking at the OOPS - that is quite surprising.  Ursinha, ^ known issue?
[19:49] <blueyed> and for another package: OOPS-1509K2052
[19:50] <blueyed> does this happen because of empty description?
[19:50] <mars> blueyed, are you trying to file the bugs with empty description fields?
[19:51] <lifeless>   Module lp.bugs.browser.bugtarget, line 702, in showFileBugForm
[19:51] <lifeless>     raise NotImplementedError
[19:51] <mars> lifeless, that is the surprising part :)
[19:51] <mars> that he found a path to that code
[19:52] <lifeless> you know what would be nice
[19:53] <blueyed> mars: not surprising for me: LP oopsed far too often when reporting bugs for me.
[19:53] <blueyed> the last time it was related to the package name or whatever.
[19:53] <lifeless> to include a link to bazaar.launchpad.net/~launchpad-pqm/launchpad/prod|edge?/<revno>/lib/...
[19:53] <blueyed> yes
[19:53] <blueyed> (simple testcase ^^)
[19:54] <mars> blueyed, well, I am looking through the list of known issues.  There are ones in malone for filing descriptions with a single whitespace char
[19:54] <Lord-Readman> when launchpad tar-balls something for you in .tar.gz is it possible to request that it does .tar.xz to save bandwidth and space ?
[19:54] <blueyed> here's another one: OOPS-1509C1814 - empty desc?
[19:54] <blueyed> it's the empty desc..!
[19:54] <lifeless> Lord-Readman: when does launchpad tar things for you ?
[19:55] <Lord-Readman> lifeless, when you request a trunk translation download
[19:55] <deryck> blueyed, mars -- yeah, this is missing form validation.  It's fixed on edge.
[19:55] <blueyed> mars: can you please also add a testcase for adding a bug with an empty desc.. this should not happen in an app that is years old already.
[19:55] <mars> blueyed, want to become an edge beta tester? :)
[19:55] <lifeless> Lord-Readman: not at the moment; you might check bugs.launchpad.net/rosetta to see if there is a wishlist bug for that
[19:56] <lifeless> blueyed: 'should' \o/
[19:56] <mars> blueyed, deryck is the man to handle this
[19:56] <blueyed> deryck: ^^
[19:56] <mars> I defer to him
[19:56] <lifeless> or blueyed - you could file a patch :>
[19:56] <blueyed> mars: I've left when there was too much annoyance in there at some point. also, I could not beg you to release this fix - I would not have noticed on edge
[19:56] <blueyed> lifeless: tests!
[19:57] <lifeless> blueyed: and code!
[19:57] <Ursinha> mars: let me see here
[19:57] <deryck> mars, blueyed this is fixed already.  and it is tested, but javascript changed the behavior in a way the original test didn't catch.
[19:58] <deryck> mars, blueyed -- we're preparing to cherry-pick the change to the main site now.
[19:58] <blueyed> lifeless: I know. But it's a beast to setup.. and the first fix would be to have a "any" checkbox/toggle in the advanced search form bug status page.
[19:58] <blueyed> anyways.. please add a test.
[19:58] <mars> Ursinha, nevermind, sorry to trouble you (see deryck's last answer)
[19:58] <Ursinha> mars: no problem :)
[19:58] <blueyed> deryck: it is tested, but got through to production with the test failing?
[19:58] <blueyed> deryck: thanks.
[19:59] <goundy> hi guys
[19:59] <mars> deryck, thank you for the info
[19:59] <goundy> still no wiki support workarround in the air ? :)
[19:59] <deryck> blueyed, the test never failed.  the test was for the static pages and the failure was introduced by converting the behavior to use JavaScript.  The new path was tested in windmill for js interaction, but the failure case was not known and therefore not tested.
[20:00] <deryck> that's complicated to explain on IRC :)
[20:00] <mars> hi goundy.  Sorry, no, still no wiki that I know of.
[20:00] <goundy> mars, hehe thanks ;)
[20:01] <blueyed> deryck: I understand completely.. there was no test.. ;P
[20:02] <deryck> blueyed, if it makes you happy, yes, there was no test. ;)
[20:02] <deryck> blueyed, but have you ever tried getting JavaScript-heavy pages properly tested? ;)
[20:04] <lifeless> mars: https://bugs.edge.launchpad.net/oops-tools/+bug/523398
[20:05] <mars> lifeless, ?
[20:05] <blueyed> no, but I know that there are frameworks.. and you appear to be using windmill.. the issue is just that there a tests for testing web pages currently in place, which are not consistent between JS and non-JS.
[20:05] <lifeless> mars: filed a bug with my suggestion for lptools; checking with you that it makes sense to someone else
[20:05] <mars> ah
[20:06] <lifeless> leonardr: so, did the bugs we filed make sense ? (getting to mail now)
[20:07] <mars> lifeless, that makes sense to me, and that would be a nice feature, too
[20:07] <blueyed> yeah, I've hit code from 2006! ;)
[20:07] <blueyed> Thanks for your work. Keep up the QA! :)
[20:08] <leonardr> lifeless: yes. a lot of your problems could be mitigated by changing your behavior (no longer starting a lot of new processes from a cronjob), but we can talk about the best way to address the problems later
[20:09] <lifeless> leonardr: I think you're assuming some stuff based on your reaction
[20:09] <lifeless> leonardr: [I don't start a lot of new processes, and don't use cron :)] but thats ok, we can tease it apart later
[20:10] <leonardr> lifeless: i showed gary your bugs and he said something about people using cron, i assumed it was you
[20:10] <leonardr> yes, i'll talk with you after multi-version is done
[20:10] <lifeless> nope, not me; I'm proxying for users of bzr and the performance and network situations they report.
[20:11] <lifeless> [and when I'm phone-tethered, me]
[20:13] <lifeless> so what is 'multiversion'?
[20:13] <Lord-Readman> lifeless I created https://blueprints.launchpad.net/rosetta/+spec/export-all-translations-as-tarball-select-compression-method
[20:13] <Lord-Readman> is that ok you think?
[20:14] <lifeless> Lord-Readman: I would have created a bug myself; did you check that there wasn't a bug available already?
[20:15] <Lord-Readman> I searched, no mention, also I didnt see how a missing feature would have been a bug and not an idea/blueprint
[20:15] <lifeless> Lord-Readman: because launchpad development is primarily driven by bugs and bugs can be 'wishlist'
[20:16] <Lord-Readman> So should I request removal of the blueprint and raise it as a bug?
[20:16] <lifeless> Thats why I suggested a bug to you :). What you do know is up to you, I'm not on the rosetta team, I don't know what will get the best reaction from them.
[20:18] <willkg> i'm having trouble editing the miro project on launchpad (https://launchpad.net/democracyplayer/).  i don't have access and don't know who can give me access.  oppy (owner of pculture) is no longer around.
[20:18] <willkg> how can i go about reclaiming the project?
[20:19] <lifeless> file a question on answers.launchpad.net/launchpad, or talk to mars: the current help contact
[20:20] <c_korn> hello, are the build deamons currently broken ? http://launchpadlibrarian.net/39318250/buildlog_ubuntu-lucid-amd64.scilab_5.2.0-7~ppa1_CHROOTWAIT.txt.gz
[20:20] <lifeless> no, but the chroots they use are broken and being fixed.
[20:20] <lifeless> see #ubuntu-devel
[20:20] <willkg> lifeless: thanks!  will do.
[20:21] <Ursinha> willkg: have you tried contacting the owner?
[20:21] <willkg> Ursinha: like i said "oppy" is no longer around.
[20:22] <c_korn> lifeless: k, thanks
[20:54] <adiroiban> now that Launchpad allows anonymous access to API, why we can not just grab an url like this https://api.launchpad.net/beta/+languages/ar directly in the browser?
[20:54] <adiroiban> is a bug
[20:54] <adiroiban> or a feature?
[20:55] <lifeless> neither
[20:55] <lifeless> its because anonymous isn't 'unauthenticated' its 'authenticated as anonymous'
[21:00] <wgrant> adiroiban: Try https://launchpad.net/api/beta/+languages/ar
[21:00] <wgrant> That will authenticate you with your cookie.
[21:01] <adiroiban> wgrant: thanks! it is working
[21:03] <wgrant> adiroiban: Alternatively append ?oauth_token=&oauth_consumer_key=something to the URL.
[21:08] <adiroiban> wgrant: great. By doing this it is realy easy to do quick tests:)
[21:08] <wgrant> adiroiban: Exactly.
[21:08] <wgrant> So much quicker than firing up launchpadlib.
[21:16] <rdb> I can't find an option anywhere to retract an offer for mentorship on the blueprint page. Am I missing something?
[21:17] <rdb> I mean, blueprint pages don't seem to list anything about the available mentoring for it.
[21:19] <wgrant> rdb: The mentoring feature has been withdrawn, I believe.
[21:21] <rdb> Beh! Why?
[21:21] <rdb> The icon still shows up for the existing blueprints.
[21:21] <rdb> But why was it removed? It's one of the main reasons why I'm still using launchpad for blueprint tracking.
[21:23] <rdb> is there an equivalent?
[21:29] <sinzui> rdb: mars: mentoring was removed from launchpad 9 months ago
[21:30] <sinzui> rdb: mars: there is not action that can be taked
[21:30] <sinzui> taken
[21:30] <sinzui> I do see mentoring icons show up, I think that is a bug in the badge code
[21:36] <sinzui> rdb: mentoring was removed because it was being used by about 20 users out of a million launchpad users. It did not work as users want it to work, and the code was costly to maintain: http://blog.launchpad.net/general/removing-mentoring
[21:37] <rdb> beh, ok
[22:00] <fta> what happened to all the builders?
[22:00] <micahg> fta: they be broke :)
[23:32] <poolie> here's a nice use of Launchpad: https://lists.ubuntu.com/archives/bazaar/2010q1/066924.html
[23:53] <lifeless> any losa in the house?