[00:00] <wgrant> It should have closed an hour ago, right?
[00:02] <henninge> exactly
[00:02] <henninge> I am pretty sure it's just the subject
[00:03] <wgrant> sinzui: Looks like your branch broke the build?
[00:03] <sinzui> :(
[00:04] <sinzui> wow I played that test
[00:04]  * sinzui prepares a fix
[00:04] <henninge> ah, I had seen that on ec2 but could not reproduce it locally. My devel must be too old.
[00:05] <henninge> which means nothing will land anyway atm because of testfix mode.
[00:10] <sinzui> oh good, this looks like the test is actually wrong now. I thought I had broken one requirement while fixing another
[00:15] <wgrant> Does anybody read launchpad-error-reports?
[00:29] <sinzui> wgrant: I do.
[00:32] <wgrant> sinzui: There is so much noise.
[00:32] <wgrant> And so many errors.
[00:32] <sinzui> production's would be more intelligible if several of us fixed checkwatches + external bugtrackers to not report offsite data issues as oopses. Or we add a nominal data cleansing task before we read the broken xml
[00:33] <wgrant> sinzui: Oh, not the OOPS reports. The launchpad-error-reports ML's cron spam.
[00:33] <sinzui> wgrant: checkwatches was divorced from the normal reports because of this, but the issues were not fixed. lifeless hops to may the problem visible by spamming the report
[00:34] <sinzui> wgrant: sorry, my confusion. I specifically tag the messages that talk about teams, karma, the prf, and distributionmirrors
[00:41] <wgrant> sinzui: Ah. I silenced the DMP (except for errors) yesterday.
[00:44] <sinzui> I was certain your bug was a dupe. spm had remarked about it in the past. I could not find it though
[00:44] <wgrant> :(
[00:45] <wgrant> Damn. We could deploy that and the nightly.sh fix now, if it wasn't Friday afternoon.
[02:22] <LPCIBot> Project db-devel build (343): FAILURE in 5 hr 38 min: https://hudson.wedontsleep.org/job/db-devel/343/
[03:21] <LPCIBot> Project devel build (420): FAILURE in 5 hr 18 min: https://hudson.wedontsleep.org/job/devel/420/
[03:21] <LPCIBot> * Launchpad Patch Queue Manager: [r=allenap][no-qa] Machinery for building documentation with sphinx
[03:21] <LPCIBot> * Launchpad Patch Queue Manager: [r=bac][bug=710054][bug=530476][bug=530477] Fix the protocol in the
[03:21] <LPCIBot> gpg documentation and branch and MP api docs.
[03:21] <LPCIBot> * Launchpad Patch Queue Manager: [r=allenap][ui=none][bug=693689] Escape the token and title in
[03:21] <LPCIBot> lp.app.widgets.itemswidgets renderer.
[03:21] <LPCIBot> * Launchpad Patch Queue Manager: [r=henninge][ui=none][no-qa] Add direct unit tests for
[03:21] <LPCIBot> TM.shareIfPossible
[04:31] <lifeless> wgrant: resetting the cache would be fine, but noone knows how to do that
[07:02] <lifeless> it would be nice to fix deleting of unpublished packages to actually delete.
[07:03] <wgrant> Possibly.
[07:06] <lifeless> wgrant: so what do you think, is there some way to purge apache-openid nonces?
[07:10] <wgrant> lifeless: Probably rm /tmp/mp_sess.dbm
[07:10] <wgrant> Unless the mp config has been customised.
[07:10] <wgrant> But devpad has that file.
[07:14] <lifeless> wgrant: care to note that in the bug; do we need a graceful or will it be recreated automatically?
[07:15] <lifeless> hmm
[07:15] <lifeless> ProtocolError for xmlrpc.lp.internal:8097/codehosting: 502 Bad Gateway
[07:15] <wgrant> lifeless: I believe it should be recreated automatically, but I've never used the default dbm store before.
[07:15] <lifeless> not a great sign
[07:16] <lifeless> hmm, i think I made it oops :P
[07:16] <wgrant> Which?
[07:17] <lifeless> pushing to lp:debian/sid/python-poster/packaging
[07:17] <lifeless> 8boom8
[07:17] <wgrant> Hah.
[07:17] <wgrant> That should blow up. But nicely, not with an OOPS.
[07:18] <lifeless> just gives a 502
[07:18] <lifeless> I'm sure its booming in the background
[07:18] <wgrant> Awesome.
[07:18] <wgrant> Does it tell you which method it was executing?
[07:19] <lifeless> no
[07:19] <lifeless> robertc@lifeless-64:~/source/debian/poster-0.7.0$ bzr push lp:debian/sid/python-poster/packaging
[07:19] <lifeless> Working tree "/home/robertc/source/debian/poster-0.7.0/" has uncommitted changes (See bzr status). Uncommitted changes will not be pushed.
[07:19] <lifeless> bzr: ERROR: Server sent an unexpected error: ('error', '<ProtocolError for xmlrpc.lp.internal:8097/codehosting: 502 Bad Gateway>')
[07:19] <wgrant> Ah. Fun.
[07:19] <lifeless> btw ^ - lets you upload to +storeblob
[07:20] <wgrant> Anyway, add a username.
[07:20] <wgrant> Hmm, why would we want to do that?
[07:20] <lifeless> wgrant: subunit stream storage
[07:20] <lifeless> wgrant: also, shrink the code in apport
[07:21] <lifeless> I'm getting round to an old tuit of mine - subunit-upload -> lp
[07:21] <wgrant> Ah, right.
[07:25] <lifeless> ok, two bugs for one action with lp.
[07:25] <lifeless> good times.
[07:27] <lifeless> wgrant: ok, so openid -  we can experiment I guess; if you can record what you *do* know, I can pester a sysadmin on monday
[07:32] <wgrant> lifeless: Commented.
[07:36] <lifeless> thanks
[07:36] <lifeless> !
[07:41] <LPCIBot> Project db-devel build (344): STILL FAILING in 5 hr 19 min: https://hudson.wedontsleep.org/job/db-devel/344/
[07:41] <LPCIBot> * Launchpad Patch Queue Manager: [r=bac][ui=none][bug=548][incr] Mute self-generated bug notifications
[07:41] <LPCIBot> for accounts that set that option (iow, mute themselves).
[07:41] <LPCIBot> * Launchpad Patch Queue Manager: [r=bac][ui=none][no-qa] Add "selfgenerated_bugnotifications" as a
[07:41] <LPCIBot> global option for people so that we can work on bug 548. Only
[07:41] <LPCIBot> in the DB at this time,
[07:41] <LPCIBot> since the actual implementation is not yet done.
[07:41] <_mup_> Bug #548: Launchpad sends change notification updates to the person who requested the change <email> <lp-bugs> <story-better-bug-notification> <story-better-notification-sending> <Launchpad itself:In Progress by yellow> < https://launchpad.net/bugs/548 >
[08:20] <lifeless> what is webservice:register in configure.zcml for
[08:24] <lifeless> figured it
[09:35] <LPCIBot> Yippie, build fixed!
[09:35] <LPCIBot> Project devel build (421): FIXED in 6 hr 13 min: https://hudson.wedontsleep.org/job/devel/421/
[09:35] <LPCIBot> * Launchpad Patch Queue Manager: [testfix][rs=sinzui] Print the escaped text of the structured string.
[09:35] <LPCIBot> * Launchpad Patch Queue Manager: [r=henninge][ui=none][bug=697294] Refactors create_questionreopening
[09:35] <LPCIBot> to be called as a helper, rather than a listener to notifications;
[09:35] <LPCIBot> this avoids the need for a slew of guards on the method, and keeps it
[09:35] <LPCIBot> from doing work when we don't want it to.
[12:59] <LPCIBot> Yippie, build fixed!
[12:59] <LPCIBot> Project db-devel build (345): FIXED in 5 hr 17 min: https://hudson.wedontsleep.org/job/db-devel/345/
[12:59] <LPCIBot> Launchpad Patch Queue Manager: [rs=buildbot-poller] automatic merge from stable. Revisions: 12330,
[12:59] <LPCIBot> 12331, 12332, 12333, 12334, 12335 included.
[13:03] <wgrant> jelmer: It's emailing you for successful binary builds?
[14:00] <jelmer> wgrant: actually, I just have a lot of failing builds at the moment :)
[21:15] <elv> Hi, how can i copy  a package fromubuntu primary archive to my ppa? i just saw some packages marked as Copied from ubuntu natty in Primary Archive for Ubuntu .
[21:15] <elv> thanks
[23:22] <cjohnston> This bug (Bug #340640) says there is no plan of fixing this "at the moment" but that was a year ago... Any chance we can revisit this? It is currently breaking the LoCo Directory.
[23:22] <_mup_> Bug #340640: Standard way of finding mugshot url, default if not set <api> <lp-registry> <Launchpad itself:Triaged> < https://launchpad.net/bugs/340640 >
[23:30] <lifeless> cjohnston: we have 6000 bugs and close ~ 1000 a year
[23:32] <lifeless> cjohnston: why is it breaking the loco directory now ?
[23:33] <cjohnston> we are trying to pull the profile images.. but if there isnt one, it gives a broken image
[23:34] <lifeless> what makes it broken? Why don' you catch the exception and use the default in that case?
[23:35] <cjohnston> it tries.. i dont know if maybe the LD code is broken in that aspect or what
[23:35] <lifeless> I would assume it is
[23:35] <lifeless> anyhow, in the set of things we're trying to focus on, this really seems appropriate as low
[23:35] <cjohnston> ok
[23:36] <lifeless> if you need it escalated, the stakeholder process provides a forum for the different stakeholders to convince each other about the relative importance of things
[23:36] <lifeless> -or-
[23:36] <cjohnston> http://bazaar.launchpad.net/~chrisjohnston/loco-directory/edge/revision/384
[23:36] <lifeless> you can supply a patch
[23:36] <cjohnston> that is the code we have.. i dont know if you know anything about how it would be broken or not
[23:37] <lifeless> well
[23:37] <lifeless> its hard coding api urls
[23:37] <lifeless> thats pretty much guaranteed to run into trouble
[23:37] <lifeless> specifically -
[23:37] <lifeless> beta is deprecated
[23:37] <lifeless> you should use 1.0 for clients that are packaged
[23:37] <lifeless> and devel for clients where you're tracking bugfixes etc
[23:38] <cjohnston> ok
[23:38] <lifeless> secondly
[23:38] <cjohnston> thanks
[23:38] <cjohnston> oo.. more
[23:38] <lifeless> issuing those urls back to browsers will be very very slow
[23:39] <lifeless> your website should resolve the real url using lp apis and then output that.
[23:39] <lifeless> otherwise you're forcing directory viewing to do an HTTPS handshake with launchpad.net
[23:40] <lifeless> (when you're already logged in and have handshaked already)
[23:41] <lifeless> both of the urls you are returning are going to do a full appserver request and then issue a 302
[23:44] <lifeless> what would be most efficient would be for whatever collection you're asking Launchpad for to return things with preresolved urls (which we can do for public assets).
[23:44] <cjohnston> ok.. thanks
[23:45] <lifeless> I'd consider a bug to enhance LP to do that high (which means 3-9 months away) priority, because we're trying to make Launchpad snappy across the board
[23:45] <lifeless> in the meantime, lets see if we can work around your issue
[23:46] <lifeless> got a link to a broken page on the diretory?
[23:49] <cjohnston> http://loco.ubuntu.com/events/team/666/detail/
[23:49] <cjohnston> sorry.. just finished dinner
[23:50] <lifeless> how do we land things on lazr-js
[23:51] <lifeless> cjohnston: ok
[23:51] <lifeless> so in that page you're outputting this:
[23:51] <lifeless> <img src="https://api.edge.launchpad.net/1.0/~drubin/logo" title="David Rubin" valign="top" width="32" height="32">
[23:52] <lifeless> cjohnston: but this is a little crazy because you've done an API call to determine the team membership anyway.
[23:52] <cjohnston> ok
[23:53] <cjohnston> i dont know very much about the api and such... so bare with me
[23:53] <lifeless> whats the api call you make to find out about the members of the team?
[23:55] <cjohnston> Ronnie: do you know what the api call is that finds out the members of teams
[23:55] <lifeless> Ronnie: specifically how are you populating http://loco.ubuntu.com/events/team/666/detail/
[23:56] <cjohnston> lifeless: i know that we pull what team a person belongs to, the code is there, but i dont know if we use it for finding out their TZ
[23:57] <cjohnston> the attendees statuses are all done completly through LD... you dont have to be a member of a team to attend an event, only to create/edit an event
[23:58] <lifeless> cjohnston: do you verify their lp uids?
[23:59] <cjohnston> I believe that we do now lifeless