[00:33]  * thumper goes to make coffee
[00:34] <rockstar> thumper, pong
[00:56] <thumper> rockstar: nm
[00:56] <rockstar> thumper, :(
[03:17] <rockstar> Dude, this ec2 crap is getting out of hand...
[03:24] <rockstar> I'm 0/3 on ec2 attempts today...
[04:41] <thumper> rockstar: still around?
[04:53] <rockstar> thumper, yeash.
[04:53] <thumper> rockstar: wanna talk?
[04:53] <rockstar> thumper, sure, one sec.
[06:46]  * thumper EODs
[09:48] <jml> bigjools, has anyone talked to the losas about the upcoming poppy changes yet?
[09:48] <mthaddon> nope
[09:48] <bigjools> lol
[09:48] <jml> mthaddon, oh hi :)
[09:49]  * jml runs
[09:49] <bigjools> mthaddon: we're adding an sftp server for cocoplum/germanium
[09:50] <bigjools> it'll obviously need some admin input for ports/configs/security etc
[09:50] <mthaddon> bigjools: what will these sftp servers do?
[09:50] <bigjools> mthaddon: same as poppy, except over sftp
[09:51] <mthaddon> bigjools: which reminds me - we need to plan a day for you to come to millbank and tell me all about soyuz...
[09:51] <bigjools> indeede
[09:51] <StevenK> And not using Zope's FTP service, thank $DEITY
[09:51] <mthaddon> bigjools: any chance of that happening next week? would be good to do it before the LOSA sprint so I can spread the love
[09:51]  * bigjools thanks beer
[09:51] <bigjools> mthaddon: +1
[09:51] <bigjools> I know how you like to spread love
[09:51] <mthaddon> you do indeed
[10:24] <lifeless> jml: I want 5 minutes with you before next session, if possible
[10:25] <StevenK> Including a heavy, blunt object?
[10:25] <lifeless> no
[10:25] <lifeless> need to flip a coin
[10:27] <mwhudson> jml: he's in flamboyant
[10:28] <lifeless> mwhudson: was that for me?
[10:28] <mwhudson> ah
[10:28] <mwhudson> lifeless: yes :)
[10:29] <lifeless> where is flamboyant ;)
[10:29] <mwhudson> lifeless: one floor up from you
[10:29] <mwhudson> lifeless: there's a session going on, but you may be able to ambush him at the end
[10:29] <lifeless> On the ceiling ?
[10:31] <mwhudson> lifeless: now he's running over to ebene
[11:00] <lifeless> jml: ping
[11:38] <jml> StevenK, http://docs.python.org/release/2.5.4/lib/module-os.html
[13:21] <lifeless> jml: I would _love_ a review of https://code.edge.launchpad.net/~lifeless/testtools/bug-335816-one-outcome-per-test/+merge/25226 - this will improve bzr's CI story.
[13:36] <danilos> bigjools, twistd rotates logs in a *very* weird way... some bits disappear for a certain time, or at least it seems so
[13:37] <bigjools> danilos: it renames all the files
[13:37] <bigjools> welcome to twisted
[13:37] <danilos> bigjools, right, but not just that... what used to be generic log is neither .1 or .2 anymore
[13:38] <wgrant> It's not still in the middle of a rename?
[13:38] <bigjools> danilos: what do you mean by "generic log" ?
[13:38] <danilos> bigjools, unnumbered buildd-manager.log
[13:38] <danilos> bigjools, look at this: https://pastebin.canonical.com/32246/
[13:39] <danilos> bigjools, a bunch of "Starting template..." entries were in buildd-manager.log, then it got rotated, and they are neither in .1 nor .2 anymore
[13:39] <bigjools> wot wgrant said - and also remember that devpad has synced logs
[13:39] <danilos> bigjools, I guess it just takes ages for it to rotate all files?
[13:39] <bigjools> so it might have synced in the middle of a rotation
[13:39] <danilos> wgrant, bigjools: could be, yes
[13:40] <danilos> well, it syncs every few minutes and I had something very similar yesterday for over 30 mins
[13:40] <bigjools> every 5 mins
[13:40] <bigjools> you can always get a losa to grep on cesium to confirm
[13:41] <danilos> well, you can't _always_ get a losa, but sure :)
[13:42] <danilos> bigjools, anyway, not a big deal, probably takes a while to rename ~3k files
[13:42] <bigjools> we should dump those old logs :/
[13:42] <bigjools> where dump == trash
[13:42] <danilos> where trash == recycle bin ;)
[13:43] <bigjools> eww
[13:55] <jtv> noodles775: I'm pushing a fix for our traceback... would you mind reviewing it?
[13:55] <jtv> Meanwhile it seems the fighting's started here, so I'll be a bit distracted.  :(
[13:56] <jtv> Crap, killed by a sniper...
[13:57] <jtv> Last taxi driver I talked to said this guy was his hero.
[13:57] <jtv> Pissing off a few million more people.. great way to make peace.  TV shows nothing but game shows, ads & documentaries, which usually means it's serious.
[13:58] <noodles775> :/
[13:58] <jtv> Just got message from the Happy Smiley Superior Firepower committee: they're going to let children and the elderly out alive.
[13:59] <jtv> That leaves the women and very few others.
[14:00] <danilos> jtv, and you? :)
[14:00] <jtv> danilos: no, but it's beginning to look like they got my number because I've been in that area.
[14:01] <jtv> Nobody else I know is getting them.
[14:01] <danilos> jtv, good luck surving the next few hours then
[14:01] <jtv> Oh, I'll be fine.
[14:02] <jtv> If I had my camera I might go
[14:11] <jtv> noodles775: https://code.edge.launchpad.net/~jtv/launchpad/production-579914/+merge/25230
[14:51] <noodles775> wgrant: the db-scheme in its current state: https://code.edge.launchpad.net/~michael.nelson/launchpad/db-build-generalisation-db-changes/+merge/22527
[15:03] <cperrin88> It's pretty tricky to crawl through all the laucnhpadlib imports to finally understand what's happening to port it to java -_-
[15:30] <gary_poster> cperrin88: ack. :-/   fwiw, I know many of the imports better than launchpadlib itself, so I can potentialy help with specific questions.  I understand that doesn't help much with getting the big picture though.
[15:30] <cperrin88> gary_poster: at least I found my error with the token
[15:31] <gary_poster> cperrin88: great!  something for us to share on the help page, to help others?
[15:31] <cperrin88> nope
[15:31] <cperrin88> to stupid :D
[15:31] <gary_poster> heh, ok
[15:32] <cperrin88> I sent my request to  +access-token instead of +request-token >.<
[15:32] <gary_poster> ah, suck
[15:32] <gary_poster> well, glad you're past that at least
[15:45] <danilos> bigjools, got a minute to discuss what we need to do to get translation jobs not to molest arm builders?
[15:46] <bigjools> danilos: yes I was just talking to jtv
[15:46] <bigjools> I have a solution
[15:46] <bigjools> was about to implement it but since you're around ... ;)
[15:47] <danilos> bigjools, who is around? :)
[15:47] <danilos> bigjools, nah, go ahead, go ahead, I am all good :)
[15:47] <wgrant> Which of the many possibilities are being considered?
[15:47] <bigjools> danilos: the easiest and sanest way right now is to change TranslationTemplatesBuildJob.create() to set buildqueue.processor to  the nominatedarchindep
[15:48] <bigjools> it restricts the jobs to i386 for now
[15:48] <bigjools> when we get builder pools sorted out, it will solve this issue in a better way
[15:49] <jtv> I did it that way once, but IIRC was told leaving it null was fine... how's that for irony?
[15:49] <bigjools> go figure
[15:49] <danilos> bigjools, so, we'll want this CPed as well?
[15:49] <bigjools> hel;l yes
[15:49] <danilos> bigjools, mthaddon told me there is a workaround in place right now
[15:49] <bigjools> this is the cycle of yeehaw and CP
[15:49] <jtv> danilos: I have to go now.  Requested CP for yesterday's fix; got the one for the buildd-manager traceback approved conditionally on dogfood Q/A.  (It's in EC2).
[15:49] <danilos> bigjools, is there?
[15:49] <wgrant> Well, nominatedarchindep sucks. It just sucks less for the moment.
[15:50] <bigjools> danilos: very temporarily, the builder was moved to the nonvirtual pool
[15:50] <bigjools> danilos: but that was to stop the manager from bailing
[15:50] <danilos> bigjools, ah, ok, so the sooner we get this done, the better, and then we can get that builder back to the virtual pool
[15:50] <bigjools> danilos: we need to put it back at some point
[15:50] <bigjools> yep
[15:51] <danilos> bigjools, ok, got the point
[15:51] <bigjools> g'night jtv
[15:51] <danilos> jtv-zzz, good night, safe walk home :)
[15:51] <bigjools> mind the snipers
[15:51] <kfogel> Can anyone confirm that the 'lp_teamparticipation' table in the DB schema is obsolete and the 'teamparticipation' table should always be used instead?
[15:51] <jtv-zzz> danilos: home how?
[15:51] <jtv-zzz> I'm not going home
[15:51] <danilos> jtv-zzz, ah, ok, enjoy the night at the office :)
[15:51] <jtv-zzz> thanks
[15:52] <bigjools> danilos: ok cool - if you need some more guidance let me know, I don't think I'm going out tonight since it's a holiday in Belgium (!)
[15:52] <danilos> bigjools, oh, it's a family holiday so you are spending the day at home? :)
[15:52] <bigjools> haha
[15:53] <danilos> bigjools, anyway, sure, will probably ping you since I totally have no idea what half of these words meant (like nominatedarchindep :)
[15:53] <wgrant> kfogel: You should use teamparticipation -- lp_teamparticipation is used in the replication mess for SSO. don't go near it.
[15:53] <mars> gary_poster, is there a new bootstrap.py that I can using for the new zc.buildout beta?
[15:53] <kfogel> wgrant: ah.  Thanks.
[15:53] <bigjools> danilos: ok :)
[15:53] <wgrant> lp_* are for the use of SSO.
[15:54] <gary_poster> mars, yes.  Use 1.5.0b1, and use http://svn.zope.org/repos/main/zc.buildout/trunk/bootstrap/newbootstrap.py (that URL is transient, btw; will become http://svn.zope.org/repos/main/zc.buildout/trunk/bootstrap/bootstrap.py)
[16:02] <mars> gary_poster, my buildout keeps picking 1.4.3.  According to the bootstrap.py --help, it should be going to PyPI and fetching your version.  I am using your newbootstrap.py.  I don't have a zc.buildout version locked down anywhere.
[16:02] <mars> gary_poster,   Perhaps it is using the local zc.buildout egg first, rather than going online?
[16:02] <gary_poster> mars, yes
[16:02] <mars> gary_poster, thanks.
[16:02]  * mars fixes
[16:09] <wgrant> gary_poster: The bug #329178 that you just referred a user to is private.
[16:09] <wgrant> (and, well, I'm interested in the proposed solution)
[16:11] <gary_poster> wgrant: ack. on call.  a bug in our openid provider, so not in my (or Launchpad's) domain directly.  um...I'll ask if I can add you in as a subscriber, I guess.
[16:12] <wgrant> gary_poster: Well, c-i-p is now open source, so it probably needn't be private any more.
[16:13] <gary_poster> it's considered security, wgrant.
[16:13] <wgrant> Ah, I see.
[16:21] <mars> gary_poster, you were right.  I killed the download-cache versions, and it picked my ~/.buildout/download-cache/ version instead, 1.5.0b2.
[16:21] <gary_poster> mars, you want b1
[16:21] <mars> gary_poster, ok
[16:24] <mars> gary_poster, lots of warnings like this: /home/mars/canonical/lazr-js/1.0/eggs/distribute-0.6.10-py2.6.egg/setuptools/command/easy_install.py:198: UserWarning: Unbuilt egg for setuptools [unknown version] (/usr/lib/python2.6/dist-packages)
[16:29] <mars> gary_poster, probably because your new version does "-S", then imports /usr/lib/python2.6/dist-packages/site.py.  That system-specific site.py contains setuptools yet again.  Something in the old easy_install code doesn't like that.
[16:30] <kfogel> wgrant: two questions, one is a meta-question: I'm trying figure out the difference between teammembership and teamparticipation tables (particularly since the latter appears to do what I would have expected the former to do).  And, I'm wondering where I'm failing to dig that would have answered this question for me (if only SQL tables could have descriptive comments attached *in the database*... or can they?  We don't appear to, anyway.)
[16:30] <wgrant> kfogel: teamparticipation is the flattened version of teammembership.
[16:30] <wgrant> The latter only contains direct memberships.
[16:31] <wgrant> The former takes membership's transitive nature into account.
[16:31] <gary_poster> mars, ack, have not seen.  upgrade setuptools to c11 and everything is fine?
[16:31] <wgrant> So teamparticipation is calculated from teammembership.
[16:31] <danilos> kfogel, \d+ for comments on tables in psql
[16:31] <wgrant> kfogel: There is comments.sql, but I'm not sure how to see that.
[16:31] <danilos> kfogel, and comments.sql
[16:31] <wgrant> Ahh, I see.
[16:32] <mars> gary_poster, this is entirely using distribute
[16:32] <kfogel> danilos: awesome!  thank you
[16:32] <kfogel> wgrant: all clear now, thank you.
[16:34] <gary_poster> mars, may just be broken setuptools locally.  I use distribute sometimes and is fine
[16:36] <kfogel> wgrant: so what I'm trying to do here is track growth in the ~contributor-agreement-canonical team's membership; in this case, I think there are only ever direct memberships, but it won't hurt to just query teamparticipation, because if there ever were indirect memberships, we'd want to count them.
[16:37] <kfogel> wgrant: talking out loud only for sanity check; no question needing answering there.
[16:37] <wgrant> (Well, the first issue is that that team is wrong.)
[16:37] <kfogel> wgrant: ?
[16:38] <wgrant> It's not complete. I'm not in it, for one.
[16:38] <danilos> kfogel, btw, there's also lib/lp/registry/doc/teammembership.txt (first paragraph seems to explain the difference between the two), so checking doc/<model-class>.txt can sometimes be a good start (ideally, it would always be a good start, but ok)
[16:38] <kfogel> wgrant: ah, I see what you mean.  The team is right, but it's got a bug :-)
[16:38] <kfogel> danilos: good hint, thx
[16:38] <mars> gary_poster, looks like it doesn't like the setuptools compatability package installed on Ubuntu.  "python-setuptools" is a versionless compatability package for setuptools 0.6c9
[16:52] <Ursinha> Chex, gary_poster, rockstar, bigjools, danilos, sinzui, allenap: production meeting on #launchpad-meeting @ Freenode in 8 minutes
[16:53] <allenap> Ursinha: Someone else will have to represent Bugs; I'm a treacherous Landscaper now.
[16:53] <Ursinha> allenap, oh, cool :)
[16:53] <Ursinha> allenap, I'll find someone else, thanks!
[16:54] <allenap> Ursinha: I'm asking now.
[17:02] <Ursinha> Chex, gary_poster, rockstar, bigjools, danilos, sinzui, someone in bugs: production meeting on #launchpad-meeting @ Freenode now
[17:10] <mars> gary_poster, do you think it is safe to update lazr-source-dependencies to use zc.buildout 1.5.0b1?  It currently uses zc.buildout-1.5.0dev-gary-r110073.tar.gz
[17:10] <gary_poster> mars, yes
[17:10] <poolie> jml, if you're still here, https://dev.launchpad.net/LEP/DKIMAuthenticatedMail is now ready for review
[17:10] <jml> poolie, I am. thanks.
[22:31] <kfogel> wgrant: you're on the contrib agreement team now
[22:34] <wgrant> kfogel: Yep, I noticed. Thaks.
[23:15] <maxb> StevenK: Are you sure about marking bug 576193 as a dup of bug 568745? I admit to not understanding the issues fully, but it looks a bit non-obvious.
[23:15] <mup> Bug #576193: Debian package metadata importer getting the wrong versions <gina> <Soyuz:Triaged> <https://launchpad.net/bugs/576193>
[23:15] <mup> Bug #568745: Debian unstable record of gpt missing <gina> <Soyuz:Triaged by stevenk> <Debian:Fix Released> <https://launchpad.net/bugs/568745>