[00:04] <spm> I lashed out on the w/e and bought some vowels. now to spend 'em.
[00:11] <wgrant> dhillon-v10 has tried to most of my teams in the past 24 hours :(
[00:11] <wgrant> Er, tried to join.
[00:11] <spm> yay :-/
[00:13] <thumper> and dhillon-v10 is?
[00:13] <wgrant> spm knows him.
[00:13] <ajmitch> another overly-helpful or inquisitive person?
[00:15] <spm> ajmitch: exactly. described himself as "... *THE* launchpad answer person". it's nice they're so keen and all; but perhaps overdoing it.
[00:15] <spm> heh. even so far as to wanting admin powers on LP so he could fix stuff.... :-)
[00:16] <ajmitch> heh
[00:25] <thumper> jelmer: there has probably been no change since the 4th of Dec for https://code.edge.launchpad.net/~kiko/linux/2.6.31
[00:25]  * thumper heads for coffee
[00:25] <thumper> bbl
[02:01]  * thumper has lotsa wine now
[02:46] <thumper> spm: is staging updating?
[02:49] <spm> thumper: looks like it. it was hung in an earlier update; killed that; is now updating for ~ the past 40 mins.
[02:49] <thumper> :-| we have some imports running
[02:49] <thumper> for 4 hours so far
[02:50] <thumper> bigish bzr-svn tests
[02:50] <spm> ew
[04:33]  * thumper stops for dinner
[05:58] <mwhudson> thumper: https://code.staging.launchpad.net/~mwhudson/boost/trunk succeeded after 8 and a bit hours
[05:58] <mwhudson> thumper: yay for bzr-svn i think
[06:32] <thumper> mwhudson: http://staging.launchpadlibrarian.net/36556292/boost-trunk-log.txt are we missing saving a cache?
[06:33] <mwhudson> thumper: bzr-svn saves its caches in ~/.bazaar somewhere -- the cache will get built once per machine
[06:33] <mwhudson> thumper: which is a bit lame, but not too bad
[06:34] <mwhudson> thumper: it would be better to be better, but that requires some hacking of bzr-svn
[06:37] <thumper> ok
[07:17] <lifeless> mwhudson: I suggest filing a moderate priority bug on that: the caches will have the following issues;
[07:17] <lifeless>  - they will grow in size and not be capped
[07:18] <lifeless>  - concurrent imports from the same svn repo will interfere with each other
[07:18] <lifeless>  - they can take some time to build
[07:58] <wgrant> Can I convince a merge proposal diff to be regenerated without pushing a new revision?
[08:00]  * wgrant considers uncommit, push --overwrite, pull, push
[08:05] <thumper> wgrant: not yet
[08:06] <thumper> wgrant: at least I don't think so
[08:06] <BjornT> gmb: what's the status of the fix for the two failing Bugs Windmill tests?
[08:07] <BjornT> thumper: what's the status of the fix for the failing Code Windmill test?
[08:07] <thumper> thumper: fixed awaiting release-critical stamp
[08:08] <thumper> BjornT: thanks for reminding me, I need to put it on the release blockers
[08:09] <BjornT> thumper: cool, thanks.
[08:14] <thumper> wgrant: why is it out of date?
[08:15] <wgrant> thumper: I set a prereq branch on the MP. It generated the diff with the prereq branch.
[08:15] <wgrant> But then I pushed some more revs several days later (the prereq branch was still unmerged).
[08:16] <wgrant> It regenerated the diff, but this time it appeared not to have taken the prereq into account.
[08:16] <thumper> umm...
[08:16] <thumper> it will have...
[08:16] <thumper> unless you convince me otherwise
[08:17] <wgrant> Argh, staging is too old :(
[08:17] <wgrant> But you're right, it looks like it does do it, since now that I've repushed it has gone crazy with conflicts.
[08:18] <wgrant> They are really strange conflicts.
[08:19] <thumper> wgrant: staging doesn't really have the branches from production
[08:20] <wgrant> thumper: I know, but it probably does have the crazy diff if a dump is being restored at the moment. But since the new diff is insane too (the pre-req has been merged), I wonder if they just don't merge well (criss-cross merges, or somesuch).
[08:21]  * wgrant tries.
[08:21] <thumper> wgrant: if you have a pre-req, then merge trunk into the second, then merge trunk into the first, then the pre-req into the second, you get criss-cross
[08:22] <thumper> wgrant: I work with pipes when using pre-reqs
[08:22] <thumper> wgrant: merge trunk into the bottom unmerged pipe
[08:22] <thumper> wgrant: then 'pump'
[08:22] <thumper> all done
[08:22] <wgrant> thumper: Yeah, I normally do too. But not in this case.
[08:22]  * thumper shrugs
[08:22] <thumper> oh well
[08:22] <wgrant> I love bzr-pipeline.
[08:22] <thumper> me too
[08:25] <wgrant> Oh.
[08:25] <wgrant> I see what happened.
[08:27] <wgrant> The pre-req that was merged is slightly different from the one that Julian landed -- he merged trunk into it, then landed. I didn't notice that it had landed, so I merged trunk into it and then pushed it.
[08:27] <wgrant> So there is a horrible horrible devel criss-cross in that branch.
[08:40] <adiroiban> jtv: hi. For branch-sync, I talked with Danilo and Henning and they told me it should also work with an empty branch. bug 494762
[08:40] <mup> Bug #494762: Export to branch is not working for a fresh branch creating using LP UI <code-integration> <Launchpad Translations:Triaged> <https://launchpad.net/bugs/494762>
[08:41] <adiroiban> thanks for the email :)
[08:41] <jtv> adiroiban: it's a somewhat obscure distinction... it'll work for an empty branch, but not for a nonexistent branch.
[08:41] <jtv> adiroiban: if nothing has been pushed, it's nonexistent.
[08:41] <jtv> but you can create an empty branch & push it.
[08:42] <adiroiban> jtv: ok. so the bug is invalid?
[08:42] <adiroiban> or  is for the wishlist?
[08:43] <jtv> no, the bug is valid but subtle—it's not actually possible to "create" (in the strictest sense) a branch using the LP UI.
[08:43] <jtv> the "register a branch" option registers a branch, but does not create it
[08:43] <jtv> (although if the branch is mirrored, a cron job will then create the branch by importing data from elsewhere)
[08:44] <jtv> When you register a hosted branch (as you do for translations exports), the LP UI really just "reserves a spot" for it.
[08:45] <jtv> The solution that we have in mind for the bug is to leave uncreated branches out of the list of options in the branch picker.
[08:48] <adiroiban> OK. Can I touch that branch now?
[08:56] <wgrant> danilo[home]:
[08:56] <wgrant> Argh.
[09:03] <wgrant> bigjools: Are you likely to have enough time to re-review gina-3.0-support today?
[09:03] <wgrant> (thanks again for getting the other one landed)
[09:04] <bigjools> wgrant: yep
[09:07] <wgrant> bigjools: The diff generated by LP is full of conflicts that I cannot reproduce locally -- the intermediate diff is attached, and http://paste.ubuntu.com/341024/ is the full diff.
[09:07] <bigjools> ok ta
[09:07] <bigjools> is the branch against db-devel?
[09:08] <wgrant> bigjools: It was, but that was before 3.1.11. The MP is still against db-devel, but it's fine in devel.
[09:08] <bigjools> right, that was my next Q :)
[09:08] <wgrant> LP doesn't allow me to retarget it :(
[09:09] <bigjools> yeah, thumper should really fix that :)
[09:09] <thumper> bigjools: it is on my dashboard :)
[09:09] <bigjools> \o/
[09:21] <mrevell> Morning
[09:51] <beuno> good morning
[10:22]  * bigjools is loving thumper's change to highlight MP changes done after it was first made
[10:22]  * wgrant liked that too.
[10:23] <wgrant> Although no automatic diff intermediate diff :(
[10:24] <bigjools> gah I need to set up a WOL on my test box
[10:24]  * bigjools trudges back through the rain into the house
[10:43] <bigjools> wgrant: FWIW the diff of your gina branch is fine here, I suspect there's a bug in the codehosting stuff
[10:44] <bigjools> danilos: I need you to approve https://code.edge.launchpad.net/~wgrant/launchpad/gina-3.0-support/+merge/15478 for release-critical
[10:44] <wgrant> bigjools: It's probably because of the prereq branch. There was something horrible going on there, but I removed the revision that didn't get merged, but the diff was still generated broken.
[10:44] <danilos> bigjools, sure, looking
[10:44] <wgrant> Very odd.
[10:45] <bigjools> danilos: thanks - it was another branch dependent on the buildbot update that didn't get done on Friday
[10:45] <wgrant> bigjools: (you merged devel into my other branch, then landed it. Later that day, I merged devel into the branch, not realising it had landed. I think that confused things.)
[10:45] <danilos> bigjools, whoa, that looks huge
[10:45] <bigjools> danilos: the diff is fucked
[10:46]  * danilos "phews"
[10:46] <bigjools> danilos: http://pastebin.ubuntu.com/341071/
[10:47] <danilos> bigjools, it's still reasonably big... have you had a chance to test this already on dogfood?
[10:47] <bigjools> danilos: doing so right now
[10:47] <wgrant> It's also only gina.
[10:47] <danilos> bigjools, ok, let me know how that goes, I'll take a look
[10:48] <bigjools> danilos: if it doesn't go in, it'll end up being a CP request next week
[10:48] <bigjools> and the Ubuntu mafia will knock your door down ;)
[10:50] <danilos> bigjools, heh, I am not that easily intimidated :P
[10:50] <bigjools> danilos: that cross-dressing guy in the bar in Barcelona did a good job :D
[10:50] <danilos> bigjools, btw, that cowboy listed on LPS for expire-ppa-binaries is also landed, right?
[10:50] <danilos> bigjools, heh, I hope you are not trying to do the same :)
[10:50] <bigjools> danilos: I don't know, it was salgado's bug
[10:50] <wgrant> bigjools: I presume the dpkg upgrades for cocoplum/germanium/iron and whatever runs debdiff are scheduled?
[10:51] <danilos> bigjools, though, that'd look good on you :P
[10:51] <bigjools> wgrant: argh good point
[10:51] <wgrant> bigjools: buildds are pretty much sorted, fortunately.
[10:52] <bigjools> yes
[10:52] <bigjools> wgrant: will gina pay attention to the sourcepackageformatselection ?
[10:53] <wgrant> bigjools: No.
[10:53] <wgrant> gina doesn't pay much attention to anything, and the archive is trusted, so I didn't think it important.
[10:53] <bigjools> I guess it doesn't matter
[10:54] <salgado> adiroiban, around?
[10:55] <adiroiban> salgado: hi. yes.
[10:55] <salgado> adiroiban, hi there.  did you get a message telling you about the tests that were failing in your branch?
[10:55] <adiroiban> salgado: nope.
[10:56] <salgado> adiroiban, can you check if it was not incorrectly flagged as spam?
[10:56] <danilos> bigjools, please add that to the special rollout requirements if it's not there already
[10:56] <adiroiban> salgado: who is the sender?
[10:57] <adiroiban> maybe it's just delayed
[10:57] <danilos> bigjools, also, what about that cowboy for expire-ppa-binaries on librarian? is it landed yet?
[10:57] <salgado> adiroiban, let me check who's the sender.  it should've been sent Saturday
[10:57] <bigjools> danilos: do you mean the one to set the number of days?
[10:57] <danilos> bigjools, there are https://pastebin.canonical.com/25653/ and http://paste.ubuntu.com/339326/
[10:58] <danilos> bigjools, well, both of these, if you know anything about them
[10:58] <bigjools> danilos: the first one was sorted with a "proper" fix that landed
[10:58] <bigjools> 2nd one is salgado's auth store fix
[10:59] <danilos> bigjools, ok, cool, thanks for the update
[11:01] <salgado> jml, around?  if I use 'ec2 land' on someone else's branch, using a merge proposal, who gets email if tests fail?
[11:13] <bigjools> wgrant: did you do any local testing with the gina patches?
[11:13] <bigjools> dogfood doesn't have a Debian archive handy :(
[11:14] <wgrant> bigjools: Not for about a month.
[11:15] <wgrant> There's a small 3.0 archive around somewhere..
[11:15]  * wgrant hunts.
[11:15] <bigjools> that'd be very useful
[11:15] <wgrant> http://people.debian.org/~brlink/v3test/
[11:16] <bigjools> awesome
[11:16] <jml> salgado, everyone
[11:16] <wgrant> debmirror that, I guess.
[11:16] <jml> salgado, you and the other person
[11:16] <jml> salgado, --dry-run is your friend
[11:17] <adiroiban> salgado: can you please forward the email? Thanks!
[11:17] <salgado> adiroiban, I didn't get it
[11:17] <adiroiban> salgado: ah.
[11:17] <jml> g'night
[11:25] <salgado> adiroiban, I think I detached the thing too early so it didn't send out emails. I'm submitting it again. hopefully we'll get an email this time
[11:25] <adiroiban> salgado: well, life sucks :) . So you just know that it failed? I just started a new full test, but I ran one before pushing the changes
[11:26] <adiroiban> salgado: ok. thanks!
[11:26] <danilos> bigjools, wgrant: btw, I want to keep track of why something had to be RCed (it can be as simple as "buildbot images were not ready for this"), so what about https://code.edge.launchpad.net/~wgrant/launchpad/gina-3.0-support/+merge/15478? also, is there a bug we should add to CurrentRolloutBlockers?
[11:28] <bigjools> danilos: basically I was a slacker
[11:28] <danilos> bigjools, ok, good enough, thanks :)
[11:28] <adiroiban> salgado: ok. although don't know if I can access those logs
[11:29] <bigjools> danilos: the CP blocked everything on Friday and by the time I'd dealt with that it was very late my time
[11:30] <danilos> bigjools, right, I can see several other problems CP on Friday might have caused, ta
[11:30] <danilos> bigjools, btw, has that been CPed? is it worth CPing now (downtime and all) or can we wait for Wednesday release?
[11:31] <danilos> bigjools, it'd be a shame to cause so many issues and not even CP it if we should :)
[11:31] <bigjools> danilos: it was CPed today
[11:31] <danilos> bigjools, ok, thanks
[11:35] <danilos> adiroiban, btw, I probably won't have much time to help with bug 496361 until the rollout (I am doing release management this time around), but I'd be happy to help with the general direction :)
[11:35] <mup> Bug #496361: Register a single view (SeriesLanguageView)  for all ISeriesLanguage-implementing objects <Launchpad Translations:New for adiroiban> <https://launchpad.net/bugs/496361>
[11:36] <adiroiban> danilos: np. I still need to get the current changes landed :D
[11:36] <danilos> adiroiban, yeah, most of that is not happening before rollout either unless they are RC :)
[11:38] <adiroiban> danilos: yep. don't worry. I can wait and there are plenty of other unrelated bugs to nail
[11:39] <adiroiban> are we going to have the final release of YUI available on edge soon ?
[11:42] <danilos> adiroiban, we don't have it yet?
[11:42] <danilos> adiroiban, afaik it should be on production as well
[11:43] <adiroiban> danilos: hm... on devel it was only 3.0_p2
[11:44] <adiroiban> let me check if it was updated
[11:46] <adiroiban> ah...stupid me. it's here, but I didn't merged that branch with the latest devel
[11:46] <danilos> adiroiban, look at lazr-js/build/yui to see what YUI are we actually using
[11:52] <adiroiban> is there an annoucement list for such updated? I tried to read the bzr log, but I could not find any reference about updatind yui
[11:57] <bigjools> wgrant: gina has the most bizarrely contrived config I've ever seen
[11:58] <wgrant> bigjools: YES.
[11:58] <wgrant> But I think that's as a result of bug #1234
[11:58] <mup> Bug #1234: Gina is an unmaintainable mess of command line options, environment variables and shell scripts <Launchpad Foundations:Fix Released by debonzi> <https://launchpad.net/bugs/1234>
[11:58] <bigjools> heh
[11:58] <wgrant> It's now instread an unmaintainable mess of having to add each target to schema-lazr.conf WTF
[11:59] <bigjools> no kidding, that's a royal pain
[12:01]  * wgrant tracks down that diff, just to see if it was less foul before lazr.config.
[12:01] <bigjools> wgrant: more than likely
[12:02] <wgrant> Yeah, there was no need to add it to the schema with ZConfig.
[12:08] <bigjools> wgrant: I am singularly failing to get an import done on DF, it's hanging, did you ever see that when you played with it?
[12:08] <wgrant> No. Let me try it again locally.
[12:11] <wgrant> bigjools: How did you mirror it?
[12:12] <bigjools> wgrant: rsync to my box at home then copied a tarball up to DF
[12:12] <bigjools> s/rsync/wget/
[12:12] <wgrant> Ah.
[12:15]  * wgrant tries the remember the correct gina sacrifice sequence.
[12:15] <wgrant> s/the/to/
[12:22] <wgrant> bigjools: It's importing OK for me.
[12:23] <bigjools> ok cool
[12:23] <bigjools> I think my goat is the wrong breed
[12:23] <wgrant> Argh, some borkage in the later ones. I think my DB lacks some of the sections.
[12:23]  * wgrant adds them and tries again.
[12:24] <wgrant> I think the docs for pocketrelease and distroseries are around the wrong way.
[12:24] <wgrant> pocketrelease needs to be the source (in this case 'test'), while distroseries needs to be the destination.
[12:24] <wgrant> Whereas the docs in schema-lazr.conf say the opposite.
[12:25] <bigjools> it doesn't help when I get a traceback from os.path.join
[12:25] <wgrant> Ewww. What did you do?
[12:26] <bigjools> nothing!  line 57 in gina/archive.py causes it
[12:26] <wgrant> You must have left some config option blank?
[12:27] <bigjools> I did, but then the prod config does too
[12:27] <wgrant> bigjools: Hm, where does it hang for you?
[12:28] <wgrant> Oh, that archive will fail now. Curses.
[12:28] <bigjools> it's ok now
[12:28] <wgrant> (it uses lzma compression parts, whereas dak recently changed to not support lzma, and I was advised to follow)
[12:28] <bigjools> yeah I saw that and wondered
[12:29] <wgrant> Should I hack a better archive together, or do you want to cheat and add lzma to the regexps?
[12:29] <bigjools> I'd rather have a realistic archive
[12:30] <wgrant> I wonder if it will import an archive produced by itself.
[12:30] <bigjools> hmmm well my test run just went bang
[12:30] <wgrant> Uhoh. What died?
[12:31] <bigjools> lots
[12:31] <bigjools> I'll capture a log this time!
[12:32] <jtv> bigjools: available when you are
[12:33] <bigjools> wgrant: http://pastebin.ubuntu.com/341125/
[12:33] <jtv> adiroiban: just heard that the export-to-bzr bug was a new one, not the one I thought it was
[12:33] <wgrant> bigjools: Oh, so not actually a test run, but a run over that dodgy archive?
[12:33] <bigjools> tes
[12:33] <bigjools> yes
[12:34]  * wgrant uploads a tarball of his dev PPA with three v3 sources.
[12:34] <adiroiban> jtv: ok. np. I will not touch that branch
[12:35] <wgrant> bigjools: I'm not sure about those section errors, but the null filetype ones are because of the lzma.
[12:35] <bigjools> yeah
[12:36] <bigjools> jtv: ok, what's going on my man
[12:36] <jtv> adiroiban: sorry, I was mentally still in the middle of our earlier conversation.  I mean it's the same problem, and with the same solution on your end—it's just that a complementary way of handling it on our end has come up.
[12:37] <jtv> bigjools: hi!  was hoping we could have a call—not anything too much special, but maybe I could (1) catch up on what happened in my absence and (2) solidify my next steps
[12:37]  * jtv turn s on air conditioner
[12:38] <jtv> in the old days you needed friends & family to tell you you had heat on the brain and were spouting nonsense
[12:38] <jtv> "too much special," indeed
[12:38] <jtv> now you just see it in your own irc client
[12:44] <wgrant> bigjools: http://williamgrant.id.au/f/1/2009/wgrant-v3-ppa.tar.bz2, import from lucid.
[12:50] <bigjools> wgrant: ok ta
[12:50] <bigjools> going OTP, will do it in a but
[12:50] <bigjools> bit
[13:23] <wgrant> bigjools: Do you need anything else from me?
[13:23] <wgrant> I should sleep.
[13:23] <bigjools> wgrant: I'll just grab your archive
[14:01] <Pilky> hey, any launchpad devs available
[14:06] <james_w> did someone land a change that made patches have different icons to normal attachments?
[14:08] <Pilky> just wondering if anyone would be interested in seeing and commenting on some mockups I made for ideas on how to improve some parts of launchpad's UI
[14:11] <maxb> It must be said that the final release-week of the year before everyone goes on holiday is possibly not the optimal time for that question :-)
[14:11] <Pilky> heh
[14:12] <Pilky> well I had some free time this weekend and just had a little play around
[14:12] <Pilky> hoping it might be of interest, maybe for the new year
[14:13] <maxb> You may be lucky, and find someone who hasn't got anything release-critical going on
[14:13] <maxb> But dropping an email to the launchpad-dev list might be preferable
[14:13] <Pilky> ok thanks, I'll try that
[14:14] <flacoste> morning launchpadders
[14:51] <adiroiban> how I can aquire devmode in a View?
[14:53] <adiroiban> devmode and icingroot
[14:55] <noodles775> adiroiban: what is it that you want to do? (I've normally only needed those from the template, where they are defined in the base layout template)
[14:55] <adiroiban> I need to include the translations.js in a template/view
[14:56] <adiroiban> i saw henning doing this for the recent translations.lp.net/+langauges page
[14:56] <noodles775> Great, he'd be a good person to ask then (and should be around)... henninge ^^ :D
[14:57] <adiroiban> :)
[14:57] <henninge> adiroiban: hang on, I'll look it up
[14:57] <adiroiban> henninge: no hurry
[14:58] <adiroiban> btw. what is the policy/rule for js function? Should I add them in translations.js ? or i should create a different namespace for each view?
[15:00] <henninge> adiroiban: add to translations.js unless you are doing something really big
[15:01] <henninge> adiroiban: is this what you meant? http://paste.ubuntu.com/341215/
[15:01] <adiroiban> henninge: yes
[15:02] <henninge> adiroiban: I am not sure where icingroot is coming from
[15:02] <noodles775> it's set in the base template iirc.
[15:02] <henninge> adiroiban: this snippet is from lib/lp/translations/templates/translation-import-queue-macros.pt
[15:03] <henninge> noodles775: I suspected something like that
[15:05] <adiroiban> henninge: I saw your changes... and I have a similar setup... but I don't know why those variables are not aquired
[15:11] <adiroiban> henninge: solved
[15:12] <henninge> oh cool, what was it?
[15:12] <adiroiban> I think they are defined as global in the main-macro
[15:13] <adiroiban> so I had to add   metal:use-macro="view/macro:page/main_only"
[15:13] <henninge> ah yes, you need that.
[15:13] <adiroiban> as I in a +fragment, it was not previously added
[15:14] <adiroiban> as I was in a +fragment, it was not previously added
[15:15] <adiroiban> uh... i can not use view/macro:page/main_only in a +fragment, as it adds the header
[15:17] <henninge> adiroiban: what exactly do you mean by +fragment?
[15:17] <adiroiban> henninge: https://dev.launchpad.net/Web/TemplateCodeReuse
[15:18] <henninge> is that a fragment of HTML returned through API?
[15:18] <adiroiban> Fragment Views
[15:19] <adiroiban> I'm working on distroseries-langchart.pt for filtering preferred langauges
[15:20] <henninge> adiroiban: I guess you will have to do the including in the main page.
[15:21] <adiroiban> henninge: I will try, but I don't think it will work, as it looks like they are done in different requrest
[15:21] <henninge> adiroiban: that does not matter to the browser, though. It's the browser that will evaluate the js.
[15:22] <adiroiban> henninge: just a second . what should I include ?
[15:22] <adiroiban> :)
[15:22] <adiroiban> the macros is already included
[15:22] <henninge> http://paste.ubuntu.com/341215/ ?
[15:22] <adiroiban> in the main page
[15:23] <adiroiban> henninge: true. thanks!
[15:23] <adiroiban> silly me
[15:57] <noodles775> danilos: when is devel closing (it's not mentioned on the CRB page...)
[16:01] <jtv> henninge: love what you did to the +languages page
[16:02] <danilos> noodles775, it's already closed, I didn't know one had the option to not close it when we enter RC mode
[16:02] <henninge> jtv: me, too ;)
[16:02] <noodles775> danilos: ok, thanks.
[16:11] <sinzui> Chex: ping
[16:12] <Chex> sinzui: hello there
[16:13] <sinzui> Chex: A user just identified https://edge.launchpad.net/~vicodin-online-35 That account must be SUSPENDED
[16:14] <Chex> sinzui: ok, looking at it now.
[16:17] <sinzui> mars: flacoste: anyone looking into SPAM: search for "buy" in review projects" https://edge.launchpad.net/projects/+review-licenses?field.search_text=buy&field.active=True&field.active-empty-marker=1&field.license_reviewed=False&field.license_reviewed-empty-marker=1&field.license_approved=False&field.license_approved-empty-marker=1&field.license_info_is_empty=&field.license_info_is_empty-empty-marker=1&field.licenses-empty
[16:22] <sinzui> Chex: This user registered 12 project creating spam: https://edge.launchpad.net/~azartok. Please SUSPEND this account.
[16:25] <Chex> sinzui: both of those accounts are now suspended.
[16:25] <sinzui> Chex: you rock.
[16:34] <Chex> sinzui: no problem :)
[16:38] <salgado> adiroiban, we're having problems with our tools that run the test suite on amazon's ec2, but you might be able to find all test failures by running just './bin/test -vvm translations'
[16:40] <adiroiban> salgado: ok. thanks. I just ran ./bin/test -vvc -m translations --layer PageTestLayer and everhing was ok. I will try with the full module
[16:46] <adiroiban> salgado: or maybe I should merge again with devel
[16:48] <salgado> adiroiban, that shouldn't be needed; just running all translations tests should tell you which ones are failing
[16:49] <adiroiban> salgado: ok. I have started the test. it should be ready in about 1h
[17:26] <deryck> sinzui, ping
[17:32] <sinzui> hi deryck
[18:07] <sinzui> bac: how goes the QA of mimetypes? Can you prove that I was incompetent?
[18:29] <MTecknology> Could some brilliant mind please look at bug 496360 and tell me what you think?
[18:29] <mup> Bug #496360: Can no longer log in with OpenID modules <Drupal OpenID module:New> <Launchpad itself:New> <https://launchpad.net/bugs/496360>
[18:30] <mars> gary_poster, ^ ?
[18:34] <MTecknology> The standard OpenID still works for Drupal. I drupal-teams still seems to work. I don't know how to verify drupal-launchpad. If there is an issue in drupal-launchpad it's likely the cause of a change in Launchpad. If this is the case I need to figure out what changed so I can correct this in the module.
[18:36] <gary_poster> MTecknology: To my knowledge, that code has not changed for > 1 month.  I'll go verify after I send this message.  Another team is in charge of the openid code now, but we deploy it for Launchpad, so I can check for revision numbers.
[18:36] <gary_poster> I don't quite understand this part of the original report: "However.. If I use the Drupal OpenID module and used the launchpad openid string. Obviously creating a different user in the process. This worked fine."  Does that imply there could be a change to the db alone (not the code)?  I don't understand what the person did in that report.
[18:37] <bumba> Hello everybody! can anybody tell me why I can't post BB message type of forum, although I am rezistered with them? is it because of my java plug-in?
[18:38] <gary_poster> MTecknology: mm, there were some changes for Py 2.5 support, but that should not be pertinent...
[18:39] <bumba> Hello gary_poster can I ask you a question?
[18:40] <gary_poster> bumba: sure, though I must admit that I don't understand what you just asked.  Is it openid related?
[18:40] <danilos> noodles775, hi, I am looking at your RC
[18:42] <noodles775> Thanks danilos... I'll be back and forth for a bit, but feel free to ask any questions :)
[18:44] <gary_poster> MTecknology: I can verify that the code has not changed except for one test line difference (for Python 2.5) since 2009-10-18
[18:44] <MTecknology> gary_poster: I'm lost right now because Drupal 5 is used for the fridge and the module code hasn't changed in long time. The code for the modules in Drupal 6 changed recently but those are trivial as well.. It affected Drupal5 and Drupal6 at the same time..
[18:44] <MTecknology> gary_poster: thanks for looking at that
[18:45] <mars> gary_poster, I was just doing QA on r10025, the buildonce_eggs target.  I'm not sure that the target is doing what you expect.
[18:46] <gary_poster> MTecknology: ok, let me ask SAs if we have seen any DB changes recently.  That's the only other thing that could affect this AFAIK.  What's the most specific time frame you can give me for when this started breaking?
[18:46] <gary_poster> mars, ok, thank you.  I tested it, but certainly may have missed something.  what are your concerns?
[18:47] <mars> gary_poster, "find eggs -name '*pyc'"  and find eggs/ -name '*pyc'" return different results in trunk/.
[18:48] <danilos> noodles775, btw, can you add it to CurrentRolloutBlockers as well so we can track QA and status?
[18:48] <danilos> noodles775, oh, it's there, nm me :)
[18:48] <mars> gary_poster, since eggs is symlink in trunk, "find eggs" won't work.  You have to explicitly tell it to enter the directory, with "find eggs/", or "find -L eggs"
[18:48] <MTecknology> gary_poster: I just noticed about 2 weeks ago that something was funky. I normally stay logged in for a while so that's teh best I can do... sorry
[18:49] <gary_poster> mars, for LOSA's this will not be a symlink.  you should test this by moving your egg directory aside
[18:49] <mars> gary_poster, since this fix is for production, and assuming that in production eggs/ is a hardlink or directory, this may not be an issue
[18:49] <gary_poster> mars: right.  If the eggs directory is missing, one is created.  that's what the LOSAs use
[18:50] <gary_poster> MTecknology: eh, ok, that's pretty broad, but I'll see what the SAs can give us.
[18:50] <mars> gary_poster, ok, I'll mark it as OK then
[18:51] <gary_poster> thank you mars
[18:56] <danilos> noodles775, I'll be leaving soon; the change itself looks good, and I can see how it's pretty important; good QA is a must though :)
[18:59] <MTecknology> gary_poster: sounds like it was working for one person on the fridge on the 10th. They didn't try to log in again until today.
[18:59] <gary_poster> MTecknology: ok cool, I'll add that to the description
[19:02] <MTecknology> gary_poster: thanks. It could very well be a bug in Drupal or the modules. I'm just trying to follow the logical order.
[19:03] <gary_poster> MTecknology: ack, makes sense
[19:05] <danilos> noodles775, I'll be leaving now, and the branch looks good
[19:10] <MTecknology> gary_poster: It doesn't look like Drupal changed any in this time...
[19:11] <MTecknology> gary_poster: actually..... The moment where it seemed everything broke is after the last time LP went down
[19:11] <MTecknology> err - the time before that maybe..
[19:13] <gary_poster> MTecknology: Drupal not changed: OK.  I have my DB question floating the proper place.  That--or some certificate thing or something--is the only thing I can figure.
[19:13] <gary_poster> MTecknology: everything broke after the last time LP went down: How do you mean?  I thought you said someone had success on Dec 10?
[19:13] <gary_poster> MTecknology: will be afk for a bit.
[19:14] <MTecknology> gary_poster: I guess too much changed for me to have any specifics on the issue. I guess I'm just hoping it's somethin with Launchpad itself because D5 and D6 have the same issue when nothing changed with either one.
[19:16] <MTecknology> If there was a db change that would make plenty of sense that this exact issue come up.
[19:16] <noodles775> Thanks danilos - yep, I'll run it through on dogfood first thing tomorrow.
[19:16] <MTecknology> gary_poster: thanks for looking into this. I might be afk a little bit too.
[19:19] <danilos> ok, approved, I assume this is not something that can even be remotely testable or usable on edge solely, so let's get it in
[19:19] <danilos> noodles775, do some good QA on it, especially with the big refactorings you've had in soyuz recently :)
[19:20] <noodles775> yes, I spent quite a bit of time dogfooding the IBuilder refactoring already.
[19:20] <noodles775> Will do.
[19:44] <bac> sinzui: i was able to confirm the incorrect mimetype downloading from staging.  i'm about to do more testing on lp.dev
[19:44] <sinzui> bac: Should we revert, or try to fix it?
[19:45] <bac> sinzui: i don't see any reason to revert.  we're no worse off now than before
[19:45] <sinzui> bac: is the patch mime-type working?
[19:45] <bac> i'd like to investigate to come up with a fix...at least understand what is happening
[19:47] <bac> sinzui: http://paste.ubuntu.com/341393/  you see here that the local mimetypes modifications work, at least locally using bin/py
[19:48] <bac> sinzui: this, however, is still messed up:  wget -S http://staging.launchpad.net/sachco/trunk/0.1/+download/words.tbz2
[20:08] <thumper> danilos: ping
[20:09] <mars> leonardr, ping, is there some way to tell launchpadlib to only use cached responses?  In essence, to work offline?
[20:09] <leonardr> mars: no, because not everything is cached. for instance, collections
[20:09] <leonardr> if everything were cached, then there would be no problem using it offline
[20:10] <leonardr> it would work automatically
[20:10] <wgrant> bigjools: Thanks.
[20:10] <bigjools> np
[20:10] <mars> leonardr, ah.  My use-case is running the same spam filter script over and over again.  I don't want to keep re-fetching person objects if I don't have to.
[20:10] <bac> hi gary_poster
[20:11] <mars> leonardr, it should be easy enough to code something myself
[20:11] <leonardr> mars: well, it might be making conditional requests to see if those person objects have changed. i don't know if launchpad sends cache settings
[20:12] <mars> leonardr, yeah, but getting 304s is slow, too
[20:12] <mars> or HTTP HEAD requests
[20:12] <wgrant> bigjools: What do you know about the dpkg upgrades?
[20:13] <bigjools> wgrant: not done yet
[20:13] <leonardr> mars: what i was saying was, to avoid 304s launchpad should send cache directives so the client knows how long it can wait between conditional requests
[20:14] <leonardr> but i don't think we do that
[20:15] <mars> ah
[20:16] <mars> leonardr, alright.  Actually, in this case I think a launchpadlib caching wrapper is the right approach.  I can set it as aggressive as I want.  I'm not actually worried if the data goes stale, I just want the speed.
[20:16] <leonardr> mars: got it. what kind of wrapper were you thinking of?
[20:17] <mars> leonardr, so that looks like a custom script, not a launchpadlib feature
[20:17] <leonardr> ok, i was just wondering if i could be helpful
[20:17] <mars> leonardr, I was just going to wrap the launchpad.people object, and write out JSON to the disk (unless there is something already in Python...)
[20:18] <mars> desktopcouch....
[20:19] <leonardr> mars: so it sounds like you want to use launchpadlib once to get the data, save the data to disk, and then use the data locally
[20:20] <leonardr> unless you need to do some complex manipulation where the full launchpadlib class api would be useful
[20:20] <mars> leonardr, yep.  But I want to keep fetching more data, as I add features to the script.  Thus a cache, rather than a one-off data dump.
[20:20] <mars> right
[20:24] <mars> leonardr, come to think of it, a feature to load an object from JSON should be possible, since representations have a self_link.
[20:25] <leonardr> mars: not sure exactly what you want. you can pass a self_link into lp_load but that will go over the network
[20:25] <leonardr> do you want to create a Person from the JSON data?
[20:25] <mars> leonardr, since you can call launchpad.load('some/person'), then you should be able to call launchpad.from_json('{.... self_link, 'some/person' }')
[20:25] <leonardr> you might be able to do that with some hacking
[20:26] <leonardr> yes, you shoudl be able to call whatever code lazr.restfulclient calls to instantiate an object from json--i don't know what that is
[20:31] <wgrant> bigjools: Did my test archive import OK?
[20:31] <bigjools> wgrant: yep
[20:31] <wgrant> Phew.
[20:32] <bigjools> I have more problems now though
[20:32] <wgrant> Hm?
[20:33] <bigjools> we're doing some testing on DF and when the buildd-manager collects a build, it bails out with "Scanning failed with: 'NoneType' object has no attribute 'utf_8_decode'"
[20:33] <wgrant> I love the way it swallows exceptions.
[20:33] <bigjools> I don't
[20:33] <bigjools> I'm trying to see if there is by some miracle a traceback hiding in the error
[20:34] <wgrant> IIRC you just have to remove the try/except.
[20:35] <wgrant> Much like you need to hack the code to change the debug level :(
[20:35] <bigjools> aha I see something that might help
[20:40] <wgrant> I've never seen that exception locally, and I've gone through quite a few builds in the past week.
[20:41] <mwhudson> bigjools: 'NoneType' object has no attribute 'utf_8_decode'" ??
[20:41] <mwhudson> that was the error that caused so much fun in the last rollout
[20:41] <bigjools> mwhudson: correct
[20:41] <bigjools> really?
[20:41] <bigjools> tell me mo'
[20:41] <mwhudson> yeah
[20:42] <mwhudson> bigjools: it's probably to do with the crazy sys.modules mangling bin/run and _pythonpath.py do
[20:43] <bigjools> mwhudson: hmmm how was it fixed?
[20:43] <bigjools> it's buggering up dogfood now
[20:43] <mwhudson> bigjools: it was a bit random, it seems like it probably was fixed by accident by the python 2.5 upgrade
[20:44] <mwhudson> (the problem was restarting the appservers)
[20:44] <mwhudson> bigjools: buildd-manager will likely be a different case, slightly
[20:44] <mwhudson> bigjools: can you pastebin a traceback?
[20:44] <wgrant> Ha ha ha.
[20:44] <wgrant> This is buildd-manager.
[20:44] <wgrant> There is no traceback.
[20:44] <bigjools> mwhudson: if twistd didn't swallow the traceback ...
[20:44] <bigjools> well actually it's buried in the error object, Imight be able to get it
[20:45] <mwhudson> oh right, yay
[20:46] <mwhudson> bigjools: can you hack _pythonpath on dogfood?
[20:47] <bigjools> if it doesn't require root, yeah
[20:47] <mwhudson> bigjools: put 'print "deleting", k' before the del sys.modules[k] around line 238
[20:50] <bigjools> mwhudson: I got a traceback anyway
[20:50] <mwhudson> bigjools: ah cool
[20:51] <bigjools> http://pastebin.ubuntu.com/341426/
[20:51] <bigjools> mwhudson: is that pythonpath hack still useful?
[20:54] <mwhudson> bigjools: yes, still worth a try
[20:55] <mwhudson> if buildd-manager doesn't print anything out on startup i'd be a bit stumped...
[20:55] <bigjools> it doesn't
[20:55] <bigjools> well, searching for "deleting" shows nothing
[20:55] <bigjools> unless it's swallowed by something
[20:56] <mwhudson> hm, shouldn't be
[20:56] <mwhudson> i guess you can put a print in at module level to check that
[20:59] <bigjools> mwhudson: what do you mean?
[21:02] <mwhudson> bigjools: if you just put "print 'bleargh'" as the first line of the module, and that doesn't show up then you know it's being swallowed somehow
[21:02] <bigjools> mwhudson: which module?
[21:02] <mwhudson> bigjools: _pythonpath
[21:02] <bigjools> ok
[21:07] <bigjools> well I;ve not done that yet
[21:07] <bigjools> I just rebuilt my environment
[21:08] <MTecknology> gary_poster: You hear anything back yet?
[21:08] <mwhudson> bigjools: some background as to what is probably happening here
[21:08] <mwhudson> bigjools: when a python module is garbage collected, all the globals in that module are set to None
[21:08] <bigjools> now I get http://pastebin.ubuntu.com/341430/
[21:08] <bigjools> mwhudson: different error now
[21:09] <mwhudson> bigjools: !!
[21:09] <bigjools> the librarian is rejecting the upload!
[21:09]  * bigjools bangs head on wall
[21:10] <mwhudson> bigjools: almost all of the traceback is the same though
[21:10] <bigjools> yeah, it was failing uploading to the librarian before too
[21:10] <wgrant> Woo corrupt librarian files on production.
[21:10] <mwhudson> i don't know if that is a clue or not though
[21:12] <gary_poster> MTecknology: yes, from SA: "there was an architectural change made a little while back.  [authmaster changed] i am not sure exactly when that cutover was made though".  Not sure how it could be related to the issue you describe.  I need to escalate to two other people, one in a different timezone.  Do you happen to have access to error messages on fridge?
[21:12] <gary_poster> (other than "OpenID login failed" :-) )
[21:12] <MTecknology> gary_poster: unfortunately no - But I can get to logs on my own system
[21:13] <gary_poster> MTecknology: +1.  Could you put 'em on a pastebin?
[21:13] <bigjools> gary_poster: if you have a moment can you check http://pastebin.ubuntu.com/341430/ and tell me what you think about the error?
[21:13] <gary_poster> looking bigjools
[21:15] <gary_poster> bigjools: I'm not a librarian guy, but it looks like we are encoding the filename string incorrectly.  Do we know what the filename was?
[21:16] <MTecknology> gary_poster: I don't seem to have anything useful in apache logs..
[21:16] <bigjools> gary_poster: this is the buildd-manager, it hasn't changed in many cycles
[21:16] <gary_poster> MTecknology: :-(
[21:16] <bigjools> gary_poster: and now I get this error on dogfood
[21:17] <MTecknology> gary_poster: you have an OpenID account I assume.. You can try to log into staging.profarius.com to see what happens if you'd like.
[21:17] <gary_poster> bigjools: sure, but this is an issue on launchpad calling the librarian afaict
[21:17] <bigjools> gary_poster: did the librarian client change with regard to utf?
[21:17] <gary_poster> MTecknology: ok.  was going to try one of the openid demo sites.
[21:17] <MTecknology> sd.ubuntu-us.org if you would prefer a site with trust root verified
[21:18] <gary_poster> bigjools: not to my knowledge but will look
[21:19] <bigjools> gary_poster: ok - the other possibility is that it's related to the other error before I restarted the librarian: "'NoneType' object has no attribute 'utf_8_decode'"
[21:19] <bigjools> which mwhudson says we had on the last rollout
[21:20] <gary_poster> bigjools: ah, yes we did.  do a make clean and a make just to be careful
[21:20] <bigjools> ok
[21:21] <MTecknology> gary I can get you the link to the source code for that stuff if you'd like.
[21:21] <gary_poster> MTecknology: no way I have time for that I'm afraid.  The sd site has timed out twice.  I'll try the other one
[21:22] <MTecknology> gary_poster: I'm not sure what's up with that actually.. staging.profarius.com or profarius.com should work without issues
[21:22] <gary_poster> MTecknology: yes it does but error message is as useless as one from fridge, I'm afraid.
[21:22] <MTecknology> ya
[21:23] <MTecknology> I'll try to grab watchdog errors if any exist
[21:25] <bigjools> gary_poster: phew, that seems to have fixed it
[21:25] <gary_poster> bigjools: good.  sorry for the scare.  fix for scare in progress.
[21:25] <bigjools> gary_poster: :)
[21:27] <MTecknology> gary_poster: I don't think this helps.... http://paste.ubuntu.com/341443/
[21:27] <MTecknology> gary_poster: is 67.195.112.239 one of yours?
[21:30] <gary_poster> MTecknology: agree doesn't help; and no, I don't see myself on there, oddly enough.  Going to escalate on another channel; will let you know if I discover anything.  May not get a reply till tomorrow.  If I haven't contacted you by tomorrow, please feel free to ping me.
[21:33] <MTecknology> gary_poster: thanks. I think this may actually affect some people paying canonical. I'm not sure but it might be a good idea to ping stu or elachuni and ask them if they know anything.
[21:34] <gary_poster> MTecknology: ok, will do (and escalating elsewhere now)
[21:35] <MTecknology> :)
[22:03] <bac> hi abentley
[22:09] <abentley> bac: Hi.
[22:10] <bac> abentley: how can i make 'bzr send' create a MP against db-devel instead of devel, without editing my locations.conf to change the target for trunk?
[22:11] <abentley> bac: Specify db-devel as the first parameter to "bzr send".
[22:12] <bac> abentley: simply 'bzr send db-devel'?
[22:13] <bac> or the full URL i suppose
[22:14] <abentley> bac: You need to specify a URL that Launchpad recognizes.  lp:launchpad will work, or lp:~launchpad-pqm/launchpad/db-devel, etc.
[22:15] <abentley> You can also use the bzr+ssh or http URLs if you like.
[22:15] <bac> abentley: great.  i'm creating a dbsend alias similar to the dbsubmit you suggested a while back
[22:16] <abentley> bac: Cool.  I have one like that, also.
[22:16] <abentley> bac: You can use a local mirror, too, if the local mirror has its public_branch set.
[22:17] <bac> abentley: this page needs updating, i think.  'bzr lpsend' is deprecated, right?  https://dev.launchpad.net/WorkingWithDbDevel
[22:18] <abentley> bac: It's a little mixed up.  lpsend used "send --no-bundle", which is a good idea right now, because we seem to have some bugs applying bundles.
[22:19] <bac> ok
[22:19] <bac> abentley: thanks for your help
[22:19] <abentley> bac: No problem.
[22:38] <jml> thumper, inline propose for review pls.
[22:39] <thumper> jml: eh?
[22:39] <jml> thumper, https://code.edge.launchpad.net/~jml/launchpad/fix-release-hot-bugs-486437 -- clicking "Propose for merge" takes me to another page. I'd love the form to be inline
[22:40] <thumper> jml: one day...
[22:40] <jml> thumper, :)
[22:44] <thumper> beuno: ping
[23:05] <beuno> thumper, pong
[23:05] <thumper> beuno: hey, got a few minutes for a call?
[23:06] <beuno> thumper, I'm double late, even for a south american
[23:06] <beuno> tomorrow?  :)
[23:26] <maxb> Whoever implemented display of PPA packages obsoleted by the primary archive, you rock! :-)
[23:41]  * mwhudson lunches
[23:42]  * thumper lunches