[04:14] yay windmill tests have hung for me in ec2 [04:48] thumper, mwhudson: I think that was my problem as well yesterday. [04:48] cody-somerville: interesting [04:49] cody-somerville: and sucky at the same time [07:10] mwhudson: istm that you could usefully ask for things to be cached for a short time even without revalidation [07:10] as long as user-forced refreshes do go through to the real server [07:10] poolie: very likely [07:13] mwhudson: and can you see errors logs from codebrowse now directly, or through a losa? [07:14] poolie: i can still log in to the machine it runs on [07:14] we should hand that over [07:14] yeah [07:15] and do you use that for anything else? [07:17] i have this script i use to get the tracebacks out of core files [07:17] but that actually works now, so we should be able to get the losas to do taht [07:19] mwhudson: don't we drag the codebrowse logs onto devpad? [07:19] spm: pass [07:19] heh, looks... [07:27] good morning === almaisan-away is now known as al-maisan [09:23] wgrant: Happy birthday :) [09:24] bigjools: Thanks! [09:26] wgrant: so... 11? 12? years old? [09:26] spm: Heh. 19. [09:27] :-) Happy Birthday man, enjoy it eh! [09:49] spm: lol [09:54] wgrant: yeah, happy birthday. Facebook told me earlier today but I just clean forgot [10:08] yay for the kernel thinking I don't have any USB ports any more [10:08] Really no USB ports, or just not detecting mass storage devices? [10:08] I've had the latter problem on and off; I think I need to reinstall. [10:09] well my keyboard stopped working [10:10] Niice. [10:10] and unplugging/re-plugging makes no difference [11:00] Morning, all. [11:17] intellectronica, ping [11:22] mrevell: ahoi [11:22] intellectronica, Sorry, unping [11:22] mrevell: np, unahoi [11:22] :) === henninge_ is now known as henninge [11:43] danilos: Can you look at this and tell me if I included all the cases we discussed? http://paste.ubuntu.com/419756/ [11:43] danilos: I'll be out to lunch soon, actually. [11:47] henninge, western-plural-form can probably be '!(n>1)' or similar as well [11:48] ok, I'll add that [11:49] henninge, it'd be nice to filter out our list using this code to see what's left :) [11:50] good idea, I'll try that after lunch [11:50] henninge, cool, thanks [11:52] henninge, btw, !(n>1) is not really the same one, so ignore that === salgado-afk is now known as salgado [13:04] intellectronica, I've created a branch with a help pop-up to explain how bug heat is calculated -- https://code.edge.launchpad.net/~matthew.revell/launchpad/bug-heat-help-bug-544799 -- (lib/lp/bugs/help/bug-heat.html) -- I'd love your thoughts before I proceed. [13:10] mrevell: i'm just about to break for lunch, but will have a look as soon as i'm back [13:11] thanks intellectronica [13:12] I too am breaking for lunch === mrevell is now known as mrevell-lunch === mrevell-lunch is now known as mrevell [13:55] leonardr: i applied your patch yesterday and it worked [13:56] bac, great [14:00] BjornT, ping [14:01] deryck: i'm on the phone [14:01] BjornT, ok. I want some phone time, when you are available. :-) [14:04] perhaps gmb can help. :-) gmb, do you have time for a mid-imp type of discussion? [14:08] leonardr, seen this? http://www.mikealrogers.com/archives/695 [14:09] mars, no, glad we're using simplejson [14:09] or maybe i should have hoped we were using stdlib json so we could easily get a huge speedup [14:09] hehe [14:10] deryck, Sure. [14:10] leonardr, worth glancing at the comments as well. People explain why and where things are going. [14:10] gmb, cool, thanks! mumble? [14:10] deryck, Yep. [14:11] gmb, for reference see the diff on the mp for bug 494257. [14:11] Bug #494257: Users not notified in email when subscribed by someone else

[14:11] * gmb looks [14:20] deryck, eggs/zope.event-3.4.1-py2.5.egg/zope/event/__init__.py [14:24] heh. [14:25] zope.event.notify is called *a lot* [14:25] salgado: Your last wiki edit to PythonMigrationStatus makes very little sense in multiple ways :-) [14:26] maxb, other than the order of the link/text in [[link|text]]? [14:26] that's one :-) [14:27] the other is that I certainly didn't create a launchpad branch based on a shipit branch [14:28] oops [14:29] * salgado reverts all that [14:29] thanks for spotting it, maxb [14:29] thanks for sorting pygpgme :-) [14:35] Is anyone fixing the failures in the latest lp buildbot run? [14:41] lifeless: how much data is in the 'message' you're passing in to createContent? [14:42] you're more likely to get a timeout if you pass a huge chunk of text [14:49] for instance, i'm passing in a 68k string and it's been about 10 minutes with no response [15:00] hey intellectronica [15:01] hey mrevell [15:01] mrevell: so, i asked deryck to have a look at it too, since i can't decide about the table [15:01] Ah okay [15:01] intellectronica, I can look now. [15:01] reviewers meeting starting now [15:02] mrevell: i have another concern. how do we make sure that the help text stays current when the implementation chnages? [15:02] adeuring: if you have some time today or tomorrow, I'd like to chat (skype even?) about parallelizing work on some queries. I know this is kind of not in your regular schedule, so if you're booked please feel free to push back. [15:02] mrevell: other than that, ask anything you might want about linking it r whatever :) [15:02] kfogel: no, let's talk [15:02] adeuring: great. I'll turn on skype. [15:02] one sec [15:03] jtv: just fyi I subscribed you to bug #567689 [15:03] Bug #567689: Have a deterministic way to check OOPS generation in tests. [15:03] kfogel: sorry, could we delay the chat until the LP reviewer meeting finishes? [15:03] kfogel: oh good more mail, just what I needed :) [15:03] adeuring: sure! just ping me here when ready [15:03] I'm bad in multitasking ;) [15:03] adeuring: np [15:03] kfogel: ok, I'll ping you [15:03] jtv: hey, want me to unsub you? [15:03] jtv: oh wait, I probably can't do that... [15:03] jtv: (hmmm, asymmetry in the UI there...) [15:04] kfogel: I think I'll unsubscribe myself. Bjorn is right, but I have no _particular_ interest in seeing this through. [15:05] intellectronica, wrt you concern ... I get asked that pretty much every time I document something in this level of detail, and I don't have an answer other than that I hope that my efforts to speak regularly with the team leads will mean I pick up on such changes. With so many potential changes, I'm not sure there's a way of tracking them all. [15:06] jtv: ok sorry about that [15:06] mrevell: sounds like quite a problem to me. one thing i think is worth doing is adding reminders to update the documentation in the code and in the doctest documenting bug heat. i can help you do that. other than that i guess your alertness is the best we've got :) [15:06] kfogel: no worries... is the test itself fixed now? [15:06] and of course the alertness of anyone working on bug heat [15:07] That's a great suggestion, thanks intellectronica. Your help adding those reminder would be really useful. What would you suggest? [15:07] SOmething in bugheat.py, presumably. [15:07] jtv: no, I have not fixed the test. I don't _think_ anyone else has. I would like to. I have some kind of urgent stuff on my plate. On the other hand, this isn't so hard to fix. Sigh :-). [15:07] jtv: it'll be my bonus today if I can get this other stuff done [15:08] kfogel: don't leave it on your plate too long or someone'll eat it [15:08] mrevell: something like "# Don't forget to change the documentation of heat calculation in /path/to/file.html if the implementation changes". i'll find the relevant places for you (or i can just give you a patch). [15:08] intellectronica, At the risk of sounding lazy, a patch would rock :) [15:09] mrevell: ok, but don't be surprised if this is reflected in your 360 review! [15:09] watching mrevell and intellectronica escalate :-) [15:11] Haha [15:12] intellectronica, Who says I'll ask for your input? :) Okay, seriously, if you have time I'd be interested in learning where would be the best place to do this, rather than just getting a patch. [15:13] mrevell: I was assuming the place where the bug-heat.html doc points to? [15:14] kfogel, Yeah, that's what I thought. I was thinking maybe in the comment on the BugHeatCalculator class. Sound right intellectronica? [15:15] mrevell: yes indeed. also i think there's a doctest somewhere. let me find it for you. [15:18] mrevell: actually, it's a unit test. lib/lp/bugs/scripts/tests/test_bugheat.py (so a python comment there too) [15:18] Ah, thanks intellectronica. I'll update the branch. Before I do that, can you tell me more about you concerns wrt the table approach? Is it misleading? [15:20] intellectronica, mrevell -- did either of you have specific spots for me to look at, or just a general glance over the doc? [15:21] mrevell: no, in fact, i think it's too detailed. i'm still of the opinion that most users shouldn't care much about the underlying algorithm for bug heat, and that the help we want to give them is by abstracting away from it a bit and just explaining what they can expect (with a link to the code for those geeky users that really want to know how it works). [15:21] deryck, I just want to make sure that the help doc (lib/lp/bugs/help/bug-heat.html or https://pastebin.canonical.com/31061/) is accurate. [15:21] deryck: mainly what i just described [15:22] intellectronica, ah, ok. maybe an overview with a link to the code is ok. but remember some power users will want to work out exactly what the heat for a certain bug is. [15:22] intellectronica, I don't mind this detailed description personally. [15:22] kfogel: I'm ready [15:22] mrevell: but i accept in advance that i may be in a minority here, and if you prefer to have it spelled out then at the very least i can say that it's accurate and well presented [15:22] adeuring: ok, let me get headset on [15:23] mrevell, we do actually display the number in the tooltip for the flames currently. [15:23] intellectronica: think of it as a way to offload questions -- the minority of users who *do* care will now have something to look at, and not bug us :-). [15:23] adeuring: one sec, setting up some stuff [15:23] kfogel: np [15:24] deryck, Ah, I thought we were removing that. I'll correct that, thanks. [15:25] intellectronica, I don't see the value in not explaining it. I can't imagine too many people wanting to game the system. I do think that curious people will want to understand how it works and that many of those will be unwilling or unable to read the Python. [15:25] intellectronica, But I'm not in the bugs team, so I defer to you guys. [15:25] adeuring: just for context: https://pastebin.canonical.com/31062/ (query you dug up for me the other day) [15:25] adeuring: ready [15:26] mrevell, intellectronica -- I understand intellectronica's concerns for the average user, but I think adding this help doc is mostly for the advanced users. :-) So I would describe it completely. [15:26] mrevell: toss a coin :) [15:26] there you go [15:26] heh [15:26] Okay, so I'll add the comments reminding people that the help needs updating if the calculation changes. [15:27] so I like the mrevell's doc here is general to start with -- the normal user will ready the first few lines and bail. but goes into some detail throughout. [15:27] And also remove the inaccurate thing about not displaying the actual number. [15:28] yup, that's my only comment [15:29] Thanks deryck [15:29] intellectronica, So, adding a link to the flame images. Is that straightforward? [15:29] mrevell: should be, yes. let me find for you the place where it's rendered [15:30] adeuring: https://pastebin.canonical.com/31064/ [15:30] mrevell: see bugtask_heat_html in lib/lp/bugs/browser/bugtask.py [15:33] adeuring: you're fading [15:34] intellectronica, Thanks [15:36] mrevell: yer welcome [15:37] allenap: Hi! [15:39] henninge: Hi! [15:39] allenap: I just saw that you edited the PythonStyleGuide today about multiline function definitions. [15:39] I just found out that you just moved those line. [15:39] ;-) [15:40] henninge: Yes, I didn't change the rule, just moved it and added a comment. [15:40] allenap: I just wrote a similar piece as was discussed last week, so I'll delete that old incarnation of the rule. Just so you know. ;-) [15:42] allenap: https://dev.launchpad.net/PythonStyleGuide?action=diff&rev2=12&rev1=11 [15:43] henninge: Ah, jolly good :) [15:44] bac: I have two time zones set in gcal, but I can't seem to actually use them anywhere. Am I missing something? [15:49] allenap: perhaps. i have mine set to show UTC and US/East but i have my preference set to UTC so scheduled meetings don't move around with DST [15:50] bac: You can't set or change the time zone of a meeting, it's always in the primary time zone? [15:51] allenap: i think that is right [15:51] bac: Thanks. [15:51] allenap, who would be the best person to help me getting lp.bugs.scripts.tests.test_checkwatches to pass on python2.6? https://dev.launchpad.net/PythonMigrationStatus#LaunchpadZopelessLayer is where it fails === Guest36135 is now known as joey [15:52] salgado: Probably me. [15:53] allenap, would you have some time now? or maybe just some pointers for me to start? I currently have no idea what could've caused that [15:55] salgado: I'll take a look. === jelmer is now known as ctrlsoft [16:00] allenap, I'm stepping through the failing code path and it seems to be making requests to bugzilla.gnome.org. am I misunderstanding something or is that expected? [16:00] salgado: That should be stubbed out, so something's going wrong there. [16:02] salgado: I'm still waiting for my branch to build :) [16:07] File "/home/salgado/devel/launchpad/python2.6/lib/lp/bugs/externalbugtracker/xmlrpc.py", line 101, in request [16:07] request.get_full_url(), he.code, he.msg, he.hdrs) [16:07] ProtocolError: [16:07] ------------ [16:07] allenap, that's what I got when I aliased bugzilla.gnome.org to my IP address [16:08] so I guess the stubbing doesn't work on 2.6 [16:08] salgado: Cool. [16:08] salgado: xmlrpclib changed quite a bit between 2.4 and 2.5, so I wouldn't be surprised if it has changed again. [16:12] allenap, so I'm not confident of my test for bugnotification for subscribe someone else on the view anymore. I was doing select * from bugnotification. [16:12] allenap, but I reverted to devel before my change and the view doesn't create an entry there. but I know it worked before, since someone subscribed me via the view and I got notified. [16:15] deryck: Have you seen bugnotification-sending.txt? It uses get_email_notifications() to see what has been sent. Perhaps that's of use? [16:16] allenap, yeah, the cronscript uses that and passes in an IBugNotificationSet. I forgot that. So that means the bugnotification table has to be populated. [16:17] i.e. IBugNotificationSet is just a set from bugnotification, no? [16:17] yeah, it is. he says answering his own question. :-) [16:18] deryck: Cool :) Sorry, I've got too many things stacked up in my head to be very useful right now. [16:20] allenap, no worries. Chatting out loud helps. :-) [16:28] BjornT: in https://bugs.edge.launchpad.net/malone/+bug/130902/comments/5 you write that there should be a header in bug notification emails indicating the reporter. But I don't see any such header in (e.g.) notifications I receive for bug 558657, even though I am the reporter. What header were you thinking of? [16:28] Bug #130902: Notifications don't tell you whether you're a bug's reporter

[16:28] Bug #558657: mouse usage causes Xorg CPU usage to spike, and mouse pointer becomes less responsive

[16:29] allenap, I can't seem to find where the stubbing occurs. I can clearly see checkwatches.updater.externalbugtracker.get_external_bugtracker being monkey patched, but that doesn't seem to be what I'm looking for, is it? [16:29] salgado: Mmmm, neither can I. I'm about to ask gmb. [16:29] BjornT: (not that it should matter whether I'm the reporter or not -- every notification anyone receives should indicate the reporter) [16:29] Whut? [16:29] gmb: A test for bug 496988 is failing in the python2.6 branch that salgado's looking at. [16:29] Bug #496988: KeyError in checkwatches cause the script to abort

[16:29] Ahhhh [16:29] gmb: It's actually contacting the GNOME Bugzilla. [16:30] Now then, someone asked me about this last week. [16:30] gmb: And I have no idea why. [16:30] rockstar perhaps; I don't remember. [16:30] gmb: I mean, I have no idea why it works in 2.5. [16:30] OH! [16:30] It was wgrant that asked. [16:31] And ISTR it's because something changed in the way the XML-RPC code works from 2.5 to 2.6 [16:31] salgado, allenap: Lemme check my logs. [16:31] gmb, very likely to be wgrant; he's helping us get LP to run on 2.6 [16:31] gmb: When salgado asked I assumed that remote calls, i.e. XML-RPC, were being stubbed out, but they're not. [16:31] gmb: That was my first thought. [16:31] BjornT: oh, I think I understand: you meant that it would be good for us to implement this, not that it *has* been implemented [16:32] gmb: But I don't think it's the case. [16:32] Buggeration. I had logging turned off in xchat. Let me check the proxy logs... [16:33] gmb, I confirmed the test is hitting bugzilla.gnome.org by aliasing that host to localhost and seeing the test fail with a service unavaliable error instead of https://dev.launchpad.net/PythonMigrationStatus#LaunchpadZopelessLayer [16:35] kfogel: right, we should add such a header :) [16:37] * gmb has a log; cleaning it up now [16:40] BjornT: agreed === kfogel is now known as kfogelunch [16:41] allenap, salgado, Log of my conversation with wgrant: https://pastebin.canonical.com/31077/ [16:41] The solution I suggested was to monkeypatch get_external_bugtracker [16:42] salgado, But I think there was some suggestion that we'd have problems with headers and XMLRPC in 2.6. I'm a bit fuzzy on the remembering though. [16:44] gmb, I'd like to find what's causing the tests to connect to bugzilla.g.o on 2.6. this might be affecting other tests that are not failing just because on my system they can access bugzilla.g.o, but if that's the case they'll fail when run on buildbot [16:44] salgado, What's causing it is a shoddy test. The real question is: why doesn't it happen in 2.5? [16:45] gmb: I'm going to guess that buildbot can reach gnome-bugs. [16:45] allenap, Horrible as it is, that sounds probable. [16:45] gmb: That's why it's working on 2.5. The header thing is because 2.6 is stricter. [16:46] In which case it's just a shoddy test. [16:46] * salgado tries to confirm that hypothesis [16:47] brb [16:48] allenap, gmb, confirmed. on 2.5 it also hits bugzilla.g.o === bjf is now known as bjf[afk] [16:51] salgado: Here's a fix http://pastebin.ubuntu.com/419909/ [16:51] Right [16:53] allenap, r=me, and I'll be happy to land it on your behalf if you'd like [16:53] salgado: Do you want to land it, or shall I? [16:53] salgado: Yes please :) [16:53] allenap, will do after lunch. thanks! [16:55] gosh. We're dangerously close to actually being able to run on py2.6 [16:57] salgado: A better fix: http://paste.ubuntu.com/419914/ <-- the setUp() and tearDown() for TestCheckwatchesWithSyncableGnomeProducts and TestCheckwatchesWithoutSyncableGnomeProducts had become the same, so I collapsed them together. [16:59] allenap, cool, will use that one === salgado is now known as salgado-lunch === bjf[afk] is now known as bjf === gary_poster is now known as gary-lunch === beuno is now known as beuno-lunch === leonardr is now known as leonardr-lunch === salgado-lunch is now known as salgado === kfogelunch is now known as kfogel === beuno-lunch is now known as beuno === gary-lunch is now known as gary_poster [18:54] yay: http://paste.ubuntu.com/419962/ [18:54] not the most impressive demonstration though :-) [18:57] james_w, you get a +2 for creative script output, in the proudest of hacker tradition :) [19:09] http://jameswestby.net/scratch/offline-lp.ogv <- that's a more effective demonstration [19:50] james_w, awesome! [19:54] jml: all done with twisted for extra awesome! [19:54] james_w, is this your twisted launchpadlib thingy? [19:54] yeah [19:55] james_w, I should really have a play with that. [19:55] I wrote a caching wrapper for Agent, plus an ICache implementation over desktopcouch, and put them together for the demo [19:56] james_w, it stores the data in couch? [19:56] it's all local at the moment, as it requires a patch to twisted, and a bunch of other experimental stuff [19:56] yeah [19:56] james_w: still, very cool. [19:56] james_w, can I help you land that patch to Twisted? [19:56] it's not my, it's exarkun's [19:57] jml: you can though, it's awaiting review :-) [19:57] james_w, ticket number? [19:57] james_w: that's rather nice! [19:57] 4023 [20:05] james_w, thanks. I've just waded through the discussion. Time to look at the code! === leonardr-lunch is now known as leonardr [20:09] I am currently unable to log into staging.launchpad.net. A bug has been filed here: https://bugs.launchpad.net/canonical-identity-provider/+bug/566778. Can anyone help? [20:09] Bug #566778: login.staging.launchpad.net crashes on login === al-maisan is now known as almaisan-away [20:40] james_w, review done :) [20:40] leonardr, I think I've let some stuff that you asked for slide. [20:41] leonardr, lackwit that I am, I've forgotten what it was [20:41] jml: we were going to talk about web service performance? [20:41] ahh yes. [20:42] leonardr, how about I book a time for a call tomorrow? are you working regular East Coast hours? [20:42] jml: yes, i usually get up a little early [20:42] leonardr, you don't use google calendar, I take it [20:45] jml: not really, but if it's easier for you to schedule something with my google username i can write it down on a piece of paper [20:45] leonardr, heh, no, that's fine :) [20:45] leonardr, 1500UTC ok for you? [20:46] that should be fine [20:46] the latest is here: https://dev.launchpad.net/Foundations/Webservice/Performance [20:46] lifeless said on irc earlier today that he had some ideas [20:47] Are bug assignees treated like subscribers? [20:47] sinzui: I seem to recall you talking about this a few days ago... [20:47] jpds, they get an extra serving of stale bread with their gruel and water [20:48] leonardr, cool. I'll make sure I read up before then. [20:48] leonardr, given TZ, it'll be easier for you to talk directly w/ lifeless [20:49] jml, did you or lifeless move? i thought you two were nearby [20:49] leonardr, I live in London [20:49] leonardr, since September :) [20:49] all is clear now [20:49] leonardr, obviously you don't blogface my tweetermail feed [20:50] i prefer to spacetweet [20:57] leonardr, I hear all the cool kids atomdent [20:57] leonardr, anyway, that's in my calendar, so it will happen. [20:58] can't spacedent from a 64-character alphanumeric beeper! [21:03] jpds, no, an assignee is not an actual subscriber, so assigning a private bug to a person is cruel [21:03] sinzui: Thanks. [21:04] jpds, well Launchpad is cruel to put users in an impossible situation [21:38] kfogel: https://pastebin.canonical.com/31112/ [21:38] (untested...) [21:38] adeuring: well, then I'll test it! [21:38] adeuring: thanks [21:41] adeuring: when the person is a team, I can just put the team name for 'xxxxxxxxxxxx' and it should all still work, right? [21:41] kfogel: regarding our disscussion about multiple bug task status changes between completed/ in progress: We should already get more than one result with the existing queries, the only thing we have to do is to find the shortest and longest times [21:41] adeuring: ah, beautiful, gotcha [21:41] kfogel: yes, though i think you shold try first with a person name [21:41] adeuring: oh, I will [21:42] kfogel: I think for teams, bug comments won't make much sense. but the other queries should be fine [21:42] adeuring: yeah. a team can own, and be assigned to, and subscribe to, right? [21:43] right [21:43] adeuring: I suspect "subscribed to" is the most important relation to, here. when a team visits ~teamname/+patches, all (?) the bugs that come back are ones that team is subscribed to [21:43] right? [21:44] right [21:45] adeuring: heh, it got 0 results for 'kfogel' of course, but that's only b/c no bugs *with patches* assigned to me right now. I'm visiting ~kernel-bugs/+patches now to get some better usernames [21:48] adeuring: so, hmm, when I do the first query (person assignee) with 'mgedmin' as the person, why doesn't bug 47775 show up? It has a patch, and one of its bug tasks was marked as Fix Released *after* the patch... [21:49] Bug #47775: [dapper] xrandr freezes the system (radeon, MergedFB)

[21:50] kfogel: mgedmin is not an asignee for the tasks of this bug, or am ii missing something? [21:50] adeuring: oh duh, he's reporter, my fault [21:50] adeuring: sorry [21:57] jml: thanks [22:02] adeuring: works for ~ubuntu-x-swat. 33 rows returned in 4386.230 ms, fwiw. [22:03] kfogel: great === salgado is now known as salgado-afk [22:26] * thumper waits for wgrant [22:49] thumper: Hm? [22:50] wgrant: we need to update the buildd thingy with a new bzr-builder version [22:50] wgrant: and none of us on the code team know what to do [22:50] wgrant: I'm guessing you might [22:53] thumper: It just relies on a new package being in the archive. [22:53] There is no solution for that at the moment -- the though in Wellington was that we'd have a special PPA. [22:53] But it's not there yet. [22:53] wgrant: I don't grok it [22:54] we have new lp code that is needed I believe [22:54] to make use of the changes to bzr-builder [22:54] Oh. [22:54] So not a new bzr-builder, but new stuff in lib/canonical/buildd? [22:54] we have a branch that adds this [22:54] both [22:54] The former you cannot do. [22:54] former what? [22:55] and why can't we do it? [22:55] You can't upgrade bzr-builder. [22:55] why? [22:55] Because you cannot upload to the Ubuntu primary archive, and there is no other archive that is used by all recipe builds. [22:56] can't we just have it installed on the xen builder image? [22:56] It needs to be installed inside the chroot, which is downloaded from LP each time. [22:56] does bzr-builder run in the chroot? [22:56] right [22:57] So, we are probably going to quickly hack in support for a celebrityish PPA which recipe builds will get. Then when we need a new bzr-builder, someone just has to push it into that PPA, and recipe builds will see it. [22:57] is bzr-builder installed via apt-get for recipe builds today? [22:58] mwhudson: It is. [22:58] huh, i had no idea [22:58] See lib/canonical/buildd/buildrecipe [22:59] * mwhudson sees [23:00] Huh. [23:00] https://bugs.launchpad.net/bugs/567652 was solved quite remarkably. [23:00] Bug #567652: Launchpad Login System is not FOSS [23:02] ... except accessing the private branch 403s. [23:02] Er, public branch. [23:03] um [23:03] (, 'unique_name', 'launchpad.View') [23:03] Hum. [23:03] i wonder if it's stacked [23:04] I was thinking that, but then that Unauthorized doesn't make much sense. [23:04] Oh, right, I guess it might revoke launchpad.View in that case. [23:05] yeah it's stacked [23:05] * mwhudson fixorates [23:05]