[00:00] thumper: it's all implemented but as up until the rollout, branches deleted on production would create deletion jobs scheduled for right now, it's not running yet [00:01] mwhudson: so what should we mark the spec? [00:01] done? [00:01] thumper: i guess the "deployment" status it's in is about right [00:01] thumper: move it to 2.2.8 and i'll set it to implemented when the cronscript is set up? [00:02] ok === Znova_ is now known as Snova_ [00:07] anyone able to review https://code.edge.launchpad.net/~mthaddon/lp-production-configs/bug-287304-staging-logrotate/+merge/9097 for me? [00:09] All those private branches ;-) [00:10] Indeed :( [00:10] Fortunately all new ones are public. [00:10] Congrats on the open source! :) [00:11] I'm very happy [00:13] pkern: a lot of the old branches are stacked on old-format versions of trunk, and as they're mostly merged anyway, they're probably not very exciting [00:13] and yes, you're not getting to see our production config :) [00:14] mthaddon: done [00:14] thx mwhudson [00:24] Is Karmic supported soon as a dev platform? I've gotten various errors while doing the rocketfuel-setup. [00:25] thumper: I think you may be confused about bug 390563 [00:25] Malone bug 390563 in bzr "overly large delta creation when fetching from 2a repositories" [Critical,In progress] https://launchpad.net/bugs/390563 [00:25] blueyed: LP needs Python 2.4, which Karmic doesn't love. A port to newer Pythons is in progress. [00:25] lifeless: probably [00:25] thumper: its not affecting lp any more, not since you cherrypicked the first stage fix weeks ago [00:25] lifeless: ok, so fix released for us then [00:26] there will be a follow on patch, but its corner cases, not common, and you can just wait for a release for that. [00:33] * mwhudson wonder if there's anything we can really do to make 'make schema' less of a hog [00:33] mwhudson: Please please please. === matsubara-dinner is now known as matsubara [00:45] I was _more_ suprised how much the bzr co process took more tbh [00:45] On both RAM and cpu [00:46] thats a bug [00:47] and filed as such [00:47] (in bzr) [00:47] oh :) [00:48] bzr does much better streaming, but the expected load was high so we made lp checkouts use http for the release period [00:48] mwhudson: speaking of which, should we disable the hot spot now? [00:49] i guess so [00:49] The HTTP checkout really didn't work too well. [00:51] wgrant: it did [00:51] wgrant: we had a single component in the DC fail, but that was the same as failed saturday so was fast to fix [00:51] It lightened the load, sure. [00:51] But I mean for the users. [00:51] wgrant: and after that we spiked up to *time times* our normal request rate [00:51] s/time/ten/ [00:51] It's reasonable that you did it. [00:52] Not bad. [00:52] now, many of those are probably due to the [major] bugs in bzr to do with 2a and http [00:52] *I* think we would have handled it on bzr+ssh with no problems [00:53] but, that wasn't load tested,and we had load tested http [00:54] * wgrant will be interested to see how swift a checkout is over bzr+ssh. [00:58] it took me ~an hour last time i did it [01:00] wget + tarball would be even quicker :) [01:00] thumper: after all that, i'm not sure stub's concerns are valid [01:00] mwhudson: really? [01:00] mwhudson: an hour? that seems long [01:00] (but i've beefed up my tests) [01:01] mwhudson: stub's concerns seemed entirely reasonable to me, and I'd be kinda surprised if they were off [01:01] lifeless: i was getting ~40 kbytes/s [01:01] I think that the tests may have problems duplicating multiple database connections [01:01] thumper: i'm surprised too [01:01] thumper: the default isolation level is READ COMMITTED for scripts though [01:02] thumper: and given that i don't write to the db, i fail to see how that is different from AUTOCOMMIT [01:02] the storm cache issue i'm less clear on, but i can't make anything break because of it [01:03] hmm [01:03] thumper: actually, it's clear the storm cache doesn't matter [01:03] I'm going to have to defer to stub on the connection stuff [01:03] thumper: we never create a Branch object [01:03] ah ha [01:03] cunning plans [01:03] well [01:03] lucky [01:06] also [01:06] 2009-07-24 00:06:32 INFO '/~sabdfl/applets/wabble/' -> 'http://localhost:8080/~sabdfl/applets/wabble/' (0.001703s, cache: MISS) [01:06] 2009-07-24 00:06:39 INFO '/~sabdfl/applets/wabble/' -> 'http://localhost:8080/~sabdfl/applets/wabble/' (0.000180s, cache: HIT) [01:07] so not sub ms for hits, but not bad [01:07] grr, not sub ms for misses [01:07] (will be a bit slower in the dc i guess) [01:07] mwhudson: still better than 100 or 500ms [01:07] thumper: yes, definitely [01:08] also, in main.log: [01:08] [2009-07-24 12:06:32.534 NZST] branch-rewrite@launchpad_dev LOG: statement: SELECT Branch.id, Branch.unique_name FROM Branch WHERE Branch.unique_name IN (E'~sabdfl', E'~sabdfl/applets', E'~sabdfl/applets/wabble', E'~sabdfl/applets/wabble/') AND Branch.private = false [01:08] which is nice an minimal, i think [01:12] mwhudson: so thats 320kbits, whats your link capable of? [01:13] lifeless: what? [01:13] oh [01:14] most isps in nz don't limit you, it's determined by distance to exchange [01:14] i get 1 megabyte/s to nz.archive.ubuntu.com on good days [01:14] never internationally though [01:14] so [01:14] not today [01:14] but we need to debug this [01:15] the python DVCS eval had issues too, and we didn't figure it out [01:15] lifeless: maybe i'm just pessimistic, i don't really expect to get much more than that to london [01:15] (on good days i do, but not every day) [01:17] mwhudson: it streams [01:17] lifeless: i meant, over any link [01:17] not bzr+ssh specicif [01:18] spm: guess what! jelmer rebased dulwich again :) [01:18] mwhudson: We should find out why you get that behaviou [01:18] r [01:18] ok [01:18] not now though [01:19] mwhudson: do you think it'd be abuse of power if I suspend jelmer's launchpad account? just asking. no reason. hypothetically. [01:20] jelmer: ^ :P [01:25] launchpad/bzr.dev builder seems to be working, yay [01:31] * mwhudson lunches [01:31] spm: (-: [01:32] jelmer: I was joking? :-P [02:11] bug 666 [02:11] Malone bug 666 in malone "can't file a bug on Ubuntu" [Medium,Invalid] https://launchpad.net/bugs/666 [02:11] Bug #666: can't file a bug on Ubuntu [02:11] Bug #666: can't file a bug on Ubuntu [02:11] Bug #666: can't file a bug on Ubuntu [02:12] we shouldn't have two bots here :) [02:12] haha [02:12] yeah, I vote for keeping mup and let ubot3 go [02:56] mwhudson, we are on AMI launchpad-ec2test11 now? [02:56] mars: i think so [02:57] * mars watches the script blow up again [02:57] we really need a library for this stuff === Guest73206 is now known as EdwinGrubbs [02:59] jml: ping [02:59] EdwinGrubbs, I think he is on leave today [03:01] mars: that's sad [03:01] thumper: ping [03:02] EdwinGrubbs: hi [03:03] yes, jml shouldn't be allowed to go on leave [03:03] thumper: hi, I was wondering if you knew if people use the feature of the BranchPopupWidget where it will create a mirrored branch if you enter in a url with the hostname? [03:04] EdwinGrubbs: it was asked for, and implemented [03:04] personally I'm not a big fan of it [03:05] thumper: I'm in the process of converting all instances of the PopupWidget to use the ajax Picker, and I wanted to verify that I had to keep that functionality. [03:06] EdwinGrubbs: yes, I'd say keep it for now [03:08] thumper, btw, found a bug in the Popup widget that links branches to bugs [03:08] thumper, it returns a launchpad.dev URL from the AJAX request [03:08] mars: please file a bug :) [03:08] besides that, works beautifully :) === matsubara is now known as matsubara-afk [03:11] thumper: do you know if there are any existing tests that exercise that feature? All the +linkbranch pagetests that I found just pass in the identifier for an existing launchpad branch. [03:12] EdwinGrubbs: the code team is very fond of unittests :) [03:12] EdwinGrubbs: I'm certain it'll be tested somewhere [03:15] ironic that I just timed out three times in a row trying to view jml's branches page [03:19] thumper, bug 403839 [03:19] Malone bug 403839 in launchpad-code ""Link to a bug report" feature returns a launchpad.dev URL" [Undecided,New] https://launchpad.net/bugs/403839 [03:19] Bug #403839: "Link to a bug report" feature returns a launchpad.dev URL [03:19] Bug #403839: "Link to a bug report" feature returns a launchpad.dev URL [03:19] Bug #403839: "Link to a bug report" feature returns a launchpad.dev URL [03:19] eeewww [03:19] bot barf [03:21] heh [03:21] mars: ta [03:37] * wgrant boggles. [03:37] RC changes aren't having much luck this release. [03:38] db-devel r8318 has another extra unintended change. [03:38] another one? [03:38] Yes. [03:38] rocketfuel-get is broken again. [03:40] odd, db-devel isn't listed on https://code.edge.launchpad.net/launchpad [03:41] ah, it's just the mainline [03:41] Right. [03:42] as are you - that is two this release [03:42] Yes. That doesn't normally happen, does it? [03:42] thumper, ^ should we rc-unbreak rocketfuel-get? [03:43] wgrant, no, this is quite odd [03:43] I can't find an MP for that branch, so I am wondering if it showed up in the diff or not. [03:44] great, so I managed to popular bin/ by removing the i18n section in buildout.cfg which was giving me version conflicts on hardy [03:45] how is rocketfuel-get broken today? [03:45] ajmitch: Same as last time. [03:45] the same glitch as before snuck into a release-critical landing [03:45] The fix was accidentally RC-reverted. [03:46] * ajmitch doesn't know what broke it last time [03:47] what's broken? [03:47] * wgrant kicks loggerhead. [03:47] Ah, there we go. [03:48] thumper, rocketfuel-get, http://bazaar.launchpad.net/~launchpad-pqm/launchpad/db-devel/revision/8318 [03:48] thumper: Observe http://bazaar.launchpad.net/~launchpad-pqm/launchpad/db-devel/revision/8318 [03:50] how'd that happen? [03:50] no idea [03:50] wgrant: woops [03:50] that is the second one I've seen this release though [03:50] * thumper shakes his head [03:50] I'm pleased I don't use the rocketfuel scripts :) [03:51] both issues came in via db-devel rc landings [03:51] which is quite strange [03:51] that's reverting the fix from http://bazaar.launchpad.net/~launchpad-pqm/launchpad/db-devel/revision/8304 [03:51] mwhudson: Right. [03:51] i guess bong merge conflict resolution again? [03:52] mwhudson, ? [03:52] mars: someone merged from trunk, got conflicts in rocketfuel-get and resolved them incorrectly? [03:52] it's a theory [03:53] ah [03:54] the other one killed some app links in sourcecode/, but it appears the test suite somehow missed it. [04:15] victory, I have launchpad running on the laptop even with a hacked buildout [04:16] ajmitch: Why did it need hacking? [04:17] because there was some version conflict in getting z3c.recipe.i18n, so I thought I'd see how far I'd get by disabling that [04:17] to my surprise I've got it running [04:17] ajmitch: did you have a copy of lp-launchpad-dependancies as ./download-cache? [04:17] thumper: yes [04:17] hmm... [04:17] wee, basically all the codehosting tests fail with bzr.dev [04:17] mwhudson: I saw that there were some failures [04:18] mwhudson: what's changed this time? [04:18] 50 failures, 19 errors [04:18] not sure yet [04:18] argh, of course it's the acceptance tests so the failures are useless [04:19] oh [04:19] argh! [04:19] thumper: I won't pretend to know why buildout failed, apart from that I'm running on hardy, not jaunty [04:20] & that I'd rsynced branches from my desktop previously [04:20] because these tests weren't run out of an actual egg, there is no file at /home/buildbot/slaves/launchpad/bzr/build/bzr.dev/EGG-INFO/scripts/bzr [04:20] grump [04:20] :( [04:21] eggs are such a heinous invention [04:27] The new license headers in current(-dev).sql have to be added manually and make 'make lint' whine. Should newsampledata be fixed to add them, or should they be removed? [04:27] lifeless, so you are a pip fan then? [04:29] mars: I'm a deb an [04:29] *fan* [04:29] deb and source [04:29] either install normally [04:38] * thumper hurls [04:38] I've just seen a truely (sp?) ugly page [04:39] and the worst bit is - I think I wrote it [04:39] :( [04:39] There are a few around. [04:39] thumper: truly [04:40] mwhudson: I was tossing up whether or not to add the "e" [04:40] but neither looked right to me [04:40] I just suck at spelling [04:40] english is ridiculous [04:45] thumper: want to review a branch that will work better at this? [04:45] better at what? [04:45] testing against bzr.dev [04:46] mwhudson: sure [04:46] here's an unrelated question [04:46] when "fixing" a bug on a page [04:46] and you notice something "wrong" [04:46] how hard should I resist fixing the "wrongness"? [04:48] almost always better to do it in another branch, i'd say [04:48] thumper: if you have time for yak shaving, shave the yak [04:49] thumper: I'd do it in a separate thing to be reviewed, to be nice to the reviewers [04:49] hmm... [04:49] ok [04:52] thumper: https://code.edge.launchpad.net/~mwhudson/launchpad/bzr.dev-help/+merge/9234 [05:31] how do you unassign via email [05:32] "dear losas, pls to be doing XYZ. kthxbye" ??? [05:32] assignee nobody [05:32] hah [05:32] thanks wgrant [05:33] wow [05:33] I so want to create an account nobody now [05:33] it's not a reserved name? [05:33] wgrant: you're sure about this? [05:33] https://help.launchpad.net/Bugs/EmailInterface needs an update if True [05:34] lifeless: It's there... [05:34] oh ugh [05:34] There was a Person named nobody ages ago. [05:34] But it got renamed. [05:34] there are two sections on assignee [05:34] >< [05:35] https://help.launchpad.net/Bugs/EmailInterface#Assigning%20and%20targeting%20the%20bug <- seems very confused [05:36] lifeless: It's shown in the reference, but not the short introduction. Seems reasonable. [05:36] wgrant: the short intro links to 'affects' not 'assignee' [05:36] which makes one think that the short intro is all there is [05:37] thanks though [05:43] * thumper just deleted two page templates \\o/ [05:45] mwhudson: want to review a branch where I clean up the codereviewcomment +fragment and +macro views to use services/comment +render? [05:45] mwhudson: it is pretty boring [05:45] mwhudson: good for a friday afternoon :) [05:46] mwhudson: I'm just running all the branches stories to make sure I didn't break anything [05:46] mwhudson: tested locally :) [05:47] thumper: boring sounds good [05:47] oh FFS [05:48] _someone_ keeps playing with sample data [05:48] grr [05:48] conflicts? [05:49] no, just a failure since I've not done "make schema" in a while and someone added Sample Person to another team [05:51] ah [05:51] that's sucky [05:51] * wgrant finds a bug in the malone email interface, thanks to lifeless' "assignee nobody" email. [05:54] ? [05:56] lifeless: It didn't announce in the email or activity log that you had added a launchpad-code task. Very odd. [05:56] heh [05:57] I _really_ want a 'change what this task affects' [05:57] rather than adding a task [05:57] if you do that, Jelmer and I will be extremely happy [05:57] Maybe a bit late, but congrats on open source! [05:57] sanxiyn: it's never too late to be loved! [05:58] lifeless: That doesn't seem too hard. I wonder if it would be better to have a command that takes two args, or a command to follow the 'affects' that reassigns the current task. [05:58] Now I can stop evangelizing how evil Launchpad is and how bad Canonical is. Thanks. [05:58] (Been doing that for last couple years.) [05:59] wgrant: reaffect or someting [05:59] affects instead [05:59] lifeless: retarget? [05:59] yeah [05:59] Although I guess that's not the UI term. [05:59] it should be:P [06:00] Indeed. [06:00] It's in the webservice API. [06:01] mwhudson: I'm just pushing the branch up now [06:01] mwhudson: although I'm about to EOD [06:01] mwhudson: shall I bother to assign the review to you? [06:01] mwhudson: or just wait till Monday [06:02] thumper: oh assign it to me [06:02] thumper: it might wait until monday, but i'll probably slog through it [06:02] bzr sending now [06:10] mwhudson: sent [06:14] sanxiyn: so you like Launchpad now? [06:15] thumper: I should try it, and then I would be able to tell :) [06:15] thumper: I like e.g. tracking of bugs in remote bug trackers. (That's why I am a contributor to Debian's bts-link.) [06:16] :) [06:16] bts? [06:16] * thumper doesn't know much about debian [06:16] thumper: http://bts-link.alioth.debian.org/ [06:17] The BTS is the Debian Bug Tracking System [06:17] ah [06:17] And bts-link is a component of that system that can track remote Bugzilla, Trac, GForge, RT, whatever. [06:17] anyway... [06:17] * thumper EODs [08:04] noodles775: Hi. [08:05] Hi wgrant :), Congrats on getting your MP approved! [08:05] noodles775: Thanks. [08:05] noodles775: Are there any designs for the new DSP page with its version listing yet? Bug #144620 would probably be fixable in the redesign. [08:05] Malone bug 144620 in soyuz "Some displayed sourcepackagerelease changes files don't have attribution" [High,Triaged] https://launchpad.net/bugs/144620 [08:05] Bug #144620: Some displayed sourcepackagerelease changes files don't have attribution [08:05] Bug #144620: Some displayed sourcepackagerelease changes files don't have attribution [08:05] Bug #144620: Some displayed sourcepackagerelease changes files don't have attribution [08:05] Oh dear. [08:07] wgrant: Only what's in those images attached to the blueprint... I think bigjools is keen for us to work on the DSP refactoring asap... [08:07] wgrant: I'm just looking at the new 3-0 style refactoring (and having a go at applying it to the archive-index) [08:08] noodles775: I also don't really like the elimination of the D(S)BP(R) pages. They're very useful for looking at the sometimes complicated histories. [08:08] wgrant: The idea (which may not be clearly expressed) is that the pages would still exist, but not be needed when the user has JS on. [08:09] noodles775: Ah, I see. That is fine, then. [08:09] wgrant: ie. that info would be available on the DSP page (after drilling down), but would still need to be there for non-js browsers. [08:09] Cool! [08:09] Some of the info isn't appropriate for the DSP page. [08:10] (eg. BPRs with the same name but from a different source) [08:10] Why would a BPR that wasn't from that source be displayed on the page? [08:11] ah, I see [08:11] noodles775: It wouldn't, which is a good reason for the DASBP page to exist. Otherwise I can't find the history of what the name mapped to. [08:12] wgrant: Hmm... it'd be great to capture these as stories on the spec, could you add one? [08:12] * noodles775 checks the (practically empty) spec [08:12] noodles775: Under Use Cases? [08:12] wgrant: yep [08:12] Thanks! [08:13] wgrant: just add any positive or negative use cases, we can evaluate them when we've collected lots. [08:14] noodles775: Sure. [08:19] good morning [08:19] Morgan adeuring [08:19] hi noodles775! [09:21] Morning :) [09:27] morning all [09:28] Morning bigjools. [09:28] hey wgrant [10:00] Hi! [10:00] what is that page again that shows all of the used icons on launchpad? [10:01] /+graphics [10:01] wgrant: cheers [10:14] mrevell: morning! [10:14] hey jtv1! === jtv1 is now known as jtv [10:15] jtv: I'll just get a cup of tea, if that's ok [10:15] thunderstorm alert here, I might go netless any time [10:15] bigjools: really? [10:15] mrevell: sure, I'll have one too thanks [10:16] mrevell: yes, my power goes out usually [10:16] bigjools: Wow but I'd still rather live in Kempsey, heh [10:17] mrevell: the house next door is up for sale :) [10:17] bigjools: interesting -- PPA and Soyuz help whenever I want it :) [10:18] mrevell: I have guard dogs, barbed wire and feral children [10:18] :) I bet I have more dogs [10:20] i used to be able to say I had more children :) [10:22] heh [10:22] jtv: Skype or IRC sir? [10:23] mrevell: I think irc for this one, since it's all in writing... good way to throw notes at each other :) [10:23] jtv: sounds great [10:23] mrevell: I sent you an email with the text I have. [10:23] jtv: I have it here, I'll read it then PM you [10:23] mrevell: perfect [10:23] * jtv reaches for coffee === henninge_ is now known as henninge [11:02] Morning, all. [11:39] I'm really sad that my new Dell Mini 10v, with Ubuntu pre-installed, has got a key with a Windows logo on it :/ [11:44] bigjools: wierd? [11:44] crass [11:44] Maybe canonical could provide Dell with similar keys? [11:56] adeuring, Hi! May I ask you a question? [11:56] andrea-bs: sure :) [11:57] adeuring, thanks for your review for my branch "blueprint-whiteboard-diffs"; what should I put instead of 72 for get_unified_diff()? [11:58] andrea-bs: nothing ;) As Björn pointed out, my question about the diff formatting was a a bug in the review. [11:59] adeuring, oh, sorry... I should read all the comments before asking :) [11:59] andrea-bs: no prblem ;) [12:00] Who owns canonical.zeca? [12:02] good question, cprov-afk do you know? [12:03] wgrant: foundations, I guess [12:03] wgrant: ideas ? [12:04] wgrant: or bugs ? [12:04] Isn't that the GPG thingy used by Soyuz? [12:04] I don't think anyone in foundations has ever messed with it... [12:05] cprov-afk: I needed a keyserver.launchpad.dev, so I patched it to work properly. Previously IGPGHandler would be unable to re-get keys from it. http://williamgrant.id.au/f/1/2009/zeca-get-key-id.diff is the ugly patch, and I wonder what to do. [12:05] Or knows enough Portuguese to even understand what the name means :-P [12:05]