[02:05] anyone around today? [02:10] ... maybe [02:12] hey spm. dumb question. for some reason when i run up a local instance of lp, i can't connect to http://launchpad.dev but must use http://launchpad.dev:8085 [02:12] that gets me the home page but most everything else is broken [02:12] i can't see what's changed [02:12] wallyworld: Sounds like your Apache config is broken. [02:12] Is Apache running at all? [02:13] wallyworld: " most everything else is broken" <== would be my guess at what's changed ;-) [02:13] wgrant: i checked the configs but they are uncganed since months [02:13] spm: i meant that other lp urls don't work unless i hack in the 8085 port [02:13] wallyworld: Maybe Apache has stopped starting automatically. [02:13] wgrant: i'll check [02:15] spm: Did you fix edge, or has it mysteriously rectified its own issues? [02:15] wgrant: not touched it [02:15] wasn't aware there was/is a problem? [02:15] Ah, it looks like it was fixed about 8 hours ago. [02:16] It spent all weekend returning truncated responses, 502s and 503s. [02:16] orsum [02:16] although.... that surprises me. I was under the belief that as of late last week edge == prod [02:17] wgrant: yeah, apache not running, restart fails but error log is 0 bytes. something weird going on :-( [02:18] wallyworld: apache2ctl configtest ? [02:19] wallyworld: How are you restarting it? [02:19] wgrant spm: i found it. i moved my checkout dir and the branch rewrite path needs to be changed in the apache conf [02:19] Aha. [02:20] I was going to suggest that, but didn't think it would be fatal. [02:20] i had a working dir called sandbox but i now need to work with db-devel hence took the opportunity to rename [02:20] oh well, a trap for noobs like me [02:21] wallyworld: now multiply that over ... lots of servers - welcome to my world :-) [02:21] and it get the error print i had to /etc/init.d/apache2 stop and then start even though it wasn't running in the first place [02:22] spm: you truly have a job i would not want to do and am very grateful that you do it :-) [05:59] * thumper waves [07:07] thumper: can you hear the sound of crickets? pretty quiet today [07:11] too quiet even for crickets. [07:12] what's that? you'll have to speak up so i can hear you over all the noise :-) [07:14] heh [07:24] * jpds listens to the airco. [08:44] good morning [08:46] hi adeuring [09:03] Hello === almaisan-away is now known as al-maisan [12:15] Hey, I'm doing a rocketfuel-setup on a fresh Maverick install. It's complaining that launchpad-soyuz-dependencies is broken. I don't see anything on the launchpad-dev list about this. Any ideas? [12:21] mrevell: Not launchpad-dependencies too? [12:21] mrevell: You need to install Lucid's python-psycopg2 first, then run rocketfuel-setup. [12:22] wgrant, Thanks. When I tried launchpad-depends it said that python-psycopg2 was broken. However, installing it manually works. Cheers. [12:23] Oh, it's complaining about the version ... launchpad-dependencies wants 2.0.14 but 2.2.1-1unbuntu1 is the installation candidate. Bah. [12:23] I have the PPAs in my sources.list. [12:24] trying a dist-upgrade [12:25] mrevell: You'll need to manually install 2.0.14 from Lucid. It won't install an older version automatically. [12:25] Ah okay [12:26] Oh, I see Maris' fix on the list [13:11] Unfortunately there is no nice way to fix this (other than making LP work with 2.2) [13:12] I am deliberately not suggesting uploading a 2.2.1really2.0.14 package, because I don't want to encourage delaying the fix [13:15] Don't give anyone any ideas... [14:06] hello from UDS [14:07] oh hai [14:08] rockstar: why is there a wifi network named after you? [14:08] Because he's a rockstar [14:16] hi [14:21] cd 15 [14:22] heh, and morning. [14:30] noodles775: Are you able to do a review for a very short shipit branch? === adeuring1 is now known as adeuring [14:47] mthaddon: is the pqm chroot on praé lucid now ? [14:48] lifeless: it is, yep [14:48] mthaddon: are you aware of any pre-lucid machines or chroots or vms in the lp arena now ? [14:49] lifeless: nope, we should be all lucid now [14:49] awesome. [14:49] mars: ^ [14:53] mthaddon, cool! lifeless, so we can make the announcement? (I'll give you some time to think about it when you are not in a keynote) [14:53] allenap: Where's the branch? I'm at UDS but if it's tiny, shouldn't be a problem. [14:54] noodles775: https://code.edge.launchpad.net/~allenap/shipit/remove-propertycache-adapters-bug-628762-shipit/+merge/39251 [14:54] Thank you! [14:54] mars: if there are no pre-lucid environments, we're fine. [14:54] mars: if there are, we aren't [14:54] mars: I can't tell if there are other thank asking tom :) [14:55] allenap: I think you need to subscribe me to the branch. [14:57] mthaddon, so just to confirm, we are now running Lucid on all machines, virtual machines, chroots, and so forth? [14:57] noodles775: Done. [14:57] mars: yes [14:58] mthaddon, awesome, thank you. [14:59] gary_poster, lifeless, then we can make the announcement - Launchpad is officially running Python 2.6 and Lucid [15:00] great mars, go for it. thanks [15:00] mars, gary_poster, lifeless: Should we also add PostgreSQL 8.4? [15:01] mars: By the way, does ec2 use 8.4 now? [15:01] StevenK: to the announcement? yes [15:01] yes it does [15:01] cool, I'll add that too [15:01] what's the launchpadlib hack where you can grab a resource from a URL directly? [15:01] thanks for the reminder StevenK [15:01] is it just .open? [15:01] * gary_poster forgets but that's what comes to mind [15:01] I'll try it, thanks [15:02] bigjools, check the launchpadlib recipes/hacking page? [15:02] didn't know we had one! [15:02] bigjools, https://help.launchpad.net/API/Examples [15:03] bigjools, I think "Fetching an object's raw JSON" is what you want? [15:06] bigjools: I don't remember, but I think I made hydrazine use it [15:06] though lplib has the relative open feature itself now [15:08] mars: that's not what I want - I want to navigate to an object's URL directly [15:08] open() is the thing I think [15:10] lp:hydrazine, or if not that then its in lp:lptools [15:11] jml: one option that's available to you for testing testtools on various python versions is daily builds. [15:16] bigjools, maybe you could you add an Example for what you do want to do then? It sounds like more than one person has needed it [15:16] yeah will do [15:16] trying to remember on which object open() exists :) [15:16] Launchpad [15:17] AttributeError: 'Launchpad' object has no attribute 'open' [15:17] mmm [15:17] really, look in hydrazine/lptools [15:17] I think its lptools milestone stuff now that I think about it [15:24] noodles775: Thanks for the review. Can you mark the proposal (as a whole) as approved; I do not have permission to do so. [15:34] if I have launchpad.things['thing'] what gets called to get 'thing' on the 'things' collection? [15:34] IPersonSet didn't give me much inspiration === almaisan-away is now known as al-maisan [15:51] allenap: done. [15:51] noodles775: Cool, thanks. [16:00] leonardr: hi, around? [16:02] jml: what is this Launchpad roundtable session at UDS? [16:02] bigjools, yes === mtaylor is now known as 52AACIHR4 [16:04] leonardr: I am exposing a new top-level collection in the API and sinzui says that you know how to make indexing of it work [16:04] we used to have to hack lplib but apparently there's something else I can do in LP now [16:04] bigjools: what do you mean by indexing? [16:04] so I am adding lp.builders [16:04] and I want to do lp.builders['bob'] [16:04] right now that blows up :) [16:05] bigjools: ok, you need to change launchpadlib [16:05] :( [16:05] gah [16:05] look at launchpad.py, eg. BugSet [16:07] will do, thanks === flacoste changed the topic of #launchpad-dev to: login problems on qastaging | Launchpad Development Channel | Week 1 of 10.11 | PQM is open | firefighting: - | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting [16:22] leonardr: are there any instructions for monkeys on how to update launchpadlib, make an egg, release it etc etc ? === flacoste changed the topic of #launchpad-dev to: login problems on qastaging, matsubara and LOSA are investigating | Launchpad Development Channel | Week 1 of 10.11 | PQM is open | firefighting: - | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting === deryck_ is now known as deryck [16:26] gary_poster: perhaps you know? :) [16:27] bigjools: let's try benji before I wade in ;-) [16:27] benji, do you know? [16:28] * benji reads. [16:28] - [16:35] gary_poster/bigjools: the only ones I know about (and follow myself) are in https://dev.launchpad.net/HackingLazrLibraries [16:36] thought as much [16:36] thanks benji [16:37] * bigjools reads, thanks [16:39] this is Too Hard just to make a simple API change :( [16:40] bigjools: I think it's just due to the fact that IBuilder is top-level [16:40] StevenK: IBuilderSet being a collection is the problem [16:40] Ahh [16:40] top-level collection [16:46] Project devel build (149): STILL FAILING in 4 hr 6 min: https://hudson.wedontsleep.org/job/devel/149/ === lifeless changed the topic of #launchpad-dev to: Performance Tuesday | login problems on qastaging, matsubara and LOSA are investigating | Launchpad Development Channel | Week 1 of 10.11 | PQM is open | firefighting: - | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting === mordred_ is now known as mtaylor [17:27] Project db-devel build (96): SUCCESS in 4 hr 1 min: https://hudson.wedontsleep.org/job/db-devel/96/ === benji is now known as benji-lunch === Ursinha is now known as Ursinha-lunch === matsubara is now known as matsubara-lunch === salgado is now known as salgado-lunch [18:14] bigjools, sorry i disappeared. if you give me a launchpadlib branch i can do a release. i'd rather do the release myself atm === benji-lunch is now known as benji === beuno is now known as beuno-lunch === matsubara-lunch is now known as matsubara === Ursinha-lunch is now known as Ursinha === beuno-lunch is now known as beuno [19:16] rockstar: hey [19:17] * mwhudson writes on the merge proposal instead === matsubara is now known as matsubara-afk [20:07] lifeless: it's not tuesday yet! [20:08] hmm, well, ok, maybe it is now down under [20:08] flacoste: It is in New Zealand :-) [20:08] (It's been Tuesday in Australia for 6 hours) === flacoste changed the topic of #launchpad-dev to: Performance Tuesday | login problems on qastaging solved | Launchpad Development Channel | Week 1 of 10.11 | PQM is open | firefighting: - | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting [20:09] flacoste: its tuesday [20:09] flacoste: 0809 in nz [20:10] Attending UDS while in a NZ timezone would be terrible [20:12] wgrant: pls don't move louvi or lakoocha between architectures until further notice. [20:13] wgrant: just in case the idea occurs to you [20:14] You can learn so much interesting trivia by googling canonical machine names :-) [20:15] the fruit theme is getting fairly stretched now [20:16] we should start using cuss words [20:16] there's an endless supply [20:16] i bet lamont can supply quite a few just by himself :) [20:17] ISTR that next up is 'star names' [20:18] mwhudson: stretched is not the same as done [20:18] lamont: indeed [20:19] I may have to start using tarmac & merge queues. [20:19] though we have definitely moved out of the yummy ones and into the toxic ones [20:20] jml: for? [20:20] self flagellation [20:20] lifeless: testtools [20:20] jml: why? [20:21] actual, face time may be best [20:21] lifeless: because supporting multiple Pythons is really hard [20:21] I need to focus on this session [20:21] lifeless: ok. [20:21] jml: hmm, my hudson should be live again shortly after I get home; the machine is turned on again. But in the break find me [20:22] lifeless: I've got hudson on mumak.net too [20:22] lifeless: just had to get backports for 2.7 etc. [20:27] Project devel build (150): STILL FAILING in 3 hr 41 min: https://hudson.wedontsleep.org/job/devel/150/ [20:27] Launchpad Patch Queue Manager: [r=lifeless][ui=None][no-qa] Fix local failures. [20:34] jml: woo [20:34] fixtures-0.3.tar.gz (md5, sig)Source tarball. 2,136 [20:34] download count s;) [20:34] lifeless: nice :) [20:35] I wonder what testtools has [20:35] less than fixtures :) [20:35] wow [20:35] yes [20:36] lifeless: I'm patching testtools now to have a subunit runner and testr.conf. feel the need for a review? [20:36] delighted to if you want one [20:36] I think its time to add useFixture to testtools [20:36] lifeless: thanks. lemme just find a .testr.conf to cargo cult :) [20:36] lifeless: yes. [20:36] lifeless: 0.9.8 is going to be a big release [20:36] I've just added details to fixtures [20:37] in 0.3.4 [20:38] actually... testr.conf is too hard. I'll just land the subunit thing without review [20:39] hmm [20:39] toss the branch up [20:39] I'll do testr.conf for you [20:39] jml: why do you want a subunit run thing though? Isn't it just [20:39] python -m subunit.run testtools.test_suite ? [20:39] hmm. [20:40] oh yeah, that works. [20:40] no point in patching then :) [20:40] I mean, we have testtools.run [20:40] but I don't really care about adding a --subunit option to that [20:40] right, in fact the subunit run extends the testtools one I think [20:41] so older us though that wasn't the best way forward [20:41] s/gh/ght/ [20:47] lifeless: :) [20:47] jml: testtools is packaged though; whats its popcon count [20:47] lifeless: I don't really know how to find out [20:48] popcon.ubuntu.com [20:48] http://popcon.ubuntu.com/ [20:48] StevenK, lifeless: thanks, I just went there. I meant I don't know how to navigate popcon.ubuntu.com [20:49] 21544 python-testtools 1153 40 352 73 688 (Unknown) [20:49] so 1153 installs [20:49] jml: click on the white page beside 'all' in the 'inst' column [20:53] just heard from poolie, he has now landed in Orlando [20:53] Crumbs [20:53] \o/ [20:54] jam: did beuno get hold of you? [20:54] yep [20:54] cool [20:54] lifeless: i have fixed the ppr report! [20:55] flacoste: \o/ [20:55] expect a merge proposal soon [20:55] trying to generate the failed reports on devpad for confirmation [21:01] Project db-devel build (97): SUCCESS in 3 hr 34 min: https://hudson.wedontsleep.org/job/db-devel/97/ [21:01] Launchpad Patch Queue Manager: [r=lifeless][ui=none][bug=665301] [21:01] SourcePackageRecipeBuild.destroySelf clean up referenced objects. === Ursinha is now known as Ursinha-afk [21:03] morning [21:04] gary_poster: what is new in storm 0.18? [21:05] thumper: hi. have not sent out message because stuck with permissions on uploading docs, but message will come out. But anyway: see release notes https://edge.launchpad.net/storm/+milestone/0.18 === al-maisan is now known as almaisan-away [21:06] jml: ping [21:06] thumper: we care about (https://code.edge.launchpad.net/bugs/659078 and https://code.edge.launchpad.net/bugs/620615), and one we had cherry-picked into our version of 0.17 (https://code.edge.launchpad.net/bugs/620508) [21:47] bigjools: I'm pretty sure I didn't have to change launchpadlib last time I added a top-level collection. [21:48] Ursinha-afk: how is updating the tagger going ? [21:48] wgrant: the ['foo'] stuff is special cased [21:48] wgrant: but he didn't need that, so is not doing it now, [21:48] Right, exactly. [21:48] Great. [21:49] wgrant: you do if you want to add a __getitem__ to it [21:49] good morning anyway [21:49] what's the future like? :) [21:50] bigjools: Well, yes, but that's an evil unnecessary hack. [21:50] wgrant: Please send cookies, lottery numbers, and stock tips to those of us stuck in the past. [21:50] It is futureful. [21:50] But busy, being the last week of classes, and all that. [21:52] How goes UDS? [21:52] I don't know, what day is it? [21:53] i think it's monday [21:53] Heh. [21:55] grar, intel gpu lockup [21:55] Not X crash when typing? [21:55] That's all I've had lately. [21:55] wgrant: uds seems to be fine so far [21:55] there's a known bug I think [21:56] but ScottK assures me the developer is here so I can bug him :) [21:56] Hm, I don't get X crashes while typing [21:56] i've had x hang on me a few times [21:56] Then again, I'm using evil-eats-kittens-nvidia drivers. [21:56] nvidia is far more reliable for me.. Go figure. [21:56] * gary_poster looks about for someone to do a quick librarian diagnostic log review. Potentially very easy review--I just add twisted.python.log.msg calls. Anyone up for it? :-D https://code.edge.launchpad.net/~gary/launchpad/bug662912/+merge/39315 [21:57] StevenK: It's sort of like that old bug where X didn't completely steal the keyboard from the TTY, so pressing Ctrl+C killed X. But it's not that. [21:57] I can look [21:57] thank you bigjools [21:57] np [21:59] bigjools: Has Soyuz had another five years of work requested yet? [21:59] not yet, but I feel the wind changing [22:00] I'm looking for a pile of sand so I can stick my head in it [22:00] I hear Orlando has things that are called 'beaches', but they don't look like one. [22:01] gary_poster: I've made a note on your review re: the log target. [22:01] gary_poster: did you look at addSystemEventTrigger() [22:01] just wondering why you're using callWhenRunning in particular, but no big deal [22:02] bigjools: eh, it seemed like a good idea at the time ;-) [22:03] gary_poster: ok it's not a problem, just wondered if aSET would do something better in this case [22:04] gary_poster: why are you changing the dev log target? [22:04] is it for "make run" or test logs? [22:05] jml: ping [22:05] lifeless: given that he's put it in the development config not the testrunner config it won't affect tests? [22:05] bigjools: aSET better: semantically/idiomatically, or...? It does the trick practically, so this is a idiomatic/semantic thing. My Twisted is a bit rusty, so if you think it's better I'm happy to swotch [22:05] switch [22:06] gary_poster: no need to switch, I was just wondering if it would log at a more convenient timing for your needs [22:06] bigjools: dev log target is just for make run [22:06] and before I believe the default is "-" [22:06] so it won't affect anything that lifeless worries about I think? [22:06] right [22:06] gary_poster: whats the current behaviour ? [22:06] gary_poster: it should be logging to a sane place without that setting [22:07] - is stdout [22:07] lifeless: currently I can't find the output. stdout is lost when this is a daemon [22:08] thumper: abentley rockstar: standup? [22:08] lifeless: and as bigjools said, this affects make run, not tests [22:08] gary_poster: yes, I did misunderstand that [22:08] bear with me a sec [22:08] np [22:09] if you find the output, perhaps it is a documentation issue--but I'd argue a "librarian.log" next to our other devel logs is still a pretty good place for it. [22:10] +1 from m [22:10] e [22:11] thanks bigjools [22:11] gary_poster: welcome! [22:13] * wallyworld actually wishes logs when running under development were written outside the source tree === Ursinha-afk is now known as Ursinha [22:15] lifeless, just put the reviewed version in prod [22:15] wallyworld: interesting. I typically have two or three branches going at once, so having the logs separate can be convenient for me. It's also nice that when I delete a branch, I'm also automatically cleaning out the logs. If you have a compelling argument for it, though, I'd be interested. [22:15] For later :-) [22:15] gary_poster: otp. will respond in a minute [22:16] wallyworld: np. need to step out myself. [22:16] lifeless, https://devpad.canonical.com/~lpqateam/qa_reports/deployment-stable.html [22:17] lifeless: need to step out. Will be back for your approval or suggestion on the MP when I return. [22:18] lifeless: since edge is not updating now we should remove the redirect I think? [22:25] gary_poster: ok, I see that it is running with -i [22:25] bigjools: https://code.edge.launchpad.net/~lifeless/launchpad/edge/+merge/39233 [22:25] gary_poster: doing that librarian.log thing is ok; I echo wallyworlds sentiments about out of tree, but lets not do it willy-nilly [22:25] Ursinha: thanks! [22:26] gary_poster: sorry for leaping to conclusions [22:26] lifeless: \o/ [22:26] lifeless: this is a monumental occasin [22:26] occasion, even [22:26] yes [22:27] gary_poster: https://devpad.canonical.com/~lpqateam/qa_reports/deployment-stable.html [22:27] gary_poster: you're up! [22:30] lifeless, he's blocked in another bug [22:31] Anyone around to review a small test-change-only branch? [22:31] Ursinha: how so? [22:31] StevenK: shoot [22:31] lifeless: https://code.edge.launchpad.net/~stevenk/launchpad/cronscript-idsjob-testfix/+merge/39321 [22:39] Ursinha: can you bump qatagger if its not running already ? [22:41] lifeless, just did it [22:42] \o/ nice long list. [22:42] gary_poster: re: logs - one reason is that when grepping for stuff from the project root, you often get 1000s of false +ves since the log files are searched as well. plus they show up in my ide since they are in the root dir and not a sub folder that can be excluded easily. could we at least have a single sub dir for logging/trace and other similar stuff (eg the millions of thread-xxxx.request files or the (new) generated [22:42] appserver config files)? [22:43] lifeless: https://code.edge.launchpad.net/~flacoste/launchpad/ppr-constant-memory/+merge/39324 [22:43] the branch name is a lie [22:45] heh [22:48] lifeless: librarian.log: ack, thanks. [22:48] flacoste: times_array[:,0] <- what? [22:49] wallyworld: would be +1 on subdir; I'd like that [22:49] lifeless: that's numpy, that's getting the first column of the matrix [22:50] the matrix is app_time, sql_statements, sql_time [22:51] I would love to see more tests of this [22:51] gary_poster: cool. it's nice to keep all that sort of messy stuff away "under the rug" so to speak. want me to do the code for it? [22:51] automated tests you mean? [22:51] wallyworld: sure, thanks! You can hit me for the review if you want (tomorrow; leaving soon :-) ) [22:52] (and don't feel like you need my review; just offering) [22:52] flacoste: yes [22:52] gary_poster: i've got a few things on my plate today but will get to it in the next day or so [22:52] I'm glad you qa'd thoroughly [22:52] cool [22:52] its sad that you needed to [22:53] gary_poster: also, thanks for offering the review. i don't mind who does it, so long as it is done :-) [22:53] :-) [22:53] understood, wallyworld [22:58] lifeless: it's a lame excuse, but there were none [22:59] yeah, manual qa sucks [23:01] flacoste: yeah, I understand :) [23:05] lifeless, can you come find mtaylor and I in the Bonaire hallway? [23:05] lifeless: thanks [23:11] rockstar: omw [23:25] flacoste, lazr-js is not landed because I cannot get ec2 to actually complete a test run. I've talked to deryck about it at length. [23:40] flacoste: ping [23:40] gary_poster: ping? [23:41] * thumper sometimes wishes that there were a few less time zones between NZ and NY [23:55] I'd just like to say that ORMs can really suck for efficiency