[12:01] daf: thats why there are build from source instructions right there. [12:01] I might modify it to 'ix86 debs are autobuilt, the rest by volunteers that we trust - but we still cannot vouch...' [12:02] I think a "it's not our fault if Bazaar ruins your life notice" might be less scary :) [12:02] s/Bazaar/these Bazaar packages/ === elmo [~james@82.211.81.249] has joined #launchpad [12:10] lifeless: what's your email. I'm confused as to why all the tests pass but I still get a failure. Maybe if I forward this you can give me a good reason why it fails. [12:10] robertc@robertcollins.net [12:11] spiv: what's your email? [12:11] andrew@canonical.com :) [12:11] sent [12:14] Heh. [12:14] test_on_merge.py is now *too* strict. [12:15] That final ignored exception broke the test for the last line of stderr being "OK\n" :) [12:15] The return value is sufficient... I'll fix test_on_merge.py :) [12:16] cool, thanks [12:16] heh [12:16] stdout is crack. [12:26] Merge to rocketfuel@canonical.com/launchpad--devel--0: star-merge from kiko. (patch-729) [12:26] that's ME! === kiko lands view-hackage into tree [12:30] Merge to rocketfuel@canonical.com/launchpad--devel--0: Make test_on_merge.py rely only on exit code, to avoid spurious failures. (patch-730) === bob2 [rob@bob2.user] has joined #launchpad [12:32] BradB: Try now :) [12:32] stub, around? === BradB will commit one more change locally then try spiv's change [12:34] kiko: he went to bed about four hours ago, I think [12:34] ah, pity. [12:35] um, no [12:35] timezone confusion :) [12:35] make that 9 hours ago [12:37] hah [12:37] it's 1037 in stub-town [12:39] yo, bob [12:39] how's things in bob2-town? === BradB thinks Launchpad needs a Mao plugin [12:41] X-) [12:43] daf: a bit hot, but not too shabby...how's ye olde wales? [12:44] daf's in the country that just re-elected bush, i thought [12:45] they did, huh [12:45] yeah :( [12:46] well, they didn't re-elect him, since they didn't elect him in the first place [12:47] well "elect" him, whatever :P [12:49] bob2: here is cold [12:49] bob2: here is also New York [12:51] daf: do you like it there? [12:51] daf: oh, wow...for work? [12:51] daf: also, wherever you are seems cold and/or wet [12:52] "everywhere you go, you always take the weaaaaaather with you" [12:52] New York sure is Crowded, Bob [12:53] BradB: it's interesting :) [12:53] kiko: am now [12:54] daf: hehehehe === BradB is all about Montreal [12:55] stub: I fixed that db problem; you owe me a sloppy wet kiss. [12:56] I should warn you about my halitosis [12:56] heh heh [12:56] stub, you da man! :) [12:56] See kiko - DVD's just don't cut it. [12:56] stub, so, if you would like to update mawson you'll get my patches (in want of the views you'd add at the same time) [12:56] stub, that's depressing, though. [12:57] spiv is such a weenie he actually accepts and then invalidates the offer by slacking. good grief. === stub starts on updating mawson, before email distracts him for a few hours [01:02] thanks. [01:02] stub: you should wait dude [01:03] i've got a bigass commit to land, and one last failing func test (that i just recreated, because i broke the existing one) [01:04] stub: could you also configure email to send real emails out? my next commit will start sending global notifications to dilys as well. [01:05] and adds the product widget to bug add forms, and make source package not-required [01:05] BradB: will do. Was brought up at the meeting last night, but I wanted to make sure you were happy with email being switched on [01:05] yeah, might as well [01:06] I got branches to merge and configs to backport and update scripts to write anyway ;) [01:21] lifeless: launchpad is like 350MB [01:21] "Ican't believe you only had 80MB free", indeed. [01:21] I can't believe you only had 350MB. [01:22] daf@mabli ~/src/canonical/dists/launchpad> du -sh . [01:22] 165M . [01:22] ? [01:22] hm, I guess I mean launchpad + buildbot [01:22] buildbot is 200M? [01:23] (nyah, bob2 is smoking something) [01:23] probably a pristine tree. [01:23] oh [01:23] touche [01:23] with an accent === daf hands bob an [01:29] I have no compose key and I want to die. [01:30] mac keyboards have a distinct lack of extra keys [01:38] make run on current launchpad: [01:38] File "/home/rob/projects/warthogs/launchpad/launchpad/lib/psycopgda/adapter.py", line 259, in ? [01:38] DATE = psycopg.new_type((DATE_OID,), "ZDATE", _conv_date) [01:38] zope.configuration.xmlconfig.ZopeXMLConfigurationError: File "/home/rob/projects/warthogs/launchpad/launchpad/site.zcml", line 17.4-17.64 [01:38] ZopeXMLConfigurationError: File "/home/rob/projects/warthogs/launchpad/launchpad/package-includes/launchpad-sql-configure-normal.zcml", line 6.2-9.34 [01:38] AttributeError: 'module' object has no attribute 'new_type' [01:40] urg [01:48] bob2: which psycopg are you using? it's worth verifying. [01:49] the config didn't get one, which zope also complained about [01:49] so I checked out... [01:49] rocketfuel@canonical.com/psycopgda--test--3.0 [01:50] that's not psycopg [01:51] erm, right [01:51] try importing psycopg from the command line [01:51] works [01:51] it would [01:51] I'll go back and check out the original error [01:52] be see what version and .new_type [01:52] >>> import psycopg [01:52] >>> psycopg.new_type [01:52] [01:52] >>> psycopg.__version__ [01:52] '1.1.15' [01:52] >>> [01:52] removing psyopga: [01:52] File "/home/rob/projects/warthogs/launchpad/launchpad/lib/zope/configuration/fields.py", line 141, in fromUnicode [01:52] raise schema.ValidationError(v) [01:52] zope.configuration.xmlconfig.ZopeXMLConfigurationError: File "/home/rob/projects/warthogs/launchpad/launchpad/site.zcml", line 5.4-5.35 [01:52] ZopeXMLConfigurationError: File "/home/rob/projects/warthogs/launchpad/launchpad/lib/canonical/configure.zcml", line 76.4-76.44 [01:53] ZopeXMLConfigurationError: File "/home/rob/projects/warthogs/launchpad/launchpad/lib/canonical/database/configure.zcml", line 5.4 [01:53] ConfigurationError: ('Invalid value for', 'class', "Couldn't import canonical.launchpad.database, No module named psycopg") [01:53] easy enough [01:53] hrm, after remocing psycopgda I can't import psycopg anymore [01:54] surely it should be the other way around? [01:54] you need to install psycopg is all [01:54] bah, I deleted the wrong thing [01:56] ok, let me start over [02:12] stub: oh well, go ahead and roll out a new lp, there's now six failures on the server again and i don't have time to do any more on this tonight [02:13] ok. sounds like whack-a-mole [02:15] argh, this is annoying [02:16] oh well, i gotta take off === BradB is now known as BradB|out [03:58] where can I find LP's psycopg? [03:58] it doesn't seem to be in rocketfuel at all [03:58] bob2: apt-get [03:58] yeah, lifeless just said that [03:58] silly me [03:59] The LaunchpadDevelopmentEnvironment wiki page (or whatever the name is) is reasonably up to date. [04:00] ah, forgot about that, thanks [04:04] File "/home/rob/projects/warthogs/launchpad-2/launchpad/lib/canonical/launchpad/database/project.py", line 98, in __iter__ [04:04] for row in self.table.select(): [04:04] File "/home/rob/projects/warthogs/launchpad-2/launchpad/lib/sqlobject/main.py", line 1198, in __iter__ [04:04] return conn.iterSelect(self) [04:04] File "/home/rob/projects/warthogs/launchpad-2/launchpad/lib/sqlobject/dbconnection.py", line 507, in iterSelect [04:04] select, keepConnection=True)) [04:04] TypeError: iteration over non-sequence [04:04] 127.0.0.1 - 16 [4/Nov/2004:14:03:13 +1100] "GET /doap/projects/ HTTP/1.1" 200 292 "http://localhost:8085/doap" "Mozilla/5.0 (X11; U; Linux ppc; en-US; rv:1.7) Gecko/20041013 Firefox/0.9.3 (Ubuntu)" [04:04] (when viewing the projects) [04:04] also when adding one [04:07] Update your SQLOjbect. [04:07] Make sure you have the latest sqlobject--test--0.6 [04:11] erk, oops [04:12] (Make sure the rest of your build config is up to date while you're there...) [04:12] this is the buldbot config, which seems a tad...broken [04:12] out of date is more accurate. [04:12] but yeah [04:13] hm, and now it works [04:13] but I'm sure it didn't before [04:18] great. [04:18] now you need to setup a simpe cvs repo on your hard disk, and then crate a 'revision control system' entry for that in launchpad [04:20] ok [04:25] Merge to rocketfuel@canonical.com/dists--devel--0: launchpad dogfood config and snapshot helper (patch-31) === doko [doko@dsl-082-082-066-197.arcor-ip.net] has joined #launchpad [04:43] bob2: how is that going ? [04:43] lifeless: just cerating the cvs repository [04:44] which is suckage on the order of tla [04:44] cvs init DIRECTORY [04:44] or pydoc CVS [04:45] nope [04:45] -d or $CVSROOT [04:47] import CVS [04:48] oh [04:48] good point [04:50] lifeless: er, where do I tell LP about it again? [04:50] project, product, create revision control system [04:53] ah [04:54] it requires a tarball in addition to the repository url? [05:02] no [05:02] just leave that field black. [05:02] *blank* [05:02] and give it :local:/path/to/rep in the rep field [05:02] the red dot next to it says "you have to fill me in", fwfw [05:02] its bullshitting you [05:02] ah [05:02] and the module? [05:02] whatever module you ahve created in there [05:03] (mkdir creates a module, FWIW) [05:03] is an empty module ok for buildbot? [05:03] yes, but not for the test you need [05:03] errr [05:03] oh yeah, I need actual stuff to import [05:03] yeah [05:03] so. mkdir foo in the repo [05:03] ok, mkdir to make the module, co it out, "cvs add" some stuff, cvs ci [05:03] then checked with -d..yep [05:04] ok, cool [05:04] chieck in a few different things with diff log emssages [05:19] Merge to thelove@canonical.com/hackerlab--devo--1.1: merge in chth@gmx.net--tla2004/hackerlab--pika-escaping-fix--0 which extends the character classes to allow bugfixes for tla (patch-3) [05:26] what *is* pika-escaping? [05:26] unicode escaping, NIH style. [05:26] ah [05:27] you know you have too many background processes when you have to grep the output of "jobs" to find the one you want [05:27] launchpad doesn't whine very hard when I fuck up an import specification [05:27] daf: garh [05:27] bob2: what did you do ? [05:27] lifeless: didn't enter the arch branch details [05:27] heh. [05:28] and there's no "sync me harder" button [05:28] right, its fully separate. [05:28] and ? [05:29] oh, I just want a button that says "sync me harder" [05:30] troll [05:31] you gotta keep people on their toes [05:31] a consistent UI is a bad UI [05:43] File "/home/rob/projects/warthogs/launchpad-2/launchpad/botmaster/master.cfg", line 95, in ? [05:43] clause="product.id = sourcesource.product AND project.id = product.project AND NOT project.name='do-not-use-info-imports' and NOT product.name='unassigned' and sourcesource.processingapproved is not NULL") [05:43] File "./jobstuff.py", line 19, in processDB [05:43] SQLBase.initZopeless(connectionForURI('postgres://' + dbhost +'/' + dbname)) [05:43] exceptions.AttributeError: type object 'SQLBase' has no attribute 'initZopeless' [05:43] lifeless: ^ [05:44] you dropped in my jobstuff.py file ? [05:44] run up python with the right PYTHONPATH [05:44] in the botmaster dir? yes. [05:45] oh, fraggit. [05:45] I know [05:45] my fault? [05:48] no [05:48] .msged you [05:48] yeah [05:48] Bad PQM! Bad! [05:48] 'Package rocketfuel@canonical.com/launchpad--dogfood--0 not found in config' [05:48] hmm. [05:49] do a debug attempt [05:49] mail me the output [05:49] I think it is because the line in the config is actually: ./dogfood rocketfuel@canonical.com/launchpad--dogfood--0--base-0 [05:50] Hmm... I know what I want it to do, but I see why it is bitching... [05:51] I think I need configs 'dogfood' which just lists the versions, and config 'dogfood-snapshot' which lists the explicit patch levels. [05:52] stub: no, you don't. [05:52] thats not how the code works. [05:55] I'm thinking I need a 'latest patches' config which pqm uses to checkout a dogfood release to run the tests, and a 'snapshot' config which specifies revisions of all packages that were rolled out. [05:56] Maybe this snapshot stuff is YAGNI - we probably don't actually care what version of sqlobject we were running last week. [05:57] stub: I'm not tracking your needs here. AIUI a merge failed? please send me the debug run output I requested, and I'll fix it for you [06:00] I think the debug submit is on the way. I can see why it is failing though. [06:02] I'm specifying an explicit version in the build-config file (rocketfuel@canonical.com/launchpad--dogfood--0--base-0 atm). So when I tell PQM to merge my local changes of launchpad--dogfood--0 into rocketfuel@canonical.com/launchpad--dogfood--0, it fails. PQM isn't intelligent enough to truncate the --base-0, and I don't know if that would be correct in all situations anyway. [06:03] Merge to rocketfuel@canonical.com/arch-pqm--main--0: implement a debug command, which when seen causes all output to be returned in email (patch-14) [06:24] bah - now pqm just isn't talking to be. === stub goes to book plane tickets [07:51] lifeless, elmo: look like I killed pqm [07:54] booya [08:00] Don't know if that is any healthier... [08:09] lifeless: Still spinning [08:10] boo [08:40] lifeless: Still no responses. Some lock not cleared? [08:40] no, locks are fine. [08:40] you had killed it though. [08:40] there are no messages in its queue [08:41] oh bah [08:41] it doesn't barf anywhere nearly loud enough on a locked lock. [08:50] Merge to thelove@canonical.com/bazaar--devo--1.0: merge chth@gmx.net--tla2004/tla--taglineescaping--0 which fixes a lack of escaping in inode signatures (patch-60) [09:23] Merge to rocketfuel@canonical.com/dists--devel--0: seperate dogfood.snapshot config (patch-32) === lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad === Kinnison [~dsilvers@haddenham.pepperfish.net] has joined #launchpad === cprov [~cprov@200.158.100.251] has joined #launchpad [11:45] morning cprov [11:59] Kinnison: morning, how are you ? [12:00] Not bad thanks === bob2 [rob@bob2.user] has joined #launchpad [12:02] Kinnison: do you know any news about the lp_dogfood code update ? isn't it supposed to be yesterday night ? === Kinnison has no idea; sorry === Kinnison has been too busy running through pretend domination exercises ready to write up that bit of lucille [12:04] Kinnison: ok. === Kinnison thinks he's ready to write canonical.lucille.Dominator [12:13] Merge to rocketfuel@canonical.com/launchpad--devel--0: get more gina bits merged (patch-731) [12:15] thanks babe === debonzi [~debonzi@200.158.100.251] has joined #launchpad [12:28] hey debonzi. === carlos [~carlos@69.Red-80-33-181.pooles.rima-tde.net] has joined #launchpad [12:32] hi [12:33] hi Kinnison [12:35] So - how do I restart that twisted thing I had to kill on Mawson to update the database schema? [12:35] cd into the /srv/launchpad.ubuntu.com/librarian dir [12:36] run ./run-librarian.sh [12:36] And that somehow telepatically picks up the correct ports and addresses? Or are the defaults fine? [12:36] the defaults are correct [12:36] and it overrides the librarianroot etc. === elmo [~james@82.211.81.249] has joined #launchpad [12:37] Well.. it appeared to do nothing so I'll assume it worked :-) [12:38] process running at least [12:40] heh [12:40] tail the log in that dir and see if it thinks it started up okay [12:42] seems to have started up happily on port 8000 and set its uid & gid. No errors :) [12:42] I like it when software works. === Kinnison has a brainwave; replaces the word "binarypackage" with the word "build" in a whole pile of his mental processes and feels as the concepts intrinsically simply and a corner-case falls away for free [01:59] s/simply/simplify/ === kiko [~kiko@200-206-134-238.async.com.br] has joined #launchpad [02:28] hey kiko [02:29] hey kinni [02:30] what's on? my views ruining the morning? === Kinnison hasn't looked at 'em :-) === Kinnison is writing domination code [02:31] heh [02:31] spiv, you spivster! [02:35] kiko, you... kikoster! [02:36] now I'm actually going to have to *work* on this holiday [02:36] waaaah [02:38] Heh. === stub [~stub@dsl-246.248.240.220.dsl.comindico.com.au] has left #launchpad [] === ddaa [~ddaa@nemesis.xlii.org] has joined #launchpad === ddaa has done some queries on the taxi data [02:59] spiv, lifeless: that stuff looks sane [03:00] The involvement of the branch table was not obvious at first, so I got some weirdness initially, though. [03:01] I can get the revision listing, the hashes for log files, and there are no hashes for checksum files. [03:01] ddaa: Excellent :) [03:02] i would not bet my job that it is correct though, but it has passed the "is not obviously borken" test as far as I am concerned. [03:02] not tested for sync though. [03:02] spiv: are different code paths exercised for sync than for import? [03:09] No idea. [03:10] taxi is working - spiv has fixed the issue that was breaking us. [03:11] bob2 is fixing the logic issue that he left behind in oxford. [03:11] lifeless: you are telling me there is not point in being paranoid? [03:11] yes, exactly. === ddaa puts --no-paranoid in his argv [03:12] taxi itself is a thin wrapper around the database arch stuff, which was heavily tested. === ddaa reloads [03:12] Hello, my name is cindy. === ddaa purge the cache and reload. [03:12] and I'm a material girl o/ [03:12] night [03:12] Material? [03:13] song reference [03:13] Ha. Weak reference then ;-) [03:14] g'night. [03:16] spiv: do you have some pointers/keywords to find the magic stuff to get a python backtrace from gdb? [03:22] ddaa: I have it in my .gdbinit back in .au :/ [03:22] I remember it being difficult to google for, too :/ [03:22] yup... google did not turn up anything useful... [03:22] trying gmane, now [03:23] kw: python gdb backtrace [03:23] Ah: http://cvs.sourceforge.net/viewcvs.py/python/python/dist/src/Misc/gdbinit?view=markup [03:23] Well, the python term is usually "traceback" :) [03:24] Looks like it's in python cvs these days. [03:24] this can explain that :-) [03:24] I assume that twistd will use whatever python is first in the path, right? [03:24] Alternatively, there's: "call PyRun_SimpleString("import sys, traceback; sys.stderr=open('/tmp/tb','w',0); traceback.print_stack()")" [03:25] (from http://www.upfrontsystems.co.za/Members/jean/zope-notes/debug-spinning-zope) [03:25] that looks totally unsafe [03:25] the python interpreter is all but reentrant. [03:26] The debian package of twistd's #! is #!/usr/bin/python2.3 === ddaa sighs... [03:26] You could do "/path/to/my/python `which twistd` ...." [03:27] ha, good. Thanks. [03:27] If you run Twisted out of a SVN checkout, "/path/to/my/python bin/twistd ...." would work. [03:28] Do i look like a guy to run Twisted out of a SVN checkout? :-) === debonzi -> lunch === BradB|out is now known as BradB === stub [~stub@dsl-246.248.240.220.dsl.comindico.com.au] has joined #launchpad === BradB is now known as BradB|brb === debonzi [~debonzi@200.158.100.251] has joined #launchpad === kiko is now known as kiko-fud === BradB|brb is now known as BradB [05:05] Merge to rocketfuel@canonical.com/launchpad--devel--0: Added some code to detect the po directories (patch-732) === BradB notes how difficult it is to find the place to say "I forgot my password" on ubuntulinux.org [05:07] BradB: there is a bug report open about that feature, because seems to be failing [05:07] that's what i'm looking at :) [05:07] :-P [05:09] debonzi: Do you have access to the error log for the forgottenpassword functionality? It's not part of the regular Plone site, so I have no idea where to look to find out what the error was. [05:15] BradB, I have no idea about it.. sorry [05:22] BradB: I think it's in our production launchpad server [05:22] but I'm not sure [05:23] debonzi: Who knows, Steve? === salgado [~salgado@200-206-134-238.async.com.br] has joined #launchpad [05:26] BradB, probably.. [05:27] BradB: not sure, but I think spiv was working on it [05:29] I know nothing about how the forgottenpassword thing works. === Kinnison -> out for a bit; got mobile if you need me [05:29] spiv: I thought you worked on the integration of launchpad's logins with our plone site.. [05:30] carlos: Yes, I did. :) [05:30] But not the forgottenpassword bit :) [05:30] carlos: It was debonzi and SteveA [05:30] oh, ok O:-) [05:31] BradB, Ive implemented it, but I have no idea where it is running and how it was inserted in the ubuntulinux site === BradB punts this one to the eggman === carlos goes to the university [05:36] later [05:51] BradB: and elmo only added a single redirect to macquarie, so it looks like SteveA is the man to sort this! [06:12] yep === kiko-fud is now known as kiko [07:26] BradB:ping [07:27] c'ya guys [07:27] time for me to head off for the evening [07:27] kinnison: ciao [07:29] lulu: pong [07:30] Bradb: :o) hey! Once RSS is enabled on the site, will it allow users to add an feed themselves? [07:30] a feed... [07:34] lulu: not afaik [07:35] oh, I was hoping they could! another item in the "add new item" drop down list..... [07:36] ok - thanks for that :o) [07:37] no prob. i don't know the sin_tool really well, but it's pretty simple, so perhaps we can look at that again when we know for sure which feeds we'll show, and then maybe we can see if there's some way to do what you want to do. [07:50] BradB: yup - perhasp we should do a test run on ubuntu_test :o) [07:51] I'm outta here - have a great weekend :o) === lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has left #launchpad [] === cprov [~cprov@200.158.100.251] has joined #launchpad [08:08] kiko: you're evil. [08:08] class SourcePackageRelease(SQLBase): [08:08] implements(ISourcePackageRelease) [08:08] _table = 'VSourcePackageReleasePublishing' [08:09] :( [08:09] no wonder i couldn't figure out why the table sourcepackagerelease has two rows, but my vocab was only spitting back one term every time. [08:12] kiko: can you look into this? this is the last thing preventing me from checking in now (the solution seems to either be to fix your view to outer join as needed to make absolutely sure that it returns a row for each row in sourcepackagerelease, or to set _table = "SourcePackageRelease", and create another class for this. === BradB goes for lunch === BradB is now known as BradB|lunch [08:38] BradB|lunch, I am indeed evil, you got that part right. [08:38] BradB|lunch, can you elaborate in details what your problem is? [08:39] Merge to rocketfuel@canonical.com/launchpad--devel--0: improves on soyuz people/person pages (patch-733) [08:45] stub: ping [08:49] I can only imagine stub's asleep === kiko is now known as kiko-afk === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #launchpad === carlos [~carlos@69.Red-80-33-181.pooles.rima-tde.net] has joined #launchpad [09:18] Ho my... vte looks terribly like it used arch back in 2002 [09:18] the first revision is removing a .arch-ids directory... [09:19] lifeless: what do we do of that one? Contact upstream to figure out what's up? === carlos [~carlos@69.Red-80-33-181.pooles.rima-tde.net] has joined #launchpad === BradB|lunch is now known as BradB [09:41] kiko-afk: the problem is that you made SourcePackageRelease a view. [09:41] kiko-afk: this creates two issues (correct me if i'm wrong): [09:42] 1. it doesn't behave the same as if the table were SourcePackageRelease (as mentioned above, there's two rows in SPR, but I only get one result now that your view has been swapped in place of what used to be the SPR table) [09:42] 2. people who expect to assign to SPR attributes are in for a surprise. === debonzi -> dinner [10:11] BradB: I haven't received any Malone announcements yet === debonzi [~debonzi@200-158-107-151.dsl.telesp.net.br] has joined #launchpad === debonzi is now known as debonzi|dinner [10:14] daf: i haven't been able to land my changes yet [10:14] daf: and i have to wait for kiko-afk till i can [10:15] oy oy [10:15] woo, the man who can save the dya [10:15] BradB, you're right on the second count. I'm not *entirely* sure that you're right on the first one [10:16] i am right on the first one, i'm seeing it happen, and my tests are now breaking because of it :) [10:16] heh [10:16] if you want to reproduce it, you can check against your sample data. [10:16] sure [10:16] can you describe it more precisely? [10:16] the real fix is to use a separate class for the view, btw [10:16] and not hitch on sourcepackagerelease for it [10:16] indeed it is [10:16] problem is that's not a trivial task [10:17] can you by any chance work around me for today? I've got salgado on it [10:17] here's the exact problem (one sec): [10:17] Ouch, go to https://launchpad.ubuntu.com/malone/bugs, click on #5. [10:18] that's a dead chicken anyway [10:18] it was a bug mdz entered when i demo'd malone to him [10:18] so the data has since gotten blown away it looks like === BradB resets his db to show kiko the spr view prob [10:20] kiko-afk: [10:20] launchpad_dev=# select count(*) from sourcepackagerelease; [10:20] count [10:20] ------- [10:20] 2 [10:20] (1 row) [10:20] launchpad_dev=# select count(*) from vsourcepackagereleasepublishing; [10:20] count [10:20] ------- [10:20] 1 [10:20] (1 row) [10:20] I wonder how that happens. [10:20] can I see the rows themselves? [10:20] missing an outer join surely [10:22] kiko-afk: http://paste.husk.org/1931 [10:23] a quick hack might be to do whatever you tell me to do with that row to make it join, but then, it's possible that'd break something else [10:23] BradB, can't you just comment out the test? [10:23] alternatively, to just put the table name back in _table, but that too might cause tests to fail [10:23] kiko-afk: nope, it's about 7 of my page tests [10:23] this needs some looking into -- it seems that you're using broken sample data [10:23] kiko-afk: i'm using *the* sample data :) [10:23] kiko-afk: get a move on :) [10:24] it may be broken -- it's not a gina run -- because the essential join with release is in the publishing tables and I suspect yours is empty. [10:25] kiko-afk: in any case, i'm not concerned about the details of the view itself, SourcePackageRelease's table needs to be the correct one again (i.e. SourcePackageRelease) without everything blowing up. I can try simply making that change, but I'm dubious. [10:25] Is adding a comment to a bug expected to fail with an IOError. [10:25] r=kiko on trying that, as anything I would do would take more time. [10:25] seriously [10:25] I have salgado on it but I would be producing the "correct" fix [10:25] not a reversal hack. [10:26] spiv: yes! heh, no :P [10:26] you can revert to the original table and comment out the extra columns, I guess [10:26] BradB: OError: [Errno 13] Permission denied: '/var/tmp/launchpad_mailqueue/tmp/1099602988.9384.rosetta' :) [10:26] Er, IOError :) [10:26] wee, mail delivery doesn't have the correct perms [10:26] that's stub's problem [10:26] stub: Whee! === BradB tries the SPR hack === spiv files a bug... [10:27] spiv: thanks [10:28] BradB: Except I can't -- not in malone... [10:28] ah, good point [10:28] Because of the same error :) [10:28] yes, urgh === spiv resorts to bugzilla [10:30] New bug 2157 for Launchpad/Malone: Can't add new bugs or comments due to IOError [10:30] https://bugzilla.warthogs.hbd.com/bugzilla/show_bug.cgi?id=2157 [10:38] kiko-afk: From my end, I'll change ISourcePackageRelease and SourcePackageRelease to be suitable for the SPR table. From your end, all you should need to do is make a new interface and class for the view, and then change whatever code's using it to use that view class instead of SPR. [10:39] BradB, will do. salgado, sounds okay? [10:40] kiko-afk: yes [10:41] kiko-afk: What do we call SQLObjects that are views then? VFoo? [10:41] Maybe the same as the view name in the DB, who knows... [10:44] kiko-afk: Also, could you give me meaningful descriptions of each of the following please: [10:44] builddepends = Attribute("") [10:44] builddependsindep = Attribute("") [10:44] architecturehintlist = Attribute("") [10:44] dsc = Attribute("") [10:46] those 4? [10:46] kinni would be a better person to ask. [10:47] builddepends is a list of packages on which this package depends on to build (comma-separated) [10:47] dsc is the dsc file (I believe it's a librarian ID for it or something) [10:48] builddependsindep are architecture-independent packages this package depends on to buid; according to elmo this is a hack that could go away IIRC [10:48] archhintlist is beyond me, I don't know. [10:49] What's a DSC file? [10:49] I've heard the term bandied about, but I'm no package maintainer so I still don't know what it means. [10:49] DEB source description? Something like that. [10:49] ah [10:49] it lives inside a source package [10:50] is it like a manifest? [10:51] ddaa: look on google. [10:51] see if you can determine if arch is still in use. [10:51] if it is, and they have the history, we can just mirror. === BradB uses google to find out what a dsc is :P [10:51] otherwise we cscvs and use their ids [10:52] Will do. Just not today. [10:52] I checked the supermirror and did not find anything convincing. [10:52] And I know jblack has been to great lenghts to get his hands on anything google could find. [10:53] btw the "use their ids" part eludes me. How? [10:55] cscvs can use a reference tree to determine file ids [10:56] I know, I wrote that feature... [10:56] I mean, how do I put that into the import process? [10:57] How, and btw, what gave me the clue is that some cvs rvsns do put the ids in the tree, or other revisions remove them. === ddaa checks the changelog [10:58] Neverminh... [10:58] brainfart... [10:59] got confused by the presence of id files in the changelog... [10:59] New bug 2158 for Launchpad/Database: SourcePackageRelease table does not have comments [10:59] https://bugzilla.warthogs.hbd.com/bugzilla/show_bug.cgi?id=2158 [10:59] lifeless: vte looks good to go, then. [11:00] ddaa: can you do the project & product mappings for it? [11:01] Aye aye. [11:06] debian source control I think, but it's basically, a file ending in ".dsc" === netjoined: irc.freenode.net -> sendak.freenode.net === salgado [~salgado@200-206-134-238.async.com.br] has joined #launchpad === netjoined: irc.freenode.net -> sendak.freenode.net === salgado [~salgado@200-206-134-238.async.com.br] has joined #launchpad [11:40] BradB: mark asked me to create a new sampledata, and it's done now. it's a mix of a gina run and the current sampledata. can you try it, and tell me if it's ok for you? [11:40] salgado: hm? that's not the real problem here. [11:41] the fix for the Real Problem is currently being merged by pqm (at least i hope i don't get yet another failure, anyway) [11:41] BradB: I know it's not the problem. [11:42] I only want to get some feedback of the new sampledata, before merging it [11:42] salgado: the real problem from my perspective was that SourcePackageRelease wasn't a SourcePackageRelease. I'm leaving the view stuff in your guys' hands though, because it doesn't affect me once the latter problem is fixed, which it now is. [11:45] Merge to rocketfuel@canonical.com/launchpad--devel--0: built a forcefield for malone. some cleanup, and moving around of pagetest-related code. bug add form fixes. unbroke SourcePackageRelease. (patch-734) [11:45] BradB: tomorrow I'll be hacking on the VSPR bug. the new sampledata has nothing to do with it === BradB praises the lord (re: pqm accepting my merge :) [11:46] stub: time for a new dogfood rollout dude! [11:52] spiv, kiko-afk, stub: can you think of any reason why I'd care what the "nickname" of a bug is? === kiko-afk is now known as kiko [11:52] it's an alias, it's often useful when you have gobs of thousands of bugs. [11:52] (I'm now cleaning up the bug listing to rename Owner => Submitted, add Assigned To, and remove "nickname" probably) [11:52] instead of bug 2314213211 [11:52] hm [11:52] you can have bug "sourcepackage-views" [11:52] kiko: bugzilla uses numbers though [11:53] I'd rather have "2113" [11:53] kiko: realistically nicknames'll go the way of hotmail addresses: "sourcepackage-views-69" [11:59] BradB, bugzilla has both numbers and aliases, for the record. [12:00] I should know -- I helped review the original code ;) [12:00] fair enough :P