[00:22] hmm, newbie question I guess, but I have to make changes to the config and try to change it using "export LPCONFIG=tempconfig; make run" and "LPCONFIG=tempconfig make run" but it keeps starting up with LPCONFIG=development ... a hardcoded restriction somewhere? [00:24] motin_0: LPCONFIG=tempconfig make run [00:24] * thumper thinks [00:25] thumper: that's exactly what I try. it still starts as for instance utilities/shhh.py make -C sourcecode build PYTHON=python2.4 \ PYTHON_VERSION=2.4 LPCONFIG=development [00:26] motin_0: have you set up a tempconfig? [00:26] there is a directory called tempconfig, yes [00:26] it that's enough? [00:26] a copy of development directory [00:27] tried `make run LPCONFIG=tempconfig`? [00:27] I can't remember how to make make work properly [00:29] ey thumper that did it :) [00:29] so make has it's own way of using env vars [00:29] thats new to me, thanks for the suggestion! [00:30] Well, they're make variables, not envvars. [00:30] They just happen to use an unfortunately similar syntax. [00:31] wgrant: ah. in configs/README.txt it says "This directory stores configs. The config used is selected on startup using the LPCONFIG environment variable." [00:33] motin_0: for scripts yes, not for make [00:34] kinda confusing [00:34] patches welcome :) [00:34] I thought about it, but then I would probably commit some change that itself was wrong, since I have yet not understood the greater picture === _thumper_ is now known as thumper === matsubara is now known as matsubara-dinner === mup_ is now known as mup === matsubara-dinner is now known as matsubara === mup_ is now known as mup [02:15] spm: can you take a look at the buildbot? it seems stuck [02:15] I can't get sendmailMailer to work. despite it being a development box, I really want it be able to send mail to me and my colleagues that will work on this together... is it hard-coded to prevent this in all cases or can anyone show a config-file that would actually allow external mails? [02:15] thumper: not atm ; am mid a rollout. hopefully in the next 15-30 [02:16] spm: ack [02:17] motin_0: most development instances are careful not to send real email [02:17] motin_0: it will be configurable [02:17] but I'm not sure where it it [02:18] thumper: right, I've been following zope docs and that won't cut it, so I guess launchpad devs has taken some extra measures [02:18] motin_0: almost certainly [02:19] ./configs/development/launchpad-lazr.conf:211:default_recipient_address: root@localhost [02:19] yeah I can configure the stub-mailer [02:19] ./configs/development/launchpad-lazr.conf:210:default_sender_address: root@localhost [02:19] to send to a specific address [02:19] hmm, I could hack stub-mailer... [02:19] you shouldn't need to hack IMO [02:20] I never got farther than this using zope example confs: ConfigurationError: ('Unknown directive', u'http://namespaces.zope.org/mail', u'sendmailMailer') [02:20] feel like hacking is the only way to go at this stage [02:22] hmm I think I found something [02:22] I'll get back with results [02:22] look in mail-configure-normal [02:24] thumper: yep, problem is that there is no mailer named sendmail [02:24] and defining it by means of standard zope yields the above error [02:28] hmm seems I have been looking in to old docs or something [02:28] http://mail.zope.org/pipermail/zope3-checkins/2005-July/025405.html [02:28] sendmailMailer was removed in 2005 [02:29] sorry about that [02:32] Has anyone else observed lots of garbage-warnings in the lp tests, concerning bzrlib.smart.medium.SmartSimplePipesClientMedium ? [02:37] thumper: yep, that solved it. sendmailMailer is dead end. smtpmailer is the way to go. simple as that. just some bits of doc-info (in the code) left still referring to sendmailMailer made me go down the wrong path here [02:37] thanks for your support thumper [02:37] motin_0: we have been using quite an old zope [02:37] motin_0: we are actively updating it now [02:38] ok that might explain some discrepancy [02:57] thumper: have kicked the slaves; if that still fails to shake tailfeathers, I'll try the master as well [02:58] spm: thanks [03:09] spm: you might need to give the master a shove [03:09] spm: the db_lp builder and the bzr builder are stuck [03:09] yeah was coming to that conclusion too... [03:09] spm: and the lp builder hasn't started again :( [03:10] see if that does it [03:10] yay, that looks like it === matsubara is now known as matsubara-afk === Ursinha is now known as Ursinha-afk [04:48] sinzui: I think your branch may have problems [04:49] sinzui: my ec2test has some weird failures that seem unrelated to my 3.0 changes [04:49] thumper: what are the errors [04:50] sinzui: https://pastebin.canonical.com/21035/ [04:51] thumper: that is me, [04:52] thumper: Those are the links I had to add to the menu, so I will update the test. [04:52] sinzui: ok [04:53] thumper I recall the test and thought I fixed it. [04:54] sinzui: did you commit? [04:54] thumper: I was very tempted to remove the SimpleViewClass part of the test, maybe even the name of the template. Do we care about the link or the specifics of the implementation? [04:54] thumper: I think I did something very stupid, revert, because I really think this test is stupid [04:55] sinzui: delete the test then [04:55] thumper: The check_menu_links link appears to be do some very cumbersome work to prove the links works. [04:57] sinzui: check_menu_links isn't really adding anything here [04:58] sinzui: it should really just validate the links [04:58] sinzui: we don't care if people add some [04:58] sinzui: we do care if they are wrong though [04:58] perhaps a new unit test that validates the menu [04:59] I think that was the intent. EdwinGrubbs struggled with this test a long time ago. yes, a unitest that verifies every menu is the correct answer. [05:00] sinzui: I could add in the extra bits into the branch I'm about to land if you like [05:01] sinzui: or remove that chunk [05:01] * thumper wonders about the testfix demon [05:01] thumper: adding chunks is my immediate intent. [05:01] ok [05:01] sinzui: I can do that if you like [05:02] sinzui: or should we wait for the failure and testfix? [05:02] thumper: There are some equally idiotic translations menu tests I want to read before doing something clever like, get every menu, instantiate the link for the menu's context, and report any named adaption failures [05:03] I just branched. I can put the test fix in in a few minutes. [05:03] Oh! PQM is still frozen [05:04] sinzui: if you put testfix in, will it preempt testfix mode? [05:04] spm: unfreeze pqm? [05:04] thumper: oops. will do. [05:05] sinzui: ok, I'll leave that bit to you then [05:05] sinzui: I'm about to land generic-edit.pt [05:05] sweet [05:05] sinzui: already deleted half a dozen redundant page templates [05:33] thumper: product-menus is no longer brittle. check_menus reports adaption errors, not state. It is now a good candidate for a real test that verifies every menu. [05:33] sinzui: cool [05:33] need a review? [05:34] No, not for this [05:34] ok [05:34] but I think I need to talk to salgado about how to get every menu in module without instantiating it. [05:46] * thumper helps with dinner, back later [05:56] hi folks [06:27] hey jtv [06:28] hi spm [08:29] good morning [08:46] Morning Launchpadders. [09:08] * thumper is back wrestling with +activereviews [09:11] Hi Launchpad-land [09:22] good morning Wolves [09:23] hi uk poeoples [09:23] ETOOLATE [09:23] eyup thumper [09:45] Err. [09:45] Anyone have any idea why I'm getting "ImportError: No module named buildout.buildout" this morning when I run make [09:45] ? [09:46] * gmb re-runs rf-get for giggles [10:08] gmb: I think gary landed an updated buildout egg [10:09] thumper: Yeah, looks that way. Running bin/buildout manually seems to have fixed it. [10:09] * gmb hates at buildout at sub-11-am [10:24] ah crap [10:25] I need to change the MenuAPI tal formatter for my code to work nicely :( [10:36] thumper: would it be easy to use the webservice api to check for new revisions, but only for an lp sub-project? (soyuz in my case of course) [10:36] no [10:36] * bigjools 's idea goes down in flames [10:36] bigjools: it would be quite hacky bzrlib [10:37] thumper: I will talk to diogo, they do it for the qa pages [10:37] unless it's a dirty hack based on who landed it... [10:38] It is. [10:38] I saw the code once. [10:38] And contributors appear on LaunchpadTestPlan. [10:45] * maxb wonders: what in Launchpad uses lxml / pysqlite2 [10:46] Because I've not rebuilt those to be available with python 2.4 in karmic, yet nothing's breaking [10:47] maxb: sqlite is probably bzr-git. [10:47] ISTR we wanted lxml to die a horrible flaming death a while back, but I might just be projecting. [10:49] Isn't lxml the XML lib that *shouldn't* die? [10:50] lazr.restful's tests use lxml, but grep doesn't show anything else. [10:51] maybe it's just gratuitously included for people who hack on lazr.restful then [10:51] Probably. [10:53] Oh well. When I get through another full test run tonight, we get to see how many of the problems I'm seeing are due to running on Karmic, vs. how many due to Python 2.4 - because even running Py2.4 on karmic, the testsuite is not entirely happy [10:54] If we're lucky it'll just be the "it breaks if the system bzr is new enough for buildout to use it" problem [10:54] we usually get lots of breakage when moving to a new distroseries [10:56] Particularly in the Soyuz bits that use dpkg bits? [10:58] that's rare actually [10:58] Morning, all. [10:59] hi deryck [11:48] * gmb waits has firefox tries to do > 1 thing at once and melts into a steaming pile of XUL. [11:52] gmb: This is why I'm starting to like Chromium. [11:52] Although LP bug listings are a bit broken in it. [11:52] wgrant: Yeah, I'm very much of the same school. Haven't seen the bug listing problem yet... can you point me to an example? [11:56] gmb: The importance icons are absent. [11:56] Oh. That's suboptimal. [12:01] wgrant: does chromium have anything like firebug, noscript and adblock? [12:01] wgrant: you mean the sorting? i thought it was fixed [12:02] bigjools: it has tools equivalent to firebug built in. i think extensions are now possible but there isn't that much choice out there yet [12:02] okay - ta [12:02] seeing pages now without adblock is....painful [12:03] * intellectronica actually reads the scrollback [12:03] There's an adblock-like thing around now, apparently. [12:03] I will stick with Konqueror as my alternative browser [12:04] But it's all still a bit unfinished. [12:04] And does strange focus things. [12:05] wgrant: so, there are still quite a few problems with sprites in webkit-based browsers. if you find anything it would be great if you can file a bug (or at least give a shout here) because little icons are so easy to miss (and you can't test that automatically either) [12:06] intellectronica: I've noticed a few. I'll file bugs in future. [12:06] thanks === henninge is now known as henninge-lunch === kiko is now known as kiko-phone [12:26] * gmb lunches === henninge-lunch is now known as henninge === salgado-afk is now known as salgado === mrevell is now known as mrevell-lunch === mrevell-lunch is now known as mrevell [13:51] cprov1, bigjools, call in 15? [13:51] yep === cprov1 is now known as cprov [13:51] ok [13:53] 14 [13:53] Rats [13:58] anyone from bugs around? [13:58] gmb: ? [13:58] sidnei: Howdy. What can I do for you? [13:59] gmb: so, i think we found a bug with 'super projects' or whatever is the term for that [13:59] gmb: we have landscape-project, which is a super-project for landscape and landscape-client [13:59] Right. [14:00] gmb: if someone is an admin on '~landscape' and files a bug directly into 'landscape', he gets all the extra options [14:00] gmb: if this person files a bug in 'landscape-project', but picking 'landscape' from the drop-down of 'child' projects, he only gets 'tags' in extra options [14:00] beuno: skype? [14:01] sidnei: Right. I *think* this is something we're kind of aware of. Let me just take a look... [14:01] gmb: ok [14:02] bigjools, yes, searching for headphones [14:03] sidnei: So, this is essentially bug 375144. The problem is that we work out who can see the extra options based on what context they're filing the bug in (i.e. project, project group, front page, etc.) rather looking at what they're trying to file the bug against. I'll update the bug to include this case. [14:03] Bug #375144: "Extra options" section doesn't show for my projects on the frontpage +filebug form, even though I'm the bug supervisor [14:03] sidnei: Thanks for the heads-up. [14:03] gmb: cool, thank you! [14:05] beuno: I'm surprised you need to search, given the number of calls you're on ;) [14:06] bigjools, it's usually my way of saying "I'm going to get some tea" [14:06] bigjools, what's your skype id? [14:06] beuno: I'll give you no guesses [14:07] bigjools, cprov, ready when you are [14:07] beuno: noodles is joining us too, to listen in [14:07] I guess I'l host [14:09] beuno: https://dev.launchpad.net/SoyuzBuildIndexPage [14:17] cprov, https://devpad.canonical.com/~beuno/LP_new_design_Bugs_v3.1.png [14:20] bigjools: can a user not remove his/her own ppas? [14:21] henninge: they can't, they need to file a question [14:21] bigjools: ok, assign it to you, then? [14:21] and the losas will take care of it [14:21] ah, losas [14:21] no, losas [14:21] ;) [14:21] cheers [14:21] np [14:24] cprov, https://devpad.canonical.com/~beuno/LP_new_design_Home.png [14:26] http://people.canonical.com/~michaeln/3-0-builder-mechanical/0-builder-index-v2.png [14:27] someone please shovel more coal into staging [14:34] http://people.canonical.com/~ed/dsp_mockup2.png [14:34] https://dev.launchpad.net/VersionThreeDotO/Soyuz/NavigationRedesignUI [14:41] https://edge.launchpad.net/ubuntu/+source/firefox-3.5 [14:43] https://dev.launchpad.net/SoyuzPackageUI?action=AttachFile&do=get&target=sample_layout_for_ff.jpg [14:50] beuno: ping [14:50] henninge, hi [14:50] hi beuno! [14:51] beuno: We'd like to have a call/chat with you about the translate page that I made suggestions for. [14:51] beuno: we have a new mock-up! ;_) [14:51] henninge, wooo [14:51] beuno: When would be a good time? [14:51] I'm on the phone with the soyuz guys [14:51] and have 3 more calls today === danilo-afk is now known as danilos [14:52] how's your schedule? [14:52] beuno: tomorrow suits you better then, I guess. [14:52] henninge, maybe today as well [14:52] beuno: meetings, meetings, meetings ... let's see if I can fit you in ... [14:52] ;-) [14:53] henninge, whats your schedule today? [14:53] beuno: reviewer's meeting in 7 minutes. And I have to leave at aout 17 UTC at the latest. [14:54] danilos: how's your schedule today? [14:54] henninge: I am up for a call anytime except when beuno is on a call as well later in the evening :) [14:55] henninge, danilos, I'll try to in 30 minutes [14:55] beuno: cool, fine by me. [14:56] hello fellow launchpad hackers! we'll be having our weekly ameu reviewers meeting in #launchpad-meeting in about 4 minutes [15:00] reviewers -> #launchpad-meeting in 1m [15:00] bigjools, https://devpad.canonical.com/~beuno/LP_userdetail.png [15:00] barry: soyuz people will be late, sorry [15:00] Hi - I downloaded and installed LP on my Debian Lenny Server. I'm already using sbuild and schroot. Now I'm looking to integrate everything with builldd, but couldn't find any information. Anyone here, who can point me to some docs? [15:01] I saw there is an old sbuild package and some scripts in the lib/canonical/buildd directory, but it doesn't include a real buildd and the control file of the generated launchpad-buildd package doesn't depend on buildd, so I'm a little bit lost here. [15:01] bigjools: no worries. looks like it might be a short meeting anyway [15:05] http://people.canonical.com/~ed/dspr_mockup.png [15:05] https://dev.launchpad.net/SoyuzPackageUI?action=AttachFile&do=get&target=package_release_details.jpg [15:13] gary_poster: I have two machines reporting ImportError: No module named buildout.buildout [15:13] sinzui: doing what? [15:14] sinzui, gary_poster oh noes! i just hit the same error doing an rf-get [15:14] gary_poster: after rocketfuel-get and when I run make build in a pristine branch [15:15] sinzui: so (1) rocketfuel-get, (2) rocketfuel-branch? or (1) rocketfuel-get, (2) bzr branch trunk foo (3) cd foo (4) link-external-dependencies (5) make? [15:16] (the first route works for me) [15:16] jmux: documentation isn't good for lp-buildd, can you please send an email to lp-users ML asking for more info ? [15:16] gary_poster: for me, #1 [15:16] barry: bin/buildout is the workaround apparently, but I need to know why you guys are seeing this [15:16] i.e. just rocketfuel-get [15:16] uh. [15:16] gary_poster: i have a few minutes to debug this with you before i have to head out for a little bit [15:17] 2) is bzr branch rocketfuel pristine; cd pristine; make build [15:17] sinzui: you are not linking (my step 4)? [15:18] gary_poster: My script has always done that [15:18] barry: ok (I'm also trying to help mthaddon in another channel) [15:18] gary_poster: k [15:18] cprov: Ok - will do [15:18] sinzui: ok. I was just trying to understand what you were clarifying for me in your message. Were you simply identifying that you were doing the "raw" approach, not rocketfuel-branch? [15:19] jmux: cool, thanks. [15:19] gary_poster: yes, I have always linked [15:19] cool [15:21] ok barry [15:21] so in bin/buildout [15:21] gary_poster: I do see zc.buildout-1.4.0dev_gary_r102684-py2.4.egg/zc/buildout/buildout.py [15:21] what is the buildout egg linking to? [15:21] (in your sys.path) [15:22] sinzui: does make work now? i.e., is this a "do it twice and it works" sort of thing? [15:22] gary_poster: '/home/barry/projects/launchpad/eggs/zc.buildout-1.4.0dev_gary_r102684-py2.4.egg', [15:22] barry: and does that exist? [15:22] NO it does not, I did that before asking for help [15:23] henninge, danilos, give me a little bit to recover from an hour of soyuz, and I'll be right with you [15:23] gary_poster: yep [15:23] the cheek [15:23] beuno: we feel for you ... [15:24] barry: could you pastebin me the error, with just maybe 10 lines above it? [15:24] c'mon henninge, I've been looking through some translations code ;) [15:24] beuno: sure, I need to a minute as well [15:24] * bigjools is laughing a lot [15:24] https://pastebin.canonical.com/21051/ [15:25] gary_poster: ^^ [15:25] Hi, it's me again with a problem with make schema (I've been tasked to install a local launchpad on a Jaunty server version): http://paste.ubuntu.com/251949/ everything up to "make schema" is just fine and quits with no errors but make schema quits with this error. After installing Ubuntu server I installed the available updates and the latest kernel. [15:25] barry: gary_poster: That is the error I am getting [15:25] noodles775: and, weren't you amazed at the beauty and simplicity ??? [15:26] ;) [15:26] Knut-HB: I do not think you are the owner of your database [15:27] beuno, henninge: ok, I am ready whenever you are [15:28] gary_poster: I think I have test failures that may be due to using the system bzr - does your just-landed branch mean lp will always use a private bzr egg, or do I have to poke some buildout config somewhere to make it do that? [15:28] barry: does the egg look broken somehow? so, you can try ls /home/barry/projects/launchpad/eggs/zc.buildout-1.4.0dev_gary_r102684-py2.4.egg/zc/buildout and see if there is a buildout.py and an __init__.py [15:28] danilos, beuno: I am ready, too. [15:29] gary_poster: it looks fine! [15:29] maxb: the branch will make it use the private egg. You will need to merge and rerun bin/buildout (or do a make clean and make) [15:29] where the branch == now on devel, right? [15:30] sinzui: http://paste.ubuntu.com/251950/ i just got through the steps mentioned at https://dev.launchpad.net/Getting and https://dev.launchpad.net/Running [15:30] henninge: where's the latest mockup (or mockups) we want to discuss? [15:30] barry: try cd /home/barry/projects/launchpad/eggs/zc.buildout-1.4.0dev_gary_r102684-py2.4.egg and then start up python2.4 and try importing zc.buildout.buildout? [15:30] danilos, henninge, I'll be ready in 15ish [15:30] maxb: yes [15:30] beuno: ok, don't forget to wipe your a... uhm, whatever :) [15:30] excellent, I'll fire up another full testrun overnight and see what happens :-) [15:30] danilos: still https://devpad.canonical.com/~henninge/mockups/translate-page/list.html [15:31] danilos: just one [15:31] Knut-HB: try this to ensure you have the necessary permissions: `sudo -u postgres createuser -s -d $(id -un)` [15:31] gary_poster: happy as a clam: [15:31] @mission[[~/projects/launchpad/eggs/zc.buildout-1.4.0dev_gary_r102684-py2.4.egg:1012]]% /usr/bin/python [15:31] Python 2.6.2 (release26-maint, Apr 19 2009, 01:56:41) [15:31] [GCC 4.3.3] on linux2 [15:31] Type "help", "copyright", "credits" or "license" for more information. [15:31] >>> import zc.buildout [15:31] >>> import zc.buildout.buildout [15:31] /usr/lib/python2.6/dist-packages/Pyrex/Compiler/Errors.py:17: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6 [15:31] self.message = message [15:31] >>> [15:31] [15:31] sinzui: createuser: creation of new role failed: ERROR: role "launchpad" already exists [15:31] maxb cool :-) also, my zbuildout branch has all tests passing (though it uses some eggs I have not shared yet). I'm pushing some more of our changes upstream and then will be getting it all reviewed [15:32] Knut-HB: did you have pg installed before you started to setup launchpad? [15:32] sinzui: don't think so. as mentioned above: i just did what the two pages told me to do [15:32] barry: urgh. so why don't you remove shhh.py from before the problem line in the Makefile and see if it happens to give us more info when you run make again? [15:33] my current objective is to actually accomplish a clean test run using python 2.4 so I have a baseline to test branches on so I can submit them with a clear conscience :-) [15:33] gary_poster: sure thing. actually 'make SHHH= build' shoudl do the same thing [15:33] Knut-HB: okay, I think you would know if postgesql was installed [15:33] maxb: :-) +1 [15:33] barry: sure, cool [15:33] after a make clean. this crashes early [15:33] % make SHHH= build [15:33] scripts/update-bzr-version-info.sh [15:33] Creating bzr-version-info.py at revno 9098 [15:33] python2.4 bootstrap.py --ez_setup-source=ez_setup.py \ [15:34] --download-base=download-cache/dist --eggs=eggs [15:34] Creating directory '/home/barry/projects/launchpad/devel/bin'. [15:34] Creating directory '/home/barry/projects/launchpad/devel/parts'. [15:34] Generated script '/home/barry/projects/launchpad/devel/bin/buildout'. [15:34] ./bin/buildout configuration:instance_name=development -c buildout.cfg [15:34] Traceback (most recent call last): [15:34] File "./bin/buildout", line 17, in ? [15:34] import zc.buildout.buildout [15:34] ImportError: No module named buildout.buildout [15:34] make: *** [/home/barry/projects/launchpad/devel/bin/py] Error 1 [15:34] sinzui: you mean "postgresq"? [15:35] barry: will try a make clean locally 1 sec [15:35] beuno: I just updated that mockup, it unfolds plurals now, too. [15:35] danilos: ^ [15:36] sorry beuno [15:36] barry: works fine after make clean. will try a fresh eggs dir. [15:36] henninge: cool, let me take a look [15:37] Knut-HB: PostgreSQL is the relational database that Launchpad uses. rocketfuel-setup installs it if you do not already have it. Most db issues encountered after setting up launchpad relate to a previous installation of PostgreSQL, which you do not have. [15:37] henninge: one immediate suggestion -- icon should be on the top right, not after the entire text [15:37] henninge: a simple floater should do, I believe :) [15:37] barry: works fine. :-( ok... [15:37] danilos: hm, I used a table, not knowing what else to do, to get the radio button where it is. [15:37] sinzui: rocketfuel-setup did NOT install postgresql, i post-installed just a moment ago after you mentioned it [15:38] barry: echo PYTHONPATH? [15:38] * henninge tries the floater [15:38] Knut-HB: You make need to follow the instructions here: https://dev.launchpad.net/DatabaseSetup [15:38] gary_poster: % echo $PYTHONPATH [15:38] /home/barry/env/python [15:38] $PYTHONPATH I should say :-) [15:38] kfogel: ping [15:38] gary_poster: been that way for a bazillion years [15:38] sinzui: ok thx, i check that page out [15:38] mrevell: pong [15:38] henninge: right, implementation details is something we can figure out later, I am talking about green/red/whatever icon for longer texts [15:39] danilos: can I float a single or do I need to put in a
? [15:39] henninge: you can float an image [15:39] henninge: fwiw, img was the first one to allow floating in html :) [15:39] barry, let's hack bin/buildout. Please have it pretty print the pythonpath variable and sys.path, and let's see what they look like on a pastebin. [15:39] henninge: using , but you should use CSS, of course :) [15:39] danilos: oh right, I remember [15:40] danilos: but it was , wasn't it? [15:40] henninge: I think it was actually float, but I am not sure :) [15:41] henninge: guess it was align actually [15:41] barry: also, would be good to know if the bin/buildout workaround that allenap and gmb reported works for you. maybe start another branch and give it a try? or sinzui, have you already tried that? [15:41] gary_poster: I see sane building now [15:41] gary_poster: https://pastebin.canonical.com/21053/ [15:42] sinzui: you mean after the bin/buildout thing? [15:42] danilos: yeah, and the old vspace and hspace attributes ... [15:42] yes [15:42] what's the bin/buildout thing? [15:42] henninge: nothing beats bgcolor, but anyway... :) [15:42] barry: run bin/buildout; make build; [15:42] barry: Run that after linking [15:42] barry: allenap and gmb reported on the list that if you run bin/buildout this symptom magically goes away. yeah, what he said. :-) [15:44] gary_poster, sinzui bin/buildout does seem to be running just fine [15:44] barry: ok, maybe we are getting somewhere now. let's try clearing out PYTHONPATH in the Makefile so we don't pass anything to it [15:44] oops. i just ran bin/buildout :( [15:44] gary_poster: and make build is happy. [15:44] gary_poster: and i have to go afk for a little while :( [15:45] barry: well, at least you are unblocked. Try another rocketfuel-branch...ok [15:45] gary_poster: happy to continue debugging this with you when i return [15:45] sinzui: may I use you for debugging now? [15:45] if not I will take barry up on his kind offer [15:46] sinzui: I wanted to see if rocketfuel-get and rocketfuel-branch worked for you now [15:46] gary_poster: I have never used the later [15:46] sinzui: no need [15:47] sinzui: just want to know if what you did before magically works now [15:47] so you can recast my request as appropriate [15:47] I can run rocketfuel-get. I ran it on both my computers in the last 45 minutes [15:47] could you retry the thing that broke, then? [15:47] * sinzui runs rocketfuel-get [15:48] sinzui: http://paste.ubuntu.com/251967/ i did everything https://dev.launchpad.net/DatabaseSetup told me to do (both ways) and it still does not work [15:48] sinzui: oh...so, you mean, rocketfuel-get broke a while ago, but has since been working? [15:49] gary_poster: I mean I ran that, I got the buildout error because the script tries to make all the branches in the repo [15:49] OIC [15:51] Knut-HB: I do not know what else to try. stub will be online in 9 hours. He has the most experience with the database. [15:51] beuno: ok, it can't be that bad :) [15:54] gary_poster: rocketfuel-get completed successfully. [15:54] danilos, henninge, I'm ready [15:55] sinzui: ...ok. well, on the bright side, I guess that fixed it. On the not-so-bright side, I don't know why. I'll send out a note to the list to that effect. thank you. [15:55] beuno, henninge: https://devpad.canonical.com/~henninge/mockups/translate-page/list.html [16:01] sinzui: ./utilities/launchpad-database-setup root and sudo make schema and then make run: everything is ok, everything is fine and launchpad runs locally [16:02] Knut-HB: I will did not know about that script. I will remember it [16:02] sinzui: now the question is: why does everything work as sudo but not as user "launchpad"? [16:03] https://translations.edge.launchpad.net/ubuntu/jaunty/+source/gnome-games/+pots/gnome-games/es/+translate [16:03] danilos, henninge, ^ [16:03] Knut-HB: I do not think you own your database === matsubara is now known as matsubara-lunch [16:07] Knut-HB: sinzui: there's a problem with accounts called "launchpad", it clashes with a permission in security.cfg [16:07] I found this out the hard way [16:08] bigjools: you're not serious? Oo we tried to solve this problem in the last 24h Oo [16:09] very serious, unfortunately [16:09] I had to rename my local account, and fix the db permissions so it can be a PG superuser === kiko-phone is now known as kiko [16:10] there's a role in security.cfg called "launchpad" which will overwrite your superuser when you run "make schema" [16:10] bigjools: then there should be a warning at dev.launchpad.net that the user has not to be named "launchpad" [16:22] Knut-HB: yes, that would be a good idea [16:23] bigjools: or at least an entry in the FAQ [16:26] henninge, beuno: one big advantage of a two column layout: headings like 'English' and 'Translation' are not repeated with every message === deryck is now known as deryck[lunch] [16:27] danilos: yeah, that was one of my issues, too. [16:28] henninge, beuno: so, I'd still like us to experiment with both, I'll try to figure out how we can best do user testing [16:28] danilos, sure. I think it won't work, but knock yourself out! [16:29] beuno: I was quite skeptical at the beginning, but the more I think about it, the more I am in favour of it [16:30] danilos, that's the problem with UI. Thinking only gives you reasons to do it ;) [16:31] beuno: heh, right :) [16:31] danilos, I think that if the first impression is disconcerning, it usually means it won't work [16:33] beuno: yeah, it usually means that the learning curve is great :( [16:35] henninge: do keep the existing mockup as the two column layout one, we do want to be able to compare the two :) [16:46] danilos: ok === danilos is now known as danilo-bbl [16:46] danilo-bbl: hey, wait! [16:46] henninge: what? I am hungry... :) [16:47] chasing me through channels ... [16:47] danilo-bbl: how long will you be gone? [16:47] danilo-bbl: I'll probably leave in like 45 minutes [16:47] henninge: half an hour or so... I guess we can talk tomorrow then [16:48] danilo-bbl: ok, let's do that. Enjoy the food. [16:48] thanks === beuno is now known as beuno-lunch === salgado is now known as salgado-lunch [17:06] gary_poster: hi! [17:06] hi henninge :-) [17:07] gary_poster: It seems I have not yet used any other script than ec2test from lp-dev-utils. [17:07] gary_poster: at least disable-projects complains about launchpadlib missing. [17:08] me either, though apparently it has some scripts we are supposed to be using certain ones for CHR [17:08] henninge: try using bin/py from a launchpad trunk [17:08] gary_poster: oh, I thought you were doing those, too. [17:08] gary_poster: ok, I'll try that. [17:08] no :-) [17:08] ;-) [17:09] gary_poster: yup, works [17:09] cool [17:09] thank you === beuno-lunch is now known as beuno === danilo-bbl is now known as danilos === deryck[lunch] is now known as deryck === salgado-lunch is now known as salgado [18:03] speak to you tomorrow guy [18:03] s [18:15] * gmb EoDs [18:24] gary_poster: it's back :( [18:25] barry: good for me, bad for you. :-) OK, so back to what I was going to try last time [18:25] one sec, let me get a patch === EdwinGrubbs is now known as Edwin-afk [18:25] ;) k [18:27] barry: ok, so in this diff, try applying just the Makefile bit (you can do the rest too if you like but it is unrelated) : https://pastebin.canonical.com/21073/ [18:28] barry: my hypothesis is that the new ability I added to pay attention to PYTHONPATH suxors as an idea [18:28] * barry applies [18:29] barry: so I haven't actually tested if this masks the Makefile PYTHONPATH and your PYTHONPATH as I hope. so we'll have to verify (1) does this do what I hope and (2) does it help. not necessarily in that order. [18:29] #2 is definitely true! [18:30] woo hoo! [18:30] make it so, cap'n [18:31] barry: ok, then that's good enough. want to give me a review of that diff? I can switch to the review channel and then explain the other two bits (bug 389065) and then I'll push it [18:31] Bug #389065: system installed eggs overriding buildout provided packages [18:31] or maybe no explanation is needed with that bug :-) [18:33] barry: oops 1 sec [19:01] gary_poster: ping with an unrelated question [19:02] barry: pong with a cucumber [19:02] gary_poster: what about a pointed stick? [19:02] ow! [19:02] :) [19:03] gary_poster: so, what do you think the best way to determine whether the appserver has definitely started up? [19:03] gary_poster: best approach so far, is try to hit the root url for a little while and timeout if it doesn't happen [19:04] gary_poster: but that makes for a fragile-ish test, especially because what i really want to test for is that the appserver has /not/ started up [19:04] sinzui: I found an area where I'm duplicating queries used in menus [19:04] barry: hm, interesting. you could try following the log, assuming there is one to follow. oh...ok. testing negatives is hard. um. [19:04] gary_poster: yeah [19:04] sinzui: I want to be able to get non-link attributes from menus [19:05] gary_poster: check a pid file perhaps? [19:05] barry: could you verify that some error message shows up? [19:05] gary_poster: in a log file you mean? [19:05] barry: right [19:06] if it were possible that would convert it into a positive test, which would be nicer [19:06] yep [19:06] barry: test-appserver-layer.log? [19:07] hmm, okay, let me think about it. i was just looking for any thoughts that came immediately to mind [19:08] barry: there's a canonical.lazr.pidfile hanging around too [19:08] the log sounds most interesting to me, until/unless it proves unworkable [19:10] gary_poster: thanks! i might go with the pid file though since it's easy to check the file's existence and contents [20:11] beuno, what would you expect as the breadcrumbs for breadcrumbs for bugs.lp.net/ubuntu/+source/f-spot/+bug/34074? [20:11] Bug #34074: F-Spot does not warn user if the free space is low before importing photos [20:11] beuno, "Ubuntu > F-spot > F-spot bugs > Bug 34074"? [20:11] Bug #34074: F-Spot does not warn user if the free space is low before importing photos [20:12] salgado, YES. Perfect. === carlos_ is now known as carlos === Edwin is now known as Guest11851 === Guest11851 is now known as EdwinGrubbs [21:04] flacoste, bug 405327 [21:04] grr [21:04] https://bugs.edge.launchpad.net/malone/+bug/405327 [21:19] leonardr: ping [21:19] barry, hi [21:20] hi leonardr. i'm a little confused as to what's going on with lazr.restful version skew. it seems like the version we have in sourcecode is fairly different than the one in lp:lazr.restful. do you know what's up with that? [21:21] barry: yes [21:21] we cannot make launchpad use a recent version of lazr.restful until gary_poster makes some buildout-related changes [21:22] leonardr: gotcha. but the intent is that we're run lp:lazr.restful at some point (soonish)? [21:22] that is the intent, but i don't know when it'll happen. gary keeps having to do other stuff [21:22] yes [21:22] the branch tests passed two days ago! yay! so light is appearing at end of tunnel [21:23] hoorah [21:23] okay cool. i have a couple of convenience changes that would help with my branch's testing. i will work on getting them into lp:lazr.restful in the hopes that it will show up in launchpad soon [21:23] gary_poster: yay! [21:24] thanks guys [21:42] sinzui, re: bug 372925 "wooooooo" [21:42] Bug #372925: Change all pillar/person links to go to the overview page on http://launchpad not subdomain.launchpad [21:42] beuno: I am hacking on the registered info in the header now [21:42] I think I should try to do the footer too [21:43] sinzui: can I have a pre-impl call about menus? [21:43] sure [21:44] * sinzui gets head phones [21:44] thumper: I am reasy [21:45] sinzui: skype is trying to tell me you aren't on line [21:45] sinzui: can you call me? [21:55] leonardr: ping again ;) [21:58] bary, hi [21:59] leonardr: hi. zope.schema.Choice() can take a string or enum as its vocabulary argument. if it's a string (naming an email) lp works just fine, but lazr.restful cannot find the component to marshal. known bug? [22:06] leonardr: i found bug 364596 it's related but different [22:06] Bug #364596: Choice classes, and their values, should be published as resources [22:06] sinzui, I completely missed the fact that the new content I added to project-details was inside the
's div. I moved it out and it looks much better now [22:07] salgado: rock [22:07] land it [22:08] barry, i think i've encountered that bug but didn't know it worked for launchpad. can you give more details about what string you pass in? [22:09] leonardr: hang on, submitting a bug on this... [22:11] sinzui, not yet; first I'll have a hard time fixing tests [22:11] leonardr: bug 412726 see if that makes sense [22:11] Bug #412726: Choice()'s can take a string for their vocabulary, but lazr.restful can't handle this [22:13] barry, can you put up part of the stack trace that shows where the problem is within lazr.restful? [22:13] leonardr: sure, let me add that as a comment to the bug === salgado is now known as salgado-afk [22:19] leonardr: added [22:21] beuno: I have a concern about showing the registered information in the header. The design shows it next to the pillar or person. bugs and branches are under pillars, so where do we show their registered information? [22:22] sinzui, top right, on the row of the tabs, like in: https://devpad.canonical.com/~beuno/LP_new_design_Bugs_v3.1.png [22:24] beuno: okay, that design is far enough from the pillar...but there is still a line between it and the title of the bug. [22:24] beuno: should I JFDI and we will see how it works on edge? [22:25] sinzui, I don't understand what the problem is, but you've solved everything fantastically well up to now, so go for it [22:25] * sinzui does [22:26] beuno: I guess I am doing the footer too [22:30] flacoste: are we doing our call or skipping it? [22:33] rockstar: sync call? [22:35] thumper, yup. [22:39] spm: is there anything going on in production that would cause timeouts on https://code.launchpad.net/~launchpad/ [22:39] EdwinGrubbs, known issue, we're talking about it now. [22:40] rockstar: is it a known production issue or a problem with the page? [22:41] EdwinGrubbs, known production/page issue. Basically, we're adding all the source package branches. [22:42] in the last 37 minutes we have 200 new branches [22:42] of which 2 are project branches [22:42] rest are package branches [22:42] this is causing growing pains in some branch queries [22:44] sinzui, :) [22:46] EdwinGrubbs: the production problem is too many new branches and a badly performing query [22:50] um, wtf? bzr: ERROR: Public branch "bzr+ssh://bazaar.launchpad.net/~barry/launchpad/325962-makestart" lacks revision "barry@canonical.com-20090812214506-vocuvms42g9en1ow". [22:50] [22:52] yay for bzr push --overwrite === EdwinGrubbs is now known as Edwin-afk [23:23] Why do we have lp.answersbugs ? [23:24] rockstar: it should be lp.coop.answerbugs [23:25] thumper, excuse me. I meant "Why do we have lp.coop.answerbugs ?" [23:26] thumper, what I mean to say is, bugbranches and specbranches aren't in lp.coop - putting it where it is now made it hard to find. [23:26] rockstar: bugbranch and specbranches should be in lp.coop [23:26] thumper, okay. So maybe we need a bug for that. [23:27] ...or is there a bug already... [23:27] Maybe the bug needs to be moved to the launchpad-coop project? [23:28] wgrant, well, it at least needs affect the project where the code is currently. [23:29] The bug doesn't exist TTBOMK. [23:50] thumper: call me when you're ready [23:51] thumper: see what i was saying? :) [23:51] barry: :) === matsubara is now known as matsubara-afk