/srv/irclogs.ubuntu.com/2009/12/15/#launchpad-dev.txt

* jml hunches00:17
* jml brunches00:49
jmlyou know what01:16
jmlI think we should combine this channel and #launchpad01:16
maxbMhm. I'm not really convinced01:17
=== _thumper_ is now known as thumper
thumpermars: I don't suppose you are still around?02:04
thumperjml: what is your command to kill the mailman bit of make run_all?02:09
jmlthumper, it doesn't work any more :(02:10
thumper:(02:10
jmlthumper, ps x | grep mailmanctl; kill the appropriate process02:10
thumperspm: ping02:11
azopDoes anyone have any updates on the redmine + launchpad integration?02:14
azopBug #32438702:15
mupBug #324387: Support Redmine Bug Tracker for bug watches <bugwatch> <Launchpad Bugs:Triaged> <https://launchpad.net/bugs/324387>02:16
=== thumper is now known as _thumper_
=== _thumper_ is now known as thumper
* thumper thrashses local lp02:17
jmlthumper, why can't I specify the review type?02:24
=== ]reed[ is now known as [reed]
jmloh yuck. bugs.02:25
thumperjml: I don't know, why can't you?02:25
thumperjml: you have to specify it before choosing the person right now02:25
thumperkinda blows a bit02:25
jmlthumper, I'm doing a review using the new form layout that I did a review of for you recently02:26
thumperand you used the keyboard?02:26
thumperI found that this morning too :(02:26
jmlthumper, I can't hit Tab, select approve and then Tab to the review type02:26
thumperyep, bug02:26
jmlthumper, I have to actually click inside the comment field -- that's the only thing that seems to do it02:26
jmlthumper, I guess that explains the comment from Michael Nelson.02:26
thumperclicking on the review selector works too02:27
jmlabout enabling etc.02:27
thumperit seems to be on click not change02:27
thumperif you file the bug, I'll fix it02:27
thumperdeal?02:27
jmldeal02:27
jmlthumper, I'm going to file the bug as critical, since I think it's a serious regression.02:28
thumperjml: it isn't a regression02:28
thumperjml: same behaviour as before02:29
thumperyes it is a bug02:29
thumperbut it isn't a new bug02:29
jmlthumper, I'm pretty sure I've been able to hit tab before and have it work.02:29
* thumper tries prod02:29
thumperjml: it is very weird on prod02:31
jmlthumper, it's broken there too, I see.02:33
jmlthumper, I think the new behaviour is still worse02:33
jmlthumper, precisely because the order of UI widgets is better.02:33
thumperfile the bug, I'll fix it02:33
=== ]reed[ is now known as [reed]
jmlthumper, anyway, I'll contend that it's an rc bug.02:34
thumperI'm not sure our windmill tests can test that level of messing around02:35
jmlthumper, https://bugs.edge.launchpad.net/launchpad-code/+bug/49682002:35
mupBug #496820: "Review type" field is never enabled if you use keyboard only <code-review> <ui> <Launchpad Bazaar Integration:Triaged by thumper> <https://launchpad.net/bugs/496820>02:35
thumperbut I'll add the fix02:35
thumperta02:35
=== ]reed[ is now known as [reed]
thumperjml: on the plus side it looks like a one liner02:47
jmlthumper, and the down side?02:47
thumperit nees an rc02:47
thumperand I don't know how to test02:47
jmlmet02:49
jmlmeh, rather.02:49
jmlrelease it first, clean up the bodies later.02:49
thumperjml: confirmed the one line fix02:53
thumpernow...02:53
thumpera test02:53
adiroibanhi. does anyone know how to solve the "Non-english database locales are not supported with launchpad. " problem? Thanks!  http://paste.ubuntu.com/341610/03:10
* jml doesn't03:12
thumperhmm03:12
* thumper doesn't03:12
thumperit should be making the db with UTF-803:12
thumperstub: do you know anything about the above pastebin?03:14
wgrant_thumper: Charset != language03:14
=== wgrant_ is now known as wgrant
stubadiroiban: The postgresql database cluster needs to be in the C locale or your ordering will be locale specific and all the tests break. You need to rebuild that db cluster in the C locale.03:15
adiroibanI'm trying to install launchpad on a fresh Ubuntu 9.10 server. The strange thing is that it worked on my Ubuntu 9.10 Desktop03:15
stubIf you are not using PG for anything else (this will destroy your databases), pg_dropcluster --stop 8.3 main; pg_createcluster --locale=C 8.3 main03:17
adiroibanstub: thanks. I'm using 8.4 and setting a machine for running full tests03:18
adiroibanso I can destroy it... no problem :)03:18
stubadiroiban: The Launchpad test suite triggers bugs in PostgreSQL 8.4. 8.4.2 should be out 'real soon' with the fixes.03:19
adiroibanstub: thanks for the info. I'll go with 8.3 then03:20
adiroibanright now launchpad-database-dependencies for Karmic depends on 8.403:22
jmlhow do you reload a page in a page test?03:22
thumperbrowser.open03:26
thumperor03:26
thumperbrowser.reload()03:26
jmlthanks.03:30
jmlfwiw, I prefer IBranchCollection to the bug task search interface.03:30
thumper:)03:31
thumperme too03:31
jmlthumper, btw, love the new activity stuff on the code review page03:33
thumperjml: thanks, I love it too03:34
* jml afk03:41
* thumper EODs04:25
thumperwill check with danilos later for rc04:25
mwhudsonjml: hello04:35
mwhudsonjml: can you have a look at this patch and see if you think it's ok as far as it goes? http://pastebin.ubuntu.com/341643/04:35
jmlmwhudson, ok04:45
mwhudsonjml: any thoughts?04:57
jmlmwhudson, no, not really04:57
jmlmwhudson, I'm not sure what the scope of the patch is04:57
mwhudsonjml: i guess that's good04:57
mwhudsonjml: i just wanted to get a single simple test passing04:57
jmlmwhudson, ahh ok.04:57
jmlmwhudson, I think that's about right.04:58
mwhudsonjml: and then want to check that i'm not doing anything completely tasteless04:58
mwhudsonjml: thanks04:58
* mwhudson has felt like no-one else is in this particular box with him04:58
jmlmwhudson, yeah, I know what you mean05:01
mwhudsoni guess it's time to EOD05:19
=== lifeless_ is now known as lifeless
=== abentley1 is now known as abentley
adiroibancan I send merge proposals for LP using "bzr send" on the wiki there is only information about sending via web UI https://dev.launchpad.net/PatchSubmission08:12
jtvadiroiban: as I recall that was supposed to work, but we've gone through tons of changes since I last used it08:20
jtvadiroiban: IIRC the main thing I liked it for was the pretty-printed diffs, and nowadays the MP produces that anyway.08:21
adiroibanjtv: thanks. I'll try and if fails I will use the web ui08:21
BjornTadiroiban, jtv: using 'bzr send' works perfectly well. i think most people use --no-bundle, but i'm not sure whether it's necessary, or not.08:23
BjornTadiroiban: don't forget that you need to gpg sign the e-mail, though08:23
adiroibanBjornT: thanks for the tips. I didn't know about gpg requirement, but I can do it. that's not a problem :)08:24
maxbAnyone know off hand what python-pullparser is in lp-dev-deps for?09:15
maxbIt's removed from Debian and Lucid09:15
maxb("ROM; Unmaintained upstream")09:15
stubI would suspect parsing hardware database submissions09:23
stubPossibly for parsing remote bugtracker data09:24
maxbShouldn't it be in lp-deps rather than lp-dev-deps for that sort of thing?09:33
maxbgrep: zero hits09:35
maxbguess I'll try a testrun09:37
BjornTmaxb: i got some hits for pullparser. mechanize seems to use it. mechanize is used by zope.testbrowser, which is used by our page tests09:45
leonardri'm getting a segfault when i run the launchpad test suite. any ideas?09:48
leonardrno output, bin/test just segfaults09:48
leonardrbin/py runs fine09:49
wgrantleonardr: You were gone during the 2.5 migration, weren't you?09:49
leonardrwgrant: indeed09:50
wgrantleonardr: make clean; make09:50
leonardrare you sure that will help? this is a brand new branch09:50
* leonardr is trying it09:50
wgrantThe problem is in sourcecode/, IIRC.09:50
leonardrah09:50
wgrantI think the root Makefile has been taught to clean that too, but I guess you'll find out09:51
wgrantYeah, make clean should do it all.09:51
maxbBjornT: Oh, it's inside an egg, explains why I didn't see09:51
mat_tHi LP people09:56
mat_tWhen posting a comment I often get a popup message with an error:09:57
mat_tThe following errors were encountered:09:57
mat_t    * Object: , name: u'dead-ayatana'09:57
mat_tThe comment is actually getting posted, but I have to refresh the page to see it09:57
mat_tThis has been a case for quite a long time now... any ideas?09:58
intellectronicaargh, i get a segmentation fault trying to run windmill. has anyone seen something like this recently?10:39
intellectronicaBjornT: have you? ^^^10:39
bigjoolsmwhudson: hey how's the build job work coming along?10:39
mwhudsonbigjools: slow progress really10:40
bigjoolsis anything blocking you that I can help with?10:40
mwhudsonbigjools: not really, it's just boring typing at this stage :)10:40
bigjoolsyeah it's fun making new model classes :)10:41
mwhudsonit's also a bit scary in some ways, you have to put so much work in before you get to something useful10:42
mwhudsonbut well, getting through that now10:42
bigjoolsit's also a bit of a nightmare testing with our buildfarm10:42
mwhudsoni look forward to having that problem, in some ways :)10:44
mwhudsonanyway, time for bed10:44
bigjoolsg'night10:44
mwhudsongood night!10:45
stubintellectronica: I got a segfault earlier today but make clean and make build cleared it up. I couldn't see a cause - the only thing I'd changed was a .py file.10:48
stub(between the segfault and the previous successful invocation of that test)10:48
intellectronicastub: cool, let me try that10:53
=== jelmer_ is now known as jelmer
=== matsubara-afk is now known as matsubara
adiroibanhenninge jtv can you give me a quick hint about how can I import some templates in lp.dev ?11:40
adiroibanthe defaul import queue is empty11:41
henningeadiroiban: using bzr import or upload?11:41
henningeadiroiban: oh, I see what your problem is11:41
adiroibanhenninge: i don't know ...11:41
adiroibanI only want to have them there11:41
adiroibanI assume there is a script / cron-job11:41
adiroibanfor that11:41
jtvadiroiban: just a manual upload is probably easiest11:42
henningeadiroiban: if you just need something in the queue, a simple upload of a tar ball is easiest locally11:42
jtvadiroiban: did you upload already?11:42
jtvonce it's in the queue, you need to approve the entry/entries11:42
henningeadiroiban: you need to "make run_all" AFAIK. jtv?11:42
jtvyes, that's right11:42
adiroibanaha. I have uploaded a new POT file for a sourcepackage11:42
adiroibanbut I don't know where it went11:43
henningeadiroiban: so, uploads should appear in the queue immediately11:43
jtvcheck translations.launchpad.dev/+imports.  once the queue entry for the pot is approved, you run11:43
jtvrosetta-poimport.py11:43
jtvfrom the source tree;11:43
henningecronscripts/rosetta-poimport.py11:44
jtvright.  I normally keep my local lp running while I do that, though don't know right now if that's really needed11:44
henningejtv: me too11:44
henningeI mean, I leave it running11:45
henninge(not my car)11:45
henningeoh, I don't have a car atm11:45
henninge;-)11:45
adiroibanhenninge, jtv thanks. I should be good to go chacking that bug :)11:46
jtvI would ask what happened to the one you were going to get, but workrave insists I'm on a break now :-)11:46
adiroibanchasing11:46
jtvadiroiban: what bug?11:46
adiroibanjtv: bug 49689611:47
mupBug #496896: Permit distribution translation group owners to administer the import queue <ui> <Launchpad Translations:New> <https://launchpad.net/bugs/496896>11:47
jtvah cool11:47
adiroibanjtv: with the latest switch from launchpad.Admin to launchpad.TranslationsAdmin11:47
adiroibanwe need to enable them in the template11:47
adiroibanit should be trivial11:48
jtvnot sure that really should be "for distribution" and not "for Ubuntu"; we don't want just anyone to register a proprietary distro and start using Rosetta for it11:48
jtvadiroiban: "trivial" is a bit of a red-flag word for me.  heard it far too many times.  :)11:48
adiroiban:)11:48
BjornTintellectronica: what did segfault? the test runner, or the app server? (and no, i haven't seen that before)11:49
jtvlot of the circumstances have changed w bzr imports though11:49
jtvadiroiban, henninge: I'm going to try & have that typing break now :)11:49
adiroibanjtv: regarding new permission, danilo has review them11:49
jtvadiroiban: then I'm sure it's ok11:50
adiroibani hope so. basically he wrote that code :)11:50
intellectronicaBjornT: i think it's test test runner, but i'm not sure. 'Segmentation fault.' was the only output i've gotten after running bin/test. make clean and make build as per stub's advice did make it work again11:51
adiroibani just pushed for review some stupid changes and he had to fix them11:51
stubintellectronica: That was exactly the behaviour I saw.11:52
=== mrevell is now known as mrevell-lunch
adiroibanhenninge, jtv can you please help me set up a test where a need to add translations from 2 persons to a pofile. This is what I got now http://paste.ubuntu.com/341860/13:08
jtvadiroiban: we're just having a call here, hang on!13:08
adiroibanjtv: np13:08
=== mrevell-lunch is now known as mrevell
henningeadiroiban: in the first code you need to call makePOTMsgSet13:21
henningeadiroiban: and always use sequence=1 (or more if you have multiple)13:22
henningeas a parameter13:22
henningeadiroiban: ok, that is probably what you are doing in the second code, so nm.13:23
henningeadiroiban: no need to call setSequence, especially not twice13:23
adiroibanhenninge: but in factory.makeTranslationMessage, if msgset is None, it is created...just like I did ...but that way is not working13:24
henningeadiroiban: so, are you trying to create two translations for the same msgset, or do you just want to add two English strings with a translation each?13:25
adiroibanI just want to have 2 translators for es13:26
adiroibanfor a single pofile13:26
henningeadiroiban: so you don't care *what* they actually translated?13:27
adiroibannope13:27
adiroibani just want to have them in the list of contributors for that pofile13:27
henningeadiroiban: so the first code should be good if you add sequence=1 and sequence=2 to the two makeTranslationMessage calls.13:27
henningeand fix the "merged_translatort" typo13:28
adiroibanhenninge: no luck . This is what I got http://paste.ubuntu.com/341882/13:33
henningeadiroiban: this fails when trying to open a url. What's the url you are trying to open?13:35
adiroibanhttp://paste.ubuntu.com/341887/13:36
adiroibanhenninge: sorry. i found it.13:36
adiroibanthanks.13:37
=== jelmer_ is now known as jelmer
sinzuisalgado: gary_poster: Is it possible to make TestBrowser to accept a 410 HTTP status code so that I can test what the browser sees?14:13
salgadosinzui, you can try setting raiseHttpErrors to False in your browser instance, but I'm not sure it will actually work14:19
sinzuisalgado: I think that is exactly what I want. thanks14:20
=== Chex_ is now known as ChEx
=== ChEx is now known as Chex
gary_posterMTecknology: you saw stuart metcalfe's reply on bug 496360?14:26
mupBug #496360: Can no longer log in with OpenID modules <Canonical SSO provider:Fix Released> <Drupal OpenID module:Invalid> <Launchpad itself:Invalid> <https://launchpad.net/bugs/496360>14:26
gary_posterleonardr: skype?14:32
leonardrgary: argh14:39
leonardrsorry, irc was on the other computer14:39
leonardri'm on skype now14:39
=== mbarnett` is now known as mbarnett
beunointellectronica, did you send out the noted from yesterdays' call?14:50
intellectronicabeuno: sorry, not yet. will do that now14:50
beunointellectronica, thanks, I'm trying to remember what I promised I would do  :)14:51
danilossinzui, hi, do you have a few minutes to discuss https://code.edge.launchpad.net/~sinzui/launchpad/spam-eggs-bug-495250/+merge/16174?15:05
sinzuisure, since the tests fail15:05
sinzuidanilos: ^15:12
danilossinzui, right, so, even though I'd really like to see this in asap, I think it's something worth punting for the January release so we can make sure we get it nicely polished without introducing risks into this release15:13
sinzuipolished?15:13
sinzuiThe can be CPed later this week or next15:14
danilossinzui, well, QAd properly (i.e. more eyeballs on it on edge for a while etc.)15:14
sinzuidanilos: actually. I am certain this branch will be CPed since we will not have a full staff watching LP at the end of the Month. I Can land this today, or at the end of the week.15:16
* sinzui hates this branch too much to care15:16
danilossinzui, well, can you ensure at least some QA happens? can you QA it on staging or dogfood (by cowboying it in)? otherwise, I'd rather wait for a re-roll or get it CPed later15:17
sinzuiYes, wait for a reroll. Yesterday was too stressful. I am not willing to hurt myself for scope creep15:18
danilossinzui, ok, sure15:20
sinzui\o/ and I have a legitimate way to test 410 in admin tests! I might be able to do the work I was supposed to do on Thursday, Friday, and Monday15:23
danilossinzui, heh, what's the trick there? can you get headers in a browser object?15:24
sinzuiWell there are two tricks for two conditions15:24
sinzuiTo view pages that raise a 410:15:24
sinzui    >>> admin_browser.handleErrors = True15:24
sinzui    >>> admin_browser.raiseHttpErrors = False15:24
danilossinzui, ah, nice15:25
MTecknologygary_poster: ya, thanks for everything :D15:38
gary_poster:-)15:39
=== danilos changed the topic of #launchpad-dev to: Launchpad Development Channel | Week 4 of 3.1.12 | PQM is closed; RC only (https://dev.launchpad.net/CurrentRolloutBlockers, RM: danilos) | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes
=== matsubara is now known as matsubara-lunch
henningesinzui: ah. lights are back on here.15:43
henningesinzui: this is the test15:43
henningehttp://paste.ubuntu.com/341957/15:43
henningesinzui: yes, it's pushed15:43
henningesinzui: ~henninge/launchpad/bug-495126-deactivate-users15:44
MTecknologygary_poster: so - do you know when that change did happen?15:46
henningesinzui: wait, I just saw an error in the test!15:47
=== barry` is now known as barry_
=== deryck is now known as deryck[lunch]
=== barry_ is now known as barry
=== jelmer_ is now known as jelmer
=== ursula is now known as Ursinha-lunch
mwhudson_morning19:22
=== mwhudson_ is now known as mwhudson
MTecknologyCan I get a static link to a launchpad team image? Instead of say https://launchpadlibrarian.net/19775148/g2529.png I have https://launchpad.net/~ubuntu-southdakota/+image/large20:18
=== salgado is now known as salgado-afk
beunoMTecknology, there os no such URL20:18
beunobut I think it would be useful20:18
beunonot sure how much work it would be20:18
MTecknologybeuno: I could really really use something like that about now :P20:19
MTecknologybeuno: for the LoCo Map20:20
MTecknologyI'll file a bug when I get back in ~10min - see how hard it would be20:20
MTecknologyTHanks20:20
maxbSuch an URL would only be useful for pulling the images externally without consulting the API, though20:23
=== matsubara is now known as matsubara-afk
lifelessMTecknology: short answer is no, it would allow xss attacks20:31
lifelessunless its off in a different domain anyway, so you'd still not be able to just use the team url20:32
MTecknologylifeless: how so?20:38
MTecknologymaxb: can I use the api to do that?20:39
maxbXSS attacks? Even though it's presumably forced to be an image?20:41
MTecknologymaxb: you mean by uploading a harmful image?20:41
MTecknologywouldn't that still be vulnerable in the same way?20:42
maxbCan an image be harmful?20:42
MTecknologyya..20:43
MTecknologyI was just lookin in the apidoc and I don't see an image for teams - nor do I really know how to use the api anyway20:44
MTecknologynvm - mugshot_link20:44
MTecknologyNow I need to learn how to actually use the api to make use of this...20:44
lifelessmaxb: there have been many bugs found in image libraries20:51
lifelessmaxb: so yes20:51
MTecknologyUnknown consumer (None).20:52
MTecknologyI can't use it yet, can I?20:52
wgrantlifeless: Lots of files are served from URLs like that -- they just redirect to a librarian URL immediately.20:52
lifelesswgrant: gotcha20:53
MTecknologyCan I use the API in this way or is it not open for me to use?20:55
wgrantYou cannot yet easily use it anonymously.20:56
MTecknologywgrant: Is there any way I can become not so anonymous?20:57
wgrantMTecknology: https://help.launchpad.net/API21:06
MTecknologywgrant: I was looking at that21:08
MTecknologyit's somewhat outside of my comprehension :P21:08
MTecknologyok - I'm starting to get it.. this will need to be PHP in the end though... Thanks21:16
wgrantEw ew ew why?21:17
MTecknologywgrant: I'll try to use python21:18
MTecknologyThe only thing this is even doing is grabbing the image url :P21:19
wgrantMTecknology: Retrieve a URL like https://launchpad.net/api/beta/~motuscience/mugshot_link21:20
wgrantYou  should be able to do that unauthenticated.21:20
MTecknologywgrant: except for the error it's giving me..21:21
MTecknologyXML Parsing Error: not well-formed21:21
wgrantMTecknology: That's because it not an XHTML document, but a raw URL.21:22
wgrantI don't know why the mimetype is wrong, but you should ignore that.21:22
MTecknologyhttps://launchpad.net/api/beta/~ubuntu-southdakota/mugshot21:22
MTecknologythere :)21:22
wgrantAh, that redirects. Nice.21:23
MTecknologythat's exactly what I want :D21:23
MTecknologywgrant: thanks :D21:25
marsDown to running the final spam names filter.  It puts a joyful, evil grin on my face >:D21:25
MTecknologymars: kill the annoying spammers?21:28
marsMTecknology, that is for the losas to do.  I'm just pointing out where the bodies are buried.21:29
marsmaybe 1500 of them21:29
wgrantOuch.21:29
marsyay for automation21:30
MTecknologymars: you should automate a tool for them to destroy them too21:30
MTecknologyie - box pops up; they glance and verify; destroy; tool moves on21:31
MTecknology1yr later all 1.5k are gone and they have a new batch21:31
marsMTecknology, yeah, but they know the proper procedures, and I can just hand a confident list over to them.21:31
marsand there are ways to do graduated deactivation in such a way that wrongfully nixed accounts have recourse21:32
MTecknologymars: you lost me there - "graduate deactivation"21:33
marsMTecknology, erm, "graduated deactivation": deactivate, send an email stating the reason, "Go here to reactivate if we're wrong", after X days shut it down.21:33
MTecknologyoh21:34
mars"christmas-trees-for-sale-72"  Huh?21:35
MTecknologymars: don't delete me; I'm legit21:36
MTecknologyregardless of what google says21:36
marsedge.launchpad.net/~christmas-trees-for-sale7221:36
MTecknologymars: HTTP/1.1 302 Moved Temporarily21:38
MTecknology goes to google.com21:38
marsnow how the heck did that get in there.  Was it part of the same operation that is generating the pharma-spam?  Or some independent job?21:40
jmlgood morning Launchpadders22:26
mwhudsonjml: good morning22:27
jmlmwhudson, hi22:28
jml*groan*22:43
jmlhmm23:31
jmlnetsplits suck23:31
lifelessits a DDOS23:40
jmlyeah23:41
jmldoesn't make it suck less.23:41
=== mpt_ is now known as mpt

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!