sinzuiwallyworld: do you want to try mumble?00:02
wallyworldsinzui: now i have no mic. i got my sound working eventually (seems a recent kernel update killed it for others too). i ended up having to upgrade to natty. but now it won't recognise my mic anymore :-(00:04
wallyworldso i'm doing the phonon/pulse dance :-(00:04
thumperwallyworld: is your mumble working00:05
wallyworldthumper: ^^^^^^00:05
thumperi silent wallyworld is a good wallyworld00:05
wallyworldsinzui: sorry, no work :-(00:12
* wgrant deletes shipit.00:14
wgrantYay, this even lets us delete a celebrity.00:14
huwshimiwgrant: As soon as I saw the announcement I knew you'd be happy :)00:27
thumper how is it coffee time again already00:29
wallyworldthumper: at least you have a working coffee machine :-(00:34
LPCIBotProject windmill build #144: STILL FAILING in 46 min: https://lpci.wedontsleep.org/job/windmill/144/00:39
lifelessgmb: did you find an OOPS for that bug 1 timeout ?00:59
_mup_Bug #1: Microsoft has a majority market share <iso-testing> <ubuntu> <Clubdistro:Confirmed> <Computer Science Ubuntu:Invalid by compscibuntu-bugs> <EasyPeasy Overview:Invalid by ramvi> <GNOME Screensaver:Won't Fix> <Ichthux:Invalid by raphink> <JAK LINUX:Invalid> <The Linux OS Project:In Progress> <metacity:In Progress> <OpenOffice:In Progress by lh-maviya> <Tabuntu:Invalid by tinarussell> <Tivion:Invalid by shakaran> <Tv-Player:New> <Ubuntu:I00:59
lifelessgmb: I'd be happy to eyeball it and see if I can suggest a cause00:59
lifelessthumper: stacked branches01:05
lifelessthumper: I see the code is going well01:05
thumperlifeless: yup01:05
lifelessthumper: did I correctly read that private http branch access isn't supported?01:06
thumperlifeless: landing the branch id support now01:06
thumperthat is right01:06
thumperwe don't support any private branches over http01:06
lifelessthats easy then01:06
lifelessI thought they worked in loggerhead01:06
thumperthat's different01:06
lifelessloggerhead backs onto http01:06
thumperit raises permission denied01:06
lifelessinside the dc01:06
thumperand loggerhead goes to https01:06
wgrantIt uses internal by-ID HTTP, right?01:07
thumperlifeless: different rule01:07
wgrantLike, raw by-ID, not +branch-id01:07
lifelesswgrant: its mapped to that01:07
lifelesswgrant: I'm not sure how it handles stacked-on urls01:07
lifelesswhich is why I'm asking these questions01:07
thumperit should be the same way01:08
* thumper thinks01:08
thumperI'll take a look01:08
wgrantIt must translatePath, right?01:08
lifelessthumper: until we change the virtual policy file01:08
lifelessthumper: this isn't going to magically affect people01:08
thumperlifeless: yes, I'm not landing that pipe yet01:08
lifelessthumper: so I'd check it on qastaging or whatever01:08
lifelessjust worth noting 'private stacked branches should work in loggerhead still' :)01:09
thumperyes they should01:09
thumperlanding incremental support is good though01:09
thumperwe can slam it on things like this01:09
wallyworld_thumper: i got my sound working again. if you are free at some stage, i'd like to mumble to discuss some code01:14
thumperwallyworld_: sure01:14
thumpernow is fine01:14
=== StevenK changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: StevenK | https://code.launchpad.net/launchpad-project/+activereviews
lifelessI just read '    def getBatches(self):' in lazr.batchnavigator01:27
mwhudsonthumper/lifeless: loggerhead still uses that nasty transform_fallback_location hook i added to bzrlib i think01:35
thumpermwhudson: ping01:43
mwhudsonthumper: hello01:43
thumpermwhudson: I was going to ask you about how loggerhead talked01:44
mwhudsonthumper: ah, i think i may have been lying above01:48
* mwhudson tries to remember if he wrote this code01:49
mwhudsonaaah right01:51
mwhudsonthumper: basically, it uses translatePath under the hood, it actually accesses branches in a pretty similar way to how codehosting does now01:52
thumperit seems likely then that my change to the vfs to support +branch-id alias will work fine there too01:52
=== Ursinha-bbl is now known as Ursinha-afk
lifelessI wonder if we should look at openid support for bzr01:53
lifeless / oauth01:53
mwhudsonthumper: yes01:53
poolierather than ssh keys?01:53
mwhudsonthumper: it seems like loggerhead should survive almost any rearrangement of how virtual paths map to real ones, basically01:54
mwhudson(this is not true of the external http service)01:54
thumperI've tweaked the rewrite map code to work with the new alias01:54
mwhudsonah ok01:54
poolieseriously, lplib started silently sending me to staging in natty? :(01:55
mwhudsonpoolie: i realized the other week that what i wanted for openid to be useful was my browser to know to send me to my chosen OP whenever openid was offered as an option01:55
lifelesspoolie: for https access to private branches01:55
pooliemwhudson, if it had been decently integrated into browsers it would have had a better chance01:56
poolieprobably chicken/egg01:56
lifelesspoolie: yeah, I saw that blog post01:56
mwhudsonpoolie: did anyone actually try that?  it sort of seems that the standard was at the wrong level for it to be feasible01:56
lifelessmwhudson: entirely wrong level01:56
lifelessmwhudson: I don't recall /any/ chatter on http-wg about it01:57
pooliewell, exactly01:57
wgrantpoolie: lplib has always defaulted to staging, AFAIK...01:58
mwhudsonit works nicely enough, although with a bunch of unnecessary hair i guess, for canonical/ubuntu SSO01:58
poolieeven if you specify a different root01:58
pooliei understand what it is; i'm filing a bug01:59
poolieadding new parameters to python methods other than at the end is not a good idea for api compatibility01:59
wgrantWhat's changed now?01:59
lifelessDistribution:+bugtarget-portlet-tags-content needs attention02:00
wgrantYeah. It almost looks cold-cachey, but it's hard to say.02:00
lifelesswgrant: it probably just needs the countBugs aggregate function applied to it02:01
lifelesswgrant: I bet most of the queries are table scans02:01
wgrantlifeless: IIRC it's a single 10s query in a lot of cases.02:01
wgrantI looked at it a while ago.02:01
lifelesslets have a looksee02:02
pooliewgrant, https://bugs.launchpad.net/launchpadlib/+bug/75209502:02
_mup_Bug #752095: Launchpad constructor parameters changed incompatibly in natty <launchpadlib :New> < https://launchpad.net/bugs/752095 >02:02
wgrant"Organize your workflow with Launchpad.02:03
wgrantFrom 37signals.com/accounts02:03
lifeless1 5211.0 202:03
wgrantDo people not Google these things :(02:03
lifelesswgrant: 2 x 5000ms queries02:03
wgrantpoolie: Hm, is the constructor really a public API?02:03
wgrantpoolie: I've never seen it recommended, I don't think.02:03
lifelesswgrant: also nonddy queries02:04
lifelesswgrant: totally identical02:04
pooliehttps://help.launchpad.net/API/launchpadlib recommends it02:04
poolieand, its name suggests it's public02:04
lifelesswgrant: and looking at every tag when in fact I think the portlet only shows official tags02:04
pooliefor fks sake02:04
pooliehow gratuituous, after all the pissing and moaning about compatibilyt02:05
lifelesswgrant: so, kill the duplicate call, constrain by the official tags, win.02:05
wgrantpoolie: Bah :/02:08
pooliei know api stability is hard in python02:11
pooliethis one is just a bit annoying because it doesn't just give a typeerror, it happily does nearly the right thing02:11
lifelesswgrant: ok, I've analyzed it up02:15
lifelesswgrant: we can make it a 300ms portlet02:15
_mup_Bug #736002: Distribution:+bugtarget-portlet-tags-content timeouts <timeout> <Launchpad itself:Triaged> < https://launchpad.net/bugs/736002 >02:15
LPCIBotYippie, build fixed!02:29
LPCIBotProject windmill build #145: FIXED in 1 hr 2 min: https://lpci.wedontsleep.org/job/windmill/145/02:29
=== _thumper_ is now known as thumper
wgrantlifeless: I'm trying to QA the bugattachment LFC preloading thing.02:34
lifelesswgrant: e-pic fail ?02:34
wgrantThe bug in question times out on qas, even without ?comments=all02:34
wgrantLooking at an OOPS, the query take 20ms.02:34
lifelesshit up ++profile++show02:34
wgrantBut there is a 4s gap afterwards.02:34
lifelesswgrant: does it timeout on prod without ?comments=all02:34
wgrantlifeless: No, it only takes 5s.02:35
wgrantGot a 502 from qas when asking for a profile.02:35
lifelesswgrant: you've marked it ok, was that optimistic ?02:37
wgrantErm, that must have been the wrong bug.02:38
wgrantlifeless: http://paste.ubuntu.com/590009/ only returns a few hundred rows on qas?02:39
lifelesswgrant: 75094902:43
lifeless[Bug 750949] [NEW] BugTask:+index times out with lots of attachments .. ** Tags added: qa-ok02:43
_mup_Bug #750949: BugTask:+index times out with lots of attachments <qa-ok> <timeout> <Launchpad itself:Fix Committed by wgrant> < https://launchpad.net/bugs/750949 >02:43
wgrantOh, right, bug number, not row count.02:43
lifeless** Tags removed: qa-needstesting02:43
lifeless342 rows02:44
wgrantSo 4s after that is completely insane.02:44
wgrantAlso, qas seems to be dead now.02:44
wgrantMaybe it's updating slowly.02:44
wgrantThat's slightly longer than the OOPses show, but still not too terrible.02:45
wgrantAh, now it's sometimes timing out before it even gets to attachments.02:51
wgrantSo I think it might just be generally unwell.02:51
wgrantRandom multi-second sleeps in there too.02:52
wgrantSo I'm going to re-OK this and blame it on asuka being shit.02:52
lifelesscheck the asuka load graph02:54
lifelesshuwshimi: btw02:54
lifelesshuwshimi: ajax log on qastaging is bokred02:54
huwshimilifeless: Oh02:54
wgrantlifeless: Ugh, any idea what's got it so high?02:55
wgrantSpiked massively an hour ago.02:55
wgrantspm: Hi.02:56
spmdamn. why does my /ignore not work anymore. :-(02:57
wgrantspm: :(02:57
wgrantspm: Something made asuka reasonably angry load- and RAM-wise right around midnight UTC.02:57
wgrantany idea what that is?02:57
wgrant(it is somewhat recovering now, but still not ideal)02:57
huwshimilifeless: Where abouts is it not working?02:58
wgrantIn fact it appears to now have flattened out at a heightened level.02:58
wgrantWhereas it normally drops back down to normally quickly after this sort of spike.02:59
spmwgrant: hrm. not offhand. tho the load in general for asuka is pretty spiky. I'd suggest a combo of a rollout of something and/or cronjobs firing off02:59
wgrantspm: Spiky and *11*.02:59
wgrantI tried to use capital digits there.02:59
wgrantSadly they don't exist.03:00
spmwgrant: heh, I see about 15 of those over the past week. :-/03:00
wgrantIndeed, but they're fairly irregular.03:00
wgrantI guess I'll see if it's less unhappy in a few hours.03:01
spmwe should probably fork qas onto a separate box03:01
wgrantOr just toss the scripts elsewhere and see what happens.03:01
wgrantI hear we have at least two free appserver boxes now ;)03:01
wgrantMaybe we can convince gandwana and potassium to serve their sentences on qas, too.03:03
wgrantSince I doubt anybody else wants them :)03:03
spmheh, I am not typically privy to the next hop destinations of servers. :-)03:04
lifelessspm: there is a ticket to move the scripts off03:05
spm /forcenickchange lifeless ticketmaster03:05
lifelesshuwshimi: huh, iz working now03:05
lifelessit wasn't working before03:05
wgrantlifeless: It just updated.03:05
wgrantlifeless: May have been icing skew?03:05
lifelessspm: I prefer KeyMaster, thanksverymuch03:06
spmsigh. delusions of grandeur.03:06
lifelesswgrant: and w/out ?03:11
wgrantSlightly faster than prod.03:12
wgrantBut not much.03:12
lifelessnie work03:13
mwhudsonhee hee, http://bazaar.launchpad.net/+branch/linaro-android-build-tools/files/head:/ works03:22
lifelessmwhudson: of course03:27
lifelessmwhudson: or is that unexpected somehow ?03:28
wgrantIt's pretty new.03:28
wgrantAnd probably an unintended side-effect.03:28
mwhudsonlifeless: it's new-ish at least03:29
thumperI'm trying to make a unit test that does the equivalent of: http://pastebin.ubuntu.com/590022/03:41
thumperanyone got any ideas?03:41
wgrantYou should be able to add a header to a testbrowser...03:42
wgrantUsing the remarkable browser.addHeader method.03:42
thumperright now it oopses03:42
thumperwhich is what i'm wanting to fix03:42
* lifeless hates on doctests03:42
StevenKlifeless: Then remove them?03:43
lifelessStevenK: its the README03:43
lifelessanyone see a reason for start=0 to be present on the first batch url ?03:44
lifelessmay be some test fallout... shurg03:48
lifelessI've already caused that03:48
wgrantTest fallout is dealable with.03:48
* lifeless loves matchers03:50
* wgrant unbreaks publish-distro.d03:50
wgrantStevenK: You has QA.03:55
wgrantWe need a publisher on qastaging.04:00
StevenKwgrant: The QA tagger should move to someone else04:03
wgrantStevenK: Oh?04:04
StevenKIndeed, it's moved to r12746.04:04
wgrantWaiting for my fixed publish-ftpmaster.py to run before I try PPA publishing.04:04
StevenKOn DF?04:05
wgrantI've turned on the run-parts things and made them work.04:05
StevenKThat could take eons04:05
wgrantSo we might get a full run.04:05
wgrantjtv deleted PublishingTunableLoop, so it should only take about 10 minutes to generate file lists now.04:05
wgrantIndeed, now into a-f itself.04:05
StevenKI wonder how much that will speed up the publisher on cocobanana04:05
StevenKwgrant: Do you need to land a branch of fixes for the run-parts fixes?04:06
wgrantYes :(04:06
wgrantStevenK: Not by more than a minute or so, I don't think.04:06
wgrantBut let me check.04:06
wgrantSeconds at most.04:08
wgrantSince wildcherry can keep the necessary indices in RAM.04:08
StevenKSo the massive speed up is for DF's benefit only04:08
wgrantI guess it'll be around a 15-20s improvement for natty on prod.04:09
wgrantie. nothing at all.04:09
wgrantgpg: signing failed: secret key not available04:11
wgrantlifeless: What's done? Batchnav?04:11
lifelessphase 104:12
wgrantNice work!04:12
wgrantDoes it work>?04:12
lifelesswho knows04:12
lifelessbzr+ssh://bazaar.launchpad.net/~lifeless/lazr.batchnavigator/keyoffset/ if you want to peek, mp coming up now04:16
wgrantpublish-ftpmaster.py's signing hook doesn't work.04:17
wgrantSince the GPGHandler utility overrides GNUPGHOME to a temp dir.04:17
lifelessdoes this break the deply?04:18
wgrantthis is the rev we didn't merge.04:18
wgrantAnd we weren't even planning to move to this new script immediately either.04:18
lifelessI can has review?04:20
lifelessStevenK: ^04:21
wgrant1400 lines.04:21
wgrantThat's almost into Soyuz territory.04:21
lifelessStevenK: if you whine that its too large, I'll whine that its using copied code from zope04:21
lifelesswgrant: I can halve it by deleting comments and tests.04:22
lifelesswgrant: Lines of code is a terrible complexity metric04:22
lifelesswgrant: particular because of context04:23
lifeless 12 files changed, 659 insertions(+), 209 deletions(-)04:23
lifelessis a better-but-still-not-great indicator04:23
StevenKlifeless: I think you linked the wrong bug04:24
lifelessStevenK: thats merge proposals being noddy04:24
lifelessStevenK: bug 72619504:25
_mup_Bug #726195: merge proposals show closed bugs <code-review> <Launchpad itself:Triaged> < https://launchpad.net/bugs/726195 >04:25
lifelesswhich is nontrivial04:25
lifelessbecause we should capture the bugs against the MP04:25
lifelessso the branch can get more links and not alter the MP04:26
lifelessthis would also make some queries faster04:26
StevenKlifeless: I'm not sure about your choice of 'memo' as a variable name -- especially since it's shown in URLs04:26
lifelessStevenK: cosmetic and easy to change if you have a better label04:27
lifelessStevenK: its going to be base64 junk04:27
lifelessStevenK: not hand guessable or enterable04:27
StevenKWell, I'm not sure what you're trying to show with it04:27
lifelessStevenK: its the db constraints needed to get back to a particular batch, serialised.04:28
lifelesse.g. for a bug search it might be ABBD23427834EfJ04:28
lifelessmeaning 'bug heat < 43 and bugtask.id > 34'04:28
StevenKlifeless: I'm ... I don't know, half-way? Breaking for lunch, so my stomach doesn't break out and insist.04:31
lifelessStevenK: thats cool04:31
lifelessStevenK: I'm going to start on a 'lets see if this breaks lp' branch of lp04:31
wgrantAha! It no longer crashes!04:32
wgrantWe may have victory.04:32
* wgrant carefully publishes a few suites.04:32
wgrantlifeless: https://code.launchpad.net/~wgrant/launchpad/bug-752149/+merge/5650004:39
wgrantOh, actually, even simpler.04:40
lifelesswgrant: tell me when04:41
lifelessI kindof wish we'd done an april fools thing04:45
lifelesslike skinning lp like sourceforge and announcing a merge04:45
wgrantEven Debian did one this time.04:46
lifelessI know04:46
wgrantlifeless: Diff updated.04:46
lifelesssadness, make schema day04:51
lifelesswe should make that incremental04:52
wgrantlifeless: You mean like, say, running upgrade.py?04:52
lifelessmmm, or make it obvious how to do it incrementally if one knows how04:52
lifelesswgrant: well its not always safe04:52
lifelesswgrant: but yes, upgrade.py on launchpad_ftest_template etc04:52
wgrantAlso, make schema isn't *that* slow any more.04:52
lifelesstell that to my wadl compiler04:52
lifelessI'm just wearing my curmudgeon hat today04:53
wgrantlifeless: Thanks.04:58
* wgrant glares at mawson.04:59
wgrantspiv: Hi.05:03
lifelessok, launchpad bugs needs to die05:10
lifelesscan we just delete that catchall ?05:10
lifelesslet folk use the shocking mute functionality ?05:10
wgrantWe should probably blackhole it for now.05:10
wgrantI was hoping the bug subscription feature work would think about how to fix not just bug notifications, but apparently not :/05:11
poolielifeless, i'm aware there's a "bugs i reported" function05:13
poolienm, i'll comment on the bug05:13
lifelesswgrant: well, by removing the thing, I think we can subscribe things more directly and appropriately05:13
wgrantlifeless: I guess we'll also find out what's going there pretty quickly.05:15
wgrantAnd work out how to fix it.05:15
Ursinha-afklifeless, hi, did you see any other occurrences of bug 727552?05:17
_mup_Bug #727552: when first rev is bad, with a later rollback, qatagger incorrectly reports it deployable atthe top <qa-tagger:In Progress by ursinha> < https://launchpad.net/bugs/727552 >05:17
Ursinha-afkI can't reproduce that05:17
spivwgrant: hi?05:22
wgrantspiv: Your Codehosting Twisted evil needs QA.05:23
StevenKEyes bleeding ...05:25
spivwgrant: ok.  What would that look like?  Would verifying that 'bzr push' of a new branch still works, (to staging/qastaging) suffice?05:25
lifelessGoogleBatchNavigator - facepalm05:26
StevenKlifeless: Burn it with Holy Fire?05:26
StevenKspiv: Well, how did you notice the slowness in the first place?05:26
lifelessUrsinha-afk: well I saw it right before I reported the bug, unless its been specifically fixed i suspect its still the case05:26
lifelessStevenK: jam noticed it locally in a vm05:27
lifelessspiv: pushing to qastaging is good - be sure to trigger some failures05:27
spivStevenK: jam noticed it.05:29
lifelessStevenK: can't burn it trivially, it wants a refactor of the base class, for now I'll just copy and paste and cry05:31
spivStevenK: I saw the profiling data he generated, applied my knowledge of which bits of Twisted are particularly crummy, jam did a crude hack confirming our hypothesis, I wrote a proper workaround (and proper fix for upstream, plus micro-benchmark for upstream).05:31
spivlifeless: practically everything we do triggers failures05:32
spivlifeless: that's the issue! ;)05:32
spivlifeless: jam noticed this issue from a simple initialize BzrDir RPC.05:32
lifelessspiv: I know, familiar with it05:32
lifelessspiv: I mean trigger something we expect to return an error to you, the user.05:32
lifelessspiv: because if its going to break, I'd expect that to be where it does.05:33
lifelesse.g. unknown project05:33
spivSo, 1) push up a new branch to somewhere that works, ~user/proj/branch, and 2) try a push that should fail, e.g. to ~user/no-such-project/foo ?05:33
lifelessor a project you don't have permissions to push into05:33
lifelessyes, exactly05:33
spivOk.  When will the code be on qastaging?05:34
wgrantIt is now.05:34
spivGreat!  I'll do that QA now.05:34
lifelessspiv: when the qatagger comments on  abug saying 'x landed', its reached a staging server.05:35
lifelessspiv: the specific one depending on the branch it landed on05:35
spivSounds handy, although I'm sure to have forgotten that by the next time I have another LP change to land and QA.05:35
StevenKdb-stable => staging, stable => qastaging05:35
spivOr just as likely, the process is likely to have changed ;)05:36
StevenKOur QA process hasn't changed for a few months.05:36
spivStevenK: yes, my point exactly05:36
lifelessspiv: I think we'll make it faster, but not change this aspect05:36
spivStevenK: just how recently do you think I've landed something? :)05:36
lifelessthe notify a human now aspect is very useful05:36
lifelessspiv: 6 weeks?05:37
spivlifeless: and it didn't pass QA :P05:37
spiv(I'm not complaining, I think you have a good process that's working well: the fact that wgrant was able to ping me about this in a timely fashion shows this process is coping adequately with infrequent contribuors like myself)05:39
StevenKlifeless: Pass the bleach, r=me05:39
lifelessStevenK: thanks!05:40
StevenKlifeless: With a little caveat that you can ignore if you wish05:40
wgrantspiv: lifeless and I keep track of our QA status pretty closely,.05:41
StevenKI tend to watch it, I just don't nag.05:41
spivStevenK: I don't need nagging, I just need my hand held because I don't have much familiarity with your processes any more.05:42
StevenKspiv: That was a dig at wgrant and lifeless, rather than you.05:42
spivwgrant: my testing worked.  Do I just change the tag to qa-ok now?05:48
wgrantspiv: Yup.05:49
spivwgrant: done.  Thanks!05:49
wgrantThank you!05:49
wgrantMuch better than me having to do it myself tonight :)05:50
lifelesswgrant: is there anything you need from me ?05:51
poolie_lifeless, hi, i was looking at outstanding bugs assigned to me05:51
poolie_what do you think i should do on https://code.launchpad.net/~mbp/launchpad/690021-rlimit/+merge/4373305:51
poolie_your last comment seems to suggest 'land it'05:51
poolie_i think landing it and seeing what happens is probably the best option05:52
poolie_or just abandoning it05:52
wgrantlifeless: Don't think so.05:52
lifelesspoolie_: I would like to see it landed05:55
poolie_as is?05:56
lifelesspoolie_: we know the branch that causes this (linux ;))05:56
lifelessso we can see if it loops as tim fears it may05:56
lifelessI would be ok with landing it as is, as long as when we deploy this rev we know to look for this potential problem05:56
lifelesswe can rollback just the import slaves *if* its a problem05:57
poolie_ok, that makes sense to me05:58
poolie_thanks; hooray for unsticking things05:59
wgrantThe lucid non-release pockets are big :(06:02
lifelessStevenK: https://code.launchpad.net/~lifeless/launchpad/bug-752153/+merge/5650506:03
StevenKlifeless: Oh, and it copies code due to your earlier comment06:06
=== almaisan-away is now known as al-maisan
StevenKlifeless: r=me06:07
lifelessStevenK: thanks06:08
lifelessnext step, migrate a real collection, like FTBFS, to use a custom IRangeFactory06:11
thumpersimple fix: https://code.launchpad.net/~thumper/launchpad/fix-dud-referer/+merge/5650806:21
thumperI was trying to fix bug 4491906:21
_mup_Bug #44919: UnicodeDecodeError while POSTing forms with non-ascii characters. <infrastructure> <lp-foundations> <oops> <Launchpad itself:Triaged> < https://launchpad.net/bugs/44919 >06:21
thumperbut the oopses I looked at weren't related to that failure06:22
thumperand I didn't notice06:22
lifelessI have a suggested tweak06:22
thumperso I filed bug 75221806:22
_mup_Bug #752218: Referer header with non-ascii oopses on not found page <oops> <Launchpad itself:In Progress by thumper> < https://launchpad.net/bugs/752218 >06:22
thumperlifeless: for me?06:22
StevenKthumper: r=me06:23
lifelessthumper: yeah, its on the revier06:23
thumperlifeless: the whole point of getting the referer is to show a link to the user to say "go back to where you came from"06:24
thumperif we replace the dud characters06:24
lifelessthumper: oh06:24
thumperit'll fail06:24
lifelessthumper: rightyho06:24
thumperand it's bogus anyway06:24
thumpernormally spam06:24
thumperfrom the oopses I looked at06:24
lifelessthumper: yeah. the underlying bug here - which you might like to put a comment about06:24
lifelessis that we shouldn't be decoding the referrer at all06:24
lifelessurls that we did not generate can legitimately be in /any/ encoding06:25
wgrantThe moral of the story: Python 2 sucks, let's use Python 3.06:25
lifelessbecause it makes this worse06:25
thumperthe problem is that the http request has a byte string06:25
thumperbut when we try to put it into the tales06:26
thumperit gets coerced into unicde06:26
thumperand dying06:26
lifelessit has to stay a byte string the whole way through06:26
lifelessI know thats way out of scope to fix06:26
thumperit will never though to include in the page06:26
lifelessand I'm ok with your branch06:26
lifelessI'm simply noting that if someone comes along and goes 'wtf we drop it' - it would be worth a mention at that point in the code06:26
thumpermost the templating engines these days are unicde based06:27
thumperlifeless: I'll add a comment06:27
* thumper wonders what happens to the 'o' in unicode06:27
lifelessits fine that they are unicode based, but they /render/ to something defined as bytes, not as ascii or utf8.06:27
thumpereventually yes06:27
thumperbut the unicode intermediate step causes pain06:27
lifelessoh, ffs, timeouts connecting to amazzzon again06:28
=== al-maisan is now known as almaisan-away
lifelesswgrant: that ftbfs collection timeout06:52
lifelesswgrant: has teh script been rewritten06:52
lifelessbug 730396 I think06:55
_mup_Bug #730396: DistroSeries:EntryResource:getBuildRecords timeouts <timeout> <Launchpad itself:Triaged> < https://launchpad.net/bugs/730396 >06:55
wgrantlifeless: Not yet.07:00
lifelesswgrant: 20ms batches of FTBFS07:11
lifelesswgrant: except for the last one, because the table structure is gnarly, as I keep whinging about07:11
lifelesswgrant: this is about 30 times faster07:12
wgrantlifeless: Nice!07:12
_mup_Bug #730396: DistroSeries:EntryResource:getBuildRecords timeouts <timeout> <Launchpad itself:Triaged> < https://launchpad.net/bugs/730396 >07:12
_mup_Bug #730396: DistroSeries:EntryResource:getBuildRecords timeouts <timeout> <Launchpad itself:Triaged> < https://launchpad.net/bugs/730396 >07:12
lifelesspoolie: I think its good to question the canned searches we have on bugs.l.n/+me07:17
poolieme too07:17
lifelesspoolie: I may have come across as pushing back on you earlier; i didn't intend to07:17
pooliei was having a kinda irritable day07:17
pooliei'm finding unity really stunning in concept and rather flaky as actual software07:17
poolieoh, also the lplib thing was annoying07:18
wgrantYes, it's really great to use, except that it crashes.07:18
wgrantAnd when it doesn't crash it glitches.07:18
poolieits failure mode of talking to staging instead of lpnet was a big timewaster07:18
lifelesswgrant: you need to ask the devs to take the if username=='' check out07:18
poolieanyhow, i agree about reconsidering the canned searches, but on the other hand this is probably not the most important change there07:19
poolieone thing i would like a lot is to make them show most-recent-first07:19
* wgrant headdesks.07:20
wgrantlifeless: Have you looked at BugTargetBugTagsView?07:20
wgrant        popular_tags = [tag['tag'] for tag in sorted(07:20
wgrant            other_tags, key=itemgetter('count'), reverse=True)[:10]]07:20
lifelesswgrant: can you please take that blunt spoon out of my eyeballs?07:21
wgrantI knew it was bad, but wow...07:22
* lifeless considers breaking list comprehensions in lp07:22
lifelesswgrant: there are 6K unique tags in lp07:22
lifelessin ubuntu07:22
wgrantIt is stupid.07:23
lifelesswgrant: criminal07:23
lifelesswgrant: it looks like *product* bug tag portlets may show every tag07:23
lifelesswgrant: check out the launchpad portlet for instance07:23
wgrantlifeless: I don't think so.07:24
lifelesshttps://bugs.launchpad.net/launchpad -07:24
wgrantlaunchpad-project's is a little odd, because it inherits the old disabled ones.07:24
lifelesslook at it.07:24
lifelessdo we really have ~100 official tags?07:24
lifelessincluding 'arm'07:25
lifelessand 'boobytrap'07:25
wgrantboobytrap is official.07:25
wgrantarm is official.07:25
lifelessso is arm07:25
wgrantboobytrap is one of Julian's.07:25
lifelesswell, apparently we do.07:25
lifelesswe should nuke about half of these.07:25
wgrantstory-better-bug-notification is unofficial, but probably in the top 10.07:25
lifelessoh, it shows the top 10.. yay07:26
lifelessdenormalisation time?07:26
lifelesspossibly a better query will be faster07:27
wgrantWell, I was going to try to improve the query, and if that didn't work just drop the top 10.07:27
lifelessadd a rank clause for starters07:27
wgrantOr drop the top 10 if there are official tags, or something like that.07:27
lifeless6K results back is pointless07:27
wgrantORDER BY COUNT(Bug) LIMIT 10?07:28
lifelesscount(tag); yeah07:28
wgrantEr, count(tag)?07:29
wgrantWhat good would that do?07:29
wgrantOh, count(BugTag), you mean?07:29
lifelessthe point is to only return the 10 most popular07:30
lifelesswgrant: one change I think would be uncontentious07:30
lifelesswgrant: include official tags in the top 10 calculation07:31
lifelesswgrant: so that if (say) 5 of the official tags are also top 10, one sees 15 tags total.07:31
wgrantrofl, for the first 8 months it showed the bottom 10 instead of top 10.07:31
lifelesserm count(official) + 507:31
lifelesswgrant: \o/07:31
lifelessyay testing.07:31
wgrant    def _getSearchURL(self, tag):07:34
wgrant        """Return the search URL for the tag."""07:34
wgrant        # Use path_only here to reduce the size of the rendered page.07:34
wgrant        return "+bugs?field.tag=%s" % urllib.quote(tag)07:34
wgrantI've not seen that reason used anywhere else.07:34
lifelessright, where was I before I knocked the battery out07:38
lifelessthanks, I hadn't noticed07:39
wgrantThat tag cloud is calculated after taking privacy into account.07:41
wgrantBut is cached publicly.07:41
lifelessthe queries I saw had no privacy clause, just private=False07:42
lifelessor was that cause they were anonymous requests?07:42
wgrantUser: Anonymous User07:42
lifelessI don't know that memcache saves us much here anyhow07:44
lifelesssame argument as usual: if misses suck, we're in trouble07:44
lifelessif misses don't suck, do we care07:44
wgrantIf it wasn't loaded in a separate request I think it would be worth preserving it.07:45
wgrantAh, getUsedBugTags* only has twoish callsites, fortunately.07:53
wgrantBecause both of those methods are broken and need a rewrite :(07:53
lifelesswgrant: you might like my countBugs aggregat ehelper07:55
=== mthaddon changed the topic of #launchpad-dev to: Launchpad down/read-only from 08:00-09:30 UTC for a code update | https://dev.launchpad.net/ | On call reviewer: StevenK | https://code.launchpad.net/launchpad-project/+activereviews
LPCIBotProject devel build #612: FAILURE in 5 hr 41 min: https://lpci.wedontsleep.org/job/devel/612/08:03
StevenKBaaaaaaaaaah, windmill08:03
StevenKWhen did you come back?08:03
wgrantIf it's failed again, it is going to be killed again in 5 minutes.08:04
wgrantAlthough I guess it's not actually testfixed us yet, so it can stay for a while.08:04
wgrantBut the moment it fails in buildbot I am killing it again.08:04
StevenKwgrant: Not even a second chance?08:07
wgrantIt has had three already :)08:07
StevenKThe fact that it failed in Jenkins and not Buildbot makes me sad.08:08
wgrantJenkins has always been better at exposing fragile tests :/08:08
StevenKTwo time out errors makes me suspcious08:09
* StevenK runs 'bzr rm lib/lp/soyuz/pas.py' to make him feel better.08:14
lifelessbug  740750 makes me very sad08:41
_mup_Bug #740750: API timeouts broken and returning no useful data... <regression> <Launchpad itself:Triaged> <lazr.restful:Triaged> < https://launchpad.net/bugs/740750 >08:41
lifelessanyhow, I need to talk to an api plumbing person08:44
adeuringgood morning08:54
lifelesswgrant: I might grab a little of your time tomorrow to bounce ideas about glueing $stuff together08:58
wgrantlifeless: Sure.09:00
wgrantOh, it's time.09:00
lifelesswgrant: on tags09:03
lifelesswgrant: is it constrained to open bugs only now ?09:03
wgrantlifeless: Yes, but it seems to fail to exclude dupes.09:03
lifelesss/open/only open/09:03
mrevellHello, good morning09:05
lifelessI wonder if there is a poor non-api collection I could improve09:05
lifelesswgrant: is that ftbfs collection also in the web ui?09:05
wgrantlifeless: Something very like it, yes.09:06
wgrantlifeless: https://launchpad.net/ubuntu/+builds uses Distribution.getBuildRecords.09:06
wgrant/~wgrant/+archive/ppa/+builds for Archive.getBuildRecords.09:06
=== StevenK changed the topic of #launchpad-dev to: Launchpad down/read-only from 08:00-09:30 UTC for a code update | https://dev.launchpad.net/ | On call reviewer: - | https://code.launchpad.net/launchpad-project/+activereviews
lifelesswould seem to match the behaviour09:08
mptIs there a graph anywhere of how much faster Launchpad has been getting?10:00
mptlifeless? ^10:00
lifelessmpt: its a bit noisy10:01
lifelessthis doesn't show the in-dc queueing problem which I discussed in my capacity email to the -dev list10:02
lifelessfor isntance10:02
wgrantAnd the in-DC queueing was a *big* problem.10:04
lifelesswe don't yet have metrics for in-dc queuing10:06
lifelesswe have a nagios check coming for one known symptom10:06
jkakarMaybe there's already a bug filed, but sometimes when Launchpad is in read-only mode I get two of the yellow notification boxes, when viewing a bug, that tell me Launchpad is in read-only mode.  It's seems to be consistent for particular bugs, I think, but there doesn't seem to be an otherwise obvious pattern.10:07
lifelessjkakar: I want to nuke readonly mode10:07
lifelessjkakar: its a source of longer downtime, ironically enough10:07
jkakarlifeless: That sounds like an excellent fix for the issue. :)10:07
stubjkakar: Are you using two browser windows by any chance?10:08
jkakarstub: Nope, one instance of Minefield (FF4) with several tabs open.10:09
stubjkakar: Several tabs open on Launchpad? Its a known issue that if you are using two browser windows to Launchpad the race conditions can cause notifications to pop up on the wrong browser10:10
jkakarstub: Ah, that could be it, yes.10:10
jkakarstub: It happened (sometimes) when I was viewing a milestone listing and Ctrl-clicked several bugs to open in new tabs.10:10
stubThat would do it. The key to 'which notifications should I display' is a cookie, and that same cookie is used for all browser windows of course.10:11
lifelessmpt: for instance, see translations 99th percentile goes from 8 down to 4 seconds10:14
lifelessmpt: on the year-wide graph10:14
lifelessmpt: the red line is the over-everything metric, which is very slow to move (because > 50% of renders are API, and API is already generally pretty good10:15
mptlifeless, thanks, ivanka just wanted an answer to the question "Why is Launchpad down for maintenance", and that graph was the most effective rationale I could think of :-)10:16
lifelessmpt: so the the problem is the technology10:17
wgrant... that covers most problems with Launchpad.10:17
lifelessmpt: schema changes are tricksy in postgresql w/replication10:17
lifelessmpt: both those links are technical, but I'm sure you can translate ;)10:19
mptin my copious spare time10:19
lifelesswe just gave you 90minutes of time ;)10:19
lifelessmpt: anyhow, short story, we've dropped nearly 2 seconds off the 99th percentile10:23
lifelessmpt: and 9 seconds off the hard timeout cap (which is set to maintain < 0.1% failure rate)10:23
wgrantAnd we'll hopefully drop it again tomorrow...10:23
lifelesshttps://bugs.launchpad.net/ubuntu 1.91 seconds10:26
wgrantAh, great. It was 2.8 when I tried before.10:27
wgrantMust have warmed up.10:27
lifeless2.8 is still better than 11 :)10:27
wgrantVery true.10:27
lifelesshttps://bugs.launchpad.net/ubuntu/natty was 210:27
lifelesshttps://bugs.launchpad.net/bzr 1.6810:29
lifelesshttps://bugs.launchpad.net/bazaar 1.8110:29
lifelessbzr 1.27 hot10:30
=== mthaddon changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: - | https://code.launchpad.net/launchpad-project/+activereviews
lifeless2.16 for launchpad-project10:35
lifeless1.58 hot10:35
lifelessbefore this change10:36
lifelessonly 20% of page loads for the ubuntu bugs-index were under 5 seconds10:36
lifeless50% under 810:36
lifelessthis is a massive win10:37
lifelessBugTask:+text needs a fixin10:41
lifelessBug:+text - 65K hits a day10:48
lifelessProjectGroupSet:+all is half fast and half slow10:51
lifelessI thought I fixed it :(10:52
lifelessoh, or maybe it wasn't deployed10:52
lifelesswgrant: I think Distribution:+bugtarget-portlet-tags-content needs fixing before we drop the timeout10:54
lifeless12s 99th percentile10:55
wgrantlifeless: Whitelist whitelist.10:57
lifelesswgrant: I think we should wait till monday;10:58
wgrantlifeless: Do you have IP addresses for BugTask:+text? I suspect Debian.10:58
jamwgrant: fixing is better than whitelisting :)10:58
lifelesswgrant: that gives us a complete day on thurs->friday10:58
wgrantlifeless: True.10:58
wgrantlifeless: Plus we'll have an ndt before Thursday.10:58
lifelessso, logs will tell us10:59
lifelessI don't have an IP handy10:59
jamlifeless: so I saw that spiv qa'd his twisted patch, how does that work in a deployment sense?10:59
jamSince codehosting is still downtime required10:59
wgrantlifeless: lucas imports Ubuntu bugs into UDD, and I think that still uses BugTask:+text10:59
jambut you are trying to only do a db deploy during this downtime10:59
=== danilos changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: danilos | https://code.launchpad.net/launchpad-project/+activereviews
lifelessjam: we either schedule downtime, or it gets rolled out next db deploy11:00
jamlifeless: so currently codehosting *does* get rolled out during db deploy11:00
jam(even if that particular patch missed this downtime)11:00
lifelessits still a full deploy11:01
lifelessbut I'm trying to get things detangled11:01
lifelessthere are complicating factors11:01
lifelessI think we need the PPR to go into 5ths of a second11:07
wgrantlifeless: Yeah :(11:10
=== almaisan-away is now known as al-maisan
henningervba: r=me ;-)11:50
rvbahenninge: \o/11:50
wgrantbigjools: Want to land your config branch now? I'll remove the schema keys tomorrow.11:51
rvbahenninge: thank you for your patience and precious advices.11:51
wgranthenninge: Thanks for encouraging the proper use of structured()!11:51
henningewgrant: welcome11:53
henningewgrant: I was surprised, though, that .escapedtext has to be used explicitely.11:53
wgranthenninge: Yes :/ That's a bit silly.11:54
henningeit is11:54
rvbahenninge: wgrant I've put the call to escapedtext in the template though11:54
henningeI saw that. I have no preference either way.11:54
wgrantIt belongs in the template.11:55
henningervba: well done ;)11:55
deryckMorning, all.12:01
=== henninge is now known as henninge-lunch
jkakarHiya jml!12:15
jmljkakar: hello12:16
rvbadanilos: Hi! Could you take a look at https://code.launchpad.net/~rvb/launchpad/dds-add-unique-packages/+merge/56548 for me?12:20
danilosrvba, hi, sure thing12:27
danilosrvba, is it intentional that this pages uses the same template as +localpackagediffs?12:35
rvbadanilos: yes12:36
danilosrvba, if so, you'd probably want to rename the template so it better indicates what is it about12:36
rvbadanilos: that's right.12:36
rvbadanilos: actually the template name is distroseries-localdifferences.pt. The 3 pages sharing this templates are all used to display local differences ... of different types.12:40
danilosrvba, ok, then I guess it's fine to keep the name12:40
danilosrvba, the branch looks pretty good otherwise, but I have another naming question: I believe it's our practice to name all our view classes as SomethingSomethingView, so it'd be nice to change that as well12:41
daniloss/as well//12:42
danilosrvba, other than that, r=me12:42
rvbadanilos: great, I'll look into the naming thing.12:42
danilosrvba, basically, just append "View" to the class name, and, seeing there's more of them like that (for the other pages), we might want to fix those (in a different branch), or discuss this on the mailing list :)12:44
* rvba is doing that right now.12:44
jmlmrevell: could I please skype you to test out my new webcam?12:57
mrevelljml, please go ahead, caller12:59
mrevelljml, I see you12:59
mrevellbut I don't hear you jml13:00
bigjoolsshakin that ...13:00
bigjoolswgrant: around?13:07
wgrantbigjools: Mostly.13:07
bigjoolswgrant: just wanted to run something by you quickly as it seems like I have a too-easy solution to something :)13:08
wgrantbigjools: Sure.13:08
=== matsubara-afk is now known as matsubara
bigjoolswgrant: when we open a new distroseries, one of the steps taken is to run the publisher in careful mode13:08
bigjoolswgrant: I am thinking that we can avoid that by creating all the publications as pending instead13:09
wgrantbigjools: That will take many hours to publish.13:09
bigjoolsthere must be a catch13:09
wgrantbigjools: It will hash a hundred gigabytes or so.13:09
wgrant== slow13:09
bigjoolswhy is careful publishing quicker13:09
wgrantHmm. Is it a full -C publish?13:10
wgrantOr just -A?13:10
wgrant-A is indices only.13:10
bigjoolsso it avoids the file checks13:10
wgrant-C implies -P, which tries to rewrite all of the files from the DB.13:10
bigjoolsso there's the catch13:11
wgrantExplicit pocket dirtying would be nice.13:11
wgrantCan has?13:11
bigjoolsthat would also force file publishing no?13:11
wgrantIt would just mark an extra suite as dirty, requesting normal domination and index generation.13:12
wgrant(it would also make the publishable safely killable)13:12
bigjoolsmight have to do this - we need a way of making new distroseries "just work" with no shell access13:14
wgrantYeah. It's pretty simple, too.13:14
wgrantNew (archive, distroseries, pocket) table, set by the init job, unset at the end of the publisher.13:15
bigjoolsbut.... pockets ... :/13:16
bigjoolssome distros don't have pockets, so it'd need to be (archive, suite)13:17
wgrantbigjools: Right, once we have explicit suites.13:17
wgrantBut until then suite == (distroseries, pocket)13:18
wgrantI hope soon we'll delete pockets.13:18
bigjoolsthat won't work for distros other than ubuntu13:18
wgrantBut that's not how it is now.13:18
wgrantSure, but neither will SPPH, BPPH, PackageUpload, PackageBuild, or just about anything.13:18
bigjoolssome distros want to keep updating the release "pocket" even after release13:18
wgrantThey will all need s/(distroseries, pocket)/suite/, and that can be done when we have a Suite table.13:19
wgrantAnd then pockets can GTFO LP.13:20
wgrantI'm not sure whether Suite wants to include Archive... I guess we'll see what's best.13:21
bigjoolsprobably not, there could be an archivesuite table later13:21
=== henninge-lunch is now known as henninge
bigjoolsbut... use cases first :)13:22
bigjoolsbug 17437513:22
_mup_Bug #174375: Distribution drivers permissions may need redesign <distributions> <lp-registry> <permissions> <ubuntu-platform> <Launchpad itself:Triaged> <ubuntu-community:Confirmed for techboard> < https://launchpad.net/bugs/174375 >13:22
wgrantI have quite deliberately avoided trying to analyse that :)13:23
=== Ursinha-afk is now known as Ursinha
LPCIBotYippie, build fixed!13:44
LPCIBotProject devel build #613: FIXED in 5 hr 41 min: https://lpci.wedontsleep.org/job/devel/613/13:44
=== dobey_ is now known as dobey
henningederyck: I'll be relocating quickly now. I may be 5 minutes late for the call13:49
deryckhenninge: np13:49
=== maxb_ is now known as maxb
* danilos -> away14:08
deryckJust FYI for everyone, I've asked abentley *not* to be on call reviewer today.14:18
deryckWe have some work to get done on Orange Squad and I need him today.14:18
deryckI won't list myself as on call reviewer, but if anyone needs a review feel free to ping me.14:19
jmlderyck: thanks for the heads up14:21
derycknp.  resending in email, too.14:21
bigjoolssinzui: how's the allergy meds going? :)14:36
sinzuiI am stoked to have survived the night after several choking bouts14:37
sinzuiI think I need to move north or to a desert for a month14:37
bigjoolssinzui: are you feeling up to talking about distro series permissions?14:37
abentleysinzui: you can come here, but it may not be north enough.14:38
bigjoolssinzui: I want to let distro drivers add new series.  The pollen in the nose however is the Ubuntu team14:38
sinzuiyes. That is a very large team14:38
* bigjools recalls the epic warthogs email 4 years ago14:39
abentleyderyck: chat?14:39
deryckabentley: sure.  mumble?14:39
sinzuibigjools: I believe IDerivativeDistro already supports this behaviour14:39
abentleyderyck: yes.14:39
bigjoolssinzui: ah... that old chestnut14:39
bigjoolssinzui: addseries needs lp.append on IDerivativeDistribution14:40
sinzuiI believe lp.append checks for admin, owner, or driver14:41
bigjoolshow does it decide that something is IDerivativeDistribution?14:41
sinzuithere was something about lp.driver that scared me when I tried it14:41
sinzuiI wish I could remember what the problem was14:41
bigjoolsah, I have found the hack14:43
bigjools        if self.name == 'ubuntu':14:43
bigjools            alsoProvides(self, IBaseDistribution)14:43
sinzuibigjools: the same kind of hack we have in Person to be an ITeam14:44
sinzuiI thought that hack was using full_functionality with checks the celebrity14:44
sinzui^ bigjools14:45
bigjoolsin this case I might not need to worry about this14:45
bigjoolssinzui: yes, it does I think14:45
sinzuibigjools: Do you still intended to introduce IRemixDistribution so that projects like balitx do not get builds and translations14:46
bigjoolssinzui: so everything is an IDerivativeDistribution unless it's Ubuntu?14:46
sinzuibigjools: That is it14:47
bigjoolssinzui: I wasn't planning on changing that at all14:47
bigjoolssoyuz is quite happy14:47
bigjoolswell, once you add some series and architectures14:48
bigjoolssinzui: See https://dogfood.launchpad.net/deribuntu/dangerous14:48
sinzuibigjools, ah understood14:48
sinzuibigjools: you may have an awesome blog post to do. You may be celebrated by  the 101 Ubuntu derivatives.14:50
bigjoolssinzui: well, possibly.  This stuff is being done for Linaro and OEM14:50
bigjoolsthe derivative management will only be available if you get explicitly allowed to use it14:51
abentleyderyck: http://people.canonical.com/~abentley/translations-usage.png14:52
sinzuiSo a driver of baltix could create a series that is based on natty. The series will be added with everything that automatically happens with a new series, but features exposed in the derivative UI will not be available?14:53
bigjoolssinzui: right - there's a separate initialisation phase that copies the packages from the parent14:54
bigjoolssinzui: BTW I just made myself a driver on a non-ubuntu distro and the addseries link disappeared14:54
bigjoolsbut it worked if I was a maintainer14:55
sinzuiI bet the link needs to be guarded by append. I recall the link was added back to the page in haste and admin/owner may have been used14:56
bigjoolsactually I should say, the link disappeared when took myself out of the maintainer positionb14:56
bigjools+addseries gives a 403 as well14:56
bigjoolsthis is odd then, I own it and I am a driver.14:58
bigjoolsbut I can't edit it14:58
sinzuibigjools: did rvba change permission after changes distro owner rules. Distro owner cannot edit  their distro because for Ubuntu, that was hundred of people. Only admins could change the distro15:01
sinzuibigjools: I am sure we had IDerivativeDistros permissions working.15:02
* sinzui looks for tests15:02
sinzuibigjools: lp/registry/browser/tests/distroseries-views.txt shows drivers creating series15:05
bigjoolssinzui: https://dogfood.launchpad.net/deribuntu15:05
bigjoolscheck out the driver15:05
bigjoolsI own that team15:06
bigjoolsI cannot get to +addseries15:07
sinzuibigjools: the test is wacky. I see check_permission('launchpad.Driver', view)15:09
bigjoolssinzui: ah!15:09
sinzuiWe expect launchpad.Append15:10
jcsackettdanilos: do you have time to review https://code.launchpad.net/~jcsackett/launchpad/hardwaredb-should-ignore-bad-data/+merge/5647215:11
jcsacketta big part of the diff is just a new sample data file being added and some tests being moved.15:11
sinzuibigjools: ModerateDistributionByDriversOrOwnersOrAdmins is now lp.Moderate15:12
bigjoolssinzui: one sec, OTP15:12
al-maisansinzui: OMG .. that name was almost Java-level verbose ;)15:13
sinzuibigjools: so drivers, owners, and a admins can create series because they have lp.Moderate, but the views and links check lp.Append. We broke this last August when we tried to remove exceptional permissions. We may need to revert this if we find that we cannot fix the lp.Append guards on the link and view15:16
sinzuial-maisan: security.py likes to live dangerously close to the 78 character length limit15:17
al-maisanso I see15:17
danilosjcsackett, sure thing, I can handle one more review today :)15:19
jcsackettdanilos: awesome. thanks. :-)15:19
=== danilos changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: - | https://code.launchpad.net/launchpad-project/+activereviews
danilosjcsackett, is it important for test data to be this big? can you not cause the same problem with a much smaller input file?15:22
jcsackettdanilos: possibly; i cribbed that file from the "good" sample data file. almost all elements in the rng appear to be required. :-/15:23
danilosjcsackett, ok, fair enough15:23
jmlducking out to get coffee15:23
danilosjcsackett, I won't be looking at the file in detail then, but I suppose at least long comment sections like the one containing XXX can be removed :)15:24
jcsackettdanilos: sure; i'll do what i can to clean it up.15:25
jcsackettdanilos: currently, it's literally the "good" file with a single bad dmi section.15:25
danilosjcsackett, that'd be much clearer if the file was potentially constructed on-the-fly: copy the good one, change one entry15:26
jcsackettdanilos: that's certainly a thought. i'll confess to laziness. i'll see about doing that.15:26
danilosjcsackett, iow, I don't like that there's a huge file without it being clear that there's one tiny little bit inside that _is_ important15:26
jcsackettdanilos: dig. i can certainly see the objection.15:27
danilosjcsackett, if you just feel too lazy, perhaps just add a big comment at the top indicating more of what we discussed here (though, I certainly prefer a better solution :))15:27
danilosanyway, off to the code now15:27
sinzuibigjools: This is my summary of the permission issue https://pastebin.canonical.com/45750/15:27
jcsackettdanilos: in order, i'll try: modifying the data on the fly; cleaning everything possible out of the file; adding a big comment.15:28
danilosjcsackett, excellent, thanks a bunch :)15:28
=== deryck is now known as deryck[lunch]
=== matsubara is now known as matsubara-lunch
danilosjcsackett, btw, is _logError used anywhere where you'd still want it to record an OOPS?15:38
jcsackettdanilos: it's used in several places. that's why by default it still creates the oops. to my knowledge, right now the one call site that's been modified to pass create_oops=False is the only place we want to supress it, but we may discover more.15:39
danilosjcsackett, ok, sounds good then, with a minimal test case for the context manager (so nobody breaks it in the future), r=me15:40
jcsackettdanilos: dig. i'll add that.15:41
danilosjcsackett, thanks! it seems a useful feature, worth emailing the list about I'd say :)15:41
jcsackettdanilos: cool. i'll do that once it lands. :-)15:42
bigjoolssinzui: right sorry, off TP now15:49
bigjoolssinzui: I'm not sure the analysis is entirely right since when I visit addseries I get a permission error, so it's not just a UI issue.  Or is this because I am tripping up on the view?15:50
bigjoolsooo a patch, I like those :)15:51
sinzuibigjools: no one has launchpad.Append, the permission was removed for distros...but Lp lets you still use it for links and views :(15:51
bigjoolssinzui: argh :/15:51
bigjoolsI'll hack mawson and see what happens15:52
sinzuiI see maps.ubuntu.com arrived 9 months too late15:59
bachi sinzui, i'm trying to isolate a windmill test failure following the instructions at https://dev.launchpad.net/Windmill16:04
bacsinzui: trying to run a single test with 'bin/test --test=test_bug_commenting' finds 0 tests.16:05
bachas it changed?16:05
sinzuibac: I do not know  you can run a single js test. the setup I wrote created a single test suite for all the files in the directory16:06
bacsinzui: the above is the example from the wiki.  perhaps it never worked?16:06
sinzuibac: deryck[lunch] did the last round of revisions.16:07
timrcEither there is some latency, http://maps.ubuntu.com is broken, or my IP does not translate correctly but I'm not seeing an Ubuntu badge over Austin, TX16:07
sinzuibac: the magic has to be in the test harness class. it may be that some harnesses were updated16:07
bacsinzui: ok.  thought i'd try you since you're here.  i'll wait for him to return from lunch16:07
bacsinzui: i'm reading through the thread from yesterday now to see if there are clues16:08
sinzuitimrc: i see a single badge over Austin in  http://maps.ubuntu.com/map/16:10
timrcsinzui: hm not I… but it may be someone else (as I just broadcast this :))16:10
timrcsinzui: other ip geolocating say they can't figure out where I am :/16:11
bacsinzui: why did you say the maps are 9 months late?16:11
sinzuibac: I mispoke. I was hoping Ubuntu would get a OSM map implementation. That is was I proposed. Instead it got a Google implementation16:12
bigjoolssinzui: your fix worked16:13
bacsinzui: ah16:13
sinzuibigjools: \o/16:13
timrcthey should accept postal codes as well ;)16:13
bigjoolsmaps.u.c hopelessly gets my location wrong too16:15
sinzuibac: I see that test_bug_commenting is a real windmill test16:15
bacsinzui: yes16:16
sinzuibac: the windmill layer is disabled at the the moment, or at leas when I checked it yesterday16:16
sinzuibac: you need to specify both the layer and the test to convince the testrunner to run it16:17
bacsinzui: really?  i got ec2 failures for windmill tests last night16:17
bacoh, i see16:17
sinzuibac: bin/test has a windmill and mailman exclusion rule16:18
sinzuibac: i tried to remove the mailman exclusion rule yesterday and failed. I saw windmill also had the same rule16:19
bacthanks sinzui, this works:   bin/test --layer=BugsWindmillLayer --test=test_bug_commenting16:20
* bac updates wiki16:20
sinzuibac: you can just use --layer=Windmill16:20
sinzuilayer matching is pretty generous16:20
baceven better16:21
* sinzui loves multi-touch love handles16:27
jcsackettsinzui: can i ask you to take a look at https://code.launchpad.net/~jcsackett/launchpad/set-question-message-visibility/+merge/56588? it's the retarget of the set-question-visibility work to devel instead of db-devel so i can get it out and working.16:44
sinzuijcsackett: I just looked at it16:44
sinzuiI have nothing to say. I can just approve it now16:45
jcsackettsinzui: that works for me. :-)16:45
=== matsubara-lunch is now known as matsubara
=== deryck[lunch] is now known as deryck
deryckwell that took forever17:21
derycksinzui, bac -- did you guys solve your testing questions I see in the scrollback?17:22
sinzuideryck: yes17:22
bacderyck: yes17:22
bacthe wiki was out-of-date...but no more17:22
sinzuideryck: ./bin/test excludes the windmill and mailman layer by default. They must be explicitly passed as an arg17:23
bacderyck: i do have another problem.  test_duplicate_finder is failing for me in my branch (where i did make a mod) but also in trunk17:23
derycksinzui: this should be fixed now with wallyworld_'s recent landing17:23
deryckbac: and you're up to date on trunk?  With the changes wallyworld_ landed?17:24
bacderyck: it fails waiting forElement on the dupe search field17:24
deryckbac: you on Firefox 4?17:25
bacderyck: nope, 3.6.1617:25
deryckhmmm, ok.17:25
bacare we to avoid upgrading to 4?17:26
deryckbac: for Windmill yes.  wallyworld_ and I are still trying to work out something going wrong there.  See his mail.17:26
deryckbac: as for the test, I can't run it yet.  Still futzing with my natty upgrade.17:27
bacderyck: no good deed...17:27
deryckbac: you could paste me the error, though.  I can look at that and see if anything stands out.17:27
bacderyck those windmill tests are passing now.  go figure.  i'll take my chances with ec2.  time for a super-quick review?17:50
deryckbac: sure, hit me!17:51
bacderyck: i'm still waiting on the MP to show up18:10
deryckbac: no worries.18:10
=== al-maisan is now known as almaisan-away
cody-somervilleWhy is the text area for further information on the filebug page so small? :(18:15
benjiso users of Chrome or Firefox 4 feel special when they resize it?18:22
bacderyck: https://code.launchpad.net/~bac/launchpad/bug-751397/+merge/5661718:24
deryckbac: thanks.  got it and looking at it now.18:24
bacderyck: i'm going to step out for a bit18:24
deryckbac: no worries.  I've got TL call top of the hour, so if I don't finish by then, I'll catch the rest after.18:24
jmlhey, can anyone spot a bug in this script? http://paste.ubuntu.com/590375/18:51
jmlmanual inspection says 20 critical bugs were closed in the last week, the script says 1718:52
benjijml: this modification http://pastebin.ubuntu.com/590387/ gives me this output http://pastebin.ubuntu.com/590388/19:11
jmlbenji: interesting19:12
benjiI'm trying to figure out what the deal is with the date_closed on those three.19:12
jmlbenji: why add 'not t.date_closed'?19:12
benjibecause the comparison fails otherwise (I'm in hack mode at the moment so don't hold me to that)19:13
jmlbenji: ok thanks. I'll explore some more :)19:13
benjialso note that the script will take quite a while longer to run this way19:14
benjijml: the three bugs that have date_closed == None are all status:Invalid19:18
jmlbenji: are they old, too?19:18
benjijml: you could say that: 2005 and 200619:19
jmlbenji: date_closed data isn't consistent in old bugs.19:19
jmlbenji: thanks for helping, very much appreciated.19:37
* jml is off for the evening.19:37
benjiglad to help19:37
=== fjlacoste is now known as flacoste
jcsackettsinzui: can you take over #launchpad?20:12
jcsackettalso: mumble, or is our two-man standup later enough?20:13
deryckbac: r=me on the js cleanup + addition tal condition branch.20:16
bacthanks deryck20:16
bachey deryck, what woes are you having with natty in vm?20:17
bacyou running vmware?20:17
baci haven't tried it yet...20:17
deryckbac: no, parallels and parallels tools won't build against natty kernel.  So worse than 2d Unity experience.  Really crappy desktop experience without a suitable video driver.20:17
lifelessflacoste: https://lpstats.canonical.com/graphs/PPR/20100407/20110407/ is looking nice now20:23
flacostelifeless: indeed20:28
sinzuisorry jcsackett. I have lost what little sense I had because of allergies and drugs20:31
lifelesssinzui: get well soon20:31
sinzuijcsackett: lets do our talk at 5:00 since it will only be U and I about at that time20:32
jcsackettsinzui: sounds good. and i hope your allergy season ends soon. i remember having terrible allergies when i lived up your way.20:32
sinzuilifeless: I think the plants should get a room if they want to get it on20:32
lifelesssinzui: amen20:32
lifelessjkakar: http://www.rfc-editor.org/rfc/rfc6202.txt may interest you20:35
lifelessflacoste: hi20:58
lifelessflacoste: we have /huge/ numbers of official bug tags for lp20:58
flacostelifeless: we do20:58
lifelessflacoste: I'd like to zerg through and zap a tonne; any objection in principle? if not I'll assemble a list and mail -dev for specific objections20:58
flacostelifeless: be my guest!20:59
lifelessflacoste: also, have you seen how fast https://bugs.launchpad.net/ubuntu is? woot20:59
flacosteindeed :-)21:01
lifelessflacoste: https://dev.launchpad.net/PolicyAndProcess/PreReleaseQAProcess looks stale too21:05
flacostelifeless: yes21:05
lifelessflacoste: or perhaps poorly named21:05
lifelessflacoste: carte blanche ?21:05
flacostelifeless: sure, i'll watch the edits and review21:07
jkakarlifeless: Ooh, thanks.  I've added it to my "to-read" list. :)21:39
=== Ursinha is now known as Ursinha-afk
=== salgado is now known as salgado-afk
=== matsubara is now known as matsubara-afk
lifelessyay deleting old wiki pages23:04
lifelessflacoste: I'm done on the doc sweep; mailed you about some pages I couldn't decide on23:09
lifelesssinzui: hi23:21
sinzuihi lifeless23:22
lifelesssinzui: heya23:24
lifelesssinzui: I wanted to touch base and confirm my understanding that jcsackett will be continuing with a button to hide spam in the UI ?23:24
sinzuilifeless: One of us will be doing that.23:24
lifelesssinzui: \o/23:24
pooliehi all23:25
sinzuiHe is still exposing some of api goodness needed to support it :(23:25
lifelesssinzui: I am thrilled to confirm that; for some reason I had started to doubt my memory23:25
lifelesshey poolie23:25

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