/srv/irclogs.ubuntu.com/2012/10/19/#ubuntuone.txt

=== mmcc_ is now known as mmcc
karniGot lots done. Now I should get some sleep.05:07
karniNight all!05:07
mandelgood morning from london!08:39
czajkowskimandel: aloha!08:41
mandelczajkowski, hello!08:42
mandelczajkowski, hey, will you be in copenhaguen next week?08:42
czajkowskimandel: you there for the day ?08:42
czajkowskimandel: nope not next week, just the UDS part08:42
mandelczajkowski, all the weekend :)08:42
mandelczajkowski, oh, just opposite to me08:42
mandelczajkowski, I cant make it to uds because I had a tip booked for my holidays08:42
czajkowskimandel: awww :(08:47
mandelczajkowski, well, it was left to the very end...08:49
JamesTaitHappy Day-after-release Day, everyone! :-D08:56
mandelczajkowski, meh, sorry compiz just crashed..09:14
czajkowskimandel: no worries :)09:16
czajkowskifile a bug :p09:16
mandelczajkowski, is more due to working with unity un-stable09:17
czajkowskimandel: you're on 13.04 already :o09:17
mandelczajkowski, even more scary, unity trunk!09:18
* mandel reboots due to updates..09:20
=== yofel_ is now known as yofel
mandelrunning the unity staging ppa is a bad idea..09:34
mandelmeh, had to re-install the entire thing09:34
mandelkarni, ping10:56
karnimandel: pong, hi11:02
gatoxgood morning!11:02
karniMorning!11:03
karniHi gatox11:03
gatoxkarni, hi o/11:03
mandelkarni, gatox, can you pm you u1 email to me11:04
mandelto send you a share with books11:05
mandelmvo, I already added you, not all the books are there yet... is uploading slowly11:05
karniSure11:05
gatoxmandel, awesome, thx!11:05
karni:)11:06
mandelkarni, gatox, mvo, their is a fair amount of garbage there.. I need to make a list of must read/ignore11:15
gatoxmandel, yep..... i'm not sure i'm going to be able to read all of that before the sprint jeje11:17
mandelgatox, well, the qt ones I'm sure you don't need, there is a c++11 one which is really good to read11:20
gatoxmandel, i'll start with that one11:20
mandeland there beginners one, with those two people should be fine11:20
gatox:D11:20
mandelgatox, is not yet uploaded, or so I think11:20
gatoxmandel, ah ok..... yes, the qt ones i already read them11:20
mandelgatox, exactly, is just the entire collection I had11:21
mandelgatox, karni I do recommend to read the Bjarne_Stroustrup one just because of him :P11:30
gatoxmandel, yep..... i read that when i enter my other job :P11:31
mandelgatox, karni, then go for cpp11Notes2012-01-06 which explains the diffs in the new C++11 iso11:32
gatoxmandel, will do11:32
karnimandel: Sounds good, thank you :)11:33
gatoxmandel, do you know how to learn c++ in 21 days? http://3.bp.blogspot.com/_va-p5aIxayA/S5VN50k8fOI/AAAAAAAADOg/dKaw4NSvLXs/s1600-h/aprender-programar-21-dias.png11:47
ralsinagood morning!11:59
gatoxralsina, hi11:59
dobeyholas12:06
ralsinahi dobey! Figured out the linking issue :-)12:07
ralsinadobey: failed partial update to quantal, and the binary linked two libs, one that linked to each version of libpulseaudio12:07
dobeyhow would a failed partial update cause that? do qt apps all do a weird relink on install in the packages thing?12:08
ryemandel: have you found anything weird in the logs I gave yesterday?12:09
karnimorning ralsina o/ hi dobey12:10
ralsinadobey: apparently not *all* qt apps, but some12:10
ralsinahi karni!12:10
karni\o12:11
ralsinadobey: I had never seen ldd report linking to two sonames of a lib before12:11
dobeyralsina: that is, to say the least, quite insane.12:11
ralsinadobey: it's a first for me too12:11
dobeyralsina: i have, but only when gcc actually linked both at compile time12:11
ralsinadobey: but I fixed it and the binary is still the same12:12
dobeyralsina: had seen it a lot in the past for some apps when for example gtk+ or some gnome libraries broke ABI and the soname changed, and some deps were linked to the other version than the app itself used.12:12
ralsinain any case, this Q update is annoying me brutally12:13
dobeylol12:13
dobeyhttp://pages.cs.wisc.edu/~kovar/hall.html12:13
karnidobey: hahahah12:14
karnidobey: anyways, how did you find that? reddit/digg?12:19
karniI doubt you were googling germanium :)12:19
karniI'd like to see more of such pages ;D12:19
dobeykarni: mentalguy rt'd it on twitter12:19
karnidobey: Ah, kewl :)12:20
dobeyhttps://twitter.com/hirodusk/status/25908156558435942512:20
ralsinadobey: that undergrad needs to be introduced to sticky tape.12:20
ralsinabecause yes, welding cables to arbitrary things usually doesn't work12:21
karniI'll pick up my flat mate with his lcd screen from the city center, and might grab lunch on the way. Back in a while!12:25
ralsina0 upgraded, 0 newly installed, 0 to remove and 307 not upgraded. 1165 not fully installed or removed.12:28
ralsinasigh12:28
dobeyi did have an idea for pyflakes earlier today though; will need a little help from someone with a mac and win buildout set up12:29
ralsinadobey: you can do a linux buildout too :)12:29
dobeyralsina: i need platform-specific data though :)12:30
dobeyi can just run python in linux to get what i need there :)12:30
dobeyi need the result of this:12:32
dobey python -c "print('\n'.join(dir(__builtins__)))" > builtins.txt12:32
ralsinadobey: I am not @home today, so no mac, and quantal exploded vbox12:33
ralsinagatox can help!12:33
dobeyyeah i can wait for someone else to help :)12:33
ralsinaor mandel12:33
ralsinaand they are supposed to be here :)12:33
gatoxi'm here12:34
ralsinacool, gatox dobey wants that: ^12:34
dobeygatox: ^^ can you send me the builtins.txt file from running that command on mac?12:34
gatoxdobey, yep..... let me turn on the mac12:35
gatox_macdobey, http://paste.ubuntu.com/1289190/12:40
dobeythanks gatox12:41
ryegatox: oh, maybe you know that - what's the revision win32 version of 3.0.2b was created from?12:43
gatoxrye, no idea...... brian probably know that12:44
dobeyrye: i think it's the release-3_0_2 tag, plus a few patches12:45
dobeyrye: aren't the revnos in revnos.txt or whatever it is?12:45
ryedobey: oh12:46
ryedobey: you are right, wow!12:46
dobeyyay, linux and mac builtins are the same at least12:49
dobeyno weird DarwinError or anything12:49
alecuhello, all!12:49
gatoxalecu, hi!12:49
dobeyhola alecu12:51
dobeygatox: can you review https://code.launchpad.net/~dobey/ubuntuone-dev-tools/run-with-flakes/+merge/130432 please?12:59
gatoxdobey, yes, on it!13:00
dobeythanks13:00
ralsinagatox: maybe you can help here http://askubuntu.com/questions/202788/how-can-i-enable-ubuntu-one-indicator-in-12-1013:03
gatoxralsina, will do..... i already write a document about that jeje13:04
ralsinagatox: the strange thing is, on OMG they mentioned it as one of the 10 new features, and it wrked for them ;-)13:04
dobeyoh13:04
dobeyi just answered :)13:04
gatoxdobey, ok..... less work for me13:04
gatoxjeje13:04
gatoxdobey, the branch looks good, and the tests pass....... but there are some lint issues..... should i ignore them? i ask because maybe this is already in trunk or something13:09
dobeygatox: oh? what lint issues?13:09
gatoxdobey, http://paste.ubuntu.com/1289256/13:09
dobeygatox: oh, you don't have the patched pyflakes, as noted in the description of the merge :)13:10
gatoxdobey, ahhhh sorry about that13:10
ralsinadobey: so we are moving frm a patched pylint to a patched pyflakes?13:10
dobeyralsina: sort of; i patched pyflakes so we could get a move on. i'm also working on pushing the changes upstream13:11
ralsinadobey: ok, at least there is hope of getting it merged :-)13:11
gatoxdobey, +1 (upgrading)13:11
dobeyralsina: yeah, though upstream seems to be somewhat inactive (though it is the twisted guys).13:12
mandelralsina, where should I place the documentation about nux etc..13:17
mandelralsina, or better say, unity intro for u1 developers :)13:17
ralsinamandel: canonical wiki, wherever you feel like13:17
dobeyhttp://askubuntu.com/questions/202992/why-do-i-get-a-cross-domain-policy-restriction-error-when-trying-to-stream-to-my13:18
dobeyrye: http://askubuntu.com/questions/202641/ubuntu-one-error-message-stale-broker <- do you know about this?13:21
dobeyhrmm, i really need to find some time to install my new hdd, and quantal, on my workstation13:22
=== m_conley_away is now known as m_conley
gatoxdobey, the weekend is coming :P13:31
dobeygatox: in which i would generally prefer to do other things. plus i theoretically have some prior engagements anyway :)13:32
* karni back!13:49
dobeyalso i really don't like pulling my workstation out of the rack :)13:50
mandelralsina, sorry, I just read the backlog.. stupid mac.. if you need any reviews let me know, I'm just writing nux examples :)13:51
mandelrye, I did not see anything weird, it just looks like the sd is no sending the start event to the state machine13:52
ralsinadobey: also to be considered, you have a rack for your workstation.13:52
mandelrye, do you have the exception logs?13:52
dobeyralsina: well, it's mounted in the rack that's in my desk. yes :)13:54
ryemandel: nope, there are no exception logs. I will now insert a lot of logger statements to track every call after the OffloadQueue and if something comes up then I will start poking13:54
ryemandel: have fun with nux :)13:54
dobeyralsina: server/switch/patch/server raid  are mounted in the rack on the other side of my desk :)13:55
* dobey needs to clean his desk up a bit more and take a picture or two13:55
mandelrye, he, thx ;)13:57
mandeldobey, do leave the toilet paper on the table hehehehe13:57
mandel:P13:57
gatoxalecu, ping13:59
gatoxalecu, do you have a minute to discuss something?13:59
alecugatox: for you? always!14:00
gatoxalecu, jeje thx....... the thing is:14:00
* alecu closes the computer and runs to package the remaining bags... byebye!14:00
alecugatox: no, no, just kidding :-)14:00
gatoxjejjee14:00
gatoxalecu, i've been playing with what we talk the other day..... i have the proper classes creating interacting with filesystem manager and doing the search using the regular expression you mention14:01
gatoxbuttttttt14:01
alecuorsom14:01
gatoxand in the proper folders........ but....... first: it can match things that are not just in the basename....... and the other problem i have....... is that if we are going to use regular expression..... we should escape all the special characters that the thing to search could have..... for example: (, \, ?, etccccc14:02
gatoxso the thing is..... i can create a function that escape all the special characters in a regular expressoin..... but just wanted to check with you if you still think this is the proper way.... or should go back to use basename14:03
alecugatox: you are right about the escaping, yes.14:03
ryebriancurtin: hi, how do i re-create the tree for 3.0.2b - i want to alter library.zip so that it was built from the same 3.0.2 sources14:04
gatoxalecu, so.... what do you think is best? do a replace for each character in the string and escape all the special characters.... or use basename instead to avoid those string operations?14:04
briancurtinrye: i'll type up some directions, should have them in a few minutes14:05
alecugatox: I would totally advice *against* doing a hand coded function to escape chars in a regex14:05
gatoxmmh14:05
alecugatox: so, let's check first if there's something already done14:05
gatoxalecu, ok14:05
gatoxalecu, i'll look for something for it14:06
alecugatox: http://docs.python.org/library/re.html#re.escape14:07
gatoxcuack14:07
gatoxalecu, yep.... i found that in stackoverflow too14:07
gatoxalecu, yep.... that works just fine..... just test it14:08
alecugatox: awesome. This clears half of the path14:08
alecugatox: the other bit you mentioned is "it can match things that are not just in the basename"14:09
gatoxyes..... for example:14:09
alecugatox: I know there's some way to do it, using the greedy / non greedy stuff. Let me think it through14:10
gatoxi can look for "ubuntu"..... and it will match: "/home/U1/ubuntu/myfile" and "/home/u1/folder/ubuntu.txt"14:10
alecugatox: right, and we would only be interested in the latter, right?14:10
gatoxalecu, right......... if we choose that....... maybe we can leave it this way..... and accept searchs like /folder/file14:11
* gatox checks....14:11
alecugatox: while we are at it, make sure to re.compile your expression, and then use that object to do the match14:11
gatoxalecu, yes..... that's what i'm doing14:12
alecugatox: I like being able to search by including bits of the folder name14:12
gatoxalecu, like a fuzzy search14:12
gatoxalecu, yes..... it works awesome....... for example i have:14:13
gatox'/home/gatox/Ubuntu One/u1/fslogger.c'14:13
gatoxand i search for: u1/fs14:13
gatoxand i found that14:13
* gatox like it14:13
gatoxlikes14:13
alecugatox: or "u1 fs" should match it too14:13
gatoxalecu, no14:13
gatoxunless we split the search in spaces..... and join them with .* in the middle14:14
alecugatox: I was thinking of that all the time! :-)14:14
ralsinagatox: that's bad, no, please14:14
alecuralsina: why?14:14
ralsinaUbuntu One ?14:14
ralsinaWe want to search for Ubuntu.*One? :-)14:15
gatoxralsina, what?? using  the folders too.... or the splitting part?14:15
ralsinasplitting and joining with .*14:15
gatoxahhhh14:15
alecuralsina: I think we do want to split, yes14:15
gatoxyes..... i'm not in favor of that neither..... because the name of the files can actually contain  spaces14:15
ralsinatry not to be too clever :-)14:15
alecuralsina: if you have a folder called "my documents" and a file named "rapunzel.txt" within it, you want to search like "doc rapu"14:16
alecuralsina: you don't want to search like "ments/rapu"14:17
dobeyrockstar: is https://code.launchpad.net/~rockstar/ubuntuone-ios-music/fix-header-search/+merge/126811 still valid?14:17
ralsinaalecu: it's un undiscoverable feature that will make it return random things14:17
ralsinaseemingly random, at least14:17
* ralsina slaps himself for micromanaging devs and steps back.14:19
gatoxsooooo..........14:20
ralsinaoh, we lost one alecu14:20
briancurtinrye: i just realized i have it on the wiki: https://wiki.canonical.com/UbuntuOne/Windows/ReleaseChecklist -- however, i don't remember what makes the B part of 3.0.2b...14:20
gatoxralsina, you heart alecus feelings jejeje14:20
ralsinabriancurtin: a proxy fix14:22
gatoxalecu, is back14:23
briancurtinthats about all i remember, but im not sure what the particular branch was since i believe i just took the 3.0.2 setup and applied one change to it14:23
dobeymorning briancurtin14:23
alecugatox: sorry, I'm having quasselcore troubles, it seems :P14:23
briancurtinhey dobey14:23
gatoxralsina, alecu so.... we need to decide..... which is going to be the approach for this?14:23
alecugatox: I typed a lot before dropping. what was the last you saw?14:24
ralsinaalecu: in case you missed it, I slapped myself for micromanaging you, so you and gatox decide :-)14:24
gatoxalecu, this: <alecu> ralsina: you don't want to search like "ments/rapu"14:24
alecugatox: great.14:24
dobeybriancurtin: can you run python -c "print('\n'.join(dir(__builtins__)))" > builtins.txt on win32 buildout and sned me (pastebin, email, whatever) the resulting file?14:24
alecuand now quassel won't let me see the backlog of the previous quasselcore instance. awesome.14:25
ralsinateam, here's a hint of the future: https://twitter.com/njpatel/status/25929898897141760014:25
gatoxalecu, so....... splitting or not splitting..... that is the dilemma14:25
ralsinaand that's called expectation management14:25
gatoxralsina, jjejeje14:26
briancurtindobey: https://pastebin.canonical.com/76938/14:26
alecuralsina, gatox: what if instead of ".*" we used ".+" ?14:26
ralsinagatox: for later http://io9.com/5953055/the-sounds-of-cylons-as-sung-by-simon--garfunkel14:26
ralsinaalecu: .+ is better14:26
ralsinaalecu: at least doc rapu would not match docrapu14:26
gatoxyes....... that would be good14:26
alecuralsina: "do crap you". Is that an order?14:27
gatoxralsina, jejejej sounds of cylons jejeje14:27
dobeybriancurtin: thanks14:27
ralsinaalecu: please no :-)14:28
ralsinaalecu: I was using your example14:28
ralsinaanyone has a pending review?14:34
alecuralsina: it seems that the sounds of cylons will be planted in my brain, like a cylon vision, for the foreseable future. It's now on repeat here :-)14:35
ralsinaand if you don't *WHY*???? ;-)14:35
ralsinaalecu: it's pure distilled awesome14:35
ralsinaalecu: besides, half the lyrics are kept from the original, it's creepy14:35
ralsinaAlso, cylon & garfunkel: http://img.gawkerassets.com/img/182l12er7udygjpg/cmt-medium.jpg14:36
dobeyhuzzah. the WindowsError issue is fixed in trunk pyflakes already even14:37
ralsinayay14:37
* dobey throws that patch into nightlies too14:38
* gatox can't wait to see bsg again....14:40
* alecu needs to run a small errand. Will probably be back before standup14:43
mvome14:59
mvo(is too early!)14:59
mandelmvo, one min :)14:59
dobeycouple minutes, yeah14:59
karnime14:59
mandelme14:59
thisfred_me14:59
rockstarme14:59
briancurtinme15:00
gatoxme15:00
ralsiname15:00
alecu_me15:00
dobeyme15:00
mmccme15:00
ralsinamvo go!15:00
mvoDONE: lp:~mvo/update-manager/lp1068389, upload to quantal-proposed, : lp:~mvo/software-center/support-size-from-agent,  lp:~mvo/ubuntu-release-upgrader/warn-on-dropped-fglrx-support/, play around with heapy and discover #1068614, apt: look at #1062503 and provide reproduce way (*urgh*), looking at c++ books with mandel (thanks!)15:01
mvo yesterday as well?15:01
mvoor is today enough?15:01
ralsinamvo since last standup, but since we had a call yesterday, today is good15:01
mvoyesterday: DONE: lp:~mvo/software-center/data-provider-add-ratings, lp:~mvo/software-center/update-sc-cmdline-flexibility, create 5.4 series, merge outstanding branches, prepare 5.4.x and 5.5.0 uploads, work on google doc "DASH app USC consolidation technical notes", travel preparing, meetings(!)15:01
mvonext mandel15:01
mandelDONE: Wrote nux examples mainly explaining layout, buttons, views etc..15:01
mandelTODO: Talk about nux properties. Metion dash code layout and how to add new previews.15:01
mandelBLOCKED: no15:01
karni:(15:01
mandelkarni, please15:02
karni^_^15:02
karniyesterday:15:02
karniDONE: Added action bar now playing item. Required some hacks, due to particular design requirements. Will discuss that with Patricia.15:02
karnihttp://ubuntuone.com/4rpZfpmuC8BfG22101sRXs (GN portrait)15:02
karnihttp://ubuntuone.com/6W5Os3giSdTJsTVE9rumZl (GN landscape)15:02
karnihttp://ubuntuone.com/1UCSvigQRCkzb5oGrcumbu (G7 portrait)15:02
karniAdded even more home activity UI tests extercising options menu.15:02
karniBuild a staging U1M apk for Vincent15:02
karniTODO: Continue yesterdays TODO list (wire up the player and music service)15:02
karniBLOCKED: No.15:02
karnitoday:15:02
karniDONE:15:02
karniReviewed two branches.15:02
karniWired up Now Playing home screen indicator.15:02
karniImplemented and hooked up populating play queue.15:02
karniMinor styling to player UI.15:02
karniTODO:15:02
karniReview urbanape's branch.15:02
karniHook up player UI album art, titles, progress (similar to now playing indicator).15:02
karniBLOCKED:15:02
karniNo.15:02
karninext: thisfred_15:02
thisfred_DONE: cdocs fixes | backend metrics TODO: whatever comes my way BLOCKED: no NEXT: rockstar15:02
rockstarDONE: Set up vm and machine for nux work15:03
rockstarTODO: Continue on iOS Music app 2.0 integration15:03
rockstarBLOCKED: None15:03
rockstarNEXT: briancurtin15:03
briancurtinDONE: team call, installer, dusting off C++ books, reviews, testing15:03
briancurtinTODO: find a way to add cert checking with the bundled certs rather than relying on system to do it (alecu, any tips?)15:03
briancurtinNEXT: gatox15:03
gatoxDONE:15:03
gatoxImplements the search that was already in u1-cp for u1 files inside u1-client, faster and with some improves. Started working in the ipc functions to obtain this info from control panel.15:03
gatoxTODO:15:03
gatoxFinish with the ipc implementation and propose the branch.15:03
gatoxRefactor u1-cp search functions to use the one in u1-client.15:03
gatoxBLOCKED:15:03
gatoxNo15:03
gatoxralsina, go15:03
ralsinaDONE: lots of calls. TODO: new "mgmt rumble" call, reading lots of docs, perhaps reviews? BLOCKED no NEXT: gatox15:03
gatoxralsina, no, i'm not next15:03
karni:D15:03
dobeyalecu_: go15:03
ralsinaso alecu_ then15:03
alecu_DONE: many mumbles, wrapping up and preparing for vacations15:03
alecu_TODO: two weeks of vacations starting in 4 hours15:03
alecu_BLOCKED: no way!15:03
alecu_NEXT: dobey15:03
dobeyDONE: team call, bug #1068335 (dev-tools), booked sprint tickets15:03
dobeyTODO: reviews, finish bug #1068335, bug #711324, upstream pyflakes fixes15:03
dobeyBLCK: None.15:03
ralsinagatox: you were in my timeline ;-)15:03
ubot5Launchpad bug 1068335 in Ubuntu One Control Panel trunk "pylint is too complex and causes many problems" [Medium,Triaged] https://launchpad.net/bugs/106833515:03
dobeymmcc: go15:03
ubot5Error: Launchpad bug 711324 could not be found15:04
mmccDONE: music, sync menu stuff on mac/win & downloads, futile bug search15:04
mmccTODO: finish up sync menu stuff15:04
gatoxralsina, but brian says gatox go :P15:04
mmccBLOK: no15:04
* ralsina goes find the developer-proding-stick15:04
ralsinaok, mmcc was last (we all agree there ;-)15:04
ralsinaSo, comments?15:04
dobeystop micromanaging! :)15:04
karniRock on!15:05
alecu_briancurtin: "a way to add cert checking with the bundled certs rather than relying on system to do it". Can you tell me a bit more about that?15:05
ralsinaEOM for standup. Quick if a bit sloppy.15:05
alecu_briancurtin: this is not os.system("iexplore https://one.ubuntu.com"), right?15:05
ryeralsina: a proxy fix? tell me more - is there a revision i can cherry-pick the change?15:06
ralsinaalecu_: of course not! We have to use subprocess! ;-)15:06
ralsinarye: it's a vague memory for me :-(15:06
briancurtinralsina: would that be acceptable? i was under the impression there was going to need to be something added to WebClient to read from cert files or something like that15:07
ralsinabriancurtin: it's not acceptable at all15:07
ralsinabriancurtin: sorry, bad joke there15:07
briancurtini dont know much about certs/ssl, missed the joke. i take it the last part is correct, then?15:08
mvobriancurtin: I'm totally off topic and may have no idea what I'm talking about, but I did deal with that cert in webkit/soup - could you give me a code example?15:08
dobeymvo: this is more a windows issue15:08
mvoor the code that needs the cert file set?15:08
ralsinamvo: on windows, the root certs are downloaded on demand15:08
mvodobey: *weeeh* I will probably not of much help then indeed15:09
karnialecu_: Before I forget - enjoy your holiday!15:09
mvoralsina: eh, woah15:09
ralsinamvo: and we use an impopular one which lots of people never run into before using our code15:09
alecu_karni: thanks15:09
dobeybriancurtin: afaik, having qt use our certs *and* the system ones when needed, is a bit complex, but should be doable15:09
mvothanks ralsina, that gives me some background that I was missing earlier :)15:10
alecu_briancurtin: so, after looking at the qt docs, this seems to be what we would need to use: http://qt-project.org/doc/qt-4.8/qsslsocket.html#addCaCertificates15:13
alecu_briancurtin: but since we don't use the qsslsocket directly, we should find how to add it to the higher level qt objects used by the webclient and the proxy tunnel15:14
mvohttp://stackoverflow.com/questions/11941630/is-this-the-right-way-to-set-the-ssl-protocol-with-qwebpage <- might be this one15:15
mvoi.e. QSslConfiguration::setCaCertificates()15:16
=== mmcc is now known as mmcc_
briancurtini think QSslConfiguration setLocalCertificate15:19
alecu_briancurtin: are we already bundling that third certificate? I recall u1-storage-protocol used two already15:21
briancurtinalecu_: we bundle three - two godaddy and one valicert15:22
alecu_briancurtin: great then.15:22
alecu_briancurtin: I think you should try running sd and u1cp with an empty certificate database (by passing an empty list to QSslConfiguration::setCaCertificates), and checking that it fails. Then adding the three certificates only, it should work.15:26
briancurtinalecu_: and would i be correct in thinking that this goes in WebClient in SSO?15:26
=== alecu_ is now known as alecu
briancurtinsorry for extra hand holding...haven't worked with SSL much15:28
alecubriancurtin: yes, yes. It goes in webclient in sso15:30
=== thisfred_ is now known as thisfred
briancurtinalecu: thanks, will look into it now15:30
dobeyok, need to get lunch. bbiab15:35
alecubriancurtin: in fact, I think sso should be the only place needing that, since I'm looking at sd and storageprotocol, and it's already being used there.15:36
alecuin ubuntuone/storageprotocol/context.py15:36
briancurtinyeah i figured it's the lowest level spot to put it15:37
alecubriancurtin: and the webclient is sso is what's used by control panel, so yes.15:37
alecuok, lunchtime for me15:38
=== beuno is now known as beuno-lunch
* gatox lunch15:41
=== gatox is now known as gatox_lunch
mmccso, a new project for the pyobjc sync menu - anyone have a problem with 'lp:ubuntuone-cocoa-menu' ?16:29
ralsinano problem here16:30
dobeyoh16:31
dobeyralsina: btw, should we rename the lp team and set up a mailing list for it?16:32
ralsinadobey: yes16:33
ralsinadobey: could you?16:33
dobeyralsina: i think i need to be an admin for the mailing list bits. care to make me one?16:33
ralsinadobey: admin of what exactly?16:34
ralsinadobey: the team?16:34
dobeyhttps://launchpad.net/~ubuntuone-desktop+16:34
dobeyyes16:34
ralsinadobey: on it16:34
ralsinadobey: done. Also add karni mvo and rockstar16:36
karniThanks16:36
ralsinathat team has been needing maintenance for a while!16:36
karni:)16:36
dobeyyeah, our teams need some cleaning up16:38
=== gatox_lunch is now known as gatox
ralsinadobey: made alecu and mvo admins just in case both of us are away at the same time16:47
dobeysure16:48
dobeyhrmm, not sure how to make the list private16:48
* dobey pings some ops for that16:49
dobeybugger :(16:55
=== beuno-lunch is now known as beuno
karniI think I nailed that painful SQL problem (column mapping) which caused crashes. It seems, however, that test runner isn't ensuring tests are run isolated, because I sometimes get tests that cause the next one to hang, although tearDown() is done properly.17:04
dobeynoice17:06
dobeynow i don't need 2 microphones17:07
mandelkarni, are you using a single instance of sql for each of the tests?17:07
mandelkarni, maybe creating a new sqlite db in the setup and removing it in the teardown is the way to go17:08
karnimandel: No, but these tests don't depend on mock data. I'm actually allowing these tests to run on any populated data (my priavte music in this case). I do, however, extercise the sqlite db in content provider tests, where I have mock data and it behaves just like you described.17:08
karnimandel: Looks like it's not cleaning up state (finishing an activity), although I request it do do so. I'm making sure an AsyncTask (with a thread worker) doesn't leak any context, which could prevent from finishing that activity.17:09
karniI suspect that could be the case here.17:10
mandelkarni, what happens if the asyn task never finishes?17:12
karnimandel: what the test says is "wait for player launch for 5 seconds. verify if it has launched.". I disabled launching shuffle all songs asynctask alltogether, what I see is the player doesn't shut down even though I have activity.finish() at the end of my test.17:13
karnimandel: the async task is a call to a simple blocking method, which is exercised elsewhere in a test, so that should never happen.17:14
mandelkarni, I would need to look at the code, and I'm not an expert java guy.. I just fear threads a lot :P17:15
karnimandel: Thank you :) I'd have to commit that now, and the diff's pretty large, but if I continue to have that problem unresolved, I'll share the branch in here :)17:16
briancurtinmandel: not sure if youve seen jenkins in a while but https://bugs.launchpad.net/ubuntu-sso-client/+bug/1061528 seems to have broken SSO on windows17:16
ubot5Ubuntu bug 1061528 in ubuntu-sso-client (Ubuntu Quantal) "Proxy tests fail after Qt update in Q" [Critical,Fix released]17:16
karnimandel: This test doesn't seem to correctly finish the launched application. https://pastebin.canonical.com/76962/17:16
mandelbriancurtin, yes, diff qt version on windows...17:20
mandelbriancurtin, I'll refrain from using the f word..17:20
mandelkarni, no idea then...17:20
karnimandel: These tests start the tested activity from scratch in setUp and clean them up in tearDown. Check this - the interesting part is each test passes, and they do not base on data nor state. (Simple UI interaction tests)17:20
karniSo how the f can one test not clean up after itself and block the next one. MEH17:21
* karni tries adding a Thread.wait in tearDown17:22
mandelkarni, on twisted we would use a defer to ensure that everything was cleaned up..17:23
karniblargh. mandel, adding Thread.sleep(500); at the end of tearDown() so that Android settles down fixed that problem. all tests green.17:23
mandelkarni, so what you need to do is ensure that there is a way to assert that the async call is done17:23
karnimandel: I would normally use mInstrumentation.waitForIdleSync(); but it seems things are happening too fast. It didn't do the job well.17:23
karnimandel: Right17:24
mandelkarni, no idea on how to do that in java :-/17:24
karnimandel: Something like waiting for a launched activity to _really_ finish17:24
karnimandel: It's been good to exchange thoughts, thanks :)17:24
* karni may use an instrumentation activity monitor17:24
mandelkarni, yes, because I suppose that there is no guarantee when the app is going to be finished,  and therefore that is the problem17:25
mandelanyways, need too  go17:25
mandelkarni, no problem :)17:25
mandelall, laters!17:25
karnibyee mandel o/17:25
briancurtinmandel: the Qt version is (allegedly) the same, but the PyQt version is 4.9.4 versus Q's 4.9.317:25
briancurtinah, i'll look into it though, i'm in this area now17:25
dobeymailing lists are hard. oi17:25
mmccoh hey briancurtin, you mentioned that the remote-folders-fix branch doesn't work for you IRL on windows? still seeing that?17:31
briancurtini havent tried today, but as of before EOD yesterday it didnt work17:31
karniAll righty. I've been hacking until sun rise last night, I'ma EOD slightly earlier today.17:35
karniHave a great weekend guys!17:35
briancurtinyou too!17:35
karniThanks o/17:35
gatoxkarni, bye17:35
karni\o gatox17:35
briancurtinupdating all of this two-factor auth crap makes me wish i didnt even get a new phone. ugh.17:40
dobeyheh17:45
dobey2 frustrating auth17:46
dobeysorry for lp spam everyone17:50
ralsinadobey: you'll hear from my lawyer!18:10
=== lifeless_ is now known as life
=== life is now known as lifeless
briancurtinmmcc: if you have a few minutes, would you mind trying this sso branch on mac? lp:~brian.curtin/ubuntu-sso-client/nam-setproxy-windows-workaround18:49
ralsinaAlecu, you are not on the private server?18:53
dobeyralsina: isn't he on holiday?18:55
mmccbriancurtin, sure…18:55
ralsinadobey: not yet :-)18:55
ralsinadobey: or maybe he already is. Ok, not urgent, can wait a couple of weeks. Or three.18:55
dobeyoh, he's not listed for today on the site18:56
dobeybut he said "starting tomorrow" yesterday18:56
dobeyalthough now's about the time when he has to go pick up amelia, no?18:56
dobeyanyway :)18:57
* gatox needs to write a lot of tests.......19:03
mmccbriancurtin: well this is strange. I get all these failures: http://paste.ubuntu.com/1290098/ -- but I get those same ones with trunk…19:04
briancurtinmmcc: would you mind doing a small change and re-running? ubuntu_sso/utils/webclient/qtnetwork.py:110 - just change that to if to darwin, see if that makes a difference. i think there's a PyQt bug that i'm hitting on Windows - that TypeError shouldnt be happening19:07
alecuralsina: hi!19:07
alecuralsina: I'm about to leave, in 5'19:07
gatoxalecu, bon voyage! :P19:08
ralsinaalecu: have fun then! :-)19:08
dobeyok, so i hope you all have a mail now, via the new mailing list19:08
mmccbriancurtin - better: http://paste.ubuntu.com/1290116/19:10
mmcc(btw that was just s/win32/darwin/ - so not calling setProxy on darwin…19:10
briancurtinmmcc: perfect, thank you19:11
briancurtinmmcc: for the remote-folders-fix branch, is there anything i can do to get you some info? it just seems to hang "getting information" forever. i see no tracebacks or anything showing up, so nothing obvious19:12
mmccbriancurtin, thinking…19:13
mmccI'm trying it now for myself and seeing if it's working for me. I thought I had done this, but I guess it wasn't possible before, so…19:19
mmccyeah, works for me… hmmm.19:21
briancurtinmmcc: which win version?19:22
mmcc719:22
briancurtini'm on 8, but that can't be the issue. i'm going to wrap this SSO branch up and then try to poke around19:24
alecunow yes: bye all, see you in two weeks!19:25
briancurtinenjoy!19:25
mmccbriancurtin ok - I'd first look at line 314 of wizard.py, where it gets the credentials. my best guess is that yield is never coming back due to something in SSO19:25
mmccalecu - have a great time!19:25
dobeyalecu: beware of giant mice on your trip :)19:45
mmccand cockroaches! http://weblogs.sun-sentinel.com/news/specials/weirdflorida/blog/cockroach-thumb.jpg19:58
mmcchmmm, looks like I really can't use inlineCallbacks with PyObjC19:59
mmccSo, is there a reasonable way to do the equivalent of "for func in list_of_funcs_returning_deferred: yield func" without using inlineCallbacks?20:04
mmccI think I am just missing something…20:05
dobeycan i get some reviews for https://code.launchpad.net/~dobey/ubuntuone-client/fix-711324/+merge/130630 ?20:08
dobeygatox: ^^ ? :)20:08
gatoxdobey, yap.... on it20:08
dobeymmcc: [yield x() for x in list_of_funcs] ?20:09
dobeyor something like that20:09
mmccdobey the problem is that I can't use yield, since decorating a function as inlineCallbacks appears tocrash pyobjc20:09
briancurtinthat sounds fun/useful20:10
dobeymmcc: but inlineCallbacks just takes the results and creates a generator of deferreds20:10
dobeymmcc: so maybe [x() for x in funcs] then?20:11
dobeyjust return that i mean20:11
dobeyi don't know20:11
dobeybeat pyobjc with a stick?20:11
gatoxdobey, branch looks good, +120:12
dobeyalso, i am sort of tempted to just fork pyflakes20:12
* gatox goes to start the weekend..... eod here! see you on monday people! :D20:15
mmcchrm, looks like it's only when it's decorating a class method… i can avoid this problem20:23
ralsinammcc: breaking only on class methods may betray some dark magic involved20:24
mmccralsina er, I misspoke - not class methods, but methods of a class. IE, it doesn't crash if I have a module level function that's decorated by @inlineCallbacks20:26
ralsinammcc: oh, I understood that, not @classmethods anyway :-)20:26
ralsinammcc: I was accidentally correct20:27
ralsinammcc: but hey, if it works, it works ;-)20:27
mmccyeah, and it's not *too* horrible. actually making it work with raw deferreds wasn't looking pretty20:27
ralsinayes, that's spaghetti20:29
dobeygrrr20:52
dobeyspeaking of spaghetti… indirect reimporting is really bad20:52
=== m_conley is now known as m_conley_away
briancurtinthis is fun. i can make things worse (as expected) when running explicitly with no certs, but adding the certs to the config doesn't seem to work. probably have to come back to it on monday...being picked up in a few minutes21:52
dobeygrr, this weird code in ubuntu_sso is quite annoying21:53
briancurtinand they're here. have a good weekend all21:54
dobeyanyone else can take a quick poke at https://code.launchpad.net/~dobey/ubuntuone-client/fix-711324/+merge/130630?21:58
dobeymmcc: you maybe? i think everyone else is gone now21:59
dobeyand i really should be21:59
mmccdobey, sure - I'll look at it later.21:59
mmccit's on my list :) have a great weekend!21:59
mmcclate lunch time22:01
mmccman, crappy home wifi today, wtf22:05
dobeyalright, i'm off, later :)22:08

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