aquariusrye: yes, yes it would be good if the streaming URL was actually documented08:16
aquariusjamesh: we should do that :)08:16
jameshstreaming URL as part of our subsonic API implementation, or something else?08:23
karniGood morning!08:27
aquariusjamesh: we don't actually tell anyone anywhere where the subsonic API endpoint *is*, was rye's point I think :)08:29
aquariusso, y'know, oops08:29
jameshaquarius: if people don't care about the rest of the Subsonic API, couldn't they just hit updown directly?08:31
jameshe.g. our REST files API08:31
aquariusjamesh: sure, but I think this is about people who do care about the subsonic api08:31
aquariushttps://one.ubuntu.com/developer/music/stream_music/cloud should mention streaming.one.ubuntu.com at some point, right?08:32
jameshI thought you were referring to a particular endpoint, rather than the API base.08:34
jameshwe should also document what is necessary to get the API to return Vorbis files08:34
jameshI think that is what http://askubuntu.com/q/138796/12469 is about, but will have to wait for the user to reply08:35
aquariusgrr, OK, I now hate smuxi because it doesn't tell me when people talk, obviously enough08:45
aquariusjamesh: I agree on the api base, yep. Vorbis stuff... we should check with beuno about how best to document that08:46
JamesTaitHappy Friday, people! :D08:46
jameshAt the moment, the rule is that the client string == "android" or starts with "ubuntu"08:47
ryehttps://one.ubuntu.com/music/player/ - 504 Gateway Time-out ?09:04
ryeaquarius: https://one.ubuntu.com/developer/account_admin/mobile_login/cloud - syncml?09:14
popeyi get the 504 too09:15
popeybut not now ☺09:15
aquariusrye: that URL isn't ever opened in a browser; you, the app dev, are expected to trap it with its custom scheme09:21
aquariusthat the server is syncml.one.ubuntu.com is a historical wart, true, but it's not a big deal09:21
ryeaquarius: okay09:21
ryeverterok: hi, do you happen to have a code to parse tritcask outside syncdaemon (using SD libs)? If not, that's ok, will try to write one09:27
JosssseHello guys.09:50
JosssseI think ubutu one has been erasing my files. Is there a way I can be sure of this.09:51
JosssseSome sort of log files, for example?09:51
ryeJosssse: first of all, check whether these files appear in trash folder, ubuntuone does not unlink files, it moves them to trash09:52
ryeJosssse: second, the logs are in ~/.cache/ubuntuone/log09:52
Josssserye, heh, didn't even know I had a trash folder.09:52
Josssseahh, they're all in trash!!! uff, got scared a bit.09:54
JosssseAny idea why it is sending my files to trash? only the ones i'm working with.09:56
ryeJosssse: hm, what application are you using to work on the files?09:59
ryeJosssse: also, what release are you running?09:59
JosssseI'm using latexila10:00
ryeverterok: nevermind, wrote one10:00
JosssseThis is the log output from the last events which dealed with the file i'm concerned about. http://paste.ubuntu.com/993916/10:01
Josssserye, last LTS xubuntu.10:01
ryeJosssse: installing latexila to see how it saves files10:03
Josssserye, You shuold now that since i'm working in latex i save files quite often so i can compile and take a look at what i'm doing.10:04
JosssseThe same thing happened to me on my laptop yesterday which is running same xubuntu version.10:05
Jossssealso latexila10:05
ryeJosssse: do you have autosave enabled in the preferences?10:05
ryeJosssse: ok, could you please run the following script - http://people.canonical.com/~roman.yepishev/us/bin/ubuntuone-diff - download it and run with python ubuntuone-diff - it will scan the local and remote folders and print the filenames that are different10:09
Josssserye, I must leave for 30 minutes. Will you be aroudn then??10:10
ryeJosssse: in case the files that were deleted are listed there, then please paste the summary10:10
ryeJosssse: yes, at least for 6 hours more10:10
JosssseAlright, see you then. Thanks again.10:10
popeyI have deleted all the photos on my iphone bar 6, but u1 files still lists 600 odd photos to be uploaded.10:46
* popey kills and restarts the app10:46
=== yofel_ is now known as yofel
gatoxgood morning!11:08
Josssseok rye I'm back.11:14
JosssseLets wget that script.11:14
Josssserye, So I can't run the script.11:16
JosssseIn fact, when I do cat ubuntuone-difff i get lots of weird characters.11:17
ryeJosssse: yes, that's a ZIP file, you can unzip it if you like to look at python code, how does it fail?11:17
JosssseI just don't know how to run it.11:18
Josssseok, zip unzipped, now what?11:18
Josssseok, running it.11:19
Jossssedone, it finishes with two conflicts.11:19
ryeJosssse: well, python ubuntuone-diff would have done the same11:19
ryeJosssse: only 2 conflicts, no missing from local or remote storage? Are these two conflict files the ones you were working on?11:20
Jossssenope. they're all files i havent used in a long time.11:21
Josssseno files need uploading nor downloading.11:21
Josssse46 files need hasing.11:21
ryeJosssse: when was the first time you noticed the files get removed?11:25
ryeJosssse: also, could you please archive the files from ~/.cache/ubuntuone/log and publish them via Ubuntu One (and send me the link privately) or send them to ubuntuone-support@canonical.com so that I could check whether the removals are requested by the server-side11:26
JosssseYesterday on my laptot.11:26
JosssseToday on my office computer.11:26
JosssseFor some reason i can't share it through one.ubuntu.com so I'll just zip it and email it.11:32
JosssseEmail sent.11:34
ryeJosssse: i will get back to you in 10 minutes (parallel tasks)11:36
JosssseThanks. I'll be around.11:36
ryefacundobatista: should we synchronize .goutputstream-XXXXXXX files? They are created only to be later removed to a real location11:46
ryeJosssse: i guess i can reproduce it11:50
JosssseReproduce what?11:51
ryeJosssse: the removal of the files11:52
JosssseI don't understand.11:52
ryeJosssse: you were telling earlier that the files were removed by Ubuntu One, and based on the logs they were indeed removed. I tried to use LaTeXila and i was able to cause the syncdaemon to remove the files11:53
JosssseAhhh. Understood.11:54
ryewell, to queue their removal11:54
JosssseSo have we found a bug?11:54
ryeJosssse: actually I have an easy fix that can be used to prevent this from happening - by default a lot of gnome applications are not writing the file directly, instead they create a temporary .goutputstream-SOMETHING file, write the data there, remove the original file, rename the .goutputstream to real file. Syncdaemon records all these changes and relays them to the server. And something weird happens here - not all statements get to the server, upon restart 11:56
ryeJosssse: the fix I am thinking about is to alter ~/.config/ubuntuone/syncdaemon.conf to have another line in ignored files -11:57
ryeJosssse: like here - http://paste.ubuntu.com/994062/11:57
Josssserye, clever.11:57
ryebut this has to be done on all the computers11:58
JosssseNo prob.11:59
JosssseLet's try it.11:59
ryeJosssse: after changing the file, syncdaemon needs to be restarted - u1sdtool --quit; u1sdtool --start12:01
ryefacundobatista: ping12:02
ryeJosssse: definitely a bug12:02
JosssseLet's see. :)12:03
JosssseI think it's working fine.12:08
JosssseIt's also awesome that u1 doesn't unlink the files.12:08
JosssseThank you very much for your help rye .12:08
thisfredralsina: if you want to start your day on a note of great hilarity, have a look at https://code.launchpad.net/~thisfred/u1db/u1todo-design-1, and more specifically, run PYTHONPATH=. python cosas/ui.py12:10
alecuSumuinen mutta hyvää huomenta!12:11
thisfredI have had Chipaca look at it, since he's the master of python strangeness, but even he gave up12:11
ryeJosssse: i am going to file a bug once facundobatista or verterok get back12:11
thisfredalecu: are you finnish now?12:11
alecuhey there thisfred! It's very early for you too, right?12:12
mandelalecu, moi moi!!12:12
alecuthisfred, oh, sprint.12:12
thisfredalecu: :)12:12
ryethisfred: umm... the last commit message is questionable12:13
thisfredrye, accurate, though12:13
alecurye, as in "police style questioning"?12:13
thisfredalecu: have you ever seen an excursion depth exceeded runtime error in a line that is comparing two integers?12:14
thisfredI'm starting to thing Qt is short for Quaint12:15
alecuthisfred, I'm running the branch... how do I trigger it?12:15
thisfredyou have, look at the terminal12:16
ryethisfred: super().data() is the same as data() for some reason12:16
ryeso you are calling the same function12:16
alecuthisfred, and "excursion depth" sounds like some spelunkers on a cave picnic!12:16
thisfredhehe, fair point12:16
rye<bound method UITask.data of <__main__.UITask object at 0x2d6e7a0>>12:16
rye<built-in method data of UITask object at 0x2d6e7a0>12:16
thisfredrye, cool12:16
thisfredwhy that is still sort of puzzles me,12:18
ryethisfred: i wonder whether that QSomething is based on object12:19
ryethisfred: return QtGui.QTreeWidgetItem.data(self, column, role)12:20
ryethisfred: fixed12:20
ryewell, i am not sure what it should do but it does not crash :)12:20
thisfredrye, you are a star of rock12:20
ryethisfred: what should it actually do?12:20
thisfredrye, not crash is a great start12:20
thisfredit's not finished by any means12:21
thisfredrye, so it's old style classes?12:21
alecuthat sounds weird!12:21
alecu"In versions of PyQt earlier than v4.5 there were restrictions on the use of super with PyQt classes. These restrictions no longer apply with v4.5 and later."12:21
thisfredalso the line it errors in is completely wrong12:22
thisfredbut a workaround is good enough for me12:22
thisfrednow back to the less fun but still quite fun C stuff12:23
alecuthisfred, I've replaced every "super", but I still get "recursion depth exceeded" when hovering over the todo entries.12:26
thisfredalecu: o well12:27
ryeRuntimeError: maximum recursion depth exceeded while getting the str of an object12:32
ryeif role == QtCore.Qt.DisplayRole:12:34
ryeRuntimeError: maximum recursion depth exceeded in cmp12:34
thisfredrye, exactly12:34
thisfredand those are integers. I checked12:34
thisfrednor does the except seem to catch it (though that may be because pdbing qt apps is not really possible)12:35
thisfredor maybe they are pretending to be integers12:36
rye    print QtCore.Qt.DisplayRole12:37
ryeRuntimeError: maximum recursion depth exceeded while getting the str of an object12:37
alecuthisfred, both are *not* integers. Check it out:12:38
alecu(Pdb) QtCore.Qt.DisplayRole.__class__12:38
alecu<class 'PyQt4.QtCore.ItemDataRole'>12:38
thisfredalecu: right. They are very very bad people for not showing that in the repr though12:39
alecuI agree!12:39
thisfredand also for not just using an integer ;)12:39
ryethisfred: integer is not portable12:39
ryealso, bit is not portable12:40
ryei bet they have QInteger which consists of QBits12:40
thisfredI don't care, and they shouldn't make me care12:40
thisfredmake the C which is the problem weird, not the python that doesn't have the issue12:41
alecurye, lol12:42
ryeit still does not explain the recursion issue12:45
thisfredyeah, I think it's just lying about the line12:48
thisfredand to be clear, I'm also quite certain I've done something catastrophically stupid, somewhere, it would just be nice if they weren't actively trying to hide from me what that is12:49
alecuthisfred, the recursion is in the call to ".font()"12:52
alecuthisfred, try commenting the "if role == QtCore.Qt.FontRole:" and what's inside it.12:52
ralsinathisfred: the Qt API never exposes any integers for enums. It's C++ :-)12:54
ralsinathisfred: and sorry I missed all the fun! (or is it still broken?)12:54
thisfredralsina: alecu found it, I think12:54
thisfredralsina: though the backlog should still amuse you ;)12:55
alecuthisfred, so, to fix all this I would create two fonts in the constructor of UITask, then return one or the other in the "if role == QtCore.Qt.FontRole" true condition.12:55
thisfredalecu: makes sense12:55
ralsinathisfred: it's always a good idea because creating QFonts (or QImages or QPixmaps) is expensive12:56
alecuthisfred, but we should not be calling ".font()" from ".data()". I don't have a more detailed explanation as to "why not?" :-)12:56
alecuI can wave my hands for a few seconds, though.12:56
thisfredI understand12:56
thisfredI think12:57
ralsinaalecu: .font() calls data()12:57
ralsinaalecu: because font() uses data(FontRole)12:57
alecuralsina, that's a good one.12:57
thisfredI should have used a super there12:57
ralsinaso using it to implement dat(FontRole) is a nono12:57
* alecu says ".font() calls .data()" while waving a hand.12:57
thisfredif that works even12:57
ralsinaalecu: the font of an item *is* it's data for the font role :-)12:58
* ralsina just misused it's. Argh12:58
alecuralsina, its no problem!12:58
alecuthisfred, anyway, this does not explain why we were seeing the error in any other line.13:00
ralsinaalecu: if you fill the stack any function call can fail13:00
alecuthisfred, also, the stacktrace we were seeing was minuscule. Probably when entering the PyQt c++ code the python stack is lost.13:00
thisfredalecu: no, I messed with sys.rescursionlimit and sys.excepthook13:01
thisfredbut the exception in the wrong line was there before that13:01
thisfredyeah, super or calling it on the superclass directly does not work either.13:05
thisfredI should probably stop trying to think of this as python ;)13:05
alecuthisfred, I think I understand it now....13:05
alecuif the python stack has place for 100 calls, and we do recursion, it's exploding when the 99th recursion is trying to call the cmp of those Qt C++ ints.13:06
alecuso, the recursion is being done on the "font()" call. But the explosion happens on the cmp.13:07
thisfredI seem to remember that those exceptions are usually unhelpful13:09
alecuthisfred, exactly! when python recursion depth is reached, pdb won't start and apport crashes :-)13:13
=== m_conley_away is now known as m_conley
alecuhey, all: ralsina just phoned to tell me his internet died.13:35
=== balloons_ is now known as balloons
=== briancurtin2 is now known as briancurtin
dobeyi guess i need to do the backport branches myself then13:51
mandelgatox, did you have the time to take a look at the deamon MP I sent yesterday?13:54
dobeymandel: there's a comment on it from him :)13:56
gatoxmandel, yes.... did you see my comment?13:57
mandelgatox, nop, looking14:01
mandelgatox, have you done what the command is telling you: 'Error: No developer directory found at /Developer. Run /usr/bin/xcode-select to update the developer directory path.'14:03
gatoxmandel, mmmm that seems to let me change the path..... now i need to find it14:04
mandelgatox, try /Applications/Xcode.app/Contents/Developer14:06
gatoxmandel, ok..... that seems to work.... exxecuted 2 tests with 0 failures14:08
mandelgatox, just 2?14:08
mandelgatox, there are 30 or so, can I see the output?14:08
gatox_macmandel, http://paste.ubuntu.com/994242/14:11
mandelgatox, oh, you read the wrong line, everything is ok: Executed 31 tests, with 0 failures (0 unexpected) in 0.038 (0.046) seconds :)14:12
gatoxmandel, :P sorry14:12
gatoxmandel, +1 then14:12
mandelgatox, thx!14:12
mmccgood morning, everyone.14:21
mmccmandel, I'm going to install xcode 4.3 now and try your branch14:21
mandelmmcc, superb! let me know if everything works ok, you might need to set the path to /Developer like gatox just did14:22
mmcchuh it looks like the project is only expecting one review?  - there's a control tower review pending already14:22
gatoxmmcc, mandel not to /Developer...... to /Applications/Xcode.app/14:23
mandelgatox, yeah, the path is with xcode but back in the day used to be /Developer.. apple has moved things around a little14:25
thisfredmandel: Do we have windows CI running for the desktop+ stuff yet?14:25
thisfredbecause u1db would very much like to freeload on that part ;)14:26
mandelthisfred, what do you mean?14:26
mandelthisfred, I don't know what CI stands for..14:26
thisfredso a jenkins instance or something that checks trunk for failures on windows?14:26
thisfredcontinuous integration14:26
mandelthisfred, oh, yes we do :)14:26
mandelthisfred, we have an ec2 instance running windows with jenkins on it, so we can create a new project for you and get it running whenever you want :)14:27
thisfredmandel: I'll get back to you on that next week. jam has all the windows stuff on his laptop14:27
mandelthisfred, if you give me the steps to create the required env for the tests to be ran I can do it quite fast14:29
mandelthisfred, if you send me an email with it you will get it done by monday :)14:29
* dobey wonders what we're going to do with u1db in c, exactly14:29
thisfreddobey: not just c, the python version needs to be run on windows as well14:30
dobeyyes, but taht's the easy part14:30
thisfredC is handy for embedded stuff, or environments where python is frowned upon/too heavy14:31
dobeyi'm interested in the hard problem that is currently making it hard for me to package u1db :)14:31
=== mhall119_ is now known as mhall119
mmccgatox, mandel, are both of your macs on 10.7?15:00
gatoxmmcc, 10.7.415:00
mandelmmcc, yes15:00
briancurtinthisfred? anyone else?15:02
thisfredstill sprinting, but will be there monday ;)15:03
briancurtinoh yeah, i'll go ahead then15:03
facundobatistarye, ?15:03
briancurtinDONE: fixed up build_installer branch, created new installer for testing and did some testing, had a half day15:03
briancurtinTODO: another half day, testing out this installer15:03
briancurtinBLOCKED: none15:03
briancurtinNEXT: mandel15:03
mandelDONE: Added run-tests script for fsevents project. Added more tests to the code.15:03
mandelTODO: More tests.. ralsina 1-1 (if possible)15:03
mandelBLOCKED: no15:03
mandelmmcc, please15:03
mmcc 15:03
mmccDONE: reviews, fixed lint setup, got sick15:03
mmccTODO: daemon on 10.6, delint my network branch, tackle packaging15:03
mmccBLCK: NO15:03
mmccNEXT: gatox15:03
gatoxWork in the filesystem notification branch, trying to fix the broken tests (the problem are not the tests :P). Review for mandel. Working on Freaky Friday project15:03
gatoxFinish with FF and write a post about it.15:03
mmcc 15:03
gatoxdobey, go15:04
dobeyλ DONE: client 3.0.1 backports/release, weekly call, found ssl error problem, some work on installer bug15:04
dobeyλ TODO: u1cp 3.0.1 backports, finish bug #853060, 1-1 review meeting, 3.0.1 release/SRUs15:04
ubot5Launchpad bug 853060 in ubuntuone-installer "ubuntuone-installer crashed with GError in function(): Failed to execute child process "ubuntuone-control-panel-gtk" (No such file or directory)" [Critical,In progress] https://launchpad.net/bugs/85306015:04
dobeyλ BLCK: none.15:04
dobeyralsina: come back to the land of the living15:05
alecuI missed the standup.15:09
dobeyalecu: go then!15:09
alecuDONE: back deep in security fixes15:11
alecuTODO: more work on it15:11
alecuBLOCKED: no15:11
dobeyand lunch! :)15:14
gatoxlunch for me too15:16
=== gatox is now known as gatox_lunch
* rye is calling too much today15:17
ryeah, alecu piiiiing!15:17
ryei forgot that it does not work in IRC15:18
alecuhola rye!15:19
mandelfuck fuck fuck15:19
mandelalecu, since ralsina is not here I'll tell you, sso was updated to use domain sockets which broke u1-client on windows, I'll have to 'stop' the work on the root daemon to fix that asap15:20
mandelalecu, should not be a huge amount of work..15:20
alecumandel, yup, it makes sense.15:20
alecumandel, there're no "unix domain sockets" on windows, right? :-)15:21
ryealecu: summoning you to #u1-internal15:21
mandelalecu, main problem is that is using the old api (get_port) instead of get_description15:21
mandelalecu, so it is a very simple change to update the use of the api15:22
joshuahooverralsina: any news on bug #1000983 or bug #1000970 in regards to win 3.0.1?15:24
ubot5Launchpad bug 1000983 in Ubuntu One Control Panel "Windows 3.0.1 - Adding folder to sync in Wizard prevents setup from completing" [High,Confirmed] https://launchpad.net/bugs/100098315:24
ubot5Launchpad bug 1000970 in Ubuntu One Control Panel "Windows 3.0.1 - AttributeError: "'NoneType' object has no attribute 'stop'"" [Medium,Confirmed] https://launchpad.net/bugs/100097015:24
alecumandel, we still need to discuss the "buffers full" case in the root daemon. Let's schedule a mumble Monday morning with mmcc.15:24
mandelalecu, ok15:25
mandelalecu, I'm ok after the team standup for example15:25
alecumandel, great.15:25
ryeis it only me - got a weird battery indicator here, vertical one with 3 red (i suppose) lines15:25
alecurye, we don't trust your color perception. Screenshot or it is green!15:26
ryealecu: i don't trust my color perception either, green color perception anomaly15:28
ryehere: http://ubuntuone.com/6fpeJv58XL4MzkULa7aHfX15:28
alecurye, I've used your android app, so I know about your color perception already :-)15:29
alecurye, and... in the screenshot it's green lines :-)15:29
alecurye, and I've never seen that icon before!15:29
* alecu opens the laptop and updates.15:30
ryealecu: i wonder whether i updated something from -proposed... something icon-related15:30
ryealecu: anyway, back to LaTeXilla, Unlinks() after restart while the SD was idle before the restart - film at 1115:31
mmccalecu, mandel, mumble after monday standup ok.15:38
mandelmmcc, sweet! we have a date then :P15:41
briancurtinjoshuahoover: ralsina is not around today, his internet is out15:46
joshuahooverbriancurtin: ah, too bad :(15:46
briancurtinjoshuahoover: i will take a look at these before i'm out for my half day15:47
joshuahooverbriancurtin: thanks15:47
joshuahooverbriancurtin: upon further testing, these 2 bugs appear to be in 3.0.0 as well...so it's not something we introduced with this latest test release15:47
briancurtinjoshuahoover: ah, on 1000983 i wonder if that's related to the hack we currently have in there to even display the cloud sync page. it used to just sit forever at "getting information", but we ship with a small branch that comments out a line to allow it to display15:49
briancurtinit used to just hang forever and never show up15:50
ralsinaI am back15:54
ralsinajoshuahoover: I am working on fixes, they will be done late today15:54
ralsinajoshuahoover, briancurtin: if they are already in 3.0.0 I would ship without the fixes anyway15:55
joshuahooverralsina: yeah, both bugs are in 3.0.0 on windows15:59
mmccmandel - TIL your daemon won't build on OS X 10.6 unless we remove ARC. you can build on 10.7 and deploy on 10.6 but no building with ARC on 10.6...15:59
ralsinaso we move forward.15:59
mandelmmcc, so, lets make that a bug that I'll fix :)15:59
joshuahooverralsina: let's target those 2 bugs and the installer improvements briancurtin is doing for the next release, sound good?16:00
mandelralsina, in theory, we have to have a 1-1 today..16:00
ralsinajoshuahoover: yeah16:00
ralsinamandel: let's16:00
mmccmandel, OR we can fix it by me buying a laptop running 10.7, which I need to do eventually anyway.16:00
ralsinamandel: give me time to find my headset16:00
mandelralsina, sure, launching mumble16:00
mandelmmcc, we need to evaluate the need of ARC.. there are some little tricks I wanted to do with blocks and ARC would not let me :(16:01
mandelmmcc, but you do have a great excuse to buy a new laptop :)16:02
mmccmandel, ok - I'm curious what you wanted to do that arc complained about...16:02
ralsinamandel: I'm there now, can you hear me?16:02
mmccmandel I was hoping to delay the laptop until they release a new generation :)16:02
mandelmmcc, let me have the 1-1 and will show you, is very stupid :)16:03
mandelralsina, I heard you... but not anymore16:03
mmccmandel ok you have my permission to have your 1-116:03
=== gatox_lunch is now known as gatox
dobeyralsina: oh, you have returned16:19
* joshuahoover is amazed at the power mmcc has amassed in such a short period of time on the team ;)16:19
ralsinajoshuahoover: that's what happens when you hire from oracle16:19
mmccthere's a joke about support contracts in there somewhere...16:20
dobeyjoshuahoover, ralsina, briancurtin: what changes exactly are in this candidate installer, which aren't in 3.0.0?16:20
* ralsina points at briancurtin16:21
briancurtindobey: i'll link you to the branches in a second, but there are two branches merged which are not a part of stable 3.0, and only one of them is new for this 3.0.1 installer16:21
=== salgado is now known as salgado-lunch
briancurtindobey: https://code.launchpad.net/~brian.curtin/ubuntuone-control-panel/c2c-loading was included on 3.0.0 and 3.0.116:22
briancurtindobey: https://code.launchpad.net/~brian.curtin/ubuntuone-client/fix984407-windows-is_root is included for 3.0.116:22
dobeyand that c2c-loading one is hackiness related to bug #1000983 ?16:24
ubot5Launchpad bug 1000983 in Ubuntu One Control Panel "Windows 3.0.1 - Adding folder to sync in Wizard prevents setup from completing" [High,Confirmed] https://launchpad.net/bugs/100098316:24
briancurtindobey: i havent had time to debug it and i'm leaving soon, but i think its possible. something in that refresh_volumes call that is now commented out might be screwing it up16:25
mandelmmcc, so an example of what I wanted to do and you cant with ARC is the following: http://paste.ubuntu.com/994465/16:26
dobeybriancurtin: and prior to that change, it just hangs on windows with "Getting information" ?16:26
mandelmmcc, for some reason that with ARC ran in Xcode works ok but if ran wit octest gets a segfault.. I think the compiler is doing some screwed up release/retain somewhere16:27
* mandel refuses to believe it is a compiler bug16:27
briancurtindobey: yep. it would just hang forever and you would just have to hit the next button and ignore it16:27
dobeyi had that happen *once* to me on linux16:27
mandelmmcc,  I might have been abusing the blocks idea :)16:28
dobeyralsina: when should we do the 1-1 review meeting?16:28
mmccmandel, are you getting a segfault with this pasted snippet or the current branch, ?16:28
ralsinadobey: let me get the papers and we'll do it right now16:28
mandelmmcc, I was getting it with what is in lp:ubuntuone-fsevents-daemon and octest command16:29
mmccmandel, and with or without ARC? I thought you had octest working on the branch?16:30
mandelmmcc, on xcode works with the octest command-line does not..16:30
mandelmmcc, is weird..16:30
mandelmmcc, way I worked around it: remove the use of the block16:31
ralsinadobey: whenever you are ready16:31
mmccmandel: yeah that's strange. I don't know much about octest - I actually thought xcode just ran octest and parsed its output16:32
mandelmmcc, me too, but I don't think is something we should worry about atm16:33
mandelmmcc, lets just not do that and will remove ARC once the test coverage is there16:33
dobeyralsina: ah, did you not see my /msg ? :)16:33
mmccmandel: ok. sounds good16:33
ralsinadobey: sorry, now I did :-)16:34
gatoxralsina, (FF related) i'm publishing files and getting the link and showing it to the user in the ui, so the user can share the link \o/... i'm doing some ui and performance improves now16:39
ralsinagatox: yay16:39
gatoxplus the search filter that the ui already has16:39
alecugatox: in what app are you doing all that link sharing?16:40
gatoxalecu, in the freaky friday project.... i'm extending ralsina's u1-toys16:41
gatoxalecu, let me show you the screenshots16:41
gatoxalecu, ralsina when you use the app to search for a file in your u1 or shares folder, you have something like this: http://ubuntuone.com/1LuX0iMBBeWvdcp0oCzy0t16:42
gatoxalecu, ralsina after you press publish, you havve this: http://yfrog.com/z/1qi9xp16:42
alecugatox: awesome!16:43
* alecu is running the u1-toys from trunk.16:43
alecuralsina, gatox: you guys may want to remove the print of the credentials to stdout.... :-) The line that says: "creds_found" is very very wrong :-)16:44
gatoxalecu, ack16:44
gatoxalecu, to use the publish file..... you should run from my branch16:45
ralsinaalecu: hehe16:45
ralsinaalecu: 1-1 a las 2 y media?16:46
alecuralsina, I'm supposed to help my wife with some errands just after lunch....16:46
alecuralsina, can we schedule it a bit later?16:46
ralsinaalecu: I will be around until 6, just ping me16:47
alecuralsina, awesome, thanks.16:47
alecugatox: I love the idea of publishing the files. How are you getting the list of available files?16:48
gatoxalecu, i'm using the folders interface to get the folders, then i do a walk inside those folders, grab the files, and create a dict that the filter handle to show suggestion while the user is typing the file name16:49
alecugatox: also, here's a proposal for a slightly different ui: Once you click on Publish, the dialog closes. And when you get the result, you show a baloon notification that says "The published url was copied to the clipboard"16:49
alecugatox: "folders interface" -> awesome.16:50
gatoxalecu, i'm using the dbus interface for everything, getting the folders, file metadata, publishing files.....16:50
gatoxalecu, yes.... i could do that.... but i wasn't sure about removing the content of the clipboard just like that.....16:51
mmccgatox, very cool! it'd also be neat to make a way to drag a file you want to share, to avoid searching for it if you already see it -- but is it weird to drag a file to a menu item? maybe...16:51
gatoxalecu, now you can select the link below the search box.... and copy that.... i'm going to improve that in some way.....16:52
gatoxmmcc, na..... that would be easy16:52
alecummcc, probably dragging a file to the u1 control panel launcher.16:52
gatoxok..... i'm going to improve some stuff.... and see about the dragging tihng16:53
mmccalecu: that'd work! if it's already in your u1 synced folders, you get a public link, otherwise, maybe it gives you some choice of what to do with the file (copy to synced folder, sync its current folder, etc)16:54
* mmcc thinks maybe osx client should have a dock icon after all16:54
gatoxmmm good ideas.......16:55
* gatox take notes16:56
* dobey thinks we need to refresh our icons and move them around a bit16:56
* alecu will have lunch now. And then run some errands.16:59
=== alecu is now known as alecu_lunch
ralsinagatox, mmcc: upload and publish without copying it anywhere...17:02
gatoxralsina, sorry.... what?17:02
ralsinagatox: file in a non-synced folder. You upload it via the REST API and publish it17:03
gatoxralsina, ahhhhhhh17:03
gatoxok.... adding that feature too to the list....17:03
dobeywhat are you doing that in?17:03
gatoxdobey, me?17:05
* mmcc going for lunch17:05
dobeygatox: you, ralsina, alecu, whoever is doing it? :)17:06
dobeytrying to follow the backlog about this publishing stuff, but not clear what you're talking about exactly17:06
gatoxdobey, you mean in which project?? no one in this moment..... it's a separate thing for a freaky friday project.... using some code from ralsina17:06
ralsinadobey: it's pyqt with the idea that it's a proof of concept17:10
ralsinadobey: to someday rewrite either in vala or C++17:11
dobey^^ someone review that :)17:11
ralsinadobey: looking...17:11
=== teknico_ is now known as teknico
dobeyactually, let me get one more in there, that's not as big a change as i thought17:13
ralsinadobey: had approved, moved back to needs review, hopefully tarmac was not that quick ;-)17:15
dobeystuck the unstyle the spinboxes change in as well17:16
dobeyman, amazon prime is awesome17:18
dobeyguaranteed 2-day delivery for free, turned into overnight; for something i ordered last night after eod17:20
mandelok, my time to enjoy the weekend17:23
dobeyralsina: i set it back to approved now. as soon as it lands i'll get it released. and will try to get all the other tarballs released asap as well17:23
mandelall, have a great one!!!17:23
ralsinadobey: awesome17:23
dobeymandel: isn't it like 21:30 there?17:23
ralsinamandel: good weekwnd!17:23
mandeldobey, is just 7:30 and since we are close to summer it looks like 517:23
gatoxmandel, are you running cp from trunk?17:23
gatoxmandel, weird about the focus border17:24
mandelgatox, no AFAIK17:24
dobeyoh right, 19:3017:24
mandelgatox, yes, is weird, yours is ugly :P17:24
gatoxmandel, mine is something that it has been there for a while....17:24
mandelgatox, not for me..17:25
gatoxralsina, the buttons in u1-cp still has the border line in focus?? or i'm out of date.... or mandel's....... :P17:25
ralsinagatox: I'm onwindows, so, no idea17:25
ralsinagatox: I think they should17:25
mandelralsina, look at or twitter conversation for reference, we have screenshots17:25
gatoxralsina, yes, they should17:26
ralsinamandel: go eat tapas17:26
ralsinamandel: fight a bull17:26
gatoxralsina, i can see it..... but mandel says that he has something else17:26
ralsinamandel: dance a sevillana ;-)17:26
ralsinamandel: rest, too17:26
mandelralsina, better to dance with a sevillana hehehhe17:26
mandelok, I'm off, see you laters17:26
ralsinamandel: I danced with a sevillana, felt cut off ;-)17:26
gatoxmandel, bye!! enjoy your weekend17:26
dobeyoh bugger17:38
ralsinadobey: http://www.engadget.com/2012/05/16/samsung-considering-13-inch-1080p-IPS-display/18:01
dobeyralsina: https://www2.panasonic.com/webapp/wcs/stores/servlet/prModelDetail?storeId=11301&catalogId=13251&itemId=664010&modelNo=Content01092012023841959&surfModel=Content0109201202384195918:03
ralsinadobey: that's a prototype for CES, this is something samsung is considering putting in next year's line18:04
dobeyralsina: and it's still not especially amazing18:06
ralsinadobey: it's incrementlly amazinger18:06
dobeymake it at least a 10" display, then we can talk :)18:06
dobey200dpi or GTFO. :P18:06
ralsinadobey: do an iphone4 mosaic18:07
dobeyif i wanted to make a multi-screen laptop, i'd just go with a bunch of the screens in these: http://www.siliconmicrodisplay.com/st1080.html18:09
dobey3000 dpi ftw.18:09
dobey1080p at 0.75"18:10
ralsinadobey: use analog pictures. 70mm film is like 50MP and you can project it in a 3' screen :-)18:11
ralsinaof course you would have to draw UI with a crayon and then film it. But hey, the TEXTURE18:12
dobeyso dependencies changing early on is pain.18:14
dobeyi think i'll just land by hand18:20
ralsinadobey: should we have a separate tarmac or stable-3-0?18:24
dobeyralsina: we need to make a quantal one now; but it's a bit of setup work to do; started a branch to do that in the background18:25
ralsinadobey: quantal is going to be massively broke for a while, is it worth doing at this stage?18:26
dobeyhow do you mean?18:26
dobeythe only thing that's broken for us on quantal is new gcc broke the build of libu118:27
dobeywhich i need to fix and isn't terribly hard, but can't test the fix easily on precise18:27
=== alecu_lunch is now known as alecu
* dobey uses a big hammer for now18:34
=== salgado-lunch is now known as salgado
* mmcc finally back after the world's slowest waiter extended my lunch19:33
mmcchey, if I file a bug that I know should be assigned to us, but I don't know which one, I should assign it to 'ubuntuone-desktop+'? I just saw that rick went thru and did that to 3 recent bugs I filed, and wondered what I need to set so I don't create extra work...19:35
dobeyyou can assign bugs to the team, sure19:36
rmcbridemmcc: Much appreciated. If you know the team for it, like ubuntuone-desktop+ you can go ahead and assign. QA does triage passes each week (elopio and I alternate) so it will get handled if you don't know where to put it19:36
* dobey doesn't really see the point though19:36
dobeysigh. bad evil tests that are locale dependent19:38
mmccrmcbride: what do I have to fill in so you don't have to triage it? just assign it?19:39
dobeymmcc: set status to triaged and assign to a team19:39
ralsinammcc: you can always ust ask me since I assign to individual devs :-)19:39
rmcbridemmcc: pretty much what they said. QA only really looks at  the "NEW" queue19:40
ralsinammcc: which bug is it?19:40
dobeyjust assign bugs to your boss. like a boss.19:40
rmcbrideralsina: there were several bugs of his that I kicked to desktop during my triage pass a little while ago he just happened to notice and asked about it19:40
ralsinarmcbride: ack19:40
mmccok, thanks. ralsina, it's bug 1000061 and bug 1000059... and one other. about UI in control panel when there's no network connection (it's broken)19:41
ubot5Launchpad bug 1000061 in Ubuntu One Control Panel "with no network connection, control panel "account information" tab shows inscrutable error dialog box that doesn't seem to have an effect " [Undecided,Confirmed] https://launchpad.net/bugs/100006119:41
ralsinarmcbride: "triaged" is supposed to mean "the manager has assigned resources to fix it" so could you guys switch to "confirmed"?19:41
ubot5Launchpad bug 1000059 in Ubuntu One Control Panel "with no network connection, control panel allows "disconnect" and attempted "connect" for filesync but gives no explanation " [Undecided,Confirmed] https://launchpad.net/bugs/100005919:41
rmcbrideralsina: yes19:41
ralsinarmcbride: thanks!19:41
rmcbrideralsina: we discussed that last week, I've only been usign confirmed this pass19:41
rmcbride(unless someone already had somethign as triaged)19:42
ralsinarmcbride: it's ok for existing bugs, I want to be the one moving them from confirmed to triaged and from the team to a dev19:42
ralsinaand then HOPE the dev moves it to fix-committed ;-)19:42
mmccralsina, rmcbride - a suggestion that might help new folks like me - it'd be good to have a doc somewhere that says what we use those status tags for and who is supposed to set them.19:42
rmcbrideralsina: I have what I will call an "alternative" understanding of those values, but all is sorted and I won't set things to triaged anymotre19:42
ralsinammcc: oh, there is one19:43
rmcbridemmcc: I think joshuahoover has such a document somewhere19:43
joshuahoovermmcc: let me get you the link...one moment19:43
ralsinagatox: re bug #1000061 we are now checking network availability on u1cp startup right?19:47
ubot5Launchpad bug 1000061 in Ubuntu One Control Panel "with no network connection, control panel "account information" tab shows inscrutable error dialog box that doesn't seem to have an effect " [Undecided,Confirmed] https://launchpad.net/bugs/100006119:47
gatoxralsina, yes19:48
gatoxralsina, ah no19:48
joshuahoovermmcc: sorry, got a bit sidetracked...here's the link you want...if something doesn't make sense to you, let me know...it's quite possible it's out-of-date or just makes no sense at all - https://wiki.ubuntu.com/UbuntuOne/Bugs/WorkFlow#Status19:49
gatoxralsina, we are checking for network connection if the user try to create a new account or login with control panel19:49
mmccjoshuahoover: no prob. thanks!19:49
* mmcc should really just read the whole wiki19:49
gatoxralsina, we are not checking for network connection in that situation19:50
gatoxjust in the wizard19:50
mmccgatox: ralsina - aha - my situation was that I'd logged in OK, but then cut the network and saw weird behavior19:50
gatoxralsina, you can assign that to me if you want, and i can connect a signal to notify the user if we lost connectoin and disable the ui19:51
gatoxor something like that19:51
gatoxi don't know which is the design guideline for this19:51
ralsinagatox: it's tricky because some things work without network, but most don't19:51
mmccbtw, I filed 3 related bugs - if you just disable everything then all 3 bugs are really one bug19:52
ralsinagatox: and u1 is an app that *has* a disconnected operation mode19:52
ralsinagatox: so we need to think on how to degrade gracefully19:52
ralsinagatox: example: remember the user's name19:52
gatoxralsina, i understand19:53
gatoxralsina, btw..... if you want to test/merge my branch of u1-toys..... is fully functional..... i made several improves in the files filtering list, everything is working.... i'll add later some features about drag&drop as mmcc and alecu suggest.... but it's working and it's nice! :P19:54
gatox(performance improves in the filtering list)19:55
gatoxand i add some osd notifications too19:55
mmcctesting q: is there a nicer way of patching a function with something that just raises an exception? right now I'm doing this : http://paste.ubuntu.com/994770/19:57
mmcc 19:57
mmccbut that seems... wordy, what with the docstring and all19:57
mmccer, there's a bug in that paste. was mid edit. but my question is the same.19:58
gatoxmmcc, mmm i usually do that19:58
mmccgatox: ok, cool. Was just wondering if there was a method that wrapped that pattern that I didn't know about19:59
gatoxmmcc, i'm not saying there isn't...... maybe it exists and i don't know it either :P19:59
gatoxok..... eod here...... time to start the ninja-ide coding for this weekend! :P see you!20:01
dobeymmcc: lambda?20:03
gatoxdobey, you can't do lambda: raise20:03
dobeyor i guess you can't raise in a lambda?20:03
mmccdobey, I used e.g. "lambda True" in other spots but... yep20:03
ralsinagatox: gimme!20:04
gatoxralsina, lp:~diegosarmentero/+junk/u1-toys20:05
ralsinammcc: that docstring is not compliant. Put it all in 1 line, including a full stop before the closing quotes.20:05
dobeymmcc: well you can define raise_exc(exc_class): raise exc_class(); and do lambda: raiss_exc(Exception) or something no?20:05
ralsinammcc:  and the docstring for the inner function should use triple quotes too20:05
ralsinammcc: and +1 to using a lambda20:06
dobeyor i guess lambda _: raise_exc(Foo)20:06
dobeybut eh20:06
dobeystill fairly wordy20:06
ralsinadobey: no argument needed for the lambda here20:07
ralsinadobey: so lambda: raise_exc( FooException)20:07
mmccralsina: ack on the docstrings. so, always triple quotes, fit on one line if possible?20:08
ralsinammcc: there is a PEP for that20:08
ralsinammcc: the 1st line is always a full descriptive sentence20:09
ralsinammcc: if you need more, full descriptive sentence, blank, long description, EOL, closing quotes20:09
* mmcc goes to look at the PEP20:09
ralsinammcc: also lambda x: exec('raise(Exception(x))')20:10
ralsinabut gack20:10
* ralsina will reject any branches with exec or eval in them unless properly bribed20:10
mmccralsina: if this is a common pattern, might be nice to add a method to TestCase, like e.g. def patchWithRaise(mod, member, ExceptionClass)20:12
ralsinammcc: have not seen it much20:12
mmccralsina: ok then.20:12
mmccanother lint question - pylint wants all module-level variables NAMED_LIKE_THIS, apparently...20:13
ralsinammcc: YES HE DOES20:13
ralsinammcc: or YES_HE_DOES20:13
dobeyour pylintrc does20:13
mmccbut I did a bunch of setting up ctypes wrappers that were named closely to the C function they wrap... so like "CFRunLoopGetCurrent" and (egads) "SystemConfigurationNetworkReachabilityScheduleWithRunLoop"20:14
mmccok, I abbreviated the last one20:14
dobeywhee, 4 (+1) more tarballs to go20:15
ralsinammcc: feel free to sprinkle liberal doses of #pylint ignore: C060320:15
ralsinammcc: same thing happens when extending Qt WhichUsesCamelCase20:16
mmccC0103... ok, if there's precedent to that then I'm happy pylint-ignoring.20:16
dobeymust replace pylint20:18
mmccone more style question: I have a couple of method names with underscore prefixes, because they're not intended to be public, but the test code calls them, which lint doesn't like. OK to #ignore those too?20:20
rmcbridedobey: replacement for pylint should be pyreallynastyfelt20:21
rmcbridemade from compressed and starched lint. ew.20:21
rmcbride"it's made from hemp"20:22
dobeyit's light and breathable20:22
mmcccomfy cool and breezy, +1 pylinen20:22
dobeymmcc: # pylint: ignore= usage is basically at your choosing20:23
mmccdobey, ok thanks20:24
* mmcc goes to put #pylint ignore=all at the top of all his files20:24
* rmcbride looks forward to seeing that mp get reviewed20:25
rmcbridenot that I haven't considered doing that myself at times20:25
mmccanyone know of a way to have # pylint ignore apply to multiple lines? The documentation on the pylint website is ... could be better20:29
dobeymmcc: it does automatically, if they are in the same scope20:30
mmccdobey: oh, well what do you know. thanks20:30
dobeyand you should # pylint: enable=<the things you disabled>, after the block of code you're ignoring them in20:31
ralsinammcc: if you go too crazy on the ignores it will be called out on review20:37
ralsinahe, ok, only now finished reading the backlog20:37
mmccralsina: understood :) trying to use sparingly here.20:37
ralsinammcc: the _ thing should not happen20:37
mmccdobey, ack on the reenable20:37
ralsinammcc: if you are calling it from outside the class, it should not have _ in the name20:38
ralsinammcc: unless it's only called in a test and you have good reason20:38
mmccralsina: right, it's intended to be private, hence the _. but the tests have to call it...20:38
ralsinammcc: k then, ignore it20:38
dobeyralsina: do we need to backport any of these 3 trunk branches on windows-installer to stable-3-0?20:40
ralsinadobey: which ones?20:40
dobeythe last 3 revisions on lp:ubuntuone-windows-installer20:41
dobeyi guess probably no for the mac-env branch from diego20:41
dobeybut less sure about the other 220:41
* ralsina looks20:42
dobeyhttps://code.launchpad.net/~ubuntuone-control-tower/ubuntuone-windows-installer/trunk for browsing easiness20:42
ralsinadobey: since we don't *ship* that, it doesn't matter, really20:43
ralsinadobey: feel free to merge them in 3.0.2 or whatever20:43
go8765Hello. What command run ubuntuone-client-gnome ?20:43
ralsinadobey: or never. I don't know if stable-3-0 even means aything for -installer20:43
ralsinago8765: ubuntuone-client-gnome installs the options in the right click menu in Nautlius, it's not a standalone program20:44
dobeygo8765: i don't quite understand the question. ubuntuone-client-gnome is a couple of plug-ins for GNOME20:44
go8765how i can use ubuntuone in xfce4?20:44
go8765is any client for this?20:45
go8765*with tray integration20:45
dobeyralsina: well, stable-3-0 there makes sense if it makes sense to release tarballs, and we're going to keep it around, and we're going to have separate stable/unstable builds and such20:45
dobeygo8765: sudo apt-get install ubuntuone-control-panel-qt20:45
ralsinago8765: start ubuntuone-control-panel-qt --with-icon20:45
ralsinadobey: we'll keep it around, but it's mostly a set of scripts to build installers and setup dev. envs nowadays20:46
ralsinadobey: plus, the buildout there is being used on mac :-)20:46
ralsinadobey: maybe rename it ubuntuone-build-tools or something20:47
dobeyright. so the name is wrong, again. :)20:47
ralsinadobey: names are so labeling20:47
go8765hm... so only qt version?20:47
ralsinago8765: yes, on precise, only qt20:47
dobeygo8765: you can just not use the control panel if you don't want the qt bits20:47
go8765ok. but why he said that he make sinhronization, but not use traffic?20:47
ralsinaok, gotta go, will work late tonight, so mail me review requests and whatnot. Good weekend people!20:50
dobeygo8765: i'm not sure i understand what you're asking there.20:53
dobeyralsina: do i even need to bother making a 3.0.1 for windows-installer then?20:54
go8765dobey: I mean that Im not shure that sinhronization work good( I launch ubuntuone-qt client . HeIt show that sinhtonization start but in real - dont shure that it end.21:01
dobeywhat does it say in the control panel?21:02
go8765dobey: http://storage7.static.itmages.com/i/12/0519/h_1337375130_3751240_1015ab0f31.png21:05
go8765It said that it start, but I dont see that it use internet connection now21:06
dobeyit might be doing local rescan or something. i can't tell from here21:07
dobeygo8765: you can "tail -f ~/.cache/ubuntuone/log/syncdaemon.log" in a terminal to get more detailed info21:07
dobeyit is possible that you are also hitting a bug for which we are about to do an SRU, as well21:08
go8765dobey: http://paste.ubuntu.com/994895/21:08
dobeygo8765: oh, hrmm. looks like your hard drive is full?21:09
go8765dobey: thks)21:09
go8765its right)21:09
go8765i forgot21:09
=== m_conley is now known as m_conley_away
dobeyalright. need to run. may be back later to finish up a couple things. have a good weekend all21:17
* alecu EOWs22:56
alecuhave a nice weekend, all!22:56

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