tntcare playlists still broken on android?02:27
duanedesigntntc: i am not sure. Most the people who could help answer that are not on right now02:37
=== m_conley_away is now known as m_conley
=== m_conley is now known as m_conley_away
=== m_conley_away is now known as m_conley
=== m_conley is now known as m_conley_away
Tbruff13hey can someone help me i cannot paste pictures into my ubuntu one folder why when i try to move them over from another folder they say permission denied03:18
duanedesignthat is odd03:20
duanedesignTbruff13: is the folder you are moving from in your HOme directory?03:20
Tbruff13it is my Home03:21
Tbruff13if i am root it works hey how long after they are in ubuntu one folder does it take to put them online03:23
duanedesignTbruff13: sounds like their is a permission issue03:34
duanedesignTbruff13: i would run the Terminal commands:  ls -la ~/Ubuntu\ One03:35
duanedesignand run the same command on the folder that has the pictures03:35
duanedesignshould look like:03:36
duanedesign-rw-r--r--   1 duanedesign duanedesign   77789 2010-08-02 11:17 ubuntu_reference.pdf03:36
duanedesignwith your username03:36
Tbruff13duanedesign, i am ok03:36
duanedesignTbruff13: the files should start syncing right away.  How long it takes after it starts depends on size and nmber of files03:37
Tbruff13drwxrwxr-x  2 lindseyjacks lindseyjacks  4096 2011-08-03 15:08 .03:40
Tbruff13drwxr-xr-x 35 lindseyjacks lindseyjacks 12288 2011-08-04 22:05 ..03:40
Tbruff13lrwxrwxrwx  1 lindseyjacks lindseyjacks    48 2011-08-03 15:08 Shared With Me ->03:40
Tbruff13i am on my  friends laptop03:40
Tbruff13hey duanedesign i have an idea be back03:42
duanedesignthat folder looks good. Must be the other folder03:57
rogerk_Anyone know where I can find the logs for the windows client?07:20
JamesTaitGood morning everyone!08:16
feastyDoes anyone know why my UbuntuOne account would be showing that I have 74GB of data sync'd when my local machine (using du) only says I have 64GB of data?08:27
feastyIt's ok, I found it on the FAQ. It's a known issue.08:50
mandelHello!1 sorry for the late arrival, internet issues at home :(09:20
mandelduanedesign: forgot to tell you, wellcom!09:20
faganoh welcome back to europe mandel09:56
mandelfagan: hello, took a while  but I'm back09:57
mandelcrazy airplanes...09:57
faganmandel: well its argentina even if the flights were all on time it still would have taken nearly a day of travel10:00
gatoxhi mandel10:04
faganMorning gatox10:04
mandelgatox: ha, I wa looking for a diff nickname… you changed it?10:05
mandelfagan:  yeah, but I ended up in an airport over 10 hours… major pain in the ass10:05
faganmandel: yeah that sounds bad :/10:05
faganI would just go drinking or shopping or something10:05
mandelgatox: do you know what is the problem with the RO shares? I saw a mail thread about it but I have no idea of what is going on10:06
mandelfagan: yeah.. but I had not much money or energy to do anything10:07
gatoxmandel, me neither.... :S10:07
gatoxmandel, are you talking about roberto last night e-mail?10:08
faganWow I really dont have any look with Ubuntu on my computers at the moment10:16
faganubiquity keeps crashing on the only disk I have in the house10:16
faganMight go down to buy a usb stick so I can install it10:17
faganActually I will be back in a few minutes10:17
* mandel back110:24
Chipacamandel: no you're not, we can't read what you type10:24
mandelChipaca: you cannot?10:27
mandelreally :P10:27
mandelChipaca: on question, do you know what is the bug that nessita has regarding RO shares?10:28
Chipacamandel: not beyond what was in the email10:29
Chipacamandel: so i can't answer your questions from your email, either10:29
mandelChipaca: :(10:31
mandelChipaca: 'cause we might be trying to do more than we need, since RO is something we create and no the user10:32
Chipacamandel: yeah, and we ignore file properties anyway, so ... dunno10:35
mandelChipaca: I'll wait to talk with nessita, right now we are simply expeculating.. at some point our os layer is going to be bigger than sd itself :(10:36
ralsinaGood morning!11:06
ralsinamandel: what's that you replied to alecu about txnamedpipes? Is that an ancient mail that just got sent?11:07
faganmorning ralsina ill have that result in about half an hour ubiquity kept crashing on the CD I was trying to reinstall from so got a usb stick to try a later version of ubuntu11:10
ralsinafagan: ok11:11
ralsinagotta take the kid to school, will be back in 50 minutes11:11
faganralsina: cool should be ready by the time you get back11:11
mandelralsina: is probably and old email since I have not sent anything that I know of...11:20
mandelralsina: by the way, what is the problem with the RO shares, is this just because of the test dir not being deleted?11:21
mandelralsina: because, if it is that they are looking to the completely wrong direction11:23
mandelcan_wrtie is the issue, no the attr11:23
duanedesignmandel, thank you!11:46
faganWow this week went by really fast11:49
faganit feels like tuesday11:50
* fagan really needs to get some lunch but will wait till after standup since I have to go out and get some lunch type food 11:56
Balaji_TVRMy UbuntuOne account is not connecting at all. I get this error >>>'NoneType' object has no attribute 'makefile'<<<. Any help on how to fix this?11:58
faganBalaji_TVR: im pretty sure thats in the FAQ ill go grab a link11:59
faganBalaji_TVR: are you behind a proxy server?12:00
faganBalaji_TVR: yeah we dont support proxies yet12:00
faganBalaji_TVR: its on the list of things to do but its a complicated feature12:00
Balaji_TVRI see.12:01
Balaji_TVREven from home, If is set my connection to "Direct connection" (ie, no proxy), still its not honoring my proxy settings.12:02
Balaji_TVRso, without proxy also not working :(12:02
Balaji_TVRSorry, I'm going for some meeting will go thru ur link for more details. Thanks for your input.12:03
nessitahello everyone12:05
faganhey nessita12:05
nessitahi fagan!12:05
gatoxnessita, hi12:08
faganI completely ballsed up my Ubuntu install12:08
faganLost my VM12:08
faganI can set one up pretty fast to test on though since ive done it so many times12:09
faganI had an image locally that I had as a backup for my VM but didnt copy it properly :/12:10
faganOk so ill have my vm up after lunch12:13
faganthats not so bad12:13
tntcfagan: what are you using for virtualization?12:14
tntcWhat's the deal with the playlists? Any idea what caused the server error that's preventing some users from syncing?12:21
fagantntc: virtual box12:24
tntcfagan: no snapshots?12:24
fagantntc: I had one to use but what happened was I was running into a weird ubiquity bug that stopped the install so I had to do a clean install instead but messed up backing up the files12:25
tntcyikes! That sucks.12:25
faganits cool just a little set back12:25
tntcI keep bouncing back to windows because my friends want to game, but I really really dislike it.12:26
faganI havent heard of any issues with playlists maybe duanedesign can diagnose whats going on12:26
tntcoh, yeah it's a known outage or something I think12:27
tntcI just can't save them12:27
fagantntc: I dont have a clue12:31
tntcOh, no biggie! duanedesign is the one to ask?12:35
fagantntc: well he could bring it further than I could :)12:36
mandelnessita: I believe  my connectin should be ok now, do you mind getting in mumble a sec to test it?12:43
nessitamandel: can we do it after the stand up?12:49
alecuhello #ubuntuone!12:49
mandelnessita: sure12:49
alecunessita, mandel: what are you doing after the standup?12:49
* alecu needs to run for 30' after the standup12:50
nessitaalecu: mandel wants to try if his mumble works12:50
mandelnessita: when do we have mumble, is it a .45 after the stand up, right?12:50
alecuah, ok.12:50
nessitaalecu: will you make it for the mumble?12:50
nessitamandel: yes12:50
nessitamandel: we can do it later if you need12:50
alecunessita, I think I will, yes.12:50
mandelnessita: time is ok for me, I'm more interested in testing the connection, that is all :)12:50
faganOk windows installed and python + pywin32 installed so im ready to test that script I think12:51
mandelnessita: lets test the connection 10 mins before the mumble, is that ok?12:51
mandelalecu: I'm going to ignore the RO issues, yet take a look at this: http://msdn.microsoft.com/en-us/library/aa446637(v=vs.85).aspx12:54
mandelalecu: take a look at DisplayAccessMask implementation12:54
alecumandel, let's see...12:55
faganOk the script is working as expected I think12:56
faganIll pastebin it12:56
mandelalecu: ideally that is how access and can_write should be working (not at the moment) yet I don't know the tests you guys did wih set_readonly and set_readwrite12:56
alecumandel, I tested it from the python prompt12:57
alecumandel, first, I imported os_helper12:57
alecumandel, then I used with some freshly created files and folders on disk12:57
alecumandel, first I tried on the E: drive, but since it was mounted by VBox, it didn't work right (it's like samba)12:58
alecumandel, then I tried on the C: drive, that is fat12:58
alecumandel, and that's how I got to see the different authorizations that are detailed on the mail12:58
faganI added in a couple of line breaks to make it more readable but you get the idea12:59
faganalecu: ^12:59
faganralsina: On that script you need to add a import os but other than that its working as expected I think13:00
mandelalecu: yes, we do not have a roll back to the setting, yet that is a problem with the default setting we use right? why would we want to have a roll back to previous attrs?13:00
ralsinafagan: I need the outputm though13:00
faganralsina: I pastebinned it look above13:00
alecume too13:01
faganIll go13:01
fagan* Found and reported a bug in Ubiquity that was blocking me actually testing dir script13:01
fagan* More of that report (still not done but is complicated so im taking my time on it)13:01
fagan* Finish the report13:01
fagan* What ever anyone else needs13:01
fagan* not now13:01
faganOh and tested the script forgot about that13:01
gatoxInstaller Page 19 UI complete. New "Sign in or Sing Up" UI style complete.13:01
gatoxFinish Installer Page 18. Fix UI bugs once Page 18 and 19 are complete.13:01
gatoxnot yet... :P13:01
gatoxmandel, go13:01
mandelDONE: Fixed remarks for https://code.launchpad.net/~mandel/ubuntuone-client/fix-file-shelf-windows/+merge/70021 worked on getting fsm pass all tests on windows by skipping old metadata tests. Realized that due to RO issues more tests should be skipped until that is fixed13:02
mandelTODO: skip tests on test_fsm that depend on ro shares and state the reason. Find a correct working implementation for access and can_write that takes file attr into account.13:02
mandelBLOCKED: no, but internet seems to suck big time.13:02
mandelCOMMENT: I'll be getting back to the morning report emails on Monday, I just realized I forgot to do it.13:02
mandelnessita, please!13:02
nessitaDONE: epic branch fixes and finally landed, UI improvements to control panel, reviews13:02
nessitaTODO: more reviews, more UI improvements13:02
nessitaBLOCKED: nopes13:02
nessitaNOTE: I miss the freaky friday :-(13:02
nessitaNEXT: alecu13:02
alecuDONE: done huge review; worked on bug #820674: branch for review: https://code.launchpad.net/~alecu/ubuntuone-client/fix-add-watch-raceconditions/+merge/7055913:02
alecuTODO: review permissions and R/O with mandel, work on new issues with bug #80367213:02
alecuBLOCKED: no13:02
alecuNEXT: ralsina13:02
ubot4Launchpad bug 820674 in ubuntuone-client "The second call to add_watch should return the same deferred that the first returns (affects: 1) (heat: 382)" [Undecided,New] https://launchpad.net/bugs/82067413:02
ubot4Launchpad bug 803672 in ubuntuone-client (Ubuntu) (and 2 other projects) "Do not start two instances of this process (affects: 2) (heat: 22)" [Undecided,Invalid] https://launchpad.net/bugs/80367213:02
nessitaralsina: go?13:04
fagandobey: go id say13:04
ralsinafagan: looks nice, thanks!13:04
ralsinagatox go13:04
gatoxralsina, i already went :P13:05
nessitaralsina: you have huge lag13:05
ralsinaDONE: pushed branches for bug #814686, bug #820945 worked more in bug #809873, calls, IRL testing of latest stuff13:05
ralsinaTODO: more IRL tests, merge pending branches, review diego's work, see if I can start on u1cp's bytes/unicode problems, ship build13:05
ubot4Launchpad bug 814686 in ubuntu-sso-client "QT UI: 'Sign in' button can be clicked even if not values were entered, same happens with verification code (affects: 1) (heat: 6)" [Medium,In progress] https://launchpad.net/bugs/81468613:05
ubot4Launchpad bug 820945 in ubuntuone-client "On windows, syncdaemon needs to find the right default config files, and store settings in the right place in the correct location (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/82094513:05
ubot4Launchpad bug 809873 in ubuntuone-windows-installer "Add a script to create a bundle and installer for all the related software (affects: 1) (heat: 8)" [High,In progress] https://launchpad.net/bugs/80987313:05
ralsinaBLOCKED: no13:05
dobeyλ DONE: finished initial shim work13:05
dobeyλ TODO: releases, packaging, etc13:05
dobeyλ BLCK: None.13:05
gatoxralsina, nessita about my branches... one of them has 1 approval, and the other is pending... do you think you are going to be able to review that today?13:06
ralsinaI first lagged then got dumped out of IRC, thus my weird comment ;-)13:06
nessitagatox: yes, before lunch13:06
gatoxnessita, awesome13:06
alecumandel, you said "roll back to previous attrs", but these are not attrs, these are permissions. aka "access control entries (ACEs)" from the "access control list (ACL)"13:06
gatoxnessita, yes....... i'm annoying13:06
ralsinanessita: I have complained to the ISP already :-(13:07
alecumandel, and I found they are a bit tricky... :-(13:07
mandelalecu, nessita, ralsina: can I get a re-review for https://code.launchpad.net/~mandel/ubuntuone-client/fix-file-shelf-windows/+merge/7002113:07
mandelalecu: sorry, true permissions for the DACL :P13:07
mandelalecu: and yes, DACL are terribly hard to understand and map to unix13:07
nessitagatox: can you please confirm if the test suite for https://code.launchpad.net/~ralsina/ubuntu-sso-client/validate-harder/+merge/70320 is passing for you? if so, ralsina, I'll approve13:10
gatoxnessita, ok13:10
nessitagatox: is failing miserably for me, but is very likely i my env13:11
* gatox testing...13:11
alecumandel, to see this problem: open the windows file explorer, create a folder in your homedir, then "right-click, New, Text Document", then right click the file, "properties->security->advanced"13:12
alecumandel, then close those dialogs and use os_helper.set_readonly on that file, and check the ACEs again.13:13
alecumandel, and finally use os_helper.set_readwrite and check the ACEs13:13
alecumandel, the problem would be that the permissions are slightly off. And we should not be hardcoding this permissions, but instead be inheriting them from the parent...13:14
alecuand changing them slightly when setting read only (but not creating a new set of ACEs, just copying and restricting the existing ones)13:14
alecumandel, does this make sense?13:14
alecumandel, ralsina, nessita: I'll be out for 30'13:15
ralsinaalecu, mandel, nessita, gatox: confirmed os.rename() works different on windows and linux, that makes tritcask fail13:15
nessitaalecu: ack13:15
nessitaralsina: ouch. different how?13:15
mandelalecu: we can talk in the mumble about it, I think I understand the issue in an slightly diff way13:15
ralsinaon linux it happily overwrites existing files. On windows, that's an exception13:16
nessitaralsina: mandel is working on doing that13:16
mandelralsina: excuse me?13:16
nessitamandel: right? for the commit_file partila thingy?13:16
nessitamandel: you said you will tweak the rename to overwrite files13:17
nessitagatox: approving the ussoc titles branch, it looks great!13:17
ralsinanessita:: OTOH, shutil.move works just fine on both platforms, so it may be a search/replace fix13:17
nessitaralsina: we're having exception while using shutil,move13:18
mandelnessita: yes, that is done with the MoveFileEx and is working, the problem we had in the commit_file is that it uses shutil.move in case we are in a diff fs but it will not be our case on winows13:18
ralsinanessita: but tritcask can't use os_helper, right?13:18
mandelralsina: the issues is not use shutil.move, which we have to do on linux, but to use it in windows as a rename13:18
gatoxnessita,  great13:18
mandelralsina: the but is not in tritcask, at leas the one 'm looking at, tritcask is doing a rename of an opened fd which brakes13:19
nessitamandel: is anyone working on that?13:19
mandelgot to close it and then move it, which I dont know if it brakes tritcasks logic13:19
nessitamandel: that == not renaming opened files13:19
mandelnessita: I though verterok was on it, but we might want to remind him13:19
gatoxnessita, about roberto's branch... tests are still running... but it is everything ok by now13:19
nessitamandel: ok, I'll talk to verterok13:20
nessitaralsina: can you show me a trace of the rename error you're having?13:20
nessitaverterok: ping13:20
mandelnessita: yeah, verterok should be your man for tritcask, I know the issue but I would not dare to say how to correctly fix it13:20
verteroknessita: pong13:20
nessitaverterok: hi there! we're having issues when running syncdaemon in windows since tritcask tries to renames opened files13:21
ralsinamandel: it's in my email from last night13:21
nessitaverterok: mandel mentioned you may have a fix for us13:21
verteroknessita: if it's the make_immutable call, I'm sure the file it's closed13:21
mandelralsina: yeah, that is tritcask, those tests are even failing atm13:22
verteroknessita: wasn't able to find the problem13:22
nessitaverterok: yes, is in that function. The thing is that we're having this problem:13:22
nessita  File "ubuntuone\syncdaemon\tritcask.pyc", line 654, in _rotate_and_merge13:22
nessita  File "ubuntuone\syncdaemon\tritcask.pyc", line 844, in merge13:22
nessita  File "ubuntuone\syncdaemon\tritcask.pyc", line 390, in make_immutable13:22
nessitaexceptions.WindowsError: [Error 183] Cannot create a file when that file  already13:22
nessita exists13:22
verterokoh, that's new13:22
ralsinamandel: exceptions.WindowsError: [Error 183] Cannot create a file when that file already exists. It happens in a os.rename() call :-)13:22
ralsinaThe line is in  File "ubuntuone\syncdaemon\tritcask.pyc", line 390, in make_immutable13:22
ralsinaDamn LAG13:22
verteroknessita: file exists?13:22
nessitaverterok: how can we know?13:23
gatoxnessita, confirmed... all the tests are ok13:23
verterokralsina, nessita, mandel: the file shouldn't exist....at least that's the whole idea13:23
nessitagatox: thanks!13:23
mandelverterok: a nice way to test that is to get to the point where you close (or believe it was closed) and make the code wait there, then use process explorer to find who has a handle to the file13:24
verterokmandel: it's being close 1 line before the call to os.rename13:24
nessitaverterok: but maybe close is failing?13:25
verteroknessita: failing?13:25
verterokmandel: anyway, this looks like a different issue.13:25
nessitaverterok: yeap, some other <something> is accessing it an can't close it13:25
verteroknessita: only the tritcask instance is accesing the files...not sure what other <something> means :(13:26
nessitaverterok: me neither, I'm trying to look for options13:26
verteroknessita: also, each test case run in it's own directory...so it should be a problem of cross test case issues13:26
nessitaverterok: this is not a test issue13:27
nessitaverterok: is IRL13:27
verteroknessita: this problem isn't related to the file not being closed13:27
verteroknessita: so, we have 2 issues13:27
verteroknessita: 1) "the file is being used by another process.." (yes, and that process might be myself :p)13:28
verteroknessita: 2) this issue about os.rename I just know about13:28
nessitaverterok: do we have fixes for 1 and 2?13:29
mandelverterok: what up with rename?13:29
mandelthat you cannot replace the file?13:29
nessitamandel: there is a trace in roberto's emails from last inght13:29
verterokmandel: no idea...you were talking about that13:29
verteroknessita: I don't have a fix for anything13:30
mandelverterok: I was saying that we could not d a reame because the fd is used by another process (or mine :( ) and that is the error13:30
nessitamandel: did you see the trace?13:30
verteroknessita: I need find out the reason of 1)  (seems that only happens on tests)13:30
mandelnessita: yes, that i do not see as a big issue, verterok look at http://msdn.microsoft.com/en-us/library/aa365240(v=vs.85).aspx13:31
mandelthat is the way we solve it in os_helper13:31
nessitamandel: how come is not a big issue since syncdaemon crashes IRL? :-)13:31
mandelwe just have to use the MOVEFILE_REPLACE_EXISTING flag...13:31
verterokmandel: the file shouldn't exists13:32
mandelnessita: I mean, is a big issue yet not a hard one to fix, bot me is a big issue if we hav eno idea what to do :)13:32
nessitaverterok: right, so self.temp_name somehow exists13:32
verterokmandel: it's really bad (tm) to replace an existing tritcask data file13:32
nessitamandel: we don't have to replace, like verterok is saying13:32
verteroknessita: are you sure it exists? :)13:32
mandelverterok: well, that is why I pointed at you for this :)13:33
nessitaverterok: no. We can add a log entry, though13:33
nessitaralsina: can you reproduce the windowserror issue?13:33
nessitaverterok: assuming the files does exists... any idea how we reached that situation?13:34
ralsinanessita: I bet I can13:35
mterryvds, I have another 500 Oops-ID if it would help: 2043appserverZJHAICCCIedIbEfIfbFJcAHHaFbfFaGcG20744413:35
verteroknessita: time travel13:35
nessitaralsina: before you do13:35
verteroknessita: really, time travel is the only way to have a duplicate inactive tritcask data file13:35
nessitaralsina: can you please add a log entry before the rename in the line is failing? something like: in line 38913:35
ralsinanessita: sure13:36
ralsinawhat should I log?13:37
nessitaralsina: logger.info('make_immutable: about to rename %r to %r (dest exists? %r).', self.filename, new_name, os.path.exists(new_name))13:37
ralsinaoh, right13:37
ralsinasure, on it13:37
nessitaverterok: well, it will not be te first time we have a PICNIC13:37
nessitaverterok: or a thread issue13:37
verteroknessita: tritcask isn't threadsafe13:38
nessitaverterok: when does that code run?13:38
nessitaverterok: only at startup?13:38
verteroknessita: it's only accesed from fsm and volume manager13:38
verteroknessita: the merge, yes13:38
nessitaverterok: fsm uses threads for watchs in windows13:38
nessitaverterok: not directly related, I know13:38
verteroknessita: not fsm13:39
nessitaverterok: filesystem_monitor I mean. You mean filesystem_manager?13:39
verteroknessita: fsm == fileystem_manager :)13:39
mandelverterok, nessita: all process of event occurs in the main thread of the reactor by using callFromThread, so it should be ok13:42
mandelunless we found an issue with that...13:42
nessitagatox: why you moved TITLE_STYLE out of alphabetical order?13:43
nessitagatox: this is the installer titles branch13:43
verterokmandel: this is only executed at startup...it shouldn't be affected by any threading issues13:43
nessitagatox: actuall, can you revert all changes to ubuntuone_installer/gui/qt/setup_account.py? I think they are not needed13:44
mandelverterok: ok, nice, looks ike the only way to know is my logging13:44
gatoxnessita, ok13:44
mandelverterok: what is the file name policy used by tritcask?13:44
mandelI hope it has nothing to do with any stupid case of the windows fs...13:45
nessitagatox: the branch looks good, I'll test IRL and then very likely you have an approve13:45
verterokmandel: <file_id>.<live|inactive|dead>.tritcask-<version>.data13:45
verterokmandel: file_id is a timestamp13:46
nessitamandel, ralsina, Chipaca, alecu, gatox: mumble?13:46
mandelverterok: can you show me an example?13:46
verterokmandel: in windows is: str(int(time.clock() + 100000))13:46
verterokmandel: in windows is: str(int(time.clock() * 100000))13:46
nessitamandel: you can look in your data dir13:46
* mandel looks13:47
verterokmandel: 131177171640160.live.tritcask-v1.data13:47
nessitamandel: ls .local/share/ubuntuone/syncdaemon/tritcask/13:47
verterokmandel: and 131177136785606.inactive.tritcask-v1.data13:47
nessitaralsina, Chipaca: mumble?13:48
gatoxnessita, changes reverted in setup_account.py13:49
mandelverterok: I cannot think a case when that happens to be the same...13:49
nessitagatox: ack13:49
verterokmandel: yup, that's the idea of the timestamp * 100000 :)13:49
verterokit should never ever, ever, ever, ever...ever be the same13:50
nessitaverterok: what if the clock in windows is not that accurate?13:50
verteroknessita: in linux we use time.time() and in windows time.clock()13:51
nessitaverterok: I know. How much we trust time.clock?13:51
ralsinanessita: in 5'13:51
nessitaverterok: I'm trying to seek alternatives...13:51
verteroknessita: I don't trust in windows, but according to the docs, time.clock should be fine13:52
mandelverterok, nessita: http://stackoverflow.com/questions/85451/python-time-clock-vs-time-time-accuracy13:52
nessitaverterok: what if the user has the time wrong?13:52
nessitaverterok: does time.time/clock depends on that?13:52
verteroknessita: yes13:53
nessitaverterok: what if: the user has the time correct, we create tritcask files, the user changes the time to th past, restart syncdaemon13:53
nessitaI know the chances are little, but we can have issues there13:53
ralsina_nessita, verterok: https://pastebin.canonical.com/50803/13:54
nessitaoh, what a nightmare13:54
nessitaverterok: I think we may need to have another algorithm to create filenames, such as use a random number, check filename does not exists, if so pick another random number, etc13:55
verteroknessita: tritcask use the order of the files13:55
ralsinaNo, we are just using time.clock wrong13:56
verterokralsina: ?13:56
nessitaverterok: ah13:56
ralsinatrivial fix13:56
ralsinaThe first call to clock always returns "almost 0"13:56
verterokralsina: please elaborate13:56
mandelverterok: look http://docs.python.org/library/time.html#time.clock13:56
verterokahh, ok13:57
verterokmy fault13:57
nessitaverterok: is there any chance you help us fixing this?13:58
ralsina_time.time() works just fine on windows, BTW13:58
verterokso, it doesn't work for what we need :(13:58
verterokralsina: no it doesn't...at least not for the tests13:58
ralsina_verterok, ok, we can use clock() and add a random huge integer13:58
nessitaralsina_: time.time() may return the same number13:58
mandelverterok: nah, it does if you are measuring how long the process was running13:58
mandelverterok: ad if you call it always very early..13:58
ralsina_call time.clock() early, then use time.time()+time.clock()13:59
verterokralsina_: I was thinking of that13:59
nessitaralsina_: mumble now?13:59
ralsina_nessita: sure13:59
mandelverterok: fancy to use http://msdn.microsoft.com/en-us/library/ms724390(v=vs.85).aspx from pywin32?14:00
ralsina_BTW: the stalled uploads in the IRL tests? same tritcask error as this one14:00
mandelverterok: is UTC so no problem with time zones…14:00
nessitaChipaca: you coming?14:01
verterokmandel: I'll propose a fix using time module14:01
nessitaverterok: thanks!14:01
verterokmandel: I'm not going to learn how to use pywin32 extension today :)14:01
ralsina_time.time()+time.clock() should be unique, monotonic, and *fast* (2 seconds per second ;-)14:01
nessitaralsina_: can you please file a bug for verterok?14:01
mandelverterok: just import and use :)14:01
ralsina_and also, work on both OSs14:01
mandelralsina: time.clock depends on when you called it14:01
ralsina_nessita: of course!14:01
ralsina_mandel: yes, but it's  always increasing. monotonic+mnotonic == monotonic14:04
ralsina_then again, here is a tiny chance i will repeat between runs14:05
mandelralsina: estan hablando en mimble?14:06
* mandel suspects his went a la mierda...14:06
ralsina_yes it did14:07
ralsina_vertero: bug #82148814:12
ubot4Launchpad bug 821488 in ubuntuone-client "windows: tricask is using a bad timestamp for unique filenames (affects: 1) (heat: 6)" [Critical,Triaged] https://launchpad.net/bugs/82148814:13
mandelralsina: oh, did I die in mumble again?14:13
mandelbad internet...14:13
verterokmandel: GetSystemTime doesn't work for this case14:15
mandelverterok: no? what about using something in tempfile?14:17
verterokmandel: it doens't have enough resolution14:17
mandelverterok: I would need more time to think, so I leave it to you :)14:18
=== m_conley_away is now known as m_conley
ralsina_verterok, why not just use a UUID?14:20
verterokralsina_: we need order14:21
verterokralsina_: tritcask use the filename to read from old to new14:21
ralsina_oh, ok14:21
ralsina_verterok, you can take time.time() at startup, and then add clock() to it14:23
ralsina_that should be increasing even between sessions14:23
ralsina_of course you need a global variable ;-)14:24
verterokralsina_: or just an attribute :)14:24
ralsina_verterok: of course :-)14:24
mandelnessita, ralsina, alecu, gatox: I'm of 5 min 'til I find the dog, apparently he likes to walk around scaring the tourists, mainly the german ones...14:34
gatoxmandel, :P14:34
ralsina_ze germans!14:34
alecumandel, btw: you may have already seen it, but the R/O bug is this: https://bugs.launchpad.net/ubuntuone-client/+bug/82035014:35
ubot4Launchpad bug 820350 in ubuntuone-client "WIndows: either set_dir_readwrite and/or set_dir_readonly are not doing what they should (affects: 1) (heat: 6)" [High,Triaged]14:35
mandelalecu: I'm back, shall we take a look at that before I get to my bugs?14:42
alecumandel, the read only stuff?14:45
mandelalecu: si :)14:46
alecumandel, I agree with nessita that we should postpone it for next week: it's important, but not critical14:47
mandelalecu: ok, then I move to my own little branches :)14:47
mandelalecu, nessita, ralsina: please can you take a look to https://code.launchpad.net/~mandel/ubuntuone-client/fix-file-shelf-windows/+merge/70021 I'd like to know if we will ignore it or not so that I base my branch on trunk or on this one14:49
ralsina_mandel: I am tempted to say ignore, but I am not sufficiently familiar with the code, I am just basing it in what I have heard.14:49
mandelralsina: look at the diff, there are not many change14:50
ralsina_mandel, I have seen them, but I don' t know what they are trying to *do* :-D14:51
nessitamandel: looking14:51
mandelralsina: use the os_helper code :)14:51
verterokmandel, nessita, ralsina_: https://code.launchpad.net/~verterok/ubuntuone-client/win32-no-more-time_clock/+merge/7058014:51
mandelverterok: I take one review!!!14:52
verteroknessita, mandel: yes, not a single test. I don't have my laptop (with the windows vm) around14:52
nessitaverterok: can you add some tests and  volunteer to run them?14:54
verteroknessita: I can't write tests that I can't execute :(14:55
ralsina_verterok, I am doing a IRL test in 5'14:55
ralsina_and I can run tests for you if you want ;-)14:55
verteroknessita: I'ld love to write tests, and will do it. but I can't do it today :(14:55
nessitaverterok: when would you be able to add the tests?14:56
ralsina_Basically, at least test that the first call is not 0 is enough for the bug we had14:56
verterokralsina_: I can add that one14:56
verteroknessita: as soon I get some stuff out of my todo/work queue14:57
mandelverterok: tine time.clock() addition in the __init__ is not really needed since if it the first one ever to be executed it will always be 0, or so I understand by this the docs: http://docs.python.org/library/time.html#time.clock14:57
ralsina_verterok: that test is what TDD would require. "hey, timestamp returns 0 in every first run", test that is not 0, fix, done.14:57
verterokmandel: it will not be 0 :)14:57
verterokralsina_: clearly I didn't fix it using tdd :)14:57
ralsina_verterok: yes, I am talking about faking it ;-)14:58
nessitaralsina_: I propose the following: let's test this IRL, and if it works, you cna merge it in the installer, but let's wait for the merge into trunk until the tests are added14:58
mandelverterok: maybe adding it in a diff line and giving a comment like #start clock or something like that :)14:58
verterokmandel:  # start counting from now isn't goof enough? :)15:00
nessitamandel: your branch fsm looks good, but can you please restore the removed line at the end of the file ubuntuone/syncdaemon/file_shelf.py15:01
nessitaralsina_, verterok: I propose the following: let's test this IRL, and if it works, you cna merge it in the installer, but let's wait for the merge into trunk until the tests are added15:01
verteroknessita: I'll add a test, tdd like15:01
ralsina_nessita: I am building the installer now, should have word about it in 5'  or so15:02
mandelverterok: hehe I mean remove the + time.clock() that is all, but just because I had to find something, for me is more than ok15:02
mandelnessita: sure15:03
verterokmandel, nessita, ralsina_: test and fixed comment pushed15:03
ralsina_verterok: awesome15:03
ralsina_verterok: runnin test now15:03
verteroknessita: there, it haz a test15:03
nessitaverterok: :-)15:03
mandelverterok: self.assertNotEqual(timer.time(), timer.time())15:06
mandelself.assertTrue(int(timer.time()) > int(timer.time()))15:07
mandelbut is not a must :P15:07
verterokmandel: that will probably fail :)15:08
verterokmandel: (in linux)15:09
mandelverterok: @skipIfOS('linux2') but I don't know… as I said, not a must15:10
ralsina_verterok: confirmed IRL, filenames are now sane15:10
verterokralsina_: cool15:10
nessitamandel: let me know when you push the line restore so I can run the suite15:13
nessitaalecu: branch looks good! running tests now15:15
mandelnessita: waiting for the push to finish...15:18
ralsina_nessita, mandel: a traceback from testing: https://pastebin.canonical.com/50820/15:24
nessitaralsina_: testing what exactly? :-) (I have 3 branches running tests ATM)15:24
mandelnessita: push is done15:24
ralsina_nessita: sorry. From running u1cp IRL15:24
nessitaralsina_: doing what exactly? is "everything" (sync, sso, etc) running?15:25
nessitamandel: ack15:25
nessitaalecu: ping15:25
ralsina_nessita: yes, everything is running, and it starts saying this hundreds of times and fails to connect to SD15:25
alecunessita, pong15:25
nessitaalecu: trunk -> (skips=13, failures=40, errors=146, successes=1982)15:26
nessitayour branch ->  (skips=13, failures=40, errors=135, successes=1996)15:26
nessitaralsina_: can you give me details about what "it starts saying this hundreds of times and fails to connect to SD" means? not sure how to help debug otherwise15:27
ralsina_nessita: I have SD running, and working, uploading files. If I start u1cp, it just starts printing this over and over, and the UI never displays ay data from SD.15:28
ralsina_nessita: sadly, this goes so fast I can' t see the beginning of it15:28
nessitaralsina_: weird, CP is workign from trunk running everything 'manually'15:28
nessitaralsina_: 2 things15:28
nessitaactually, 315:28
ralsina_nessita: ok15:28
nessitaralsina_: can you kill you evil twin? autocompleiton is killing me :-D15:28
ralsina_nessita: talk to either of us, I don' t care ;-)15:29
alecuralsina_, can you try running it and capturing the output to some file? it would be useful to have the start of those errors15:29
ralsina_alecu: I can try, of course15:30
alecuralsina_, and also they'll probably show up as well on .cache/logs/somewhere15:30
nessitaralsina_: is difficult talk to any of you, both are hard to autocomplete having the other around15:30
ralsina_nessita: ok, will kill the other one15:30
nessitaralsina_: second, yes please, capture output as alecu says15:30
nessitaralsina_: and third, is your tests using installed stuff or running all manually?15:31
ralsina_nessita: installed stuff15:31
alecuralsina_, anyway: it looks like on_request_queue_removed is not defined as it should15:31
nessitaalecu: right, but why does this happen using the installed stuff only (apprently)?15:32
mandelralsina: seems like a remove method is missing in the ipc of sd on the windows side, would be nice if you check that the method (without remote_) is present.. (I'm on a diff thing right now :P)15:32
ralsina_nessita: I don't think it's just with installed stuff. It's also because I am the only one that has a working syncdaemon beyond 5 minutes (because of the timestamp fix)15:32
nessitaralsina_: I have syncdaemon running all the time since MOnday15:33
nessitaralsina_: but not from installer15:33
ralsina_nessita: it runs. But does it do anything?15:33
nessitaralsina_: from time to time, yes15:33
nessitaralsina_: you suggesting that we need to coordinate opening the control panel with some sync stuff? maybe...15:33
nessitaralsina_: can you reproduce not using installed stuff?15:34
ralsina_nessita: not trivial to reproduce that way, because of how I setup things. I am going to try to get better traces and take a short look myself.15:35
alecuralsina_, is it perhaps because of the 965 files your syncdaemon still has as pending upload?15:35
alecuralsina_, or have you deleted all the metadata?15:35
ralsina_alecu: I deleted the metadata, started uploading that again (and seems to be working)15:35
ralsina_then tried u1cp, got this. I will try to research a bit better, so I will be away a bit, on the other session.15:36
alecuralsina_, so, I'm tracing the "emit_signal" method that's printing this, and it seems to be harmless15:37
ralsina_alecu: good.15:37
alecuralsina_, the thing is that syncdaemon is not finding that method on the ipc client, and it's just printing this as a warning15:37
alecuralsina_, perhaps we should make this a bit more silent15:37
alecuralsina_, but it's certainly harmless15:37
ralsina_alecu: or implement a dummy so it doesn' t fail ;-)15:38
alecu_emit_failure() in ipc.py is the errback for this errors, and it's just logging the warnings.15:38
alecuralsina_, I'm not sure it makes much sense to implement a dummy15:38
alecuralsina_, since these are signals15:38
alecuralsina_, and they should be ignored if there's nothing subscribed to them..15:39
ralsina_alecu: it's because it's missing on StatusClient. It's trivial to fix.15:39
alecuralsina_, I think we should handle the "NoSuchMethod" exception, and log the rest15:39
alecuralsina_, since this will happen for all other signals that are not implemented, I think the sensible thing would be to ignore missing signal handlers15:41
ralsina_alecu: ok, makes sense15:41
ralsina_alecu: it really makes following problems in u1cp difficult because of the noise15:41
alecuralsina_, I'll make a branch to fix this and I'll push it in just a few minutes.15:42
ralsina_alecu: awesome15:42
* alecu loves pluging the monitor in the laptop and it *just working*15:44
nessitaalecu: in linux i'm getting some lint issues15:45
nessita    755:  local variable 'd' is assigned to but never used15:45
nessita    1208:  local variable 'd' is assigned to but never used15:45
nessitaalecu: can you fix and let me know?15:45
alecunessita, will do15:46
alecuque grande lint!15:48
alecuI mean:15:48
alecuque grande, pyflakes!15:48
alecunessita, pushed15:50
nessitaalecu: thanks15:51
alecuralsina_, I think that instead of logging the huge stacktrace as a WARNING, we should be just logging a line with the name of the method that was not found in the DEBUG log level.15:53
alecuralsina_, would that be too noisy too?15:53
ralsina_alecu: it would be a 25x improvement at least :-)15:53
alecuralsina_, I think it might help for debugging if an issue like this ever happens backwards15:54
alecuralsina_, ok, I'm doing it like that.15:54
ralsina_alecu: +1 on that15:54
ralsina_confirmed that with the timestamp fix, syncdaemon uploads like a champion. Huge folder tree with a ton of small files in it, uploaded perfect.15:55
* mandel wants to kill himself due to his connection...15:57
alecuralsina_, !!!!!!!16:02
alecuralsina_, wowowowo!16:02
nessitaralsina_: can I ask you to do another test?16:02
ralsina_nessita: sure!16:02
* alecu remembers that "a champion" means "a standard tennis shoe" in .uy16:03
alecuand shoes do lousy uploads.16:03
nessitaralsina_: setup 2 clients for the same user, in one do bzr branch lp:ubuntuone-client -> wait for all the uploads and downloads in the other client to finish, run make check in the second client16:04
ralsina_nessita: ok16:04
ralsina_will try that, brb16:04
alecunessita, you sick sadistic you. You can't stand champions, right?16:04
nessitaalecu: :-D16:05
nessitaalecu: you should wait for the next test I have prepared after that :-P16:05
* ralsina_ is about to write a thingy that creates random folders with random names with random content in it.16:06
nessitaralsina_: the thing is that we need to compare those random stuff against something to check nothing is corrupt, and using random data is difficult16:07
ralsina_nessita: it's actually rather trivial.16:07
dobeylunch time, bbiab16:07
nessitaralsina_: yes? please teach me!16:07
ralsina_nessita: create the structure, do a os.walk recording pathnames and checksums, wait until upload, stop everything, remove metadata, move everything out of ~/Ubuntu One, start syncdaemon again, wait, run os.walk again, compare them.16:08
ralsina_or use two profiles, f course16:09
nessitaralsina_: that is a good tests, but there is something that we need to test that that doesn't cover16:09
ralsina_nessita: it' s called fuzzing, it's meant to test " stuff we didn't think about"  :-)16:09
nessitaralsina_: that syncdaemon acts properly to server notifications, and for that, we need the 2 syncdaemon clients working simultaneously16:10
ralsina_nessita: then that same thing using 2 profiles.16:10
nessitaralsina_: I would like to do that but having 2 syncdaemon working at the same time (2 vms will require, yes)16:10
ralsina_nessita: no, just two user profiles in one16:10
nessitaralsina_: or that, yes :-)16:10
* ralsina_ creates another user then16:10
nessitaverterok: approving your branch!16:11
nessita(ran tests in all platforms)16:11
verteroknessita: cool, thanks16:12
ralsina_nessita: I currently have two sessions, one uploading the bzr branch the other downloading... seems to be evolving correctly so far :-)16:21
nessitaralsina_: that's GREAT news16:21
ralsina_and it' s not even making the notebook slow down. f course I have a 4-core i5 ;-)16:22
nessitamandel: I'm about to approve your file shelf branch16:24
nessitamandel: when the l;inux make check finish, you get my +116:25
nessitaalecu: you got my +1 already16:25
nessitaalecu: while running review tests I found a couple of trivial thing to fix, so, can you please: https://code.launchpad.net/~nataliabidart/ubuntuone-client/bar-baz/+merge/7059016:26
nessitaalecu: is pretty trivial16:26
mandelnessita: cool, I'm right now pushing the next with the fix on the commit_partial method, there are some tests that fail on windows due to ro, I'll write the expected failures in the MP16:26
mandelralsina_: with that ^ you can try to do an other IRL that will allow to download big files :)16:26
alecunessita, will do, thanks16:26
ralsina_mandel: will do after I finish the current torture test :-)16:27
gatoxralsina_, now that the qapplication is back in the windows.py (installer), do you prefer that i moved the stylesheet and font setup from gui.py to windows.py?16:27
* nessita votes "yes please"16:27
mandelralsina_: will ping you, and _nessita when ready16:27
ralsina_gatox: yes16:27
gatoxralsina_, ok!16:28
alecunessita, "Author: Alejandro J. Cura <alecu@canonical.com>" ???16:28
ralsina_gatox: also, I get *occasional* errors about "can't parse the stylesheet"  when using binaries. I hae no idea whatsoever about why.16:28
nessitaalecu: with that branch, I get: (skips=13, failures=40, errors=93, successes=2033)16:28
nessitaalecu: yes, is a copy of what you did16:28
nessitaalecu: so you are the author :-D16:28
gatoxralsina_, do you have any particular example where i can take a look?16:29
ralsina_gatox: no. it just happens every once in a while.16:29
ralsina_And says exactly that in stderr16:29
alecunessita, 135 -> 93... nice!16:29
gatoxralsina_, weird... i'll take a look at that16:29
alecunessita, I'll run tests on it right away16:29
ralsina_gatox: don't worry about it, really, it may happen only here for all I know.16:30
ralsina_gatox: looks like a problem loading the resource to me, but I don't have any rationale for it16:30
gatoxralsina_, ok... let me know if it becames a problem16:31
* mandel Error 15 (net::ERR_SOCKET_NOT_CONNECTED): Unknown error.16:35
mandelnessita, ralsina_, alecu: please: https://code.launchpad.net/~mandel/ubuntuone-client/dont-use-shutil-move/+merge/7059516:38
nessitamandel: on it16:38
ralsina_nessita, alecu, mandel: update on the torture test. Upload and download progressing. Only errors I saw are in the uploader about files being removed before they were hashed, which is, I think, normal.16:38
mandelI have not run it on linux, I'm trying at the moment but I'm getting some issues :(16:38
mandelralsina_: ein?16:39
mandelralsina_: why is it normal?16:39
mandelis it realted to the tests?16:39
ralsina_mandel: I am doing a bzr branch inside u1. The limbo files are there only for a fraction of a second.16:39
ralsina_mandel: so, sd gets the file creation, puts it in the queue to be hashed, but when it gets to it, it's gone already16:40
ralsina_mandel: and it's a nicely formatted error, not a thrown exception or anything ;-)16:40
mandelralsina_: ah… nice :)16:40
ralsina_u1-client is 58 freaking MBs?????16:41
mandelralsina_: yeah… does bzr use desktopcouch as the backend...16:42
nessitaralsina_: why are files being removed? I mean, did you remove those files?16:42
ralsina_nessita: no, just the rdinary way bzr works. Files are created and removed during a branch.16:43
nessitamandel: file shelf approved16:43
mandelnessita, ralsina_, alecu: EOD here, please take a look at the branch and let me know, any comments will be fixed early my morning on monday16:43
nessitaralsina_: ok, please once all the torture-nessita-tests is done, send me all the logs (probably you will get more than one)16:43
mandelhave all a great weekend!!!! that includes gatox too :P16:43
nessitamandel: ack16:43
ralsina_mandel: you too!16:43
gatoxmandel, byeee!16:43
nessitaralsina_: if you have some slot, could you give a second review to https://code.launchpad.net/~mandel/ubuntuone-client/fix-file-shelf-windows/+merge/70021 ?16:44
ralsina_nessita: expect a tarazillion megabites of logs. I am running n --debug16:44
nessitaralsina_: --debug does not add extra debug, but logs to terminal and to a different file16:44
ralsina_nessita: ok then16:44
nessitaralsina_: but yeah, huge log file is welcomed16:44
=== beuno is now known as beuno-lunch
alecunessita, with your new branch, I get: FAILED (skips=13, failures=47, errors=105, successes=2006)16:51
nessitaalecu: may be the action queue random failing tests16:51
nessitaalecu: you branch has not yet landed, afaik16:51
alecunessita, that's right!16:52
ralsina_nessita: +1 on bar/baz16:59
nessitagreat :)16:59
ralsina_looks like u1 is really not optimized for creating and then deleting a bazillion small files in u1. The hashing of a ton of unexisting files takes forever. Then again, killing sd and starting it again clears that up.17:17
alecuralsina_, gatox: if you can, please review https://code.launchpad.net/~alecu/ubuntuone-client/fix-add-watch-raceconditions/+merge/7055917:17
gatoxalecu, on it17:17
alecuralsina_, the average user is not a developer, so it does not have a bazillon tiny sourcecode files. Nor a sysadmin, having a morzillon emaildirs.17:19
alecuralsina_, our average user has lots of music and images, though :-)17:19
ralsina_alecu: agreed. Just that for the test, it got stuck there for [whoknows] hours, so I had to kill it to purge that :-)17:19
alecuuhhhh.... I got to go to the baaaaank17:19
ralsina_alecu: the many files is not even the problem. It's create/delete really quick17:20
gatoxalecu, +117:22
alecugatox, grax!17:22
gatoxalecu, ralsina_, nessita small branch for review: bzr push lp:~diegosarmentero/ubuntuone-windows-installer/icon-signin17:22
ralsina_gatox: got the merge url?17:23
gatoxclean some styleSheet and font setup stuff... add application icon, and fix the ui style for sign in page17:23
gatoxralsina_, sorry... wrong paste :P17:23
gatoxralsina_, nessita, alecu https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/icon-signin/+merge/7060317:23
ralsina_gatox: got it17:26
alecugatox, I'll do it after bank and lunch17:26
dobeyeverything going well for you guys on tarmac front? just checking :)17:30
ralsina_nessita: the trace I gave earlier about the unimplemented method was caused by an accepted share being deleted from the website, so that will show up again when we test shares.17:31
nessitaralsina_: great(ish)17:45
nessitagatox: on it!17:45
ralsina_nessita: the torture is finished. And the trees are identical.17:46
nessitaralsina_: make check worked?!?!?!17:46
ralsina_nessita: will have definitive answer to that in a few minutes17:46
ralsina_nessita: but a tree comparison tells me this looks just like trunk17:47
nessitathat's great!17:47
ralsina_nessita: However... I had to restart syncdaemon once on the receiving side for it to finish downloading. It was sitting with an empty queue.17:47
nessitaralsina_: sitting with an empty queue? not sure what that means17:48
nessitaralsina_: you mean it will not download anything?17:48
ralsina_nessita: it had downloaded about 70% of the files, and was not downloading anymore.17:48
ralsina_nessita: then, I restarted it, and downloaded the rest17:49
ralsina_and when it was not downloading it was logging " Queue: 0"17:49
nessitaralsina_: hum, you should report this to chicharreros, be sure to save those logs from the 2 runs: while sitting, and after17:49
nessitaralsina_: did you wait "long enough"? sometime the notifications from the server takes some time17:49
ralsina_nessita: probably not long enough17:50
ralsina_nessita: I wanted to check for corruption more that I wanted to check that17:50
ralsina_nessita: but as soon as I started it, it started downloading, if that means anyhing.17:50
nessitaralsina_: may mean that the notification wasn't there yet, so I will not worry for now17:50
ralsina_nessita: but I think we have reason to be happy, this was pretty hardcode testing of the core, and it worked well17:51
ralsina_nessita: make tests works/fails the same as the original.17:56
ralsina_nessita: also, our tests fail miserably when you run them on a removable vfat drive :-)17:56
nessitaralsina_: got a trace? Im curious17:57
ralsina_nessita: http://pastebin.ubuntu.com/659452/18:00
ralsina_I think that's because on vfat there are no ACLs18:00
nessitaanother reason to ignore read-only stuff18:02
nessitagatox: ping18:14
gatoxnessita, pong18:16
nessitagatox: can you please change the qss loading code to use QResource instead of QtCore.QFile(":/ubuntuone.qss")?18:16
gatoxnessita, ahhhh okok18:16
ralsina_nessita: good catch18:21
=== beuno-lunch is now known as beuno
gatoxnessita, ralsina_ branch updated: https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/icon-signin/+merge/7060318:47
nessitagatox: ack18:48
gatoxralsina_, also... this branch is approved... but is not being merged.. (https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/titles-on-pages/+merge/70169)18:48
gatoxi don't know why18:48
gatoxnessita, ^18:48
nessitagatox:  no commit message18:48
gatoxnessita, ohhhhhhhh..... fixing it18:48
ralsina_nessita: http://ubuntuone.com/p/18bT/ -- it's justa  rough draft18:53
nessitaralsina_: looking18:53
ralsina_for example, I am not thrilled by the spacing between the two areas in the tab, and it's missing a horizontal line (which is easy to add, though)18:54
nessitaralsina_: you broke the 2 group boxes18:54
ralsina_nessita: indeed I did18:54
nessitaralsina_: each group is "independent" from the other18:54
ralsina_nessita: if you want the buttons to be the same size, they need to be in a single layout18:55
ralsina_and a layout can't get into two groupboxes18:55
ralsina_that "independence" is what made it look bad18:55
nessitaralsina_: semantically they are 2 different, idependent groups, so merging into one seems wrong to mw18:55
ralsina_you are confusing presentation and logic18:55
nessitaralsina_: aux le contraire mon ami :-)18:56
nessitaralsina_: at least from my POV, presentation can not break logic18:56
ralsina_if we were to have an arbitrary number of independent groups? I would have them. For two? It's not important, IMHO.18:56
ralsina_nessita: you are getting all that data from a single dictinary. You split them yourself in u1cp ;-)18:57
nessitaralsina_: I "highly" disagree re "is not that important" :-). I prefer to hard code a button width18:57
ralsina_nessita: NO18:57
ralsina_I will disapprove any branch that hardcodes a button width18:57
ralsina_sorry, but you have your standards, I have mine ;-)18:57
nessitaralsina_: start reverting what's in trunk then :-)18:58
ralsina_nessita: I probaby should eventually18:58
ralsina_nessita: hardocded widget sizes are evil.18:58
nessitaralsina_: conceptually, they are two different groups of information. We even had them in 2 tabs, so we can't just go an merge them visually18:58
nessitaralsina_: I agree, that was I was seeking another clean option18:59
ralsina_nessita: add a line in the middle.18:59
nessitathis is not clean either18:59
ralsina_it looks a ton better18:59
nessitaralsina_: the looks doesn't matter when the insights have cellulitis :-P18:59
ralsina_nessita: there is no logic change at all. Just a change in the tree structure of the widgets.19:00
nessitaralsina_: you're abusing the widget set to accommodate your needs. I will not do that, not easily at least19:00
ralsina_nessita: and you are pushing your internal logic into the presentation. That's not good.19:01
ralsina_"abusing"? Whoa.19:01
ralsina_I just moved 5 widgets outside of two boxes, and removed the boxes :-)19:01
nessitaralsina_: conceptually they are 2 independent groups, we should be able to grab a "widget" (either profile or services) and move it to another tab, without loosing formatting nor layout19:01
ralsina_nessita: and you can.19:02
ralsina_nessita: just as easily as before19:02
nessitaralsina_: no, before you grab a groupbox and dump it wherever you like, now you have to pick a button from here, a label from there and a title from somewhere else19:02
ralsina_nessita: and it looks misaligned everywhere.19:03
nessitaralsina_: think about each grouping to be a widget itself19:03
nessitaactually, each grouping should be a widget itself19:03
ralsina_nessita: you are showing the internal logic in the UI. That is actually *bad*.19:03
nessitaralsina_: you keep saying that :-), and I don't agree to that19:03
ralsina_The UI has to be guided by the user experience, not by the design of your code.19:03
ralsina_You can not agree and be wrong, you know.19:04
nessitaralsina_: 100% agreed. And this is not leacking logic into the UI, is using the right widget for the right needs19:04
ralsina_nessita: it totally is leaking logic.19:04
nessitaralsina_: and of course, you and I and everyone can be wrong, we're humans :-)19:04
ralsina_nessita: you say " they are separate groups". Says who? Your app logic.19:05
nessitaralsina_: that's what you think, so it would be better if you prepend "I think you're leaking logic"19:05
ralsina_nessita: indeed. I think it's obvious you are leaking logic. And you also hink leaking logic is not bad :-)19:05
nessitaralsina_: they are conceptually 2 different groups of info, like a folder list is, oh exactly that, a list of folders19:05
nessitaralsina_: I agree leaking logic is bad. I don't think I'm leaking logic.19:06
ralsina_nessita: yes. Now, does a list of folders consist of 50 different misaligne widgets? No.19:06
nessitaralsina_: exactly, that's why we use a listsomething19:06
nessitaso, we have 2 pieces of info that has independent goals and transmit independent information19:06
ralsina_nessita: good. Now, a difference is that a list of folders is of variable length.19:07
nessitano matter how we "logic" that19:07
ralsina_This is of fixed length.19:07
nessitaralsina_: nopes19:07
ralsina_In fact, it's displaying a fixed set of fields.19:07
nessitaralsina_: I agreet this is a list of 2 items: profile info and services info19:07
nessitathat's why this is a vertical layout19:07
nessitawith 2 items in it19:07
ralsina_Problem 1: That way of laying it out looks horrible.19:08
ralsina_"I think"  ;-)19:08
ralsina_Now, you can also consider it as a display of 2 things. Not a list of two items, because the items are heterogeneous.19:09
nessitanot sure what you mean19:09
ralsina_ie: it's not like a list of folders, because thing A (email, name) is not like thing B (list of plans)19:09
nessitaralsina_: when thinking of the outter layout you should not look "into" each group content19:10
ralsina_nessita: we know what we want to display.19:10
ralsina_nessita: ignoring that is not conducing to clean design of the UI19:10
nessitaralsina_: this is where I think you're "abusing" the toolkit: you depend on what each grouping has to propose this layout19:11
ralsina_I intentionally flattened it so there is no " outer"  and "inner" layouts. That's the whole idea.19:11
ralsina_nessita: but we know that.19:11
nessitaralsina_: then that's where we disagree the most19:11
ralsina_so, yes, I confess, I designed the layout considering what it has to display.19:11
nessitaralsina_: you designed the outter layout considering what each inner group has to display ;-)19:12
nessitawhich is not the same19:12
ralsina_I designed the layout in its entirety that way. Ther only "inner" layout is grouping email and name.19:12
ralsina_Which I would be happy to replace with a single label that accepts newlines19:12
nessitaralsina_: by inner layout I mean the whole content of each group19:13
nessitano the grouping of email and name19:13
ralsina_nessita: I didn't design it in groups. I designed it in a table with one empty separating row19:13
nessitaralsina_: think each grouping as a widget that shows info and offers an action thru a button19:13
ralsina_which I am again happy to replace with a line19:13
nessitaso we can have 2 instance:19:13
nessitathe profile instance, and the services instance19:13
nessitaralsina_: how would you solve this if each grouping is a widget19:14
ralsina_nessita: it would be very very difficult. That's why it makes no sense to do it that way.19:14
ralsina_or, I would use a form layout19:15
nessitalet me check what's a form layout19:15
ralsina_but it's  a tradeoff. If you create megawidgets (things containing other widgets), combining them in windows is very difficult. That's the bad side of them.19:16
ralsina_ok, gotta pick the kid, be back in 15'19:16
gatoxnessita, can you approve this branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/icon-signin/+merge/70603 please??19:28
nessitagatox: I'm still reviweing, trying to run stuff in windows (I'm having some issues)19:29
nessitacan't run the installer, I'm debuggin19:29
gatoxnessita, ok..... let me know if i can help with something19:30
nessitagatox: I have a conflict now, with trunk19:32
nessitaText conflict in data/qt/ubuntuone.qss19:32
nessita1 conflicts encountered.19:32
nessitagatox: can you fix please?19:32
gatoxnessita, on it19:33
gatoxnessita, ready!!! it's easy when i know both versions of the file :P19:35
nessitare branching19:35
nessitagatox: you added an empty line insde19:36
nessita QLabel#title_label{19:36
nessita    font-size: 20px;19:36
gatoxnessita, ....crap... sorry.....19:37
gatoxnessita, now should be ready19:38
ralsinaI'm back19:41
nessitagatox: dude19:42
nessitaI have this in the qss19:42
nessita    border-style: dotted;19:42
nessita    border-color: #939389;19:42
nessita    border-bottom-width: 1px;19:42
nessita    border-top-width: 1px;19:42
nessita    font-size: 20px;19:42
nessitathere is a missing }19:42
gatoxnessita, omg...19:42
gatoxnessita, that's for trying to fixing it really quick19:43
nessitaralsina: I'm doing something in between you're proposal and my standards, but I have a question19:43
ralsinanessita: sure19:44
nessitaralsina: how can I set style so the dotted line is in place when using a grid layout?19:44
ralsinanessita: add a "horizontal line" widget, set it to dotted style, and make it span all columns19:44
nessitais not like I can have a frame or similar around the profile box and set border bottom ro it19:44
nessitato it*19:44
nessitaralsina: but this is only style, I'm not sure we should add a "line" widget. Can we set style so, for example, each row in a grid line has a border that we can customize?19:45
nessitain a grid form*19:45
ralsinaNo, it's not a grid widget but a grid layout, and grid layouts are not visible19:45
ralsinalayouts in general are not visible. For visible things you need to use widgets19:46
ralsinaIf you add a container widget, you are back in your old layout19:46
ralsinaI you want semantic separation, the widget for that is a line19:47
nessitaralsina: how can I make a widget span columns?19:47
nessitadoesnt matter, I'll google19:48
ralsinadrag the edge you want to span19:48
ralsinaok, that was unclear. Drag the side you want to extend towards the column you want to cover19:48
ralsinamaybe that made a bit more sense :-)19:48
ralsinaYou can only span one column on each drag, so it may take a few tries to get it right19:48
nessitaediting the xml was easier :-P19:49
ralsinanessita: hehe19:50
dobeythisfred: ping19:56
gatoxnessita, done... (sorry): https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/icon-signin/+merge/7060319:56
thisfreddobey: pong19:56
thisfredwhat did I break now?19:56
dobeythisfred: mterry proposed a branch against u1couch a couple weeks ago. could you review it please? :)19:57
gatoxnessita, ralsina also i've just submited this branch that adds some missing titles: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/missing-titles/+merge/7062419:57
nessitagatox: ack19:57
ralsinagatox: got it19:57
thisfreddobey ah sure. I have been filtering pretty much all merge proposals, recently. I wish lp would put the project name in the subject19:57
dobeythisfred: it puts it in the headers19:59
dobeyand technically in the subject, given it is part of the branch location strings20:00
thisfreddobey, yeah, in its own custom headers. Almost no email clients can filter on those20:00
dobeyi thought notmuch was magical20:00
dobeyyou use notmuch+emacs. you don't get the "it can't filter" argument :)20:01
thisfreddobey: it is, and that feature is planned, but not landed ;)20:01
dobeyyou can't regex arbitrary headers?20:01
ralsinagatox: looks good to me20:02
gatoxralsina, ok! :D20:03
thisfreddobey: since the search is xapian, it has to index those headers before they can be searched, and currently it only indexes the usual headers20:04
nessitaralsina: you sure the line gets visible drawn? I get nothing IRL20:04
thisfreddobey: for mps as you say, the info is in the subject anyway. Where it's really annoying is bug mail.20:05
ralsinanessita: they should unless something weird is done using qss20:07
ralsinanessita: do you see it in the preview in designer?20:07
nessitaralsina: preview is just the designer itself?20:07
ralsinanessita: ctrl+r in designer20:07
nessitaralsina: yes, the line is there20:08
ralsinanessita: or the menu form -> Preview, or form -> Preview in -> a style20:08
ralsinanessita: then I blame gatox20:08
dobeythisfred: right, but you can just ignore bug mails :)20:08
thisfreddobey: belieme I do :)20:09
ralsinanessita: did you remember to compile it?>20:09
nessitaralsina: good point, but yes, I've learn to do compile always before running the ui20:09
nessitaralsina: I'll play with the qss then20:09
ralsinanessita:  can you push it? I can take a look20:09
nessitaralsina: I'll try for a couple of minutes and then I'll seek your help if I don't succeed, sounds ok?20:10
ralsinanessita: cool20:10
nessitaralsina: did you run tests for gatox's branch re: icons? I get failures... a lot20:10
nessitaralsina: but maybe is my silly env20:11
ralsinanessita: let me re-check, I think I did20:11
ralsinanessita: got the link handy?20:11
nessitagatox: other than the tests failing for me, +120:12
ralsinahe, I have like 100 devices configured, and the devices tag seems to call find_credentials once for each20:13
gatoxnessita, what failures?? that branch doesn't touch any code (except for setting the stylesheet... that is working...)20:13
ralsinaMy account is becoming a collection of corner cases20:13
ralsinanessita: indeed that's why I didn't run the tests.20:14
ralsinaI am running them now though20:14
nessitagatox: I know, that's why I will approve if they run ok for ralsina :-)20:14
gatoxnessita, ok... anyhow... weird :S20:15
gatoxnessita, can i move both branches to approved?20:15
nessitagatox: both? no! I haven't look the second20:16
alecuralsina, you'll like this branch: https://code.launchpad.net/~alecu/ubuntuone-client/ignore-missing-signal-handlers/+merge/7062620:16
gatoxnessita, ahhhhh ok20:16
ralsinanessita: tests pass just fine20:16
nessitaralsina: ack approving20:16
ralsinaalecu: yes I will!20:16
ralsinaalecu: anyway, that signal is triggered by removing an accepted share from the web, so it has to be implemented eventually20:17
nessitagatox: you can now approve the icon-signin20:17
gatoxnessita, awesome20:17
ralsinaalecu: OTOH it only makes a difference if you had control panel open on the folders tab when you removed the share and expect it to update, so not really important at all ;-)20:18
ralsinaalecu: makes a *big* difference, thanks :-)20:19
alecuralsina, great. It seems like the cleanest solution so far.20:19
ralsinaalecu: indeed . +1 after IRL testing from me20:20
ralsinaalecu, nessita, gatox: I will build the installer late tonight. I would love if each of you sends me an email with branches you want in it. For example, nessita, you may want your improve-ui branch even if it's not merged, so design can look at it and have feedback for you on monday.20:21
nessitagatox: can you please add tests to the last branch you proposed?20:21
nessitagatox: so we ensure titles are set now and always in the future20:22
gatoxnessita, ok20:22
gatoxralsina, ok20:23
ralsinaso, alecu, nessita, gatox: consider that as part of your EOD email reports :-)20:24
alecuralsina, I've got only the branch you've just reviewed to include in today's installer20:25
ralsinanessita: also, remember https://code.launchpad.net/~ralsina/ubuntuone-client/find_config_files/+merge/70472 please20:26
nessitaralsina, gatox: can I set style to a Line?20:26
alecuralsina, and right now I need to go fetch Amelia from kinder, and do some errands; will work a bit more tonight, but lateish20:26
gatoxnessita, a line is a qframe20:26
nessitaralsina: what do you want me to prioritize? ui or the review (I have little time left)20:26
ralsinaalecu: tha's ok20:26
gatoxnessita, do you mean a line separator?20:26
ralsinanessita: I' d say the UI, I can put my branch in anyway20:26
nessitagatox: I mean the thingy that is called "Horizontal Line" in designer, but saying that is a qframe gave some ideas of what's wrong20:27
gatoxnessita, yes, that line is a qframe20:28
nessitaralsina: there was a border: none for QFrame.... who on earth will think that a line is a qframe?20:28
ralsinanessita: well, there is no Line widget. That's weird20:28
ralsinaand QLine is something completely different. Welcome to the part of Qt that is not well documented :-(20:29
nessitaralsina, gatox: I'm using this qss20:33
nessita     39 QFrame#account_separator {20:33
nessita     40     border-style: dotted;20:33
nessita     41     border-color: #939389;20:33
nessita     42     border-width: 1px;20:33
nessita     43 }20:33
nessitafor the line, but it looks ugly, see http://ubuntuone.com/p/18cS/20:33
nessitathe line is  "too tall"20:33
* gatox looking...20:34
ralsinause border only on the bottom edge20:34
nessitashall I set a border-height? that may make no sense20:34
nessitathat makes sense20:34
gatoxnessita, yep20:35
nessitaralsina: all screenshots updated in the shares20:40
ralsinanessita: cool20:41
gatoxnessita, tests added20:41
nessitaralsina: let me know if that works for you, and I'll push the changes20:41
ralsinanessita: looking...20:41
ralsinanessita: much nicer20:41
ralsinalisette is probably going to complain about the vertical alignment but let's wait until she does :-)20:42
ralsinaAnd the labels are misaligned on the left side, if you have time to fix it. If you don't, I said nothing ;-)20:42
nessitaralsina: I was hoping to get feedback before and have a talk with her, I have more question about padding and aligmens20:43
nessitaralsina: so, before applying any style to that, I prefer talking to her20:43
ralsinanessita: yes, those things have to be reviewed by them, but I meant alignment between the email/name and the plan20:44
ralsinawhatever the padding those three are going to be aligned, I think20:44
ralsinaBut yes, I'll mention in the mail for lisette that we need padding/alignment reviews all across u1cp20:45
ralsinanessita: but hey, good work!20:46
* ralsina is the crazy widget alignment guy, if that makes you feel any better20:46
nessitagatox: I'm not so happy with the adding of the CurrentUserControllerValidationMockTest, but I will not make you change it20:47
nessitagatox: IRL tetsing and approving20:47
gatoxnessita, i did it in that way because all the other test use mocker to test that kind of thing... :(20:47
nessitagatox: is ok :-)20:49
nessitagatox: you up for a review?20:49
ralsinagatox: in general we are trying to not use *more* mocker, but one more won't hurt. Too much. ;-)20:49
gatoxnessita, ok..... lesson learned20:50
nessitagatox: you up for a review?20:50
gatoxnessita, yes20:50
nessitagatox: https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/improve-ui/+merge/70611 thanks!20:51
gatoxnessita, did you approved the branch?20:52
nessitagatox: the missing-titles? running tests in windows noq20:52
gatoxnessita, ahhhhhhhhhhhh okok20:52
nessitagatox: hum that branch will not go thru tarmac, see:20:53
nessita    245:  [E1101, CurrentUserControllerValidationMockTest.test_setup_ui] Instance of 'CurrentUserControllerValidationMockTest' has no 'get_backend' member20:53
nessita    248:  [E1101, CurrentUserControllerValidationMockTest.test_setup_ui] Instance of 'CurrentUserController' has no '_set_up_translated_strings' member20:53
gatoxnessita, let me check20:53
nessitagatox: and I'm getting this failure in windows, not if it's legit though20:54
dobeyhave a good weekend everyone!20:55
nessitadobey:  you yoo20:56
nessitayou too20:56
nessitaralsina: no need to change, but for future reference, there should be no blank line between these twoL21:00
nessita45+from twisted.trial.unittest import TestCase21:00
nessita47+from ubuntuone.devtools.testcase import skipIfOS21:00
ralsinareally? ok21:01
nessitaralsina: what I will ask you to change is the duplicated import of "import os"21:01
ralsinanessita: sure, missed that one21:01
nessitaralsina: yeah, when importing modules, you need to use this layout: http://pastebin.ubuntu.com/659540/21:04
nessitaralsina: and in this case, devtools is a 3rd app for us21:05
ralsinaok, treated it as "part of the project". Will fix it.21:05
gatoxnessita, i've uploaded some changes in the branch for missing titles... all tests are passing for me21:07
nessitagatox: ack21:07
=== m_conley is now known as m_conley_away
gatoxnessita, have you updated the style of the buttons in your last branch?21:09
nessitagatox: nopes, I will leave that you probably :-)21:09
gatoxnessita, ok..... just checking21:10
ralsinanessita: pushed with the import fixes21:10
nessitaralsina: ack, I will have to finish the review on Mon21:11
nessitaralsina: I will finish gatox's and then I have to run to a birthday party :-)21:11
ralsinanessita: if that's all you had, can I get the approve? If you didn't finish testing it's ok, we'll leave it for monday21:12
nessitaralsina: I didn't even finish looking at the diff21:12
nessitaralsina: I stopped at the double import21:13
gatoxnessita, i'm finishing with your branch21:13
ralsinanessita: ok, monday it is21:13
gatoxnessita, +121:13
nessitagatox: thanks!21:13
nessitagatox: sdid you see the failing test I have? does it make any sense?21:13
gatoxnessita, yes, i saw the pastebin... but It doesn't make sense at least for me... i'll try to find what that is21:14
nessitagatox: is ok, do not waste time21:15
nessitagatox: my env may be funky21:15
nessitagatox: approving21:15
nessitaralsina: if you review/approve https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/improve-ui/+merge/70611, we can land this soon21:16
=== yofel_ is now known as yofel
ralsinanessita: I am looking at it. I see weird stuff in those buttons we talked about earlier.21:17
nessitaralsina: weird stuff such as?21:18
ralsinain designer the buttons are not the same width21:18
nessitaralsina: have a screenshot?21:18
ralsinanessita: uploading...21:19
ralsinanessita: http://screencast.com/t/regrVTSrVPkR21:19
nessitaralsina: wait! I was missing a commit + push :-/21:19
nessitaralsina: that's the old code... sorry21:20
ralsinaok! np21:20
nessitaralsina: I just saw I had the commit-editor opened :-/21:20
nessitaPushed up to revision 198.21:21
ralsinanessita: on windows it's worse, the commit editor goes to background automatically, so the commit message is never saved. It's -m or nothing :-)21:21
diverse_izzueis the evolution contact backend currently supposed to work in oneiric or not?21:21
ralsinanessita: I usually use a QGridLayout and then a vertical layout on the window, because if you use a grid layout on the window, you can't "see" the layout (or do things like change spans)21:23
ralsinabut that's just a personal preference21:24
gatoxralsina, +121:24
ubuntunoobi am having problems with the nightly ppa21:25
ubuntunoobcan someone help?21:25
ralsinanessita: also, I like putting placeholder texts on the widgets, even if it means cleaning them in code, because it makes manipulation in designer easier21:25
ralsinanessita: so, other than those style issues, everything looks ok.21:27
nessitaralsina: ack to your personal preferences. Not sure if you're asking me to change something though...21:28
ralsinanessita: I think it makes it easier for the next person who has to touch it, so yes, I would prefer if you changed it.21:28
ralsinaotherwise, there are invisible widgets in an invisible layout :-)21:29
ubuntunoobis anyone here?21:29
ralsinaubuntunoob: sorry, the support staff has left for the day21:29
ubuntunoobwhat do i do then21:29
nessitaralsina: add text to the labels, you mean? I think is out of the scope of this branch, and also my personal preference is not to set text to labels, so I would like to have that conversation in other moment, if possibe21:29
ralsinaubuntunoob: and we are the developers and are trying to wrap up too. Give me 5' and I will try to help21:29
ubuntunoobok thanks21:30
ralsinanessita: yes, since that was already that way, I'm ok with not changing it. But the layout should change21:30
nessitaralsina: I think adding a grid on top of the form  adds a layer of complexity that is not needed...21:30
ralsinanessita: the grid is there, it's just invisible21:31
nessitaralsina: right, so why adding another one?21:31
nessitajust to be able to see it? :-/21:31
ralsinanessita: because it's invisible. I don't say have two grids21:31
ralsinaI would use a grid (visible), put the bottom expander outside of it, and use a vertical (invisible) layout on the window21:31
nessitaralsina: if I understand properly, you're telling me to have the form layout vertically and add a grid on top of that, right?21:32
ralsinanessita: yes. Because there are things you can't do with the invisible grid21:32
ralsinasuch as change spanning21:32
nessitaI don't see any gain to that, actually I saw that stuff before and that confused me a lot21:32
nessitaralsina: anyways, I gotta go, I guess we'll keep talking about this on Monday21:33
ralsinanessita: ok, monday.21:33
ralsinabut how an invisible grid is less confusing than a visible one is bvaffling to me :-)21:33
nessitaralsina: I use the tree widget a lot21:33
nessitaralsina: having an extra layer there is confusing, is like is adding some funct5ionality but then you realized is not needed21:34
nessitais not strictly needed, I mean21:34
nessitaso it took me a long time realize that the extra layer was not needed21:34
nessitaand that the form itself can have a layout according to the widget's need21:34
ralsinanessita: not having it is bad for maintainability of the thing. Could you trust me that have been doing Qt stuff for 12 years this once? ;-)21:34
nessitais not lack of trust, really. I don't see any gain, but confusing stuff.21:36
ralsinaok, I suppose you can't, I'll try to convince you on monday. Have a nice weekend!21:36
ubuntunoobwow 12 years?21:36
ralsinamore like 15 but don't want to brag21:36
ubuntunoobthats amazing21:36
ubuntunoobwell i think it s21:36
nessitaralsina: anyways, I tried to review your branch while you try to convince me, and I see that you added windows specific stuff to a directory that is multiuplatform21:36
ralsinano, it's what happens when you are *old* :-D21:36
ralsinanessita: ok, put a needsfixing, will see it on monday21:36
nessitaralsina: TestBaseDirectoryWindows shoudl be inside tests/platform/windows/<something>21:37
ralsinanessita: ok, that's an easy one, at least21:37
ubuntunoobi have a question for you guys21:37
ralsinaubuntunoob: shoot!21:37
nessitaok, I'm off21:37
nessitabye all!21:38
ubuntunoobwhy do you only make it for windows ?21:38
ralsinaubuntunoob: what thing?21:38
ubuntunoobwhy not make a mac client21:38
ubuntunoobfor ubuntu one21:38
ralsinaubuntunoob: we haven't even released windows yet! :-D21:38
ubuntunoobi know its a beta21:38
ubuntunoobbut i was just curious21:38
ralsinaubuntunoob: and a mac client is considered. But there are over 400 million windows users. And about 35 million mac users.21:39
ubuntunoobso bigger userbase21:39
ralsinaubuntunoob: yes, pretty much21:39
ubuntunoobdo you guys support the nightly ppa here21:40
ubuntunoobbecause im having some problems installing it21:40
ralsinaubuntunoob: what kind of problem21:40
ubuntunoob  ubuntuone-client-gnome: Depends: ubuntuone-client (= 1.2.2-0ubuntu2) but it is not going to be installed21:40
ubuntunoobE: Broken packages21:40
ralsinaubuntunoob: probably a broken build21:41
ralsinaubuntunoob: you should ask dobey but he already left21:41
ubuntunoobwell is there another ppa for the nightlies i could use?21:41
ralsinaubuntunoob: nope, nightlies is what it is, and occasionally it breaks21:41
ubuntunoobok i understand21:42
ubuntunoobneed to backup my ipod music somehow, any suggestions?21:42
ralsinaubuntunoob: so, if it's not better tomorrow morning, come here, and ask for dobey, rye or duanedesign :-)21:43
ralsinaubuntunoob: you can still use the web ui21:43
ralsinabut of course uploading files one at a time is a pain21:43
ubuntunoobyes it is21:43
ralsinafor a backup you could zip them21:43
ubuntunoobi just about to say that lol21:43
ubuntunoobi think thats what i will do21:44
ubuntunoobfor all your help21:44

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