/srv/irclogs.ubuntu.com/2012/06/25/#ubuntuone.txt

mandelmorning all!08:52
ryec_smith: hello, are you still here? Could you provide more information about the crash?09:50
=== teknico_ is now known as teknico
=== zyga_ is now known as zyga
gatoxgood morning!11:01
gatoxbrb11:10
gatoxback11:24
alecuhello all!12:26
gatoxalecu, hi12:28
mandelalecu, gatox, morning!12:30
mandelalecu, alecu, I have one more review and all the fsevents changes will be merged! so I'll do revies for gatox and will propose my branch12:31
gatoxmandel, ack12:31
mandelgatox, how is the change that was driving you crazy?12:31
gatoxmandel, i had some issues, it's already working.... you can see it in the MP12:32
mandeloh, and my internet is super crap today.. so expect me to disappear every now and then..12:32
mandelI'm off to lunch and will be later back from 3g... puto movistar de los huevos!12:38
gatoxack12:39
* mandel goes to have lunch cursing..12:39
dobeylol12:56
=== yofel_ is now known as yofel
dobeyand it's release day13:10
alecumandel, gatox: send reviews my way13:18
dobeyhaha13:18
dobeybug #1017220 is awesome13:18
gatoxalecu, https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin2-fsevents/+merge/11142713:18
ubot5Launchpad bug 1017220 in ubuntuone-control-panel (Ubuntu) "Welcome message displays incorrectly if user has not entered name" [Undecided,New] https://launchpad.net/bugs/101722013:18
gatoxalecu, https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin3-fsevents/+merge/11166613:18
gatoxalecu, and i'm fixing the test in the last one..... i learn to never say again: "this should be easy" (refactor related)13:19
alecugatox: great!. I need to go to the paint shop, because my flat is getting some fixes... and I'll do that review in 15'13:23
gatoxalecu, no problem13:23
=== zyga is now known as zyga-food
mandelalecu, gatox, ralsina, dobey I'm off some mins to move to a diff location with better connection13:50
mandelshould just take min 10 mins max13:50
gatoxmandel, ac13:50
ralsinadobey: since we still have not SRUd 3.0.2 and we barely released it on windows, what o you think of skipping a release?13:50
dobeywhat do you mean skipping a release?13:52
ralsinadobey: you said it's release day. We barely finished the previous release13:53
ralsinadobey: so, maybe there's no point in doing a new release today13:53
ralsinabriancurtin: ping13:54
briancurtinralsina: pong13:54
ralsinabriancurtin: my XP VM has decided my windows is fake13:54
ralsinabriancurtin: could you check that our latest release works on a clean XP?13:54
briancurtinralsina: checking now13:54
ralsinabriancurtin: if not, I will do it after I finish reinstalling13:54
ralsinabriancurtin: thanks. By the symptoms described, I think the VS runtime is not being installed :-(13:54
briancurtinralsina: it's packaged side-by-side now (i think that's the right term)...but im getting the VM up right now13:56
ralsinabriancurtin: ack13:56
briancurtinralsina: my VM already has the runtime installed and i cant find a way to get it off of there, so i'll need to setup a clean VM to try it14:06
dobeyralsina: well, unless there's a security update coming out of nowehre tomorrow, i'd prefer to do the release14:07
ralsinabriancurtin: ok, I'll do the same thing14:07
ralsinadobey: ok, then14:07
dobeyralsina: even with no changes, i think there is a point to doing the releases14:07
dobeyand with no changes, the release is really fast anyway :)14:07
dobeysince there's no mucking about with triage and backport branches in that case14:07
ralsinadobey: plus we may get to test the MRE with 3.0.3 instead of 3.0.214:07
dobeytoday isn't a 3.0.x release day14:08
=== mandel_ is now known as mandel
dobeyit's 3.99.x14:08
dobeyfor quantal14:08
mandelI'm back!14:08
dobeyhide your children!14:08
* mandel cannot believe that 3g is faster than the dsl connection14:08
* ralsina goes to check back the schedule14:09
ralsinadobey: ok, 3.99.x is cool14:09
dobeywe still don't have a schedule for 3.0.x releases14:10
dobeyalecu: btw, the ubuntu-sso-client stable-1-4 tests aren't running on oneiric (because we can't run tests on oneiric because of twisted); all the oneiric stable branches are being merged on natty because of the twisted/python-gobject issues that weren't resolved in oneiric14:13
ralsinabrilliant, my *physical* XP CD doesn't work anymore :-/14:13
mandelralsina, alecu, I'll be working on landing the keyring patched on pykeyring trunk f that is ok with you14:16
mandelgatox, is there anything I have to review?14:16
gatoxmandel, this: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin2-fsevents/+merge/111427  -  https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin3-fsevents/+merge/11166614:17
mandelgatox, ack14:17
=== zyga-food is now known as zyga
mandelgatox, I'd love to see how things were breaking when you added the events as class attrs14:26
briancurtinralsina: doesn't work on XP.................14:26
ralsinabriancurtin: could you try if installing the runtime fixes it?14:26
gatoxmandel, is just a comment? or you really want to see it?? because that was on friday.....  i didn't save any trace, and i should redo all the code to reproducce that14:27
ralsinabriancurtin: if it does, then we can revert the ubuntuone.xml changes, resign the installer and reupload14:27
mandelgatox, was not just a comment, but I guess I can take a look from here if that is the case (that you do not ave the traces)14:28
mandelgatox, I think it should be doable without doing any hacky things, I simply want to evaluate the problems, that is all14:28
gatoxmandel, ack14:28
mandelbriancurtin, I added a small needs fixing in one of the py3 branches, I think dobey is correct regarding the order of the imports14:28
briancurtinralsina: 3.0.1 doesn't work either14:31
ralsinabriancurtin: whoa14:31
ralsinabriancurtin: 3.0.0? ;-)14:31
dobey2.99.3?14:31
dobeyor 2.0.3 or whatever it is14:32
briancurtinhmm, and 3.0.1 was the last release where we ran vcredist as an installer step14:32
briancurtinweird14:32
ralsinabriancurtin: indeed14:33
ralsinabriancurtin: did installing vcredist fix it?14:33
ralsinabriancurtin: because it may be completely something else14:33
briancurtinralsina: i sort of did that in a way by trying 3.0.1 since it installs that and it doesnt work. i'll try 3.0.2 now that 3.0.1 installed vcredist to see what happens14:34
ralsinabriancurtin: ack14:34
mmcchappy monday, folks14:36
dobeymeh, alpha freezes still on14:39
briancurtinralsina: its gotta be something else. now that the machine has the redist installed, neither 3.0.1 nor 3.0.2 work14:43
ralsinabriancurtin: sigh14:43
ralsinabriancurtin: is it the right vcredist?14:44
ralsinabriancurtin: what size is it?14:44
briancurtinyep14:44
briancurtinralsina: 1779 kb14:44
briancurtinralsina: for vcredist_x86.exe14:45
ralsinabriancurtin: ok, let me check something14:45
briancurtinralsina: i just ran a repair install of vcredist and it claims to have worked, but u1cp still doesn't start14:45
ralsinathat seems to be the right one AFAICS14:46
ralsinabriancurtin: no, that one's too new14:46
briancurtinralsina: that's the same one ive been packaging all along14:46
ralsinawe need the 1.1MB vcredist_x86.exe14:46
ralsinabriancurtin: yeah, it's the same one I have uploaded14:47
ralsinabriancurtin: but it's not the one my notes say14:47
ralsinabriancurtin: and I have no idea why, either14:47
briancurtinralsina: i just looked back and 3.0.0 was created with the 1779kb vcredist. it looks like i deleted the 3.0.1 folders14:48
briancurtini dont have any 2.x installer folders around if i ever created them, but i only seem to have downloaded that 1779kb vcredist_x86.exe one time and used it everywhere14:49
ralsinabriancurtin: I don't have any copies of the smaller one, either14:49
briancurtini'll try to find this 1.1mb version14:49
ralsinabriancurtin: however, I even blogged about it http://lateral.netmanagers.com.ar/weblog/posts/BB955.html14:49
ralsinabriancurtin: and I specifically say "get the 1.1MB one"14:50
* ralsina hates ralsina sometimes14:50
mandelmmcc, I did an update for: https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/less-error-prone/+merge/11100014:52
mandelgatox, if you can take a look at that one ^ I would appreciate it :)14:52
ralsinabriancurtin: latest version is 4.8MB too14:53
mmcccool mandel, thanks, I'll go look14:53
gatoxmandel, sure14:53
mandelcheers14:53
briancurtinralsina: i found the SP1 version at 4.0mb14:54
ralsinabriancurtin: the 2005 version is 2.6MB14:54
ralsinaand the 2.8 is 1.7 which is probably the one we have14:54
ralsina200814:54
briancurtinralsina: we would need 200814:54
ralsinabriancurtin: let's try the 4MB one then14:55
ralsinabriancurtin: the 2.99.1 build was done with the 1.8MB one, too14:57
ralsinabriancurtin: so it may be something else entirely14:57
ralsinabriancurtin, mandel, dobey, alecu,gatox, thisfred, mmcc: stanup in 2'14:59
gatoxack14:59
thisfredsí jefe14:59
mmccmandel, +114:59
gatoxme15:00
thisfredme15:00
briancurtinme15:00
mmccme15:00
ralsiname15:00
dobeyme15:01
dobeyalecu, mandel?15:01
alecume15:01
mandelme15:01
gatoxDONE:15:01
gatoxCouple of reviews, submit some branches for review (and add some corrections), working on the last branch.15:01
gatoxTODO:15:01
gatoxFinish fixing the broken tests in the last branch and propose15:01
gatoxBLOCKED:15:01
gatoxNo15:01
gatoxthisfred, go15:01
thisfredDONE: Bug #999569 Bug #999562 TODO Bug #999562 | Dentist apt. at 4 (now +5 hours), will work a little longer to make up for it BLOCKED: no, NEXT: briancurtin15:02
ubot5Launchpad bug 999569 in U1DB "invalid json errors. ensure its a json object" [High,In progress] https://launchpad.net/bugs/99956915:02
ubot5Launchpad bug 999562 in U1DB "retry logic on 503 " [High,Confirmed] https://launchpad.net/bugs/99956215:02
briancurtinDONE: python3 research/organization, also got antsy and started doing easy ports after looking at the 2to3 diff15:02
briancurtinTODO: find out whatever happened to XP releases, fix that15:02
briancurtinBLOCKED: XP sucks15:02
briancurtinNOTE: i have a PSF board meeting in 1 hour, should last for around 30 minutes15:02
briancurtinNEXT: mmcc15:02
mmcc DONE: WWDC videos & 10.7/8 docs, uncovered some interesting FSEvents changes in Lion15:02
mmcc TODO: chat about FSEvents? more killing imports that install reactors...15:02
mmccBLOCK: none15:02
mmcc NEXT: ralsina15:02
ralsinaDONE: worked a bit more on cmake for u1db, askubuntu, canonicaladmin, misc stuff, worked on setting up a lot of VMs, Q VM  failed big time, retrying TODO: finish Q VM, try to help debug windows breakage, finish cmake'ing  u1db, BLOCKED: VBox is acting up badly NEXT:  dobey15:02
dobeyDONE: reviews15:02
dobeyTODO: release day15:02
dobeyBLCK: None.15:02
dobeyalecu15:02
alecuDONE: a few reviews, a fix for bug #1007109, tried to debug (and failed) why this branch won't merge: https://code.launchpad.net/~alecu/ubuntu-sso-client/use-pycurl-1-4/+merge/11063615:03
alecuTODO: heed dobey's warning re: merge, catch up with py3k with brian, do more mac reviews15:03
alecuBLOCKED: by non-merging branch15:03
alecuNEXT: mandel15:03
ubot5Error: Launchpad bug 1007109 could not be found15:03
mandelDONE: reviews, reviews, reviews. Look at the state of pykeyring, not happy. Update fsevents last branch (we stay with sbjson mmcc :) )15:03
mandelTODO: more reviews. What to do with pykeyring.15:03
mandelBLOCKED: no yet stupid ISP left me with no internet at the office.15:03
ralsinabriancurtin: it may be that we are doing it now with a new python build and the runtime is not compatible?15:03
mandelmmcc, superb, thx!15:04
ralsinabriancurtin: OTOH, in the machine where you have the runtime it works, which makes it more confusing to me15:04
mandelmmcc, we are very very close to get a sd running with no problems on mac :)15:04
briancurtinralsina: the only place it seems to work for me is 715:04
ralsinabriancurtin: ok15:04
alecummcc: "uncovered some interesting FSEvents changes in Lion" ???? tell us more!15:05
briancurtinralsina: we build the exes with a stock python 2.7 which is built with VS2008. trying the bigger/newer vcredist now15:05
ralsinabriancurtin: ack15:05
mmccalecu: the short version, since Lion, FSEvents can give you per-file events. They just didn't really tell anyone15:05
mmccI was watching the video for "what's new in the filesystem" at WWDC2012, and the guy mentioned it off-hand15:06
briancurtinthose are the best kind of changes. "here's something people want...*fixes it, doesn't document it*"15:06
mmccso I dig around, and it's listed in the API changes, but not documented15:06
alecummcc: is that on 10.8?15:06
ralsinammcc: WTF15:06
mmcc10.715:06
aleculovely15:06
mmccralsina: I KNOW15:06
* ralsina goes check if his computers makes waffles and they just forgot to mention it15:07
alecugatox: now we'll have to contribute fixes for macfsevents to do that15:07
mmccthe kicker is that they may still coalesce events in some cases. I'm digging up the little snippet of docs I found about that15:07
ralsinabut hey, good news :-)15:07
mandelralsina, mmcc, alecu so all the work we did is just for 10.6?15:07
mmccI want to run a test once I get my new system with 10.7 (a couple days) - how often do they coalesce events?15:07
briancurtinralsina: 2008 SP1 redist doesn't help...it has to be something else15:07
gatoxalecu, it seems15:07
ralsinabriancurtin: ok15:08
briancurtinor it doesnt have to, but this seems too weird15:08
mmccmandel, maybe - if they coalesce events often, it might still have a lot of overhead to use fsevents15:08
alecuralsina: btw: what was the problem with the Q vm?15:08
ralsinaalecu: it doesn't freaking boot after install15:08
alecuralsina: is that OpenGL/unity related?15:08
ralsinaalecu: also, install took 3.5 hours15:08
ralsinaalecu: what I get for using a daily ISO15:09
mandelmmcc, hm.. well, if it is the case (that we did it for nothing) we deserve it for using a not documented api15:09
alecuralsina: did you try the alpha 1 as well?15:09
ralsinaalecu: I am downloading it, 80% done15:09
* briancurtin brb15:10
mmccmandel heh, the new API is also not documented... :p15:11
mandelmmcc, lol15:11
mmccso it's in the auto-generated header documentation but not the written guide to using fsevents. Here's the headerdoc: http://developer.apple.com/library/mac/#documentation/Darwin/Reference/FSEvents_Ref/FSEvents_h/index.html#//apple_ref/c/econst/kFSEventStreamCreateFlagFileEvents15:12
mandeldobey, pyflakes and python-mocker are not installed when doing sudo apt-get build-deps ubuntuone-client, is that a bug?15:12
dobeynot installed where?15:13
ralsinammcc: that feels just like linux. Really.15:13
mandeldobey, in P15:13
dobeymandel: no, it's not a bug really. you don't have ppa:ubuntuone/nightlies added i guess15:14
mmccralsina: except with linux at least we could've looked at the source. I'm quite fond of looking at the source...15:14
mmccused to do that with opensolaris a lot, handy.15:14
ralsinayeah15:14
mandeldobey, oh..  that could be the reason, is a new set up15:14
dobeythat's it then15:14
ralsinaI think I will go have some coffee.15:14
mmccso that fsevents header comment doesn't mention coalescing. I know I saw it, but my browser session died. looking through history now15:15
dobeyneed to get lunch. bbiab15:16
mmccah, ok - it's just down the page - the flag kFSEventStreamEventFlagMustScanSubDirs tells you that you get a dir path and you have to scan it and every subdir below it15:17
mmccso if that happens a lot, it's not much better than old FSEvents15:17
mandelmmcc, they must have something for icloud to deal with this, right?15:19
mandelgatox, the only issue I see with the fact that the events are not class attrs is that we will have an interesting problem getting your work and my work together. I use diff values comming from the daemon15:26
* briancurtin back15:26
mandelgatox, and we will not know which implementation will work until we try to connect to the daemon and get an error15:26
gatoxmandel, maybe i was doing something wrong... my brain wasn't in his best shape15:27
mandelgatox, not to worry, I can take care of that :)15:27
mandelgatox, your brain I mean, you do the code!15:27
mandelheheh15:27
gatoxjajajaa15:28
* gatox lunch15:29
=== gatox is now known as gatox_lunch
mmccmandel, just saw your note - what do you mean about icloud? I don't think it necessarily uses fsevents15:30
mandelmmcc, can I have the +1 for https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/less-error-prone/+merge/111872 I had to remove the prerequisite and it remove your vote.. stupid lp15:30
mmccspotlight does, and time machine does, but icloud support seems to be built in to the frameworks, above the filesystem15:31
mmccmandel ok, just a sec15:31
mandelmmcc, they most know some how what files changed.. well to be hones I have never used icloud15:31
mmccmandel - icloud isn't a file sync service. think of it as being designed for the iphone first, where there is no user-visible file system15:31
mandelmmcc, ah.. so we are not in the same boat, right>15:32
mandel?15:32
* mandel has new keyboard..15:32
mmccright, u1 and icloud have different goals15:32
mmccvery different technically, perhaps less different to the user, depending on what kind of user15:32
mmccactually i think u1db and icloud potentially have more in common, although icloud is much bigger - they have a key-value store as one part of it, but there's also document-level stores and a core-data store, which is pretty interesting15:35
ryeralsina: ubuntuone nautilus plugin does not work properly for file publishing - the flaw is in libsyncdaemon and plugin (anybody actually depends on libsyncdaemon?) - any plans to make it work?15:36
ralsinarye: libsyncdaemon is used only for that plugin AFAIK15:36
ralsinarye: have the bug # handy?15:37
mandelmmcc, we had a key-value store using couch, we will have a new one soon :)15:38
ryeralsina: bug #753319 and bug #97598915:39
ubot5Launchpad bug 753319 in Ubuntu One Client "nautilus plugin requests public file list for every file in current UDF/Ubuntu One dir" [High,Confirmed] https://launchpad.net/bugs/75331915:39
ubot5Launchpad bug 975989 in Ubuntu One Client "[precise] PublicAccessChanged is not sent on unpublishing - TypeError: Expected a string or unicode object" [Undecided,Triaged] https://launchpad.net/bugs/97598915:39
mmccmandel are you referring to u1db or something else I'm not aware of?15:39
mandelmmcc, yes, u1db, we used to have desktop-couch but that did not work well :(15:39
ralsinarye: k, I will try to take a look myself this week, and if I can't fix it, I'll try to see if dobey has any time for it next week15:39
ralsinarye: 975989 looks easier15:40
mmccmandel right, I knew about that project before and I've seen some old discussion about it around... nice idea, bummer it was so painful15:40
mandelmmcc, yep15:40
ryeralsina: yes15:41
mandelgatox_lunch, ignore the diff size of https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/less-error-prone/+merge/111872 it seems that readding sbjson has made bzr get very very confused15:42
mandelfacundobatista, quick question, if I set a file to be chmod u-w and then I ty to sync it, what should it happen? atm I think we fail due to the rights15:44
mandelgatox_lunch, there seems to be a merge issue in the other branch, set it to needs fixing15:49
gatox_lunchmandel, ack15:50
nigelbwhere's sil?15:51
thisfrednigelb: in a meeting, away from his laptop15:59
nigelbthisfred: ouch, he has a classroom session starting now-ish15:59
thisfrednigelb: I forwarded the message best I could16:00
nigelbthisfred: thanks, I think he may have rescheduled just now.16:00
thisfredah16:00
nigelbthisfred: apparently not. the rescheduled list doesn't ahve him16:02
nigelbHe's coming in I guess :)16:02
nigelbthisfred: thanks a bunch!16:02
thisfrednp :)16:03
nigelbyay aquarius!16:06
aquariusmhall119, ping16:07
aquariusI didn't think I was doing this session :(16:07
mhall119aquarius: can you?  even if it's just an overview of the REST apis?16:08
aquariusmhall119, what do I have to do? I haven't prepped, but I can wing it16:08
aquariusdo I need to start now?16:08
* briancurtin afk for a short bit, PSF board meeting call16:08
mhall119aquarius: I'll start a hangout and invite you to it, then you just chat and screenshare16:08
mhall119aquarius: https://plus.google.com/hangouts/_/dcb85bf832b154aef74b1e5c5ac26057071f4d10?authuser=0&hl=en16:09
facundobatistamandel, we fail to sync it because we can not *write*?16:11
mandelfacundobatista, I wonder, can we not try to set the rights to rw like we do in the read only shares?16:11
mandelfacundobatista, or I think we should have a way to tell the user about it, that is, wrong rights and it wont be updated, do we have something like that?16:12
dobeyi haven't had time to move it yet, but i was planning to rewrite/move libsyncdaemon into libubuntuone this cycle16:18
dobeyrye, ralsina: ^^16:18
dobeyof course, that doesn't really help existing installs16:19
ralsinadobey: we should try to find a way to make it happen or fix the bugs. I suspect fixing *all* the bugs is more expensive :-/16:19
dobeyright. and i don't see any way to fix all the bugs without the rewrite anyway16:23
ralsinaok16:24
ralsinabut this would only help us in Q16:24
ralsinaunless we do something very difficult about P16:25
dobeyright16:25
ralsinaand we are to support P for 5 years16:25
dobeywe need to do that very difficult thing for P at some point anyway I think16:25
dobeysupporting 3.0.x for 5 years is not going to be fun16:26
dobeyjust like supporting 1.2.x for 3 years isn't fun16:26
ralsinaso let's aim for it16:27
ralsinadobey: a firefox-style SRU16:27
dobeywell, we are working toward it.16:28
ryedobey: agree with you re: rewriting.16:28
ralsinaexactly16:28
ralsinaso, we should rewrite and after Q release try to get it back into P16:28
dobeyyep16:28
dobeythere are some things that need to happen for that still16:29
=== gatox_lunch is now known as gatox
dobeyand i have already started doing some of it; trying to get the test suites running during build in Q, and having the necessary bits to build some of the packages all the way back to lucid16:29
gatoxmandel, ping16:30
ralsinarye: so, basically, short term I will try to get a fix for the easy one, the fix for the one that causes bad performance/hangs will probably not happen16:30
mandelgatox, pong16:30
dobeythere are a couple of problems with lucid not having things we need now, and such. but we're doing a good start16:30
gatoxmandel, did you fix the conflicts you mention here_: https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/less-error-prone/+merge/111000 ?16:30
dobeyi'll look more after i get the releases done today16:30
ralsinadobey: things that take longer than 6 months are starting to not affect lucid much16:30
dobeyyeah i'm not sure how many users we still have on lucid16:31
dobeyprobably not many :(16:31
mandelgatox, yes, there should be no merge conflits, is just that it was build on top of a previous branch that was removed and that is why the diff is huge16:31
gatoxmandel, ack16:31
dobeynatty is a bit bigger of a problem, but it's also going to be EOL come October anyway16:32
ralsinadobey: exactly16:33
dobeyso maybe i can do some hacking to get 3.0.x built on lucid in the stable PPA, and we just tell people to use that, who are on lucid16:34
ralsinadobey: +116:34
ralsinalucid +ppa sounds like a good idea16:34
gatoxmandel, +116:34
mandelgatox, superb! thx!16:34
dobeyi'll poke at that after i get the releases done today16:35
dobeywhich i can hopefully actually get done today, as i don't have 50 surprises coming at me this week :)16:35
mandelgatox, can I have the +1 here, that MP is superseeded: https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/less-error-prone/+merge/11187216:35
=== beuno is now known as beuno-lunch
=== beuno-lunch is now known as beuno
gatoxmandel, done16:39
mandelgatox, gracias16:39
mandelalecu, can I have a re-review for https://code.launchpad.net/~mandel/ubuntuone-client/better-encoding/+merge/11158016:40
alecumandel: sure. I'll have lunch and work on that.16:41
ryematiasb: the code in lp:~matiasb/ubuntuone-servers/accepting-tos-before-accepting-share - does that mean we'll be able to set up storage for accounts that for some reason don't have one like the lucky ones after the shard went live on 22nd?16:45
matiasbrye: that code would allow to create the root volume for a user that already have a storage user, taking the info from that storage user; this should allow to fix users without root volume related to the accepting share before TOS bug16:48
=== salgado is now known as salgado-lunch
mhall119thanks aquarius, that went well16:58
aquariusmhall119, cheers, pal!16:59
ubot5Announcement from my owner (jussi): #ubuntu-discuss can-voices17:03
leo-unglaubaquarius: according to the ubuntu one session right now. i have implemented a "flat sync" for thunderbird contacts in dropbox. with u1db i think it's very easy to port that to ubuntu one as well. anyting special i should have to keep in mind?17:05
leo-unglaubflat sync -> means i don't care in witch address book the contact is, i simply sync all of them17:06
ryematiasb: re: people w/o userstorage (2221970..2222049) - i am filing a bug, right?17:19
matiasbrye: yes, please17:29
=== salgado-lunch is now known as salgado
mandelok, EOD here, laters!"17:32
gatoxmandel, bye!!17:32
* mmcc is headed for some BBQ17:52
* gatox finally understand the logic behind this method17:54
gatoxnow is time to fix it17:54
alecugatox: you've left me wondering... which is the method that you've understood the logic?18:40
gatoxalecu, jejeej it was a really silly thing...... but i was lost with the tests18:41
gatoxlet me remember18:41
alecugatox: no prob :-)18:41
alecugatox: don't worry!18:41
gatoxalecu, i don't recall exactly, but was something checking that we are ignoring the proper path, and catching the events from another one....... the test was failing weird so i was a little confuse18:42
gatoxand i'm kind of drained this days :P waiting for pycamp!18:42
gatoxthese days18:42
joshuahooverralsina: heh, too many channels18:46
Yanchoguys, i am trying to install ubuntuone on xubuntu. however after installing ubuntuone-control-panel-gtk when i go settings > ubuntu one control panel - the application is crashing .. any idea from where to start fixing please?18:46
joshuahooverralsina: on the qa call just a minute ago we were discussing windows automated testing...we want to run automated tests against each of the major win platforms, which we don't currently do...there were 2 things that came out of that discussion:18:47
joshuahooverralsina: 1) we really need bug #984964 fixed to get testability running on windows18:47
ubot5Launchpad bug 984964 in Ubuntu Single Sign On Client trunk "SSO is not started with the -testability argument" [High,Fix released] https://launchpad.net/bugs/98496418:47
ralsinajoshuahoover: agreed on 1)18:48
joshuahooverralsina: 2) we need a basic validation that the install is good (considering bug 1017019)...we can do this but rmcbride mentioned that maybe there is a way to do this kind of validation as part of the installer build process (automated)?18:49
ubot5Launchpad bug 1017019 in Ubuntu One Client "3.0.2 not run on XP" [High,Confirmed] https://launchpad.net/bugs/101701918:49
ralsinajoshuahoover: running u1sdtool --help should be enough for that one18:49
ralsinajoshuahoover: in fact, I am in the process of trying progressively older versions on a clean XP VM until I found when it started failing18:50
briancurtinjoshuahoover: to get testability going is going to require once again that we drop everything and figure it out entirely on our own with zero documentation18:51
joshuahooverbriancurtin: that sounds delightful...sigh18:51
ralsinajoshuahoover, briancurtin: in fact, I am not even sure if anyone ever got it working on windows :-/18:52
briancurtini know we need to have it, and i want it to work, but its painful and even after however many days i spent on it, even the example Qt project didn't work18:52
joshuahooverbriancurtin: is the issue w/ testability on windows specifically? or testability in general (re: no docs/help)?18:52
briancurtinralsina: i've read that people "got it working" but it's all message board posts or mailing list posts where they ask a question, then they just figure it out on their own, then just say "oh it works now"18:52
ralsinabriancurtin: damn them ;-)18:53
joshuahooverbriancurtin: right, i hear you... elopio and rmcbride have looked at various frameworks over the months and they all have significant drawbacks :(18:53
joshuahooverlove that, "oh it works now"...very helpful insight people provide with follow ups like that18:53
ralsinaspecially since they have this one working on ubuntu18:53
rmcbridewell, given that the top google result in the last year is https://bugs.launchpad.net/bugs/1008583 ...18:54
ubot5Ubuntu bug 1008583 in Ubuntu One Windows Installer "UbuntuOne doesn't work with qt testability on Windows" [Undecided,Confirmed]18:54
rmcbrideI'm guessing having it work isn't common18:54
briancurtinjoshuahoover: the problems i faced were that getting testability built in the first place, but it ends up being pretty easy. then associating it with the Qt install isn't too hard. actually getting it to do anything productive was the part i got stuck on. running the sample Qt calculator app that is supposed to support testability didn't work no matter what i tried18:54
briancurtinrather, it worked, but all it did was add numbers...it wouldn't associate with the testability visualizer18:55
joshuahooverheh, right, got ya18:55
gatoxalecu, another one for review: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/11191418:57
gatoxthis is the last one18:57
joshuahooverelopio: as much as it pains me, maybe we need to look at alternative frameworks again for automation? or at least start looking at other alternatives for windows? at least something to give us an idea of what some options are - spending time on (possibly) getting testability working on windows, or porting all tests to a different framework that works across windows, mac and ubuntu (in that order)...or maybe there is nothing that me18:58
ralsinajoshuahoover, elopio: lots of people recommend squish but it's quite expensive19:04
rmcbrideapparently LDTP has a windows version19:04
rmcbrideas of April of this year Naggapan got permission from VMware to open source it. FWIW. I haven't done anythign with it19:05
ralsinabriancurtin: can you point me to any notes on what you tried with testability?19:05
ralsinabriancurtin: and/or a tarball of your most successful attempt :-)19:06
briancurtinralsina: i don't have any written up, i can write something quick and zip up what's there19:06
ralsinabriancurtin: cool19:06
rmcbrideelopio: joshuahoover: ralsina: FWIW LDTP is not only available for windows now, it's going to be available for Mac "real soon now" https://github.com/nagappan/pyatom/commits/master19:13
joshuahooverrmcbride: hmmm19:13
rmcbrideI haven't looked at it on Ubuntu in a while, since  we were looking for something cross=platforma nd I assumed it would not be19:14
rmcbridehttp://ldtp-soc.blogspot.com/2012/04/announce-windows-version-of-ldtp-gui.html19:14
dobeydoes qt work with it yet?19:16
rmcbridedobey: not at all certain, Jamie gave me a couple links referring to it. I'll ask rather than read19:17
rmcbridedobey: actually it does19:18
rmcbridedobey: forgot that was about the second thing he said to me when I asked about cross platform tools19:18
dobeyhmm, ok19:19
czajkowskikazade: aloha19:20
czajkowskikazade: aq isnt here atm19:20
kazadehey czajkowski19:20
kazadeyeah, I'll just hang around :)19:20
kazadehow's things with you?19:21
czajkowskikazade: good thanks enjoying the tweeting conversation19:21
kazade:)19:21
czajkowskikazade: others in here might be able to help in the mean time until aq arrives19:23
kazadeczajkowski: it's OK, I think I got enough info over Twitter for now19:24
kazadealthough, saying that, I'm sure I'll be back tomorrow ;)19:24
kazadeczajkowski: it's for this: http://i.imgur.com/tYIhX.png19:25
dobeyoh hi urbanape19:26
czajkowskikazade: ah19:26
ralsinabriancurtin: I am setting up a new XP dev vm so I can try building the installer there19:27
dobeyhmm19:27
ralsinabriancurtin: I can find anything about the buildout in https://wiki.ubuntu.com/UbuntuOne/Contribute/WindowsTesting am I looking at the wrong page?19:27
briancurtinralsina: i have a trunk installer built to run on XP, will do it right after i wrap up this wiki page for testability19:27
ralsinabriancurtin: cool19:28
ralsinabriancurtin: I also want to give testability a whack to see if it falls ;-)19:28
briancurtinralsina: https://wiki.canonical.com/UbuntuOne/Windows/ReleaseChecklist contains some buildout-y stuff, installer building19:28
ralsinabriancurtin: awesome thanks19:28
briancurtinwow this wiki software is garbage19:32
briancurtinhey heres an enumerated list, just jam it together on one line!19:32
briancurtinralsina: here's the loose form documentation for right now: https://wiki.canonical.com/UbuntuOne/WindowsTestability19:33
ralsinabriancurtin: thanks19:34
briancurtinralsina: it's probably better that you build them locally rather than me zip things up, but i can do that if you want19:34
ralsinabriancurtin: I'll try19:34
briancurtinralsina: building is the easiest part. doing anything with what's built is the kicker19:34
ralsinahe19:34
urbanapehey, dobey19:36
urbanapecrazy weater.19:36
urbanapeweather even19:36
dobeyyeah, looks like it19:37
dobeyand looks like it will be down here in a couple hours19:37
urbanapemost likely.19:38
urbanapeit's been moving in that direction19:38
mmccralsina, briancurtin: - not sure if it's relevant but when I was working on the PyQt plugin issue I got a lot of mileage out of the QT_DEBUG_PLUGINS=1 env var19:39
mmccralsina, briancurtin also if there's a dynamic loader debug flag along the lines of darwin's DYLD_PRINT_LIBRARIES=1, that was useful too19:40
briancurtinralsina: oh, and if you get everything setup, when you start the visualizer, you may here a ding if you have sound turned on, and apparently there's a popup that shows up for like 1 microsecond and goes away. i tried figuring that out last time i worked on it. maybe this env var will help, not sure19:40
ralsinabriancurtin: to build the plugin is VS express enough?19:40
briancurtinralsina: all you really need is the VS2008 SDK since you just build on the command line using nmake. Express will include what you need19:41
ralsinabriancurtin: awesome19:41
ralsinaWTF there are no binaries of PyQt available for download!!!!19:44
briancurtinralsina: i wish i got as far as even considering python/pyqt in this whole testability debacle...19:44
ralsinabriancurtin: hehe19:45
ralsinabriancurtin: but really, there is no way to install PyQt right now for people who don't have old copies19:45
briancurtinand whoa, you're right there are no binaries19:45
briancurtinoh maybe it's because that version is brand new, as in it just came out yesterday19:46
ralsinayeah, and the previous one had a bad regression19:46
briancurtin(not that creating windows installers should take this long)19:46
ralsinabriancurtin: we can't say that19:46
elopiojoshuahoover, rmcbride, ralsina: ok. I'll check the options again.19:47
joshuahooverelopio: thanks19:47
dobeyspoke too soon. clouds are starting to roll in already19:54
briancurtinralsina: not that i really thought this would work, but trunk installers also do not work on XP19:58
ralsinabriancurtin: are you building them on 7?20:01
briancurtinralsina: yeah i am, i don't currently have a source setup on XP. i could try that, though20:01
ralsinabriancurtin: please try it20:02
briancurtinalso, most everything i've found points to needing vcredist 2008 SP1...which is what ive run and rerun 100 times20:02
briancurtinralsina: will do20:02
ralsinabriancurtin: once upon a time, my installers built on 7 did notwork on xp either20:02
briancurtinralsina: hahahah. running bzr also gets the same problem20:02
ralsinabriancurtin: of course later they did but whatever :-(20:02
ralsinabriancurtin: I got "invalid binaries" with the 2.6 bzr, I had to get 2.520:03
briancurtinralsina: it's not going to start now because of the CRT issue, even though i installed it. this machine is just hosed i think...20:03
ralsinabriancurtin: oh, well20:03
ralsinaI feel like I am waving a beheaded chicken around20:04
ralsinaand not a delicious, deep-fried beheaded chicken20:04
gatoxpeople! eod for me!! see you tomorrow! :D20:07
dobeyralsina: https://code.launchpad.net/~dobey/ubuntuone-storage-protocol/update-4-0/+merge/11192920:08
ralsinadobey: looking20:09
mmccquick style question: anyone have a problem with moving 'from twisted.internet import reactor' in ssoclient's ubuntu_sso/utils/tcpactivation.py from the toplevel into the two methods that use it? It requires some test tweaks to patch twisted.internet instead of tcpactivation, but otherwise works fine (still in progress of testing on linux / win)20:13
mmccthe reason is that if we import reactor at toplevel, it'll install a default reactor too soon20:14
dobeyyes20:14
dobeyoh, maybe not20:14
dobeytwisted reactors are a fun bunch of insanity regarding imports20:14
mmccyeah, importing causes installation - not great20:14
dobeyyeah20:14
dobeyunless you already have a reactor installed20:15
mmccyep. but in this case I need to import a bunch of stuff before I install the reactor, so I can creaete the QApplication before I install the qt4reactor, whee20:16
mmcc(or else qt4reactor will create a QCoreApplication for me, so helpful)20:16
* briancurtin late lunch20:24
dobeyah right20:26
dobeymmcc: how many places is it importing reactor where that's an issue?20:27
* popey waves20:28
* popey has installed U1 files on IOS 6 beta20:28
beunooh-oh20:29
beunorun urbanape!  run!20:29
popeynicely highlighted there beuno :)20:29
popeywhen you first start it the window is titled "Root view controller"20:29
beunoI've seen that in iOS4 as well20:30
ralsinammcc: maybe moving it into a function that returns the reactor?20:30
ralsinammcc: just to avoid needless duplication20:30
ralsinadobey: sorry, +1 on that branch, forgot to do it on launchpad20:43
dobeyok20:44
dobeyralsina: also have https://code.launchpad.net/~dobey/ubuntuone-client/update-4-0/+merge/11193920:44
dobeyand https://code.launchpad.net/~dobey/ubuntuone-control-panel/update-4-0/+merge/11191520:45
* briancurtin back20:45
dobeymeh20:50
dobeypower keeps blipping here20:50
ralsinadobey: +1 on the 1st one, the second one I can't check before EOD21:02
dobey:(21:02
dobeythe last one can't land until the client one is landed/released/inubuntu/etc21:02
ralsinadobey: maybe mmcc or briancurtin can take a look? It's earlier for them21:02
mmccargh, my chat-name notification thingy died or something21:02
dobeyheh21:03
dobeymmcc, briancurtin: care to give a sanity check to https://code.launchpad.net/~dobey/ubuntuone-client/update-4-0/+merge/111939 ?21:03
mmccsure, I can do that21:03
ralsinaThere, dev-env setup on windows at last21:03
ralsinanow I can try to do stuff tonight21:04
ralsinaso EOD, will do windows tinkering very late, will report to briancurtin if I find something tomorro21:04
dobeyhrmm, i need to make a branch to fix a couple pep8 errors in protocol also21:04
ralsinaw21:04
mmccalso - re the reactor import thing - there are only two functions where i had to import it.21:04
mmccI like ralsina's idea, but in this case it's less code to just write the import in each function that uses reactor21:04
dobeymmcc: might be worth refactoring to a function, so that all the qapp/import/reactor.run stuff is in that one spot?21:05
dobeyoh, i guess that's what he suggested21:05
dobeyheh :)21:05
mmccdobey: sort of. I think he was just suggesting to replace the import with a function that imports and returns the reactor. although now I'm not sure that makes sense21:07
mmccanyway it'd be a much bigger change to put all the app/import/run stuff together, it's spread across several files right now. I assume for test-ability21:08
briancurtindobey: checking that branch now21:11
briancurtinupdate-4-021:11
dobeythanks21:12
dobeybriancurtin: mmcc apparently gave a +1. should i wait for you?21:20
briancurtini'll approve in a second, just getting to the end21:20
dobeyok, cool21:21
briancurtindobey: done, approved21:21
dobeythanks briancurtin21:21
mmccsorry, forgot to say anything21:24
mmccshould I expect pylint on windows to work? I keep forgetting if it's supposed to bleed out and die like this21:25
briancurtinmmcc: it works for me21:25
mmcchmm21:26
mmccwell, I'll make a note. just in case you have a sec and this looks familiar: http://paste.ubuntu.com/1059797/21:30
mmccbut don't spend time on it...21:30
briancurtinmmcc: i see that part too and i just ignore it. after that you would still see regular lint warnings if there are any21:31
briancurtinwhich is why i never bothered looking into that traceback21:31
mmccbriancurtin: oh, heh, ok. thanks :)21:32
mmccdobey, do you have cliffs notes for setting bugs to project/distribution/series ? my reactor change is in sso...21:33
mmcchm, that wasn't a clear sentence. I'm sure you get what I meant. I want to know how to set all the bits on the bug correctly.21:34
dobeyi don't. i still haven't gotten that e-mail written up :-/21:35
dobeyand as soon as this u1client branch lands (or fails to do so), i am heading away from the computer21:36
mmccok np, I'll just make a note to fix it later.21:36
mmccif anyone wants a short review, I'll just leave this here: https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-1017672-delayed-reaction-in-tcpactivation/+merge/11194521:43
briancurtinmmcc: approved, i saw the discussion earlier, makes sense21:47
dobeyok, i'm off. have a god evening guys21:47
briancurtinyou too21:47
mmccthanks briancurtin.21:50
mmccbye dobey.21:51
=== salgado is now known as salgado-afk

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