/srv/irclogs.ubuntu.com/2012/07/11/#ubuntuone.txt

=== txwikinger2 is now known as txwikinger
=== wgrant_ is now known as wgrant
ryealecu: when you appear here, have you seen my message regarding stuck uploads/downloads?04:16
ryeralsina - question about valicert certificate - will it be somehow installed by future installers?06:44
MarkusHrye: thanks for your answer (10/07/2012 17:54:59 UTC+2). I just got 200 and 201 and could not determine the differences06:50
ryeMarkusH: 201 is returned when a new object is uploaded/created, therefore the use of 201 (CREATED) http code06:58
MarkusHrye: got it. Thanks07:03
ryeerm, mandel is it ok that there are 2 startup entries after ubuntuone/windows install - for syncdaemon and control panel. I thought that control panel will activate SD, i wonder whether that can lead to a race condition07:10
ryebeuno: what prevents us from displaying the shareoffer URL for non-accepted share? The id is there in the code, we just need to display it07:12
mandelrye, interesting, I had the same idea, that one started the other..07:45
mandelrye, sounds like something to ask to briancurtin, he has been working on the installer07:46
mandelall, morning!07:46
JamesTaitGood morning all! :)08:35
MonkeyDusthi -- "Yes, you are right https://one.ubuntu.com/ appears to be Offline!"11:24
gatoxgood morning..... waking up was really dificult12:12
* gatox is dying12:12
mandelgatox, morning!12:12
mandelgatox, and don't die!!!12:12
gatoxmandel, hi12:12
mandelwe are nearly there :)12:12
gatoxmandel, the only reason why i wake up12:12
gatoxjejje12:12
mandelgatox, I added a small comment about darwin4, maybe re-propose the branch and add clean-fsevents as a pre-requisite12:13
gatoxmandel, ahhhhh right12:13
mandelgatox, that way we don't have a 2000 lines dif hehe12:13
ryeMoscherkobold: yes, we are already investigating the reason12:13
mandelgatox, goos news are, I got the u1-client tests are running and passing but I have dirty reactors (not waiting for the monitor to finish in the tests)12:14
mandelgatox, but the results right now are very promissing, some small failures here and there but nothing that cannot be fixed :)12:14
gatoxmandel, did you read my email? i fix a lottttttt of conflicts and tests in tests-refactoring branch....... but i still have some failures..... i'm going to try to fix that now12:15
Moscherkoboldrye: ?12:15
gatoxmandel, https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/114405 770 lines nnow12:15
mandelgatox, yes, read it, not a huge deal, I'm fixing some last little things12:16
ryeMoscherkobold: oh, sorry, autocomplete failure, i meant to answer MonkeyDust who is no longer in the channel12:16
mandelgatox, if you want we can do something similar to yesterday to get it ready12:16
gatoxmandel, let me stay in bed for a while, while i fix the windows tests.... there's really not much you can do i think.... it just a lottttt of changes in the tests..... i'll let you know when i finish..... and we can do pair programming after that, or if i'm block.....12:18
gatoxmandel, sounds good?12:18
mandelgatox, yes, perfect, I'll be finishing the merging impl branches (got some last tests to add)12:19
gatoxmandel, wait! to filesystem notifications?12:19
mandelgatox, will review you branch and then 1-1 with ralsina12:19
mandelgatox, there are some lats things, but and the very end, not to worry :)12:19
mandelgatox, it does not touch anything from your side12:19
gatoxmandel, ok..... i'll go to spain and hunt you down if you are lying! jeje12:20
mandelno no12:20
mandelgatox, but that gives me 15 hours to run away, plus your flight being canceled :)12:21
gatoxjajjjajajaj12:21
Moscherkoboldrye: np, any news about my issue?12:22
ralsinagood morning!12:22
mandelralsina, morning!12:23
mandelralsina, feeling better, want a 1-1 with me to feel worse ;)12:23
ralsinamandel: sure, let me plug the headset12:24
gatoxralsina, hi12:24
ralsinamandel: on mumble, ready when you are12:25
mandelack12:25
alecugoood morning all!12:38
alecuI love this kind of weeks!12:38
gatoxalecu, hi!12:38
gatoxalecu, short ones?12:38
alecugatox: absolutely :-)12:41
* alecu opens thunderbird, and regrets short weeks :P12:42
mandelalecu, we need reviews!12:43
mandeloh, and I need lunch..12:44
mandelI'm off to have lunch and I'll hunt down people!12:46
mandelfor reviews I mean :)12:47
alecumandel: ack!12:49
mandelalecu, gatox and I did a good change in the code to make things easier to do, I'll ping you and we can chat about it :)12:49
* mandel really away from machine12:50
* mandel back13:45
gatoxgroso manuel wirtz http://youtu.be/pjtazX3BZw413:50
mandely asi va argentina..13:51
gatoxjaja wrong channel13:52
briancurtini'm back13:52
mandelgatox, I though so..13:53
gatoxmandel, now listen this one: http://youtu.be/mk9kNEMhXn4 so you can enjoy more good music jeje13:54
* mandel throws a catalan song to gatox: http://www.youtube.com/watch?v=FhJR6OO1X8Y13:56
gatoxmandel, you already share that13:57
gatoxon twitter i think13:57
gatoxit's good13:58
mandelgatox, do you know what is a 'cani' ?13:58
gatoxnop13:58
mandelgatox, is a low class person (in the south of the spain) like choni in the north13:58
mandelgatox, something like this: http://www.youtube.com/watch?v=h6k5qbt72Os13:59
mandelgatox, now you wont be able to stop saying 'soy cani, cani cani, cani' :)13:59
ryealecu: ping, have you seen the mail I sent to you and facundo, regarding stuck uploads/downloads14:02
alecurye: I saw that; it sounds awful.14:06
ryealecu: that's nightlies and that's reproducible quite easily with my dataset14:07
alecurye: I'll ping you in a few minutes to ask for more info.14:08
=== zyga is now known as zyga-food
=== salgado is now known as salgado-brb
=== salgado-brb is now known as salgado
=== zyga-food is now known as zyga
gatoxmandel,  i finish fixing the tests for windows! \o/14:41
gatoxmandel, now i need to fix the darwin ones.... but that's the easy part14:41
mandelgatox, awesome!14:41
dobeyhmm14:48
ralsiname15:00
briancurtinme15:00
mandelme15:01
mmccme15:01
ralsinathisfred, gatox, alecu, dobey, standup15:01
gatoxme15:01
alecume15:02
thisfredme15:02
ralsinadobey is last, go me!15:02
ralsinaDONE: tracking contractors for protobuf + twisted ports, mgmt call, pre-mgmt-call-call, 1-1 with mandel, started debugging "folder list is empty on Q", misc other small stuff TODO: finish that, catchup with the mac team on where they all are, see if we can kick that puppy out the door, BLOCKED: not yet NEXT briancurtin15:02
briancurtinDONE: holiday/vacation, caught up on email this morning15:02
briancurtinTODO: finish a branch i couldn't wrap up before leaving, figure out what's left, look through the issues assigned to me while i was away15:02
briancurtinNEXT: mandel15:02
mandelDONE:updated all branches that needed a need fixing. Finished events implementation that uses the daemon if possible and if not the fsevents implementation.15:03
mandelTODO: talk with alec about landing this things. Add MPs.15:03
mandelBLOCKED: no15:03
mandelmmcc, please15:03
mmccDONE: reviews, adding dirspec-exefinding code to u1client and sso15:03
mmccTODO: finally finish exefinding stuff, package15:03
mmccBLCK: still sick. ugh15:03
mmccNEXT: gatox15:03
dobeyme15:03
gatoxDONE:15:03
gatoxFixed some branches, cause of the conflicts with the branches from yesterday, finish resolving conflicts and changes in darwin4, and tests-refactoring for windows.15:03
gatoxTODO:15:03
gatoxFinish with darwin tests-refactoring15:03
gatoxBLOCKED:15:03
gatoxNo15:03
gatoxalecu, go15:03
ralsinaNOTE: alecu, briancurtin, think about what needs to be done to do a release on windows with PyQt 4.9 so that authed proxy works better (plus make it log as 3.0.2)15:03
alecuDONE: went to pycamp. Catching up with mail15:03
alecuTODO: debug sync issue that rye found, reviews for mandel, gatox15:03
alecuBLOCKED: no15:03
alecuNEXT: thisfred15:03
thisfredDONE: Bug #1022659  TODO: Bug #1022659 + client side quota error handling BLOCKED: no15:04
ubot5Launchpad bug 1022659 in U1DB "Add document size limit to database" [High,Fix committed] https://launchpad.net/bugs/102265915:04
dobeyDONE: more review on mmcc's dirspec branch, changed mike's proposed API to be more generic, proposed new dirspec branch15:04
dobeyTODO: finish dirspec branch, u1db packages as multiarch, 3.0.3 backports/release?15:04
dobeyBLCK: None.15:04
=== mpt_ is now known as mpt
dobeymmcc: btw, just pushed update per your review, for that branch15:05
dobeyhttps://code.launchpad.net/~dobey/dirspec/add-exefind/+merge/11427315:05
ralsinaso, dobey, alecu, briancurtin: maybe do it as 3.0.3 instead15:05
mmccdobey, ok thanks. will look now15:05
mandelralsina, I have the name of a person from twisted that is doing contracting which vds met in pycon eu15:05
ralsinamandel: pass along15:05
ralsinamandel: we do have a very good candidate, the main issue is that I don't want us to be responsible for a twisted fork for 18 months15:06
briancurtinAntoine Pitrou was involved in (one of) the original porting efforts15:06
briancurtini kind of know him through working on CPython15:06
ralsinabriancurtin: he's who I've been talking to15:06
briancurtinah15:06
ralsinabriancurtin: but his py3 port looks like is never going to be merged upstream15:07
ralsinabriancurtin: which would suuuuuuuuuck for us.15:07
briancurtinralsina: that's likely due to lack of review, which glyph from the twisted team mentioned. we talked about it for a while at the Language Summit at PyCon, that Antoine's port would be more likely to be merged if it was fully code reviewed, which no one wants to do15:08
mandelbriancurtin, isn't he the one that was also commenting in my cpython patch?15:08
briancurtinmandel: yep, that's him15:08
ralsinaok,so maybe we can help there15:08
ralsinaso we may hire someone to review his code instead ;-)15:09
mandellol15:09
briancurtinralsina: it might be worth contacting glyph or some twisted-dev list if it exists, to bring it up. twisted is hardcore about code review and process stuff, and  no one was sure of a timeframe, but they thought you'd have to pay someone a good bit to sit down and give Antoine's branch a thorough review15:10
ralsinaok, if it's reviews and time, then we need to do a few things15:10
ralsina1) check if what is there now works for us. alecu?15:11
ralsina2) hire antoine to finish what's missing15:11
ralsina3) pay for reviews to help merge15:11
ralsinawe may not do 3, or we may do a part of it, or something15:11
ralsinaNot having 3 means we are left holding the port for 18 months, which we can't15:12
ralsinaSo I am considering it as part of the porting process15:12
alecuralsina: perhaps even more if nobody does the twisted port to 315:12
ralsinaalecu: exactly15:12
alecuralsina: the problem with 3 above is that the twisted folks explicitly said that they don't want a 2to3 type of port.15:13
ralsinaalecu: looks like noone does. Got the same from protobuf guys15:13
ralsinaalecu: I will put things straight in my head and we'll make a plan in tomorrow's call15:14
dobeyok, need to get lunch15:14
dobeybbiab15:14
gatoxmandel, alecu darwin-tests-refactoring is done for review15:14
briancurtini don't remember protobuf looking that hard to port15:14
alecuralsina: also, twisted has dropped python 2.4 as of very recently. I don't think they'll want to drop 2.5 just yet.15:14
briancurtin(but i don't know their minimum supported version)15:14
alecugatox: great. Link please?15:14
mandelgatox, ok15:14
ralsinabriancurtin: it was 2.4 which was a pain, just got them to say 2.6 is good enough15:14
gatoxmandel, alecu https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/11328915:15
mandelmmcc, can you re-review the clean-fsevents branch please, we are block with that one15:15
mmccmandel, look again :)15:15
briancurtinralsina: 2.4 might make it suck (exception handling between 2.4 and 3 isn't pretty). i'll run it through 2to3 just to get a picture of what it looks like. The C extension part of protobuf is a trivial port15:16
mandelthx15:16
ralsinabriancurtin: we don't even need the C extension15:16
ralsinabriancurtin: and luckily, we have a contractor for that aligned already :-)15:16
ralsinabriancurtin: but check anyway, just in case15:17
briancurtinoh, nice15:17
gatoxmandel, alecu also this one is ready too: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/114405 it has been reproposed changing the prerequisite, and i answer alecu 's comment15:17
gatoxalecu, mandel you should probably review darwin4 first15:18
mandelack15:18
mandelgatox, let me work around a problem I have and will get to that15:19
gatoxmandel, ack..... i'll keep working on another issues in the meantime15:19
briancurtinralsina: protobuf looks really easy. unicode/str, int/long, one metaclass, a bunch of exceptions, items/iteritems15:20
briancurtinit's a 437 line diff15:20
alecugatox: your message here seems to be chopped off: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/11440515:24
alecugatox: just after the "for fix in"15:24
gatoxalecu, ignore "This is marked for fix in".... is the next line15:26
mmccreview for alecu or ralsina: https://code.launchpad.net/~dobey/dirspec/add-exefind/+merge/11427315:33
ralsinammcc:  looking15:33
alecummcc: ack15:33
ralsinaalecu: I got it15:34
mmccalso, thanks dobey for spending so much time on that.15:34
ryebriancurtin: is u1 installer supposed to write both autostart entries for control panel and syncdaemon?15:41
briancurtinrye: yes15:41
ryebriancurtin: uhm, why?15:41
=== zyga_ is now known as zyga
briancurtinrye: is that a problem?15:41
ryebriancurtin: no, not really, but i was curious about possible race condition - control panel wants to start SD which is already being started, but if that's ok then it's ok with me15:43
mandelgatox, can I have the patched branch of fsevents I for got the url15:52
gatoxmandel, this one? lp:~diegosarmentero/+junk/python-macfsevents15:53
mandelgatox, bingo :)15:54
mandelgatox, does mmcc  know about it? that is very important to ensure it is packaged15:54
mandelare we going to try and put that in the fsevents project?15:54
mmccmandel, yes I know abou tit15:54
gatoxmandel, yes.... i told him the other day15:54
mandeland upstream?15:54
* gatox lunch15:56
mmccwe do need to add fsevents to the buildout eventually...15:56
=== gatox is now known as gatox_lunch
mandeldobey, have you seen this before: https://code.launchpad.net/~mandel/ubuntuone-client/clean-fsevents/+merge/114252 (error form tarmac)15:59
mandeldobey, I ran all the tests in P with no problems and the code does not touch a thing from linux15:59
ralsinaalecu, dobey: can either one of you look into backporting this? https://bugs.launchpad.net/ubuntuone-client/+bug/101340115:59
ubot5Ubuntu bug 1013401 in ubuntuone-client (Ubuntu Precise) "Files not re-uploaded when server reports empty hash, resulting in apparent data loss" [Undecided,Triaged]15:59
ralsinammcc: please get it into buildout ASAP so it doesn't get lost in the noise16:00
mmccralsina: ack.16:00
gatox_lunchalecu, sorry, i forgot to push the change removing time.sleep....... now is there!16:03
ralsinalunchtime for me16:14
=== salgado is now known as salgado-lunch
mandelgatox_lunch, I'm using inheritance, see I can do it too :)16:23
dobeyralsina: backporting to?16:23
mandelthis time makes more sense :P16:23
gatox_lunchmandel, jejeje good for you!16:23
dobeyralsina: i included it in the SRU currtently waiting to be accepted into precise-proposed16:23
mandeldobey, did you see the error I mentioned?16:25
dobeymandel: looking16:26
mandeldobey, thx :)16:26
dobeymandel: hrmm. i have seen a similar error before; not sure why that happens. maybe something is broken in the twisted in quantal16:28
dobeymandel: trunk branches are landing on quantal now, not precise16:28
mandeldobey, ouch!16:28
dobeyand quantal has twisted 12.1 i think16:28
dobeyor 12.0 at least16:28
mandeldobey, do you have a Q machine to test the branch, I don't want to bomb everyone with messages about it16:28
dobeyi guess 12.0 as the gireactor stuff is still a patch afaik16:28
mandelelse I can install Q in a vm to run the tests16:28
dobeynot yet. been having video issues with Q, so haven't upgraded anything yet really16:29
dobeymandel: i think i saw a similar error in the nightlies builds before16:29
dobeybut iirc it went away with a rebuild16:30
mandeldobey, I'll do a re-approve if if fails I'll install a vm and take alook16:31
dobeymandel: yeah it might be ok now16:31
=== gatox_lunch is now known as gatox
=== salgado-lunch is now known as salgado
dobeyoh cmake. you foul fowl.16:54
mandelok, eod here, see you all tom!17:10
dobeycheers mandel17:11
gatoxmandel, bye17:16
ralsinadobey: backport to oneiric17:20
dobeyralsina: sure17:22
ralsinadobey: thanks17:22
ralsinadobey: natty may be a nice-to-have but depends on effort and availability, lucid doesn't need it IIRC17:22
dobeylucid/natty certainly need a lot more than that one patch; but so does oneiric really17:23
mmcchooray, I figured out why the objc tests in mandel's branch weren't running for me. fiddly library path locations, AS USUAL17:25
ralsinadobey: IIRC lucid didn't need it because it was pre-generations17:27
ralsinadobey: but I may be confusing it with another patch/bug17:27
dobeymaybe, but the immediate context of the bug seemed like it was a per-file issue, not a generations thing17:29
dobeyanyway, it will either apply as-is, or won't, i suspect. and if it doesn't apply, probably doesn't affect lucid, as that code is pretty unchanged over the years17:30
ralsinadobey: ok then17:30
ralsinadobey: global +1 on multiarch17:31
dobeyyay17:32
ralsinadobey: if you are fixing that, I assume packaging is close to done ;-)17:34
dobeyralsina: well, i already had it 'done' last week, with nightlies building on quantal/precise :)17:37
dobeyralsina: but realized it needs to also be multiarch, so fixed it up to do that, which needs this hack because ugh cmake. :)17:38
ralsinacome on, it's not all that horrible. There may even be something nicer to do, but this works17:38
dobeyralsina: well, when searching for how to do this, i did see hint of the possible existence of a GNUInstallFiles.cmake or something, that creates some magic to use standard directories in some way. but i didn't really look any further in that direction, as i just wanted to get it working with what we have right now17:55
ralsinadobey: yes, it's small and easy enough that it's not worth hacking at17:55
ralsinadobey: in fact, there is a patched multiarch cmake in a PPA17:56
ralsinadobey: but again, not worth trying to get the patch into cmake first17:56
mmcclunchtime…18:27
dobeyalecu, ralsina: https://code.launchpad.net/~dobey/ubuntuone-client/update-2-0/+merge/11448819:01
ralsinadobey: looking19:01
ralsinadobey: looks clean enough, was it hard to apply?19:02
dobeyralsina: no. bzr merge -c worked without complaint :)19:02
ralsinahehe19:02
ralsinaI wonder if a fake_meth method is a good idea19:03
ralsinaunless it's a stub for a the_real_deal method somewhere else19:03
dobeywell, i guess it's better than real_meth, at least in terms of legality and health, anyway19:03
ralsinadobey: +119:04
dobeyshould wait for alecu to review?19:06
alecudobey: +119:06
dobeythanks19:06
ralsinaso there :-)19:08
dobeyit's too bad oneiric isn't about to be EOL19:26
dobeyhmm. maybe i can build the 2.0.x releases into lp:ubuntuone/stable for natty; then at least could land branches more reliably for stable-2-019:37
ralsinadobey: haha, you want us to be more fedora-like? ;-)19:37
dobeyralsina: no, it's just that oneiric is that weird point where the tests don't run on it, so we have to land branches on natty, and it's annoying to manage that now19:38
dobeyanyway, gotta run for a bit. brb19:38
ralsinadobey: right19:38
* briancurtin late lunch19:47
ralsinaHave to go on a school run and some errands. Will be back for an hour or so in about 90 minutes20:06
dobeyback20:18
gatoxok..... eod for me!! see you tomorrow people20:24
dobeyugh21:12
dobeyjust tapped the screen on my laptop and x crashed :(21:13
dobeywhile i was scping a large file21:13
dobeyfun times. :(21:13
mmccdobey, It's like tapping on the glass at the zoo - it's really loud in there and confuses the animals21:15
dobeyall i wanted to know was if i was connected on G or N to the wifi21:16
dobeyof course, i don't know if this laptop even does N21:16
dobeyanyway, done getting that file copied off, so now i can reinstall ubuntu on it, using 64 bit quantal21:18
=== salgado is now known as salgado-afk
mmcchmmm, more organic growth21:35
mmcctwo tests in sso test_credentials are identical except that one asserts (using python assert) that the fake test exe_path doesn't exist21:36
mmccoh, one of them used to test that it falls back to gtk, now they both test that it falls back to qt21:38
dobeyah21:38
dobeythat should probably be cleaned up21:38
mmccdobey: agreed. I'm doing that as part of using the new dirspec exe finder, since that broke these tests21:48
dobeyah ok21:49
dobeyman, i really need to do some digital housekeeping21:52
dobeyi have way too many files21:52
dobeywhee. kernel boom when attaching to the dock :(22:00
dobeyoh well. guess i should go get some other stuff done now.22:07
dobeyhave a good evening all22:07
mmccok, I finally have branches with sane tests that pass, trying to package it and see how it goes on darwin23:19
mmcc10.6 this time, because that worked before. putting off packaging on 10.7 for right now23:20
mmccurgh, now I get a unicode error in something that used to work... perhaps related to the recent StringIO change?23:36
mmccI'll just leave this here: http://paste.ubuntu.com/1087078/ <-- the traceback from sso dying in handling on_captcha_generated23:37
mmccoh man, this takes forever to debug23:41
mmccOK, I have to go, can't wait for it to finish packaging again. I'll check back later. I think it was just missing a .getvalue() on line 330 of setup_account_page.py23:42
mmccnope: http://paste.ubuntu.com/1087091/23:55

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