[00:06] Hi, I'm new to 12.04 and am struggling with Ubuntu One. I wonder if I can just sync just a few folders already in my cloud. [00:20] Hi. Is anyone awake here? [01:07] hey, has anyone had files go missing lately? [01:07] my account seems to have deleted some files 13 hrs ago [01:07] ... 13 hrs ago I was at work. [01:11] Jester86_Mobile, no, no known such issues [01:12] beuno, i'm wondering if one of my machines didn't do a reboot .. i had had my main desktop not running ubuntuone for a while [01:12] if it rebooted, started ubuntu one and ubuntu one screwed up.. perhaps it was trying to sync the server to my PCs dir ? [01:12] Jester86_Mobile, yeah, not sure. Support can help you recover those files if needed [01:13] maybe tose files were gone from one of the computers [01:13] so U1 assumed you had deleted them and that propagatd [01:13] well its my arduino library that I keep on ubuntu one [01:14] .. so all of my machines work off the same library [01:14] I believe I deleted the library on that machine then stopped the sync from that machine [01:14] right, so that probably propagated to the server and other machines [01:15] i was trying to replace the library entirely but that machine was connected via crappy wireless so I stopped the sync and uploaded the new lib from a hardwired machine [01:15] i'd think it'd see the newest lib as being the newest and not continue what the other machine had started doing though [01:16] but regardless, if theres any easy way to restore these files i'm all ears [01:16] Jester86_Mobile, well, I'm not a support person but I can try. What's your email address? [01:16] jesse.klaus@gmail.com [01:18] why is restore not an easy feature to access? [01:18] Jester86_Mobile, I can't see your files or any information, I can just do a blanket "restore deleted files" [01:18] which will go into a new folder [01:18] even dropbox allows for easy restoring [01:18] is that ok? [01:18] sure [01:18] yes, yes, we're working on allowing you to restore them yourself :) [01:18] that would be terrific ;) [01:19] so when my machines decide to run a muck i can fix it lol [01:19] Jester86_Mobile, ok, pressed the button and it said success [01:19] you should see a folder [01:20] if I learnt from watching support properly :) [01:20] lol [01:20] well we'll see what happens [01:20] how long do you think it will take? [01:20] Jester86_Mobile, should be there already [01:20] hmm [01:20] no dice [01:21] under ~/Ubuntu One, I'd guess [01:21] wait [01:21] hmm [01:21] "hmm" is rarely good! [01:21] it has it on the web but not pushed down to my [01:22] ah, maybe the client is disconnected? [01:22] thats what i'm looking into again [01:23] just disconnected and reconnecting... its taking a while so i'm thinking it did screw up in some fashion [01:24] I'm just glad it didn't delete my sketchpad lol [01:24] Jester86_Mobile, alright, I'll let the client do its work then, your files are there so it's just a matter of getting to them [01:24] going to call it a day then [01:24] i wrote a program last night to track the sun .. for a solar panel [01:24] didn't want to lose that lol [01:25] Jester86_Mobile, if you have any more issues, just contact support through the website, they'll help you with anything you need [01:25] alright man, thanks for your help [01:25] o/ [01:25] alright [01:25] thank you very much sir [01:25] or miss lol [01:26] Is there anyway to lock a folder to not allow changes.. essentially a read-only sync? [01:27] something you just manually deactivate [01:46] Jester86_Mobile: I don't think there's a way. And beun'o is a guy, a really awesome developer. [01:47] Now, I'm calling it a day, too. [01:47] Night all o/ [01:47] lol [01:47] i thought he was, just wanted to verify [01:47] There you have it :) [01:47] seems like it loaded my old libs though [01:47] I'm sorry, my brain's fried for today (or yesterday, almost 4AM here), I wouldn't be of much help. [01:48] * karni disappears [01:48] alright - but is there anyway for me to get the file back that were deleted as of 8am today? [01:48] you don't have to - I just want to know if there is a way [01:50] Okey, I was just leaving. Basically, there is a way to recover files. If you need that, please contact our support via the website, or hang out here tomorrow or the day after, and ping ry'e or joshuahoove'r (without the ' , they're not here now anyway). [01:50] * karni disappears for real [01:51] ok [01:51] thanks [04:19] hey neato, side effect of debug print fixes bug! My favorite! [04:29] heh. Side effects of debug print might squash the behavior, but it doesn't really fix. I think of bugs that are fixed by debug prints as a sub-classification of Heisenbugs [04:29] still, sometimes that's enought to help find the root of it [04:29] * rmcbride should be in bed an hour ago. good night [08:18] hi - I'm new here. I have problems with my ubuntuOne install. Trying to sync 5GB of data. u1sdtool --current-transfers gives me this: [08:20] pastebin.com/V3Aj3BwZ [08:20] and then it hangs ... [08:21] My nautilus also hangs a lot ... [08:32] I think I will try to ask in the #ubuntu channel [08:52] morning all! [09:01] mandel: morning [09:02] rye, hello :) [09:02] mandel: we seem to be getting a lot of package installation bug reports, such as bug #1009891 - do you happen to know anything about that? [09:02] Error: Could not gather data from Launchpad for bug #1009891 (https://launchpad.net/bugs/1009891). The error has been logged [09:03] * mandel looking [09:03] rye, which OS version, N? or is it N just in that bug report? [09:04] oh, no sorry, there is this: UpgradeStatus: Upgraded to precise on 2012-06-06 (0 days ago) [09:05] rye, do you have something more than the comment you gave? [09:09] In the File "/usr/lib/python2.5/site-packages/ubuntuone/utils/curllib.py", line 134 [09:10] except pycurl.error as e: [09:10] and i guess i saw this [09:10] mandel: nevermind, let me dig more [09:11] Happy $TODAY, all! [09:11] rye, alecu and ralsina where talking about that, or code should be installed in python2.6 and not python2.5 which does not support that syntax (Exception as e) [09:12] + except pycurl.error as e: [09:12] + raise HTTPError(request.url, e[0], curl.errstr()) [09:12] + else: [09:12] rye, although in theory the oldest python we support is 2.6 we know there are users that install 2.5 to work with google appengine [09:12] yes, that's the same "syntax from the future" [09:13] rye, exactly, and a package with that syntax should be in "/usr/lib/python2.6/ and not /usr/lib/python2.5/ [09:13] rye, does the info help? [09:14] mandel: yep, looking into latest reports now to duplicate properly [09:14] rye, cool, let me know if I can help.. I think there must be a dependency no interacting well with the fact that there is a 2.6 and a 2.5 in the same system === yofel_ is now known as yofel [09:30] aha, for lucid fix is bug #1009573 [09:30] Launchpad bug 1009573 in ubuntuone-client (Ubuntu) "package python-ubuntuone-client 1.2.2-0ubuntu2.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [High,Fix released] https://launchpad.net/bugs/1009573 [09:43] mandel: moar questions - https://pastebin.canonical.com/67625/ - ActivationTimeoutError - what does this mean - which process failed to start in time? [09:44] the only thing that sso says is 2012-06-05 18:19:38,068:68.0000782013 - ubuntu_sso.utils.webclient.qtnetwork - INFO - Querying OS for proxy. [09:45] rye, yes, so the way it works is that the tcpactivation tries to check a number of times if the process it wanted to start did indeed start [09:45] rye, if after that number it did not, we get that error [09:45] the definition of the number of times and the delay between checks is in ubuntu_sso/utils/tcpactivation.py [09:45] rye, that is on windows, right? [09:46] mandel: yep [09:47] rye, so yes, the process did not start for some reason.. is a work around to the fact that we do not have dbus to start things :( [10:04] mandel: can "Querying OS for proxy" in sso hang? [10:06] also, my yesterday tweet should have had this in it - http://files.lappyfamily.net/flood.jpg - Euro2012 - Kyiv welcomes you. I am just scared about what people are going to experience here [10:13] rye, it could but that is very very strange.. [10:16] rye, is it raining that bad over there? [10:18] mandel: due to high number of buildings and outdated/broken infrastructure of draining system (and it was built around 1992) we had enormous amount of water. subway station entrances were flooded, parking lots, you can search youtube for kiev 2012-06-06 and see those. For some reason rain is thought as of exception here [10:19] rye, lol.. in mallorca we once tried to have a subway it also looked like that :P [10:22] mandel: the other thing is that people continue to get into water even though there is electricity still up and running [10:23] rye, now that is dangerous! [10:23] okay, i suspect another thunderstorm to happen now... [11:12] good morning! [11:42] anybody had issues installing lucid in lxc on precise? [11:43] gatox, hola! [11:43] gatox, I hate dirty reactors, I hate twisted ;) [11:43] mandel, jejeje i hate twisted too.... but basically because my interaction with twisted was always fixing stuff [11:44] mandel, i'm starting to review your second branch now, before going into macfsevents again [11:44] gatox, yeah.. I don't know why the fuck in mac is not working.. [11:44] mandel, i read your twit...... [11:45] gatox, the connectionLost does not get called, wtf! [11:47] language... [11:51] frustration [11:51] releasing frustration :P [11:54] it's still not respectful toward others, you can release frustration in private (: [12:06] gatox, http://twistedmatrix.com/documents/12.0.0/api/twisted.conch.insults.insults.html [12:07] gatox, and yes, conch is concha and insults, well insults :P [12:07] mandel, it seems that is a module to not mess around jeje === eu is now known as Guest36713 [12:11] gatox, http://twistedmatrix.com/trac/wiki/FrequentlyAskedQuestions#WhyisntmyconnectionLostmethodcalled [12:11] gatox, sometimes I want to kill myself when working around twisted issues.. [12:12] mandel, the platform specific thingy was screwing your tests? [12:13] gatox, very possible, making the call for connectionLost not happen in certain cases and therefore making the test wait for ever on mac os x [12:14] gatox, which means that I have to fix devtools to ensure that we if we wait to long we are ok to continued, yet timeouts cannot be added to the main loop because then we will have a dirty reactor issues [12:14] mandel, :S === Guest36713 is now known as adorilson [12:18] gatox, so I'm running the tests with --loop=100 and sometimes passes the 2500 times other gets stuck waiting for the connectionLost method to be called.. I think I'm going to cry [12:18] well.. at least I found the error [12:19] mandel, timing issues..... always the worst kind! [12:20] gatox, the problem is mainly, I cannot trust the os to call the connectionLost and therefore I need to use a timeout, the problem is that if you do TestCase.addlCleanup(func) and func never ends all tests are blocked because TestCase.timeout is only used for the tests and not the cleanups [12:21] * mandel smells a good post somewhere there.. [12:21] mandel, i decided to start doing posts about my frustration with some ninja's bugs :P [12:22] gatox, it is a great exercise, lets you think carefully about the problem, improves communication skills and smarter people read them and correct you [12:22] gatox, I had glyph (twisted creator) sending me an email about the show() dialog + qtreactor :) [12:23] mandel, wow... cool! [12:23] gatox, is good to be in planet python just for that :) [12:37] hellhole, all! [12:38] alecu, hai [12:42] mandel, ping..... let me know when you are back === m_conley_away is now known as m_conley [13:39] good morning, was on a call forgot to say hi [13:39] ralsina, hi [13:42] oi [13:44] gatox, popey [13:44] popey, sorry [13:44] gatox, pong :P [13:44] jajajaa [13:44] alecu: any reason you didn't start with trunk, rather than stable-4-0 for the validate-ssl-cert branch? [13:44] mandel, i'm having some problems trying to run the python tests of your branch :S [13:45] dobey, no reason [13:45] dobey, alecu do you have a few mins to talk about an issue in ubuntuone-dev-tools + mac os x? [13:45] mandel, sure [13:45] gatox, which branch? [13:45] mandel: what's up? [13:45] alecu, dobey, so the issue is the following, according to the twisted FAQ http://twistedmatrix.com/trac/wiki/FrequentlyAskedQuestions#WhyisntmyconnectionLostmethodcalled [13:46] mandel, "improve python client"....... are you using the system configured with the buildout to run these tests? [13:46] alecu, dobey connectionLost is not guaranteed to be called, in the clean_up of the servers we are expecting a deferred to be called when the connectionLost method in the server is executed [13:47] alecu, dobey, we are seeing cases in which that connectionLost method is never called, and because the execution is in a cleanup call all tests are blocked (the timeout of the testcase does not affect functions added as cleanups) [13:47] mandel: this seems like a general problem, not an osx issue [13:47] mandel, you say "in the cleanup of the servers" <- what servers? [13:47] gatox, I'm just using utrial in my path and normal python [13:47] alecu: ServerTestCase i presume [13:48] mandel, ahhhhhh i think i know where the problem is..... you are using "u1lint"..... and with the env setup.... you should be using it as "$u1lint" [13:48] alecu, this guy: http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntuone-dev-tools/trunk/view/head:/ubuntuone/devtools/testcases/txsocketserver.py#L154 [13:48] that allows the env to find the proper u1lint [13:48] dobey, exactly [13:49] gatox, oh.. what is the diff between $u1lint and using u1lint? [13:50] alecu, dobey, so, while we have not seen that happening much in win or linux, it seems is 'quite' possible in mac os x.. I wonder if you have an idea on how to work around the problem [13:51] mandel, first, let's discuss the FAQ you posted above. [13:51] mandel, actually i don't know if the right thing to do will be: "python $u1lint", the difference is executing just: u1lint tries to find the dependencies in the system, not in the buildout folder [13:51] mandel: fix the code to not depend on connectionLost happening? [13:51] mandel, it should not affect us, because we are "closing actively", not waiting for a "tcp timeout close". [13:51] mandel, so i'm having some import issues, but if i open a python console and import that, it's ok [13:52] mandel, so, we will *always* get the connectionLost. [13:52] alecu, but we are not.. [13:52] mandel, we won't have the case described in the FAQ at all. [13:53] mandel, anyway, it's not the case that's being described in the FAQ. [13:53] alecu, ok, nevertheless, connectionLost is not called which makes the cleanup to hang there.. [13:53] mandel, ok, but it's a different issue. [13:54] mandel, so, is this only happening on osx with Unix Domain Sockets? [13:54] alecu, and tcp sockets too [13:55] mandel, but only on osx, right? [13:56] alecu, yes, only on os x AFAIK [13:56] mandel, are any of the twisted tests failing like this on osx? [13:56] alecu, I have not seen this happening in any other os (running same tests and same u1devtools) [13:56] alecu, I have not ran the twisted tests. let me take a look [13:56] mandel, I mean, the tests for the twisted trunk. [13:57] mandel, if the twisted tests are running, then we have to use the same "cleanup" method that they do. [13:57] mandel, if they fail... then there's something more serious. [13:58] alecu, let me run them then :) [14:01] alecu, it seems i fix macfsevents :D [14:01] gatox, branch or it didn't happen! [14:03] alecu, jeje ok... uploading the branch [14:03] alecu, and i didn't add any wait condition [14:03] need to add [14:03] awesome! [14:06] alecu: quick IRC 1-1 ? [14:07] ralsina, sure! [14:14] alecu, https://code.launchpad.net/~diegosarmentero/+junk/python-macfsevents [14:14] gatox, I'm on my 1-1, I'll review it laters [14:15] alecu, ack [14:15] * gatox go back to tests [14:18] alecu, tests pass with a number of skips.. I'm looking closer at them [14:19] mandel, twisted tests, right_ [14:19] ? [14:20] gatox, if you only push the modified branch there's no sane way to see the diff! [14:20] gatox, perhaps you can make two commits (before and after your changes) and repush? [14:21] alecu, ok..... although the change is really simple [14:21] alecu, can i show you the diff? [14:21] gatox, sure! [14:22] alecu, yes [14:26] alecu, the diff is not working.... i'm going to upload the code in 2 steps [14:33] alecu, now you can see the original file in revision1, and the modification in revision2: http://bazaar.launchpad.net/~diegosarmentero/+junk/python-macfsevents/files [14:33] morning folks [14:34] mmcc, hi [14:34] mandel: please remember to send me the proxy stuff. [14:34] alecu, sorry, i mean this: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-fsevents [14:35] alecu, ohhhh wrong paste again...... right one: https://code.launchpad.net/~diegosarmentero/+junk/python-macfsevents [14:36] elopio, true, let me finish some tests and I'll write the email [14:37] mandel: sure. Thanks. [14:51] alecu, tests from twisted in trunk seem not to have that problem, yet I need to take a closer looks because I think they sometimes leave the rector dirty and don't fail [14:52] * mandel just got a grey screen of death in the mac [14:55] mandel, awesome! [14:55] mandel, with the pretty text in many languages? [14:56] mandel, is there any mach setting to make it blue? /me wants [14:57] I have a PyQt / SIP puzzle for anyone who's interested... [14:57] alecu, haha I got the many langs but I have no idea to change it to blue [14:57] alecu, I'm asking about the error in #twisted maybe someone have seen the error before [14:58] if you call PyQt4.QtGui.QApplication(sys.argv) and ignore the return value, what happens to the C++ QApplication? [14:58] mmcc, get collected and bad things happen [14:58] mandel, so it would seem [14:58] mmcc, best is to keep a ref so that the gc does not take it away [14:58] mmcc, I have had that problem before :) [14:59] one of those bad things is that Qt no longer looks in the app wrapper for a qt.conf, thus using the default plugin path and loading the wrong plugin [14:59] s [14:59] also (i don't know mixing the qtreactor and twisted), but the common use is to bind: sys.exit(app.exec_()) [14:59] are we doing mumble or sticking with irc? [15:00] ralsina, team meating? [15:00] yes, meeting [15:00] I'll be there in 1' [15:00] team meating makes me hungry for BBQ [15:01] its never too early for bbq [15:01] i'm joining a bit late...sorry [15:01] mumble! [15:03] dobey, mandel, thisfred, joshuahoover, mmcc: mumble! [15:03] fdafdmk [15:04] MANDEL! [15:05] ralsina, is thrusday? sorry going [15:05] mmcc: I think *maybe* you can set the environment variable QTDIR or QTDIRS for that [15:06] mmcc: http://www.os2world.com/forum/index.php?topic=3809.30 [15:08] alecu, ralsina: sorry, running late...on another call [15:08] joshuahoover: ack [15:27] alecu, so at twisted they told me not to make network connections in tests... fuuuu [15:28] mmcc: maybe you should do a review of https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/add-dispatcher-tests/+merge/107766 [15:28] please [15:28] dobey, abstain? [15:28] mandel: i can't run the tests or anything :) [15:28] dobey, yes, I agree :) [15:29] mandel, dobey, I can't compile it on 10.6 [15:29] dobey, as I understand the abstain.. [15:29] mmcc, code review is ok, gatox ran the tests [15:29] mmcc, we can ask alecu to run them [15:29] mandel: ok [15:29] ok. then alecu or ralsina can run the tests i guess [15:30] dobey, I'll set up a machine (macmini) in a fix ip, when done I'll send you a user and a stupid password so that you can do it too [15:35] mandel, exactly who said "no net connections on tests"? [15:36] mandel, there are lots of network connections in twisted's tests. [15:36] mandel, grep for connectSSL and connectTCP in the twisted codebase, and see for yourself. [15:36] alecu, quote: mandel: so my initial advice is to stub out the clientFromString bits and don't make actual network connections in your test [15:36] * alecu joins #twisted [15:37] alecu, and yes, I know there are lots.. [15:38] alecu, dash is => http://washort.twistedmatrix.com/ [15:45] alecu, so I have changed the tests and the error does not happen that often (1-10) but I still want to know why that connectionLost method is not called (me reading twisted source) [15:46] mandel, I hope I'm not popping any happy balloon, but "1 out of 10" is not good enough for tests either. [15:46] alecu, you are not :) [15:47] alecu, yet it is an improvement over nearly all the time.. [15:47] ok, i need to get lunch. bbiab [15:47] * gatox lunch === gatox is now known as gatox_lunch [15:47] mandel, right :-) [15:48] alecu, are you close to the mac? (running OS X) [15:48] mandel, what I meant is that if a test works "9 out of 10" we will still have to skip it. [15:48] mandel, yes [15:48] mandel, one quick flick of the mouse downwards and synergy controls the mac. [15:49] alecu, try running the tests for lp:~mandel/ubuntu-sso-client/fix-tcp-activation [15:50] alecu, if you pass -loop=10 they'll be ran 10 times [15:51] alecu, if it does get stuck it means that the server protocol on connectionLost deferred was not called and is waiting for it [15:57] alecu, did it happen to you? (blocked tests) [15:58] mandel, I've tried both branching it, and merging it over trunk, but I always get "no module named gobject" [15:59] mandel, I might need a newer u1-devtools, right? [15:59] alecu, yes, its trying to use gi as the default reactor [15:59] alecu, the new u1devtools chooses the correct one (or simply do --reactor=qt4) [15:59] alecu, actually, do reactor=qt4 because it tests the qt ui [16:03] mandel, brb [16:04] mandel, I just ran tests on that branch and got two reactor unclean errors. http://paste.ubuntu.com/1028846/ [16:04] [16:04] and a spelling error :) [16:04] mmcc, which version of u1devtools? [16:04] mmcc, try with trunk, please :) [16:06] trunk, ok. I was using parent branch: bzr+ssh://bazaar.launchpad.net/~mandel/ubuntuone-dev-tools/darwin-unix/ [16:07] mmcc, I guessed so.. I have been fighting this problems for way too long hehe [16:09] Hi. [16:10] Can origin and destination be different on different computers? [16:10] mandel, against trunk your fix-tcp-activation branch passes the tests for me [16:11] mmcc: http://icanhascheezburger.files.wordpress.com/2012/06/advice-animals-memes-animal-memes-evil-cows-sounded-like-a-good-idea.jpg [16:11] mmcc, can you run them a couple of times.. one must fail at some point (/me going crazy) [16:11] And with that, lunch time for me! [16:11] mandel, ok. is there a repeat flag somewhere? [16:11] mmcc, for trial you can do --loop=100 for example [16:12] mmcc, u1trial I mean [16:12] mandel ok, looping 100 for your sanity [16:12] thx [16:13] How can I get an answer here? [16:13] lgc, rye can give you hand [16:14] Thanks, mandel. [16:14] rye, can origin and destination be different on different machines? [16:16] lgc: Sorry, origin and destination of what? [16:16] lgc: i am not sure I understood your question but if you refer to UDF then yes, it is possible to mount only certain UDFs on different machines. Main Ubuntu One folder will be always synchronized on all computers. [16:16] mandel, test_is_already_running hung for me the first time now [16:17] mmcc, exactly, for some reason connectionLost is not called sometimes.. [16:17] * mandel wonders if it is due to the null protocol... [16:23] ralsina, if files. Say, on machine A, with Windows, .../My Documents/Ubuntu One, and on machine B, with Ubuntu, ... /home/johndoe/Desktop/Ubuntu One, for example. [16:24] lgc: no, sorry [16:24] The Ubuntu One folder is always on the user's home dir [16:24] rye_, pardon me, but do you mean by UDF? [16:24] lgc: we have a feature in our roadmap about making the location folders configurable, but it's not going to be done soon [16:25] rye_, I mean "what do you mean". [16:26] ralsina, I see. Is it in the roadmap the prospect of deleting the space within "Ubuntu One"? It is a pain for command line users. [16:26] lgc: no [16:26] lgc: and why is it a pain? Ub adds the quotes and everything. [16:26] double whammy. [16:27] It doesn't always work. [16:27] ralsina: does the windows command shell do smart complete? maybe that's the reason lgc wants a different folder name on windows? [16:27] lgc: sorry, but no, we are not planning on removing it by default, although as a side effect of the feature we *are* planning, it may become possible [16:28] mmcc: it round-robins [16:28] ah, ok then [16:28] mmcc: So it may be Ub or something :-) [16:29] If you do a scp, for example, you need to take doble care of the blank space. [16:30] mmcc, no, I don't use command line on Windows. [16:30] lgc: UDF - user defined folder [16:30] lgc, yes, agreed it's not as nice as it could be [16:30] lgc, would a symlink with no spaces be a reasonable workaround? [16:30] lgc: but we can't change it yet [16:31] mmcc, it's not the best thing. [16:31] lgc: you may want to vote on https://bugs.launchpad.net/ubuntuone-client/+bug/581050 [16:31] Ubuntu bug 581050 in Ubuntu One Client "Ubuntu One Directories Are Hard Coded" [Wishlist,Confirmed] [16:31] ralsina, you bet. [16:33] mandel, pong === gatox_lunch is now known as gatox [16:37] alecu, pong? [16:38] and now really, lunch === salgado is now known as salgado-lunch [16:44] ralsina, thanks for the link. I posted my opinion there. On the other hand, I wonder why I get the "file sync is up to date" checkmark but at the same time the Ubuntu One folders appears as if its syncing. [17:07] ok, I'm going crazy debugging within twisted.. I'm off to take the dog for a walk, see you all tom! [17:09] ok mandel, byw [17:10] mandel, bye [17:10] mandel, enjoy while the dog walks you :P [17:10] ugh [17:10] evil allergies [17:12] ralsina: it's possible to rename it now [17:12] ralsina: we just don't document it, because it's a pain to support [17:15] I think I've found the bastard.. but I really need to walk the dog :( [17:15] adios.. [17:21] gatox, ping [17:21] alecu, pong [17:21] gatox, I'm taking a look at the branch you uploaded [17:21] gatox, I'm very curious about "PATH_SNAPSHOT.pop(path)" [17:21] alecu, yes, i need to add a better check there [17:21] gatox, is PATH_SNAPSHOT supposed to be a constant? [17:22] gatox, why is it being modified then? [17:22] alecu, no, it wasn't suppose to be a constant.... maybe i should name it path_snapshot...... it was to store the different snapshots that all the instances of FileEventCallback will use [17:23] for the case when we want the data of the folder before the thread reach there [17:24] alecu, if you want, we can have a quick mumble about that, and i can explain you the problem and why this sollution [17:24] gatox, ok, but let's do it in 30 mins, ok? [17:24] alecu, ack! [17:24] alecu, let me know when you are free [17:37] briancurtin, ralsina quick review for you guys: https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-1010102/+merge/109193 [17:37] mmcc: will do [17:37] kind of sad I missed number 1010101 for my bug [17:37] ok, leaving for lunch now... [17:45] mmcc: got it [17:46] mmcc: aquarius got bin(93) early today [18:05] mmcc, briancurtin: approved and globally approved for being trivial [18:05] mmcc: now, I will do the triaging or this bug, ater dobey sends his presentation you will have to do it ;-) [18:06] my brain is in shambles at the moment. [18:07] dobey: I am not rushing you [18:07] dobey: take a break? [18:08] it's not so much a shambles of exhaustion, but more of surprise [18:13] dobey: ok, I'll bite. Surprise? [18:13] * ralsina held for 5 whole minutes there [18:15] ralsina ping [18:15] aquarius: pong [18:16] ralsina, my mate from earlier? the one who logged out? He now has a ROOT_MISMATCH :) [18:16] should he fix it by u1sdtool -q; rm -rf ~/.ubuntuone; u1sdtool -c ? [18:16] aquarius: let me find the URL [18:16] ralsina, https://one.ubuntu.com/help/faq/what-does-the-root_mismatch-error-mean/ [18:17] is WRONG [18:17] step 1 is: just sign out and sign in again :) [18:17] which does not work :) [18:18] ralsina, and the rest of it just suggests backing up the folders.. but not deleting .ubuntuone [18:18] I would have thought that deleitn g.ubuntuone would be the best because then it's like the machine is fresh [18:19] aquarius: checking... [18:21] aquarius: the solution there assumes the problem is solved by logging in with the correct user [18:21] ralsina, indeed [18:21] aquarius: it's not .ubuntuone. it's ~/.local/share/ubuntuone [18:21] aquarius: your friend logged in first with the wrong one, which is more unusual ;-) [18:21] http://askubuntu.com/questions/31940/root-mismatch-error-after-following-faq-directions [18:22] aquarius: [18:22] ^ [18:24] ralsina, heh [18:24] mandel! don't leave :( [18:24] * aquarius reads [18:25] elopio: i think he left a while ago :P [18:25] dobey: yes, I was just making a dramatic cry. [18:26] elopio: now you run alongside his train, with a white handkerchief in your hand! [18:28] and he will return in 20 years just to find that I died leaving a son that has his eyes. [18:28] ralsina, that looks useful, thanks! joshuahoover should update the faq, perhaps? [18:28] aquarius, joshuahoover: maybe! [18:29] elopio: awwwwww snif! [18:29] aquarius: I can edit the FAQ. What do you want me to change? [18:31] elopio, https://one.ubuntu.com/help/faq/what-does-the-root_mismatch-error-mean/ is updated by http://askubuntu.com/questions/31940/root-mismatch-error-after-following-faq-directions which rye did, I think [18:31] elopio: the ROOT_MISMTCH error has 2 solutions really, [18:31] elopio: one case is, you logged with user A, then with user B, and want to go back to user A [18:31] elopio: that one is covered in the FAQ [18:32] elopio: the other case is, you logged with user A, it works, but you really want user B [18:32] elopio: that one is described in askubuntu [18:33] ralsina: in askubuntu rye says rm -rf ~/.local/share/ubuntuone/syncdaemon [18:33] that's also on the FAQ. [18:33] elopio: yes [18:33] elopio: really? [18:33] and yes it is [18:33] I'm looking at it on Option Two [18:33] https://one.ubuntu.com/help/faq/what-does-the-root_mismatch-error-mean/ [18:34] aquarius: READ THE WHOLE PAGE ;-) [18:34] And so should I [18:35] ralsina, ya, but option 1 is recommended, and doesn't work, and option 2 days to back everything up but not to delete your .ubuntuone folder [18:35] aquarius: you owe me 1 minute of my life. It was going to be the greatest minute. [18:35] I do. [18:35] aquarius: it does delete the correct folder [18:35] ah, just the syncdaemon folder? [18:35] why do we need to back up all the files? [18:35] aquarius: yes, just the sd metadata needs to be deleted [18:35] rye never mentions to delete .ubuntuone. [18:35] aquarius: plus this way you steal all the other account's MP3s [18:36] elopio: .ubuntuone is in UDF of purchased music, it's not a very special folder [18:36] what we should add is why option one sometimes is not the right thing to do. [18:38] elopio: right. Option 1 is if you want to go back to the 1st account, Option 2 is for when you want the 2nd account to work [18:38] neither one should be recommended over the other in principle [18:39] ralsina: why such a spike in ROOT_MISMATCH-related activity? [18:39] ralsina: ok, I'll add that. [18:39] I want to file a bug about the U1 sharing folder email, any hints on which project I should be selecting? [18:39] rye_: aquarius has friends! [18:39] infrastructure perhaps? [18:40] rye_: I overheard something about duplicated accounts due to the ubuntu pay stuff. [18:40] might be related. joshuahoover should know. [18:40] rick_h: for the email itself, https://bugs.launchpad.net/ubuntuone-servers [18:40] ralsina: ty [18:41] elopio: duplicate accounts was related to a bug rye found and it's fixed now...though we still need help fixing the duplicate accounts [18:41] elopio: bug #1004009 [18:41] Error: Launchpad bug 1004009 could not be found [18:41] elopio: https://bugs.launchpad.net/ubuntuone-servers/+bug/1005819 [18:41] Ubuntu bug 1005819 in Ubuntu One Servers "Merge duplicate accounts created by bug 1004009" [Critical,Confirmed] [18:42] joshuahoover: you are here. I updated the faq, I hope you don't mind :) [18:43] elopio: which one, the root_mismatch? [18:43] joshuahoover: yes. https://one.ubuntu.com/help/faq/what-does-the-root_mismatch-error-mean/ [18:43] elopio: yeah, i hate the formatting on that one...very confusing [18:44] joshuahoover: and the windows stuff is missing the \ [18:45] what's that format? markdown? [18:45] elopio: right [18:46] ralsina: and for the U1 files web UI? Same? [18:46] rick_h: yes [18:47] ok, now the backslashes are escaped. [18:59] alecu, do you want to mumble?? i'm convinced about the solution, but maybe you don't :P [19:10] gatox, uh, I forgot about your mumble. I must leave now to kinder, let's do it first thing tomorrow. [19:10] alecu, ok, no problem..... i'm moving to fix another tests..... because i need to ask mandel something for the remaining test inside test_filesystem_notifications [20:02] elopio: so i think i have everything built and in the right places...what's something i can actually do on u1cp or in the visualizer to know that this actually works? [20:04] like, before i wasn't able to get the visualizer to connect or anything, but after putting the plugin in the pyqt packages then starting up u1cp -testability...the visualizer started without issue, so i think its ready to run [20:04] briancurtin: the visualizer has a recording utility. You can click things and then reproduce them. [20:04] elopio: hm, the record menu is grayed out [20:04] briancurtin: but if you can see the tree of objects from the control panel on the visualizer, that's should be enough to check it works. [20:05] hm, i dont see that [20:05] briancurtin: let me send you the screenshot from what we should see. [20:08] briancurtin: http://ubuntuone.com/3sVbFw0mPiESbMuNzk2XP9 [20:09] elopio: do you start u1cp on your own, or do you start it from within the visualizer? [20:09] briancurtin: I start it from the control panel. [20:09] osrry [20:09] ahhh, I need a little break, jaja. [20:09] briancurtin: I start it from the terminal [20:10] ubuntuone-control-panel-qt -testability [20:10] then, on the visualizer I hit refresh. [20:10] you need the qttasserver running before starting the control panel. [20:11] i've tried the same way and it doesnt seem to be showing up in the visualizer [20:12] briancurtin: have you started the sample calculator? [20:12] i cant figure out how to get it compiled for windows [20:12] that should definitely appear on the visualizer. Then, our control panel should work just as that calculator. [20:13] but I just know what should happen. I don't know how to do that, sorry. [20:14] i'll try again to find out how to build that, then see if it works in there [20:14] ok, thanks. [20:16] briancurtin: this is how I compiled everything and had the calculator working: [20:16] http://projects.developer.nokia.com/Testabilitydriver/wiki/WindowsInstallation#a8b.Alternative2:GetTDriverusinggit [20:16] just without the checkout, I used the trunk. [20:16] EOD for me!! bye!! [20:17] tomorrow i'll be here around the standup time..... doctor appointment in the morning [20:22] elopio: i got calculator built and running yet it doesnt show up. weird. anyway, at least i now have that part... [20:25] briancurtin: yes, that's weird. That didn't occur to me. [20:27] briancurtin, elopio: maybe the plugin is not loading for some reason [20:27] briancurtin: is there something that traces files being opened on windows, like strace and dtrace on linux/mac? [20:27] i'm rebuilding everything to ensure it was all built in release mode. maybe something was in debug [20:33] I'm planning to push a branch that adds my current setup-mac.py and a couple of template files to ubuntuone-windows-installer - but I don't think I've heard a solid yes that they should go there... [20:33] it could really go anywhere [20:33] mmcc: it goes there [20:34] mmcc: we will rename it eventually, so please put it there :-) [20:34] ralsina: ok that's pretty solid, but I'd be more comfortable if there was a DTD [20:35] THERE [20:35] I'm testing the switch that tells it to use fresh branches, then I'll push it up [20:35] mmcc: awesome [20:36] then I have a fix for bug 992593, but it's kind of blocked by the test problems that mandel is working on, so I'm waiting on it [20:36] Launchpad bug 992593 in Ubuntu Single Sign On Client "SSO backend command on darwin is not correctly provided" [Medium,In progress] https://launchpad.net/bugs/992593 [20:37] so my plan is to push on control-panel platform support so the main package that setup-mac.py builds actually does something [20:39] mmcc: awesome * 2 === m_conley is now known as m_conley_away [21:23] are there any problems with reordering the test suites in controlpanel's run-tests? I'm adding darwin support and it'd make the script a little cleaner to run the dbus tests first. (fewer ifs) [21:26] generally no === salgado is now known as salgado-afk [21:36] dobey, you mentioned adding a set -x to run-tests over in dev-tools. would that be useful for eg. tarmac? [21:37] also I had to move 'set -e ' below the line that does $(which xvfb-run) or else it just exits quietly when it can't find it [21:38] well -x prints the commands being run, iirc [21:38] what happens if it can't find it after moving the set -e below it? does it exit with a useful error? [21:40] hmm, on darwin we just ignore it and don't use xvfb. [21:41] if it's on linux and can't find xvfb, I should add a useful error... [21:42] right, we definitely need it on linux [21:42] does it pop a million windows without it, on osx? [21:44] the sso client branch pops a bunch of windows, but it's fast enough that it's a blur. [21:45] in control-panel I can't run enough tests yet to see any windows [21:46] but all the non-GUI tests pass, once I tell it to use the qt4 reactor instead of the gi reactor when on darwin... [21:46] otherwise the webclient tests try to use libsoup [21:52] also, I'm aware of no eqivalent to xvfb on OS X. [21:57] right [21:57] well, there's xvfb on osx also, but we're not using X on osx, so it's not useful :) [21:58] yup. that reminds me, I should switch back to xterm [21:59] * mmcc dislikes Terminal.app, which thinks I might ever want to print my scrollback instead of using meta-P [22:00] I mean really, why does it even support printing? Come on, save a tree [22:04] heh [22:04] so dobey, in your opinion, if we don't have xvfb on linux we should abort with an error? I looked at the log for when the test for xvfb was added but it didn't enlighten me as to why it has a default "" [22:05] if that makes sense... there's a revision that adds XVFB_CMDLINE="" before the 'which', but of course if the which fails, we're done, so why the default? [22:05] i don't know that it should error, but we should print something. like do an echo "Runnint tests without xvfb." or something [22:05] it really shouldn't fail silently [22:05] for real [22:05] ok, I'll print a warning before we window-bomb with tests [22:06] oh, there's a default because we have to add an argument to it [22:06] i guess [22:07] or we should just alwyas require xvfb on linux, unless an env var is set to not use it or something [22:07] anyway, i gotta go [22:07] maybe i can actually get the grass cut today [22:07] later [22:16] my u1 has been trying to upload a file for days and keeps retrying [22:17] i have now moved that one file out of a synced folder and it still seems to be trying again ☹ [22:26] popey: :( I think you might have to ping someone directly this time of day... I'm still around but I'm not going to be much help [22:29] for those keeping track: control-panel now has 1421 out of 1432 tests passing on macs [22:30] mmcc: I'll ping someone tomorrow ☺ [23:12] end of my day. see you all tomorrow [23:12] have a good night mmcc.