=== zyga-afk is now known as zyga [09:44] Happy Day-After-Valentine's-Day! [09:45] JamesTait: where is my box of chocolates?!?!? [09:46] okay, who knows a reliable way of determining how much bandwidth is used by the process for sending data to outside world? [09:47] best if it just shows - current transfer for python - 50KiB/s [10:23] hm, it looks like bandwidth throttling is working [10:24] i.e. it transfers a bulk of data, then stops for some period of time, so it is more like interrupting transfer than the delay before every packet [10:24] anybody here saw ubuntuone bandwidth throttling not working? [11:42] morning all [11:42] rye: i did at one point [11:43] on maverick i think it was [11:43] (bandwidth throttling) [11:43] duanedesign, i am trying to find out how to make it misbehave and even though it looks like once it failed to stop streaming it no longer fails :-/ [11:44] i wonder whether it is unable to modify throttling limits durnig the transmission [11:44] interesting idea [11:44] i am trying to play catch up to all the new changes :) [11:45] been looking for a job last few months so have not had the time I have wanted to put into U1 [11:46] hi guys [11:46] rye: the controlpanel.log is replacing the preferences.log? [11:46] hey karni [11:46] nokia abandoned linux. what a fail. [11:47] uggh [11:47] duanedesign, yes, completely, and it comes in 2 packages to enable 3rd party control panels for ubuntuone (not tied to gtk i mean) [11:47] rye: oh cool [11:47] karni, nokia abandoned symbian, what a win... but what will happen to Trolltech's QT? [11:47] rye: and the credentials log? [11:48] hello everyone [11:48] rye: from what I see nokia has been promoting Qt quite intensely recenly (if that's what you're talking about) [11:49] karni, yes, exactly, since they believed QT is the magic thing that will attract the developers to the platform. They have even created a remote compile service so that you could compile code for symbian even on linux (sources are sent there, compiled, then binary is sent back)... [11:50] rye: wow, cool. and they have abandoned that already o_O? [11:52] karni, need to check, nevertheless i tried to compile some examples from the QT Creator/Nokia SDK thing with that remote compile and all failed. Probably symbian and I have a mutual hatred [11:52] rye: :D [11:52] nessita, credentials.log - whose part is that - sso, ubuntuone-syncdaemon or ubuntuone-control-panel? [11:53] rye: ubuntuone.credentials module, a thin layer on top of ussoc to provide dedicated auth services for U1 [11:54] rye: but basically is part of ubuntuone-client [11:54] * duanedesign nods [11:54] rye: would you have some time to review (with fieldtest) https://code.launchpad.net/~nataliabidart/ubuntuone-client/really-download-ro-shares/+merge/49462 ? [11:57] nessita, on it [11:57] rye: is anyone workiing on bug 718196 [11:57] Launchpad bug 718196 in ubuntuone-control-panel (and 1 other project) "Add control panel logs to apport reports (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/718196 [12:33] nessita, approved [12:33] rye: awesome, thanks [12:33] duanedesign, i had some initial bits but i also wanted nessita to tell what other logs can be collected [12:35] ahh, ok [12:36] just saw something I might beable to help with :) [12:36] be able* === yofel_ is now known as yofel [12:39] good morning everyone! [12:40] duanedesign, an idea is popping up regularly in my head about collecting not all syncdaemon.log but grepping out some parts (possibly screaning the paths) so that we have more info immediately [12:40] hi [12:40] yesterday I used the nightly ppa to get new u1* on my maverick box [12:41] there was a bug with missing platform.linux.tools python package [12:41] AFAIR someone said it's just a matter of hours till this is fixed [12:41] is there any update to this? [12:44] rye: yes that would be nice. Some of those logs you get hypnotized about half way through and start to possibly miss information [12:44] hello zyga [12:44] zyga: good question [12:44] duanedesign, hi, how are you1 [12:46] zyga, hm, is it working for you now? because my maverick sd starts properly [12:46] rye, nope, let me check my precise version [12:47] I'm on 1.5.3+r865~maverick1 [12:47] and it's not working [12:47] I can paste the traceback if you want [12:47] zyga: I was. sorry, no fix yet, I am working on it right now. [12:47] ralsina, okay, thanks, I'll check again after upgrade [12:49] zyga, hm, u1sdtool not working? [12:49] There's no way bug #572006 is for the desktop+ team right? [12:50] Launchpad bug 572006 in ubuntuone-servers "Shared File URL are Guessable (affects: 5) (dups: 1) (heat: 15)" [Undecided,New] https://launchpad.net/bugs/572006 [12:50] rye, yes, ImportError [12:50] ralsina, no and it is web/foundations/anybody since there is already a plan on how to prevent that... and it is only server-side [12:50] okay, why does mine work? [12:51] rye: I'll assign to web [12:51] ralsina, it's foundations :) [12:51] zyga, ok, rebooting the vm to re-test, my sd nightlies work on maverick [12:52] zyga could you post the traceback? It works for me too :-( [12:53] zyga@fx-maverick:~$ u1sdtool [12:53] Traceback (most recent call last): [12:53] File "/usr/bin/u1sdtool", line 33, in [12:53] from ubuntuone.platform.linux.tools import ( [12:53] ImportError: No module named platform.linux.tools [12:53] ralsina, I'm sure it's a missing dependency then [12:54] zyga: let me re-check on maverick [12:54] * zyga will have to look at command-not-found for python packages and modules that hints on import error :-) [12:57] zyga: should be in python-ubuntuone-client [12:58] ralsina, I have 1.5.3+r865~maverick1 [12:58] ha [12:58] it's there [12:58] so what could have happened? [12:59] hmm I cannot import ubuntuone.platform [12:59] zyga: output of "dpkg -L python-ubuntuone-client | grep tools" please? [13:00] /usr/share/pyshared/ubuntuone-client/ubuntuone/platform/linux/tools.py [13:00] zyga: remove and reinstall python-ubuntuone-client [13:00] zyga: ok, your python is screwed [13:00] zyga: oh, wait, no it isn't :-) === hito_jp0 is now known as hito_jp [13:00] Or rather, your site.py may be, I am not really sure :-( [13:01] so, should I reinstall to check [13:01] zyga: yes please [13:02] still the same [13:02] hmmm [13:04] tell me what you get with "python -m ubuntuone.platform.linux.tools" [13:04] python -m ubuntuone.platform.linux.tools [13:04] /usr/bin/python: No module named platform.linux [13:04] zyga: ok, I think I have a clue now [13:05] zyga: friday we changed how the python modules are imported and something is broken. [13:05] ralsina, you have a custom import hook? [13:05] please run "python -v -m ubuntuone.platform.linux.tools" and put the output in a pastebin [13:05] ralsina, I looked a the package and it looks normal to me, I'm surely missing something [13:05] zyga: pth files [13:06] zyga: or could be a stale .pyc file [13:06] http://pastebin.ubuntu.com/567301/ [13:06] zyga: ok, I'll read that. It's going to take a little while ;-) [13:07] zyga: do you have a /usr/lib/pymodules/python2.6/libubuntuone/ubuntuone/__init__.py ? [13:08] yes [13:08] $ ls -ld /usr/lib/pymodules/python2.6/libubuntuone/ubuntuone/__init__.py [13:08] -rw-r--r-- 1 root root 0 2011-02-14 13:26 /usr/lib/pymodules/python2.6/libubuntuone/ubuntuone/__init__.py [13:08] I checked that too while looking at package structure [13:09] Well, you shouldn't! [13:09] hmm, why not?' [13:09] At least I don't have it :-) [13:10] w8 [13:10] oh [13:10] dpkg -S that file? [13:11] already did [13:11] not found [13:11] hmmmm [13:11] Well, that's the problem, you have bits and pieces of some old ubuntuone thing. Before the change, they were later in the path than the oficial one. Now they are earlier, and are breaking things. [13:11] I remember one thing [13:11] just a second [13:11] let me check [13:13] hmm [13:13] I thought this might be caused by http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=479852 [13:13] Debian bug 479852 in python-central "python-central: does not remove files when upgrading to a version not using pycentral" [Important,Open] [13:13] but I don't think so anymore [13:13] it's related to older pycentral [13:13] so... [13:14] any more ideas how this could have stayed behind [13:14] zyga: no idea really [13:14] But I found the same thing in one of my VMs for another module [13:16] so it could be a more common bug [13:16] I'm trying to think if this machine was hit by any random crash but I cannot think of anything really [13:16] and this issue popped out as soon as I started using the nightly ppa [13:17] zyga: there is something broken here, it's just that I don't know what it is yet [13:17] zyga: but I have a VM with very broken imports like yours [13:20] zyga: I have a whole set of modules in /usr/lib/pymodules/python2.6/ubuntuone-client that I have no idea where they came from [13:20] Do you? [13:20] zero idea, I don't know how pycentral works, all I know is that it tries to maintain the source away from the .pyc files for each python version [13:21] it looks like stale cache, leftovers from previous installation, right? [13:21] that were not removed for some reason [13:21] perhaps doko knows more [13:21] zyga: probably the uninstall script failed badly [13:22] I am snapshotting the VM to check if removing that folder fixes the problem [13:22] k [13:23] ralsina, yet another reason why dpkg should not have so many scripts and should just use triggers and other declarative actions [13:24] zyga: still broken [13:24] I need to look much deeper, sorry [13:24] ralsina, I'm playing with pycentral binary [13:24] sure [13:29] zyga: I wish so much dpkg -S had told me those files were symlinks to files owned by a deb :-( [13:30] ralsina, what did you do exactly? [13:30] zyga: nothing irreversible ;-) [13:35] zyga: do you have a /usr/lib.python2.6/dist-packages/ubuntuone/ ? Deleting that fixed it for me. [13:35] ralsina, no I dont [13:35] hmmm [13:35] odd [13:36] zyga: grmbl [13:36] oops that 'slib/python2.6 :-) [13:36] I can't copy/paste from that VM so I am retyping and made a typo [13:39] ralsina, you mean /usr/lib/python2.6/dist-packages/ ? [13:39] I still don't have ubuntuone there [13:39] zyga: yes, it was a typo [13:42] zyga: do you have a libubuntuone installed? [13:43] if you mean libubuntuone-1.0-1 then yes [13:45] ok, uninstall it [13:45] hey all, I need some reviews here: https://code.launchpad.net/~alecu/ubuntuone-client/aggregator-fixes/+merge/49806 [13:47] alecu, looking [13:47] alecu: I'll review [13:47] alecu: small thing I noticed about notifications: starting nautilus will always cause a "file sync finished" bubble. [13:48] ralsina, gone [13:48] ralsina, weird. How are you restarting nautilus? by killing it? [13:48] that fixed i [13:48] alecu: yes, I am testing the plugin :-) [13:48] :-) [13:48] ralsina, can you please enlighten me [13:48] zyga: I have absolutely no idea why that happened :-) [13:48] ralsina, hmm :D [13:49] alecu: very likely that nautilus is trigerring a GetPublicFIles that translates to a command [13:49] zyga: but if reinstalling it does it again, file a bug ;-) [13:49] yes, a python script that calls getpublicfiles does that too [13:49] meh [13:49] nessita, right! good point. [13:49] But I don't think that should cause bubbles [13:49] ralsina, sure [13:50] alecu: try with magicicada and an IDLE syncdaemon [13:50] nessita, queueing command (0/1): GetPublicFiles [13:50] nessita, it's exactly that. [13:50] :-) [13:50] ralsina: so, the sile sync complete is kinda a true message [13:50] file* [13:50] though we may hid that? [13:51] we should only start bubbling with upload/download [13:51] +1 [13:52] BTW, can anyone tell me why this script is broken? https://pastebin.canonical.com/43304/ [13:52] zyga: might as well stick my nose in here then [13:52] ralsina: broken how? [13:52] nessita: as in "not working" and "I am clueless about twisted" :-) [13:53] I am trying to debug why nautilus is not getting anything back from getpublicfiles [13:53] And I am not being able to get anything myself. But I am not sure if it's because my script sucks or not . [13:53] ralsina, reinstalling that library did not cause a regression [13:53] zyga: ok, so it's magic. [13:53] tumbleweed, ralsina has a VM with a snapshot that shows the broken state [13:53] yes, we will only bubble with upload/download in an upcoming branch, there's a bug for it... [13:54] bug #715842 [13:54] Launchpad bug 715842 in ubuntuone-client "Percentage in notification is a bit off: Uploaded 52/52 files, 95% completed (affects: 1) (heat: 6)" [Medium,Confirmed] https://launchpad.net/bugs/715842 [13:54] tumbleweed, I got this broken state by installing a few packages form ubuntuone nightly ppa [13:54] zyga: aha, which ones, and do we know how thy are broken? [13:55] * tumbleweed is no expert here, but I have a fair understanding of the python helpers [13:55] tumbleweed: the fun part is that libubuntuone is not even about python. It's about mono [13:55] tumbleweed, so looking back at my scrollback here, there seems to be a /usr/lib/pymodules/python2.6/libubuntuone/ubuntuone/__init__.py that should be gone, it was present in an older version of a library [13:56] ralsina, (if I miss something please correct me) [13:56] zyga: no, that was not the problem. That file was a symlink to the one python-ubuntuone-client installed. [13:56] Or maybe not, I don't remember anymore :-( [13:56] ralsina, hmm, sorry, so what was the problem? [13:57] zyga: sadly the problem in my VM is only similar, but not the same. [13:57] tumbleweed, from the end user point of view a python import was failing, the package that shipped the python modules was okay but the cache was still holding older .pyc files that broke the import sequence [14:00] ralsina: well, several things. You're not using the reactor how is supposed to be used. You need to install a particular reactor (glib's, gtk's, etc). I switch to use the gobject loop, and is working. The second error is that the list is not returned, but signal back with 'PublicFileList' signal [14:01] ralsina: is there a stand up? [14:01] mandel: wow! I completely forgot [14:01] me (no notes yet) [14:02] me [14:02] me (no notes yet either) [14:02] ralsina, alecu, dobey, thisfred? [14:04] nessita, u1 control panel fails to work when you have static IP assigned, is there any way to force it to assume network is present? [14:04] zyga: it checks against network manager. If network manager reports no connection, then U1 thinks you have no connection [14:05] zyga: we have some bugs to support more net managers, but no schedule to work on those any time soon [14:05] me [14:05] sry [14:05] zyga: ok, unofrtunately ppas make digging through build history hard :) [14:05] me [14:05] nessita, network manager is not started if all your interfaces are configured in /etc/network/interfaces [14:05] which ppa was this? [14:05] tumbleweed, let me check [14:05] zyga: yeah [14:06] zyga: not sure what good solution we can provide there [14:06] tumbleweed, deb http://ppa.launchpad.net/ubuntuone/nightlies/ubuntu maverick main [14:06] zyga: networkmanager is started, it's just not used. does it work if you uninstall networkmanager? [14:06] dobey, let see [14:06] me [14:06] ah, sorry I did not have n-m installed [14:07] if u1cp doesn't work without nm installed, then it is a bug in u1cp [14:07] nessita, start! [14:07] dobey, then it's a bug I guess [14:07] DONE: day off. Last Friday I coded several branches for small style fixes to the control panel UI. [14:07] TODO: file storage API work [14:07] BLOCKED: nopes [14:07] NEXT: alecu [14:07] sorry, NEXT: mandel [14:08] DONE: IPC servercie & client on windows. [14:08] TODO: Ensure all tests pass, talk with claire about windows installer UI. Talke with ralsina about starting ubuntu sso UI on windows. [14:08] BLOCKED: no [14:08] LOVE: http://bit.ly/fBZawM [14:08] alecu, please [14:08] DONE: fought *a lot* with VirtualBox, and the xorg from alpha2 just won't work with 3d needed for unity. Installed alpha2 in the desktop, so now I'm maverick free. Finished a branch to add a config option to disable notifications (part I) that's up for review: https://code.launchpad.net/~alecu/ubuntuone-client/aggregator-fixes/+merge/49806 [14:08] TODO: finish part 2, catch up with DroidCouch [14:08] BLOCKED: no [14:08] THISFRED: you [14:08] DONE: started on https://launchpad.net/ubuntu-one-couch | tried more things to get Unity on nvidia working and failed [14:08] TODO: Bug #719039, Bug #719042, Bug #702183, Bug #702116, make UDF notifications aggregate [14:08] BLOCKED: unity/nvidia [14:08] HATE: binary drivers, and depending on stuff that isn't ready [14:08] Launchpad bug 719039 in ubuntu-one-couch "Fix ubuntuone-couch-query to work with current SSO code (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/719039 [14:08] Launchpad bug 719042 in ubuntu-one-couch "Add tests and refactor out code to a library (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/719042 [14:08] Launchpad bug 702183 in ubuntuone-client (Ubuntu) (and 1 other project) "Syncdaemon needs to open the control panel in the background and change the launcher icon to urgent when the user exceeds their quota (affects: 1) (heat: 120)" [Medium,Triaged] https://launchpad.net/bugs/702183 [14:08] Launchpad bug 702116 in ubuntuone-client (Ubuntu) (and 1 other project) "Syncdaemon needs to show and update a progressbar in the Ubuntu One launcher for pending operations (affects: 1) (heat: 120)" [Medium,Triaged] https://launchpad.net/bugs/702116 [14:08] ralsina! [14:08] DONE: lots of bug triaging, canonicaladmin, sprint planning (ask me!) [14:08] TODO: reviews, management stuff, sprint planning, triage another 6 tons of bugs, some coding (my days are all alike) [14:08] BLOCKED: no [14:09] λ DONE: bug 718806, updated servers for u1sync split, [14:09] λ TODO: banshee mustic store fixes, evaluate SRUs for maverick [14:09] λ BLCK: on-call for jury duty, server upgrade fail [14:09] Launchpad bug 718806 in tarmac "LP EDGE is no more (affects: 1) (heat: 6)" [Critical,Fix released] https://launchpad.net/bugs/718806 [14:09] alecu: you have unity working on virtualbox now, or not? [14:09] thisfred, not [14:09] comment time! [14:09] ah :( [14:09] same here [14:09] thisfred, I ended up with a full bare install on the desktop. [14:09] ralsina: computers hate me. [14:09] right. And that also doesn't work for me [14:10] ralsina: are you done with the sprint planning? [14:10] I wonder if a eee 701 is enough for unity. It'sthe only box I can do bare installs in [14:10] also, any spanish speaker should watch the link I added :P [14:10] mandel: it's advancing! [14:10] thisfred, oh, and the nVidia binary drivers won't work at all, so I ended up using the open source experimental (noveau), and it works just fine. [14:10] thisfred, I think I've got a nVidia 6600 something [14:11] alecu: really? nouveau did not yet work for me yesterday [14:11] ralsina: fell yourself asked about sprint planning [14:11] SPRINT NEWS: the new plan is 2 sprints [14:11] ralsina: i think that machine is poulsbo, so i doubt it [14:11] alecu: ah yeah, my hardware is probably older than that [14:11] dobey: no, it's pre-poulsbo. i815? 845? [14:11] thisfred, yeah, I'm pretty sure it depends on the hardware a lot... [14:11] ralsina: oh, then it might work, depends on how much GL support unity needs. the i815 isn't that great at it [14:12] dobey: it was enough for compiz a while ago. And even kwin 3d effects worked somewhat. [14:12] 2 sprints? Is that serious? [14:12] thisfred: or rather a split sprint not-in-parallel [14:12] ah ok [14:12] ralsina, how comes? [14:12] :) [14:12] One in BA with most of the team, focused on killing bugs pre-release [14:13] Another in London with Manuel, me, chipaca and UX help [14:13] alecu: i dont like u ;) [14:13] ralsina: yeah, a while ago. :) [14:13] The London sprint is about the windows port [14:13] hehe [14:13] oh, that makes sense. [14:13] dobey: a while as in "last november" :-) [14:13] dobey: of course it was using Arch, so who knows ;-) [14:14] ralsina, I don't think arch makes magical gl drivers [14:14] alecu: it's the other way around, They don't freaking touch xorg :-) [14:14] :-) [14:15] So, I am now starting the formal process to get the sprint(s) on the way, so everyone but mandel should schedule april 4th-8th as sprint time, and mandel should think about april 11th-15th [14:15] my old laptop has i815 i think, but i uninstalled unity and all that jazz and switched to metacity, because i don't like spending all my time in swap [14:15] ralsina: ack [14:15] dobey: I have 4GB of ram (and a 4GB SSD ;-) [14:15] zyga, ralsina: looks like you didn't do any cleanup after migrating from python-central to python-support [14:16] http://wiki.debian.org/Python/central2support [14:16] ralsina: you have 4GB of ram in an eee 701? [14:16] tumbleweed: could be! [14:16] ralsina: be sure to ping CardinalFang and vds (if they are coming to the sprint), since they may not paying attention to this [14:16] dobey: oh, no 2GB [14:16] nessita: sure [14:16] anyway, these days you should probably use dh_python2, which handles cleanup :) [14:16] my old laptop has 512M [14:16] which should be well more than enough to live on [14:16] nessita, thx, I was not paying attention [14:16] but alas [14:16] dobey: it came with 512MB. It now has 6hour battery, and a touchscreen. It's pimped to the limits :-) [14:16] tumbleweed, thanks for digging into this [14:17] ralsina: but no ticket buying yet, right? [14:17] zyga: that was a quick skim. I don't know *exactly* what you are seeing [14:17] dobey: see what tumbleweed said, we have been experiencing some hiccups on maverick nightlies because of that [14:17] ralsina: another 512M stick for my old laptop would be like $500+ [14:17] thisfred: not yet, I'll let you know [14:17] kthx [14:17] tumbleweed, but it could be this precisely, ralsina will look at this from their perspective [14:18] zyga: I will try at least. [14:18] i am pretty sure dh_python2 vs dh_pysupport have nothing to do with issues in nightlies [14:18] and dh_python2 is new, and not available on lucid/maverick afaik [14:18] dobey: I had a whole set of ubuntuone-client files in /usr/lib/python2.6/site-packages that shouldn't have been there. No idea why, though. [14:19] correct, not in lucid [14:19] ralsina: you did a ./autogen.sh --prefix=/usr && make install probably [14:19] /usr/lib/python2.6/site-packages is either created by install or pycentral. python-support uses pymodules [14:19] dobey: yes I did. Oops. [14:20] damn nautilus testing :-) [14:20] zyga's was really something else. [14:20] ralsina: make -C nautilus install-extensionLTLIBRARIES [14:20] dobey: didn't work because of libubuntuone [14:21] libsyncdaemon, sorry [14:21] well [14:21] So, I have the date, but what city? [14:21] make -C libsyncdaemon install-libLTLIBRARIES [14:21] CardinalFang: BA [14:21] CardinalFang: BA [14:22] dobey: it's a VM, I should just throw them away every day ;-) [14:22] also that [14:22] http://tinysong.com/jrbi [14:22] so what was zyga's problem? [14:22] One important thing: dobey, lucid nightlies are broken [14:23] because python-couchdb is not available, IIUC [14:23] please define broken [14:23] yes it is [14:23] https://edge.launchpad.net/~ubuntuone/+archive/nightlies/+files/python-couchdb_0.8-0ubuntu1%7Elucid1_all.deb [14:23] dobey, ask chad and vds, they are saying it on #desktop+ [14:24] Also, the rythmbox plugin is completely broken on natty because of plugin API changes. I will be taking a look myself today. === zyga is now known as zyga-afk [14:24] dobey, I'll tell you in 30 minj [14:24] anything else? anyone? anyone? bueller? [14:25] life is broken [14:25] Oh, please, quick report in private by everyone, because I have team leads today! [14:26] the only remaining uninstalled motherboard/cpu combo i have that i know works, is the one i have that doesn't have on-board video, and i don't have any PCI or PCIx video cards lying around :( [14:27] ralsina: the working example is http://pastebin.ubuntu.com/567329/ [14:27] nessita: cool! [14:27] nessita: the bad news is that means I have no idea why it's broken on nautilus+libsyncdaemon [14:27] ralsina: is nautilus waiting for the signals? [14:28] ralsina: I changed your script considerably, please check that nautilus is doing the proper stuff [14:29] nessita: nautilus calls libsyncdaemon and sets a callback [14:29] So maybe that's the problem, that it's trying to use a callback instead of a signal [14:29] ralsina: libsyncdaemon should be waiting for the proper signal, you should check that [14:29] nessita: will check [14:49] dobey: https://code.launchpad.net/~ralsina/ubuntuone-client/fix_712674_on_1.4 is not landing. Is something broken? [14:49] In the branch, I mean ;-) [14:50] my server is broken [14:50] dobey: well, that explains it. Good luck with that [14:51] yeah, best buy opens in ~10 min, so i'm about to head up there to see if i can get some parts or just buy another machine to use as a server [14:55] dobey: why don't we use this to ask for canonical server to run tarmac? [14:55] dobey: it's unfair to you that you have to manage that on your own hardware [14:55] because either way i still need to fix my server for other reasons [14:55] dobey: sure you have to, but it's just weird :-) [14:57] well i don't want to have to run tarmac on my machines either, so yes, i would love to get it moved to a magical VM thing. but right this second my network is in utter despair [14:57] i had to ifconfig/route add to get my workstation back on, because i have no dhcp/etc at the moment. and i can't even connect via wifi if i need to use my laptop for something :) [14:58] dobey: like the good old times :-) [14:58] and we really need to come up with a good plan forhaving the magical VM stuff, because just having losoas control everything will be horrible pain for us and them both [14:58] dobey: ok, I'll ask in team leads who I need to bug to move tarmac to official hardware [14:58] and that without l*sas [15:00] we need our own metal running UEC wtih VMs that we can actually do stuff in without having to deal with vogon beauracratic methods to make changes :) [15:00] exactly [15:00] basically more or less like toay but out of your living room ;-) [15:01] anyway, off to recover from my mini disaster. be back shortly [15:01] go ahead, good luck! === teknico is now known as teknico_away [15:07] facundobatista, is it ok that i got this exception after removing the file from the share (as another user) ? = http://paste.ubuntu.com/567345/ (file.flv was supposed to screen the filename :) ) [15:10] rye, it's not ok, it's a bug that alecu is aware, but it's just a zeitgeist detail, won't afect your syncing [15:11] I am aware! [15:12] alecu, thanks!, ah, zg_listener, missed that [15:20] thisfred, that mirah song is lovely [15:20] alecu: yeah I really like it [15:26] beuno, this music-app "prepare" error nearly has be flummoxed. I'm sorry it's taking so long. I'm going down deep in the OS source to find out the cause. [15:26] CardinalFang, cool. Lets give it another day, otherwise lets just download and tell the user it can't be streamed. Sound good? [15:27] beuno, yeah. === zyga-afk is now known as zyga === dobey_ is now known as dobey [16:05] oi === beuno is now known as beuno-lunch === ralsina is now known as ralsina_lunch === teknico_away is now known as teknico [16:49] * nessita -> lunch [16:50] alright, lunch times for me too [17:18] hi, i can't connect to ubuntu one on either of my maverick machines. ubuntu one preferences just says unknown [17:24] birdybee: hey there [17:24] hi [17:24] birdybee: hve you tried removing your tokens from the keyring? [17:24] tokens from keyring? [17:24] birdybee: system -> preferences -> pasword and encryption keys [17:25] birdybee: in there, right click to the entry that has 'Ubuntu One' in its name [17:25] and delete [17:25] ok, deleted [17:25] re open preferences [17:26] birdybee: you should be prompted to login to Ubuntu One [17:26] k, entering credentials... [17:30] hm, it started reuploading 30k files i had previously unchecked for syncing, and deleted from web interface...ugh [17:34] but why does it do that? is there a file or something with a list of places to sync? [17:34] cause in nautilus, it isn't shared anymore, and i have deleted it from web ui [17:35] nessita; (bump for attention) :P [17:36] birdybee: well, when syncdaemon starts, it performs a local rescan, where it build a match between what is on your hard drive and in your local metadata [17:37] birdybee: once that's done, it will connect to the server to actually sync up everything [17:37] that's when the syncdaemon should realize it should not upload all those files [17:37] birdybee: anyways, let me point you to someone from that particular area [17:37] facundobatista: ping === ralsina_lunch is now known as ralsina [17:38] hm, i just noticed in magicicada that it starts to upload files [17:38] birdybee: do you know if you have your logs in DEBUG mode? [17:38] isn't there like a config file or anything, that lists directories? [17:38] erm [17:38] i don't know [17:38] in maverick, without our nightlies PPA, nopes [17:39] k, that's a no then :) [17:39] pure maverick here [17:39] nessita, pong [17:39] birdybee: let me put you in contact with facundobatista, he should be able to help you debug [17:39] weird, cause both my machines were working, now, none of them :) [17:39] ok [17:39] facundobatista: birdybee is reporting that "it started reuploading 30k files i had previously unchecked for syncing, and deleted from web interface...ugh" [17:39] (about syncdaemon) [17:40] facundobatista: can you please help him/her digging into what is going on? [17:40] nessita, do you know what "unchecked for syncing" means? [17:40] well, i first did stop syncing this, on the folder, then i deleted the folder in webinterface on server [17:40] facundobatista: nopes, but you can ask directly the reporter :-) [17:40] i figured that was the way to get rid of the content [17:41] that would be me :) [17:42] facundobatista, did i do it wrong? [17:42] birdybee: how did you 'stop syncing this'? using nautilus and the checkbox in the ribbon? [17:43] i rightclicked folder, and stop syncing this folder [17:43] then i deleted the content in one.ubuntu.com web ui [17:43] facundobatista: ok, uncheck for syncing is translated to delete the UDF in the plugin [17:43] hrmm [17:44] facundobatista: I'm more worry about syncdaemon uploading the same even if the UDF was removed in the web ui [17:44] birdybee, I would need to see the logs, could you please zip your log directory and send that to me? [17:45] nessita, ok, it went back to unknown now, and magicicada is just grey, and nothing happens anywhere related to ubuntu one... [17:56] fun [17:57] almost like a statik-split [18:01] nessita, you there? [18:23] facundobatista: this is probably somewhere in the sources, could you tell me how SD handles such situation: a file appears in the synced folder. SD makes a file on the server and starts uploading content - in the mean time, it receives the notification of newVolumeGeneration. how does it know it shouldn't start downloading the file which it is uploading? [18:25] facundobatista: does it try to enqueue the download command, but since upload is still in processed queues, it doesn't enqueue it/cancels the download, or some other approach? [18:25] facundobatista: by "somewhere in the sources" I meant somewhere deep.. === ferai is now known as Guest23248 [18:27] karni, the new volume generation is because a change in that very file? [18:28] facundobatista: either because we just made the server create a new file or started uploading the content, yes. [18:28] facundobatista: it's a new file that appeared in the system (say, somebody pasted it into a synced folder) [18:29] facundobatista: I just don't want to invent the wheel for 2nd time, that's why I thought I'd ask how you guys handle this. === Guest23248 is now known as jefferai [18:39] dobey: hey there. did you see bug #717829? [18:39] nessita: Bug 717829 on http://launchpad.net/bugs/717829 is private [18:39] no [18:40] karni, the "magic" for that is in handle_AQ_DELTA_OK in Sync [18:41] facundobatista: thank you [18:41] nessita: I saw something similar this morning on maverick and it was because of old pyc files left around [18:41] facundobatista: I'm sorry I don't have more, I owe you a beer or couple more ;) [18:42] ralsina: the bug has zillion of duplicates [18:42] nessita: ugh [18:43] karni, we'll meet somewhen, and I'll remember that :) [18:43] facundobatista: :) [18:46] nessita: I am now switching to those versions to see if it happens to me too [18:48] nessita: nope, doesn't happen [18:50] ralsina: I believe you, but I also believe the several bug reports :-) [18:50] ralsina: how are you testing? [18:51] nessita: dobey just answered the bug as "you are missing an update" it seems [18:51] nessita: uninstalled ubuntuone, and installed the versions listed in UbuntuOneClientPackages.txt [18:52] well [18:52] It happened to a guy 20 minutes ago, though [18:52] UbuntuOneClientPackages.txt doesn't list all the ubuntuone related packages that might be installed [18:52] dobey: yeah, but it was just something I could try [18:54] sure [18:54] It's reported on L M and N :-( [19:08] I am now installing a fresh natty VM to see if it happens, just to be sure, but I really doubt it. [19:08] Oh, great, here I was trying to fix the rhythmbox music store plugin and now all python plugins for rhythmbox are broken. [19:09] What a crappy day am I having. [19:17] ralsina: i think bug #718760 is a dupe of one you were already fixing the other day, no? [19:17] Launchpad bug 718760 in ubuntuone-client "problems with already published files in nautilus (affects: 1) (heat: 502)" [Undecided,New] https://launchpad.net/bugs/718760 [19:20] dobey: yep, that would be... bug #701557 I think [19:20] Launchpad bug 701557 in ubuntuone-client "In nautilus and the ubuntuone-indicator "Copy the link" of a published file disappear after time (affects: 2) (heat: 10)" [High,Confirmed] https://launchpad.net/bugs/701557 [19:20] dobey: but calling what I did "fixing" is wildly optimistic. [19:20] well [19:20] arranging to be slightly less broken, if you will [19:20] CardinalFang, any eureka moments? [19:20] I don't think I'm eating lunch === beuno-lunch is now known as beuno [19:20] dobey: not even that [19:20] mostly, I have no idea what's broken. I think it's libsyncdaemon but I am not sure yet. [19:21] ralsina, nessita: I've been discussing bug #715842 with lucio, and I think I found a better way to display that [19:21] Launchpad bug 715842 in ubuntuone-client "Percentage in notification is a bit off: Uploaded 52/52 files, 95% completed (affects: 1) (heat: 6)" [Medium,Confirmed] https://launchpad.net/bugs/715842 [19:21] I will use nessita's and lucio suggestion, and will count *all* operations, even those that are not Upload and Downloads [19:22] and we will not show individual file progress, but just the percentage. [19:22] something like: [19:22] alecu: fine [19:22] "Uploading 53 files, Downloading 30 files [19:22] 40% completed." [19:23] alecu: that looks pretty good [19:23] alecu: and it simplifies things for you, right? [19:23] yes, it means less code :-) [19:27] :-) [19:27] less code is better code. [19:46] beuno, no "Eureka"s yet. [19:46] beuno, I'm merging upstream's updates also. [19:46] CardinalFang, ah, cool [19:46] did we send Sindre the playlist work karni did? [19:47] beuno, Sinde? [19:47] alecu, Sindre, the subsonic upstream dev [19:48] beuno: I was wondering if anybody tested that IRL, but I heard jblout doesn't hang around any more [19:48] oh, I thought "Sinde", as in the anti-internet spanish law. [19:48] karni, we did plenty, yes [19:48] beuno: oh.. good. was it okey? [19:48] alecu, heh [19:48] karni, yeap, lovely [19:49] beuno: I think that Sindre may come up with a cleaner solution as an upstream developer, but that did the trick :) [19:49] cool [19:50] karni, question for you: I'm working on adding Ubuntu One support to DroidCouch, and I was wondering if you already use OAuth in android to talk to Ubuntu One web servers (for publishing, and similar operations) [19:51] alecu: indeed, we do [19:51] karni, cool. Can you point me at some of that code? [19:52] alecu: sure. I can also tell you that there'll be an Account Authenticator on the way, so all U1 android apps will probably use that [19:52] karni, that's "ultra-mega-cool" (tm) [19:52] alecu: since it's not ready yet, I can point you at what we've got atm [19:52] hehe [19:55] karni, thanks for the pointers! [19:55] alecu: more coming ;D [20:49] can I have a couple of reviews for https://code.launchpad.net/~nataliabidart/ubuntuone-client/async-credentials/+merge/49866 ? [20:54] alecu_, thisfred: I assigned bug #718171 to alecu_ but any of view can tackle this (and is important! :-)) [20:54] Launchpad bug 718171 in ubuntuone-client "Untranslatable strings in ubuntuone/status/aggregator.py (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/718171 [20:55] nessita, ack [20:58] * dobey fixes the nightlies package versions [21:00] nessita: I just finished a clean natty + updates install and everything works great. [21:00] ralsina: bug reports keep coming in [21:00] nessita: no idea about those really :-( [21:00] ralsina: not sure what to say. dobey, do you have any ide what is going on (with platform.tools bug)? [21:01] other than people not having the latest updates for all packages, no [21:01] i would need more information to say anything else. [21:07] my guess is an older version of python-ubuntuone or python-ubuntuone-control-panel is installed [21:11] nessita: if I may ask, grep "check_conditions" * under /syncdaemon folder returns only action_queue and volume_manager. is VM the only one calling check_conditions in case of SV_FREE_SPACE ? [21:13] karni: I have absolutely no idea :-) I can look in the source code if you need some help [21:14] nessita: no it's ok. I don't think grep would like to me. I was just suprized that all this checking was implemented only for 'got free space!' event ;) [21:14] nessita: I just expected more conditions on which commands could be waiting [21:15] but since markers are treated separately, there's not much left for *conditions* [21:15] karni: right. facundobatista is the master nehind all that magic ;-) [21:15] behind* [21:15] * facundobatista takes a pidgin out of his hat [21:15] nessita: I didn't want to overuse his helping hand [21:15] hahah [21:16] karni, don't worry, I'm glad you ask these stuff [21:16] * facundobatista has a nice beer counter [21:16] facundobatista: deal is a deal, true! but I should be getting some for the lack of documentation, too ;) ! [21:16] karni, so, which was the question? [21:17] karni, you write that documentation using my responses, and I *surely* will buy you beer :p [21:17] sometimes I think that this is not documented more then comments and /docs because somebody could start reversing this and write an OS server ;d [21:17] facundobatista: hahah [21:18] facundobatista: it's fine. the question was about who's calling check_conditions() on action queue - grep only showed volume_manager so I was a little suprized [21:18] I expected more conditions, but there aren't any more than SV_FREE_SPACE or whatever it was exactly :) [21:18] karni, the check_conditions call is a kind of "hey, commands, check if you have your conditions to run ok because something may have changed" [21:19] the only "condition" to stop something, today, is free space in the server, that stop uploads [21:19] facundobatista: right. but that's only invoked by the volume_manager in one place. am I right? [21:19] aha, right [21:19] so, when VM notices that the free space changed, it calls check_conditions [21:20] right. and if theres no network, StatesManager (through NetworkManager or whatever it was) shuts down the queue of commands [21:20] correct? [21:20] RequestQueue.stop() runs .cleanUp() on every command, from what I recall [21:20] karni, you still with unleashed-queues-6? [21:21] facundobatista: I'm looking at free-the-queue but I haven't looked at the diff exactly [21:21] facundobatista: anything I should be aware of? major changes? === AlanChicken is now known as AlanBell [21:23] karni, more than shut it down, it stops it, and the queue stops the commands in it [21:24] hmm RequestQueue.stop(self) is a one-liner now, self.active = False [21:27] facundobatista: interesting. there's no call to check_conditions on a cmd in RequestQueue.check_conditions(self). just _run_command(cmd) - does that check for conditions? [21:28] karni, btw, all this changed heavily in unleashed-queues-7 [21:28] karni, the command life cycle is simpler now [21:29] I'm fcked. I can't keep up with you (which isn't your fault) [21:29] I wanted to port the SD to have a very cool queueing logic in U1F [21:30] I ended up in such a huge machinery I wasn't sure it was a good idea any more. And day by day, you simplify things. [21:30] If only we would code U1F in April or May, instead of know when there are so many chages :( [21:30] *changes [21:30] * karni cries in the corner === lifeless_ is now known as lifeless [21:31] karni, see it this way: you can ask me what changed, I can not ask anyone what I will do in the future :p [21:32] ok, I'm gone [21:32] see ya guys tomorrow! [21:32] facundobatista: I don't blame you, don't get me wrong. I'm blaming the timing U1F was called into living (d'oh) [21:32] bye nessita [21:32] nessita, chau [21:35] bye all! [21:37] * karni gets back to work [21:38] dobey: that platform.tools bug has stopped being reported 3 hours ago. I am betting on a package being stuck as you said, but let's check it tomorrow, ok? [21:38] ok [21:40] facundobatista: free-the-queue came after unleash-the-queues-7, yup? [21:41] so there's no 'canceled' in AQC anymore huh.. okey [21:41] karni, nop [21:41] facundobatista: wait. that was considering 'canceled', not the branches question ;) ? [21:42] karni, IIRC, I did unleash-the-queues-7 and unleash-the-queues-8, and didn't touch the client until today (unrelated) [21:42] aha [21:42] what was free-the-quueue then? [21:42] * karni gets -8 branch [21:43] I just want to make sure I'm looking at the lastest sources [21:43] karni, well... the latest source is trunk [21:44] karni, but yes, all the heavy changes stopped in the -8 branch [21:44] facundobatista: perfect, thanks [21:44] (until next week or the other, where I resume these changes) [21:45] facundobatista: I'll be happy to work with -8 till that time. Please ping me when you roll out some bigger changes :) [21:45] karni, ok! [21:46] facundobatista: huge thanks! [22:39] later all