/srv/irclogs.ubuntu.com/2010/05/03/#ubuntu-devel.txt

* jdong just found one of the most amusing quotes about the Linux sound stack01:33
jdong"Linux's audio architecture is more like the layers of the Earth's crust than the network model, with lower levels occasionally erupting on to the surface, causing confusion and distress, and upper layers moving to displace the underlying technology that was originally hidden."01:33
psusiOH HELL YEA01:44
psusiwow... my first attempt at having defrag batch up requests and issue a bunch at once with readv() and writev() made it 3 times faster01:45
RAOFpsusi: Sweet.  Your stuff sounds so interesting :)01:46
psusihehe, do I have beta tester number 2? ;)01:47
ajmitchpsusi: it's all safe, right? :)01:53
psusilol, of course not ;)01:54
psusibut that's why you have a recent backup right? ;)01:54
ajmitchsure, only about 2 weeks old01:54
psusigood old dump01:54
psusi2 weeks?  man... gotta dump every day ;)01:54
* ajmitch would like to see this laptop boot up in < 30 seconds or so :)01:54
psusi10 seconds ;)01:55
ajmitchI'd be surprised if it could manage that, it starts up a few heavy things like postgres01:56
psusiit's amayzing what you can do when you get rid of the typical 90+% time wasted waiting for disk rotations01:56
ajmitchgiven that I've probably got nice slow 5400 rpm drives in the laptop, it could help01:57
psusiI finally wrapped my head around the tower of hanoi dump backup pattern... and it is nice02:03
* imbrandon perks up, make a 5400rpm hdd based laptop fast ?02:03
imbrandonheh02:03
psusiimbrandon, yep... trying to make some improvements to ureadahead and combine it with defrag02:50
temugenKeybuk: I updated http://bradmisik.com/trunk/fragraph/ (bzr) a bit. Still want to add some tests, comments, and possibly new graph types, but it's a lot better02:53
RAOFpsusi: Your work doesn't apply to btrfs at all, does it?  That's really what I want to lose my data with at the moment :)02:58
psusihehe, no03:02
psusijust ext2/3/403:02
ScottKRAOF: It's unlikely you need psusi's help for that.03:03
RAOFScottK: I might want dpkg's help, though :)03:04
* psusi is still grumbling about the dpkg sync slowdown to fix ext4 bustedness03:04
=== xnox is now known as dima
=== dima is now known as xnox
psusioh wow, a larger hash table made a big difference...03:57
=== almaisan-away is now known as al-maisan
Keybuktemugen: cool, pulled! :-)05:33
=== Bernardo is now known as Bernardo|away
pittiGood morning06:29
pittiYokoZar: ah, I see06:30
pitticjwatson: there was no definitive decision yet (something for tomorrow's TB), but my gut feeling for maverick is to use unstable; but it can certainly stay at testing for now06:32
nxvlpitti: are you suggesting that we move permanently to testing?06:34
pittinxvl: no, I'm not; as I said, I'd prefer unstable for maverick06:35
pittibut I don't want to decide that on my own only06:35
nxvlmy vote goes for sid06:36
nxvl:D06:36
nxvlbtw, is archive re-org going to be in place for maverick?06:37
arandHow does one debug mountall?06:47
Keybuk--debug06:48
arandKeybuk: Hmm, that works at boot? (It's the plumouth/mountall fsck stalling issue..)06:51
Keybukisn't that supposed to be a plymouth bug?06:51
arandKeybuk: It's both mountall and plymouthd that eats cpu, depending on if you are in the graphical plymouth or jump out with arrowkey, or boot with quiet splash...06:53
arand*without "quiet splash"06:53
Keybukyes, but that can mean it's a libplymouth bug06:53
arandBug #571707 by the way, it seems to be rather common (I've been messing about with it quite a bit lately, there's at least plenty of bootcharts)06:55
ubottuLaunchpad bug 571707 in mountall "fsck progress stalls at boot, plymouthd/mountall eats CPU" [High,Confirmed] https://launchpad.net/bugs/57170706:55
Keybukcertainly by the e-mail volume, there are a number of people affected, yes06:55
arandKeybuk: If you want any debug/testing I've got a free virtualbox which see the issue...07:01
KeybukI think slangasek is looking at that bug07:01
Keybuknot me07:01
arandAh, ok.07:03
dholbachgood morning07:17
KeybukDEBUG:root:Downloading http://ftp.uk.debian.org/debian/pool/main/k/kde-l10n/kde-l107:58
KeybukIOError: ('http error', 404, 'Not Found', <httplib.HTTPMessage instance at 0x769bb00>)07:58
Keybukwell07:58
Keybukscrew you kd07:58
Keybuk+e07:58
Keybuk:)07:59
Keybukhmm08:04
Keybukthat's impressive08:04
Keybukstarted dist-upgrade before bed last night08:04
KeybukDPKG IS STILL RUNNING08:04
=== tkamppeter__ is now known as tkamppeter
lifelessKeybuk: fs? hardware?08:10
Damascenebug 57277608:10
ubottuLaunchpad bug 572776 in ubuntu "Ubuntu should provide update packages for download and use for offline users" [Undecided,New] https://launchpad.net/bugs/57277608:10
Keybuklifeless: ext4 on hard drive08:10
Keybukdpkg is calling fsync over and over and over again08:11
lifeless\o/08:11
DamasceneI hope someone looks in this bug because it's important to many users in many developed countries08:13
lifelessDamascene: do you mean developing ?08:13
DamasceneI don't know. the third world countries or something like that :)08:14
Damascenewindows spread there because you get what ever you want on cds and from friends08:15
Damascenein ubuntu you should have internet connection on your machine08:15
pittijdong, slangasek, cjwatson: does either of you have some time to review by lucid-proposed uploads? (gvfs, gnome-keyring, gdm, totem); gdm might be controversial, the rest should be fine, but even though we have this gnome 2.30.x exception I'd rather not self-approve them08:47
Damascenebug 57277608:48
ubottuLaunchpad bug 572776 in ubuntu "Ubuntu should provide update packages for download and use for offline users" [Undecided,New] https://launchpad.net/bugs/57277608:48
DamasceneI see new people have come in08:48
macosame people as above, Bug #550145  permission requested08:48
ubottuLaunchpad bug 550145 in anki "Anki 0.9.9.7.8 in Lucid suffers irreparable loss of functions. sync with debian to 0.9.9.8.6-2 needed." [Undecided,Confirmed] https://launchpad.net/bugs/55014508:48
macoDamascene: bring that up on ubuntu-devel-discuss@lists.ubuntu.com08:49
macoDamascene: that seems like the sort of thing that would take some planning at UDS to get done08:50
Damascenemaco, ok08:50
Damasceneyou mean the meeting coming up?08:50
macoyes08:50
macowhere the release is planned08:50
nxvlDamascene: that's kind of a trivial thing09:00
Damascenetrivial? why?09:01
nxvlwell, when you run apt-get upgrade all the packages get downloaded and stored in /var/cache/apt/archives/09:02
nxvlso you only need to copy all of them into a CD/USB/HD and take it to an offline machine, copy again in /var/cache/apt/archives and be happy09:03
nxvlbasically09:03
Damascenenot really09:03
Damascenewhat if you want to install new software. or what if you have cleaned the folder?09:04
nxvlif you install new software the packages get stored in the same folder09:05
ghostcubeis there a reason that jackd is still not in main repo in lucid?09:05
ghostcubeshouldnt this be changed?09:05
nxvlnow, if you runned apt-get clean, that's a problem09:05
\shDamascene, so how do the windows people in developing countries get their "updates" on CD? someone needs to "download" the crap right?09:05
nxvlalso you will probably need to copy the content of /var/lib/apt/lists/ to do the "apt-get update" trick09:06
nxvl\sh: yes, but i think the bug report is asking for a way to that09:06
\shnxvl, right...it's a hen <-> egg problem09:07
Damascenewhat I'm suggesting is more simple09:07
nxvlDamascene: now, how would you know which updates you need09:07
ChipzzDamascene: "don't do that then"09:07
Chipzz"Docter, my files are gone when I remove them"09:07
Chipzzduh09:07
nxvlwhat i mean is: let's assume i've a hardy system offline, last may i updated it using this magic script that gets me a CD with the updates so i can use it09:07
nxvlhow would you make me not to download the updates i already installed this may09:08
sorenapt-offline - offline apt package manager09:08
DamasceneChipzz, do you mean the person who need to get updates should tell his friend not to clean their system?09:08
sorenDamascene, nxvl: ^^09:08
nxvlsoren: yup, i knew about that one, still is not UI09:08
nxvlGUI*09:08
Damascenenxvl, try that feature in synaptic and you will know09:09
nxvlDamascene: huh?09:09
sorennxvl: Ah.09:09
Damascenesynaptic will sort things out09:09
kklimondaghostcube: something in main have to depend on it in order to get it pulled into main. fox example libjack0 is in main because it's a dependency of libasound2-plugins (and maybe other things)09:09
nxvlno it won't09:09
ChipzzDamascene: what else do you propose? redownloading them again?09:10
ghostcubekklimonda: so the dependencies for libxine to build with jackd support are in main?09:10
nxvlin that case you will need to generate some kind of report in the offline machine, then go with that to an online machine and then download the packages09:10
ghostcubethen its ok09:10
Damascenenxvl, if you have updated software list. synaptic will just mark the new packages09:10
nxvli'm taking about wasting my bw downloading things i don't need09:10
nxvltalking*09:10
kklimondaghostcube: I don't know - apt-cache rdepends shows that libxine1-misc-plugins depends on libjack0 - is it enough?09:10
ghostcubeyep thx very much should do the trick :)09:11
nxvlsoren: i heard i will see you next week? is that correct?09:11
Damascenehttps://bugs.launchpad.net/ubuntu/+bug/572776/comments/309:12
ubottuLaunchpad bug 572776 in ubuntu "Ubuntu should provide update packages for download and use for offline users" [Undecided,New]09:12
Damascenenxvl, I've updated the report. please look at it09:12
Damascenethat might help you understand what I'm saying better09:12
nxvlDamascene: let's say i've offline machine (A) and online machine (B), in machine A i already have udev up-to-date, then i go to machine B and download the new updates, how i tell machine B not to download udev?09:12
nxvlok, got it09:13
nxvlhow will you do step 2 without an internet conection?09:14
Damascenenxvl, this step need internet connection09:15
Damascenefrom any machine09:15
nxvloh, ok, so i think i got it, you will do step 2 in machine B, so will need to go from machine B to machine A then back to B then back to A09:15
nxvlright?09:15
Damascenethat is better than having outdated system with not being able to install the software you want on it09:16
nxvlwell, you still have apt-on-cd and lots of options09:16
Damascenetell me how to use apt-on-cd on windows?09:17
\shdebmirror <release>-updates + <release-security> => burn it on a dvd => share the dvd => add the dvd to your sources list => repeat the steps every now and than, at least you need one internet connection from someone09:17
nxvlmvo: ^^ did you see that possible, to have a downloadeable Packages.gz or something like that?09:17
Damasceneand again you shouldn't clean your system09:17
nxvl\sh: i think i got the idea and it's kind of good actually09:18
macoa webpage where you can upload your current dpkg --get-selections and itll spit back a list of packages you need in order to upgrade with download links... though thatd still involve right click -> save as on every single link...09:18
nxvlmaco: not if you do a tiny python script09:19
macoyeah ive heard this a /lot/ from people who go to the library, use windows computers to grab a deb from packages.ubuntu.com and put it on a flashdrive, get home, and discover they fail at manual dependency resolution09:19
mvonxvl: its possible to generate a download script like this, yes, should not even be very hard09:20
Damascenemaco, right09:20
maconxvl: or if you get the server to make a zip of the debs and pop it out as an autodownload thingy... though there'd be a loading bar involved in that09:20
nxvlso, what i'm thinking of is synaptic to generate a python script or something you can actually run on windows, mac or linux independently in step 409:20
nxvlso you don't need to do the right click save as09:20
macoa portable app type thing maybe for the windows case09:21
macobecause window doesnt have python by default09:21
mvonxvl: ideally it would provide multiple languages, something like C#/mono, python, java09:21
mvonxvl: depending on what the user has availalbe on the windows system (or other system)09:21
nxvlmaco: the zip in the server would be kind of not what we need with this idea actually, every case will be a different set of packages for the zip09:21
mvonxvl: but yeah, the idea of a script is certainly appealing09:21
Damascenehi, mvo09:22
maconxvl: which is why i said itd involve a loading bar...because itd have to sit there and zip stuff up for *every* person...which is ugh09:22
mvohey Damascene09:22
* maco raises a cup o' tea to mvo09:22
mvohey maco!09:22
nxvlmvo: yeah, i said python because it was the first i thought it can be used in several systems, maybe a compiled python script for windows (i readed once that was actually possible)09:23
\shmvo: if you just concentrate on windows, it would be c#/vb .net ... or you provide several executables made from C (for windows, linux etc. because we don't know which OS the user is using;)09:23
* mvo toasts back with a cup of sencha09:23
nxvlDamascene: are you going to be at uds?09:23
kklimondaor maybe write it in C with gtk+ ? sounds like the most portable solution. or replace C with Vala09:23
mvonxvl: oh, cool idea, I did not think of the compiled ones09:23
macoso um i think im going to be learning c# soon09:23
nxvlthis is actually an idea i kinda like09:23
macoand that ill be having a windows 7 vm living on my laptop soon09:24
\shmaco, c# is not that complicated ;)09:24
Damascenenxvl, no, I'am afraid I cant09:24
macoso that might actually fall into the category of things i could do09:24
maco\sh: yeah i know java so it should be easy09:24
macokklimonda: windows doesnt have gtk+ by default. its another one of those things you have to tell people to install09:24
nxvl\sh: it not hard, just HORRIBLE09:24
nxvl:D09:25
macokklimonda: which makes it not-ideal for the case where youre using a public machine on which you lack admin09:25
kklimondamaco: there is no way to distribute gtk+ library with the executable itself?09:25
nxvlmvo: so, that can be a software center plugin?09:25
Damascenemyself I used downthemall to get the packages from windows machine09:25
macokklimonda: i guess you could staticly link it....itd be HUGE09:25
Damascenefor windows you can just generate text file with links09:25
* \sh -> meeting09:26
nxvlDamascene: and then click on every link?09:26
Damasceneno09:26
nxvlDamascene: did you have any idea how many links that could be?09:26
Damascenein downthemall you can make it download every thing on the page09:26
DamasceneI did it and yes I know09:26
DamasceneI used that way when I had slow internet connection09:27
nxvlDamascene: ok, let's take maco's example of people in the library, usually the users are not able to install anything on the machines09:27
Damasceneok. they have firefox09:27
nxvlmaybe not09:27
Damascenejust install downthemall09:27
macothey *maybe* have firefox09:27
macothey have internet explorer09:28
nxvli will say we will need to think of an o-o-t-b windows system without the posibility to install anything else09:28
nxvlsame for mac09:28
macoagreed09:28
Damascenethey can use freedownloadmanager which it has portable version and can import links from txt files09:29
macoim willing to take on figuring out the windows stuff as i will soon be required by my employer to learn C#/.NET and have a windows vm09:29
nxvlmaco: mono09:29
nxvlmaco: :D09:29
Damascenethe problem is not with getting the package in windows09:29
Damasceneit's with getting the updated software list09:29
nxvlnot really, that's kinda trivial09:30
Damascenenxvl, which thing? using freedownloadmanger?09:30
nxvli just thing is completely wrong to tell a user: go download this closed source software to be able to update your open source system09:30
macoDamascene: run "dpkg --get-selections > current_packages" and then just put the current_packages file on the flash drive with the program09:30
nxvlDamascene: nop, the getting the list09:30
macohave an import option in the program to look at the current list of installed packages and from there it can figure out what packages will need to be upgraded to get from karmic to lucid09:31
Damascenenxvl, freendowloadmanager is opensource09:31
nxvloh ok then :D09:31
macoactually, better: dpkg -ll09:31
macobah -l09:31
nxvldoes it need to be installed?09:31
nxvlmaco: again, it's not GUI :D09:31
nxvlmaco: apt-offline will be better09:32
maconxvl: hmm i dont know about this09:32
macoi was just thinking if you need a way to fnd out whats installed so that you can tell the other-os-downloady-thingy that info, dpkg -l would do it09:32
nxvlmaco: we need to think on our user base, we don't expect them to be techical users09:32
nxvlmaco: and CLI can really scare non tech users09:32
Damascenefor now I would be thankful if you just give me the software list in zipped file. and I'll take care of the rest09:32
maconxvl: true09:33
maconxvl: does apt-offline have a gui?09:33
kklimondawhat are actual use cases we are discussing? installation of the new software, updating software in a distribution and updating distribution to a newer release?09:33
macokklimonda: pretty much09:34
nxvlmaco: well, actually the idea is to go to the ubuntu web page, get what apt-get update gets you in you system, but for offline usage, then put that on you system, then generate a list of what needs to be updated (a download-packages script) then take that to library again and be happy09:34
nxvlmaco: no it doesn't, but what i mean is that  CLI tools are a lot out there09:34
kklimondawhat happens if users use alternative cd to perform a dist upgrade and they don't have internet connection?09:34
nxvlkklimonda: actually all of them :D09:35
macokklimonda: shortcut ;-)09:35
macokklimonda: but remember that wont upgrade the bits that were installed from universe, if they exist09:35
nxvlkklimonda: actually i will take dist-upgrade out of the picture, because you can basically use an ubuntu CD to do that09:35
macoi had some brokenness when i upgraded to edy that way :(09:35
maco*edgy09:36
kklimondamaco, I know - that's why I am asking :)09:36
nxvlin that case you will need to use both, the idea we are discussing for universe, and the ubuntu cd for the core system09:36
mvocdrom upgrades should work just fine, it may become a bit of a problem if you had a cdrom install, then did a lot of package installs with network and then later upgrade without net, that may lead to it being not able to perform the upgrade09:37
tkamppeterpitti, thank you for your quick passing through of the SRU on s-c-p, can you do the same with the Ghostscript SRU, it is segfaulting of GS when using CUPS Raster drivers (bug 539708).09:38
macomvo: yep thats what i did from dapper to edgy :P09:38
ubottuLaunchpad bug 539708 in ghostscript "/usr/lib/cups/filter/pdftoraster failed " [High,Fix committed] https://launchpad.net/bugs/53970809:38
macomvo: i had a dadgy system for a bit there :P09:38
mvohihi09:38
StevenKAt one UDS I had a Gardy system for a while -- it was stuck between Gutsy and Hardy09:39
nxvlmvo: that's why this need to be a way to use offline "apt-get" what i was thinking is dowloading the Packages.gz and all that stuff in a format the plugin/app/* can read it and pass it to apt (/var/lib/apt/lists/) then generate the needs-update list or to-install list09:39
mvoDamascene: what was the bugnumber of the report?09:39
mvoDamascene: for this feature I mean09:39
StevenKnxvl: A tool like that already exists, I just can't remember what it's called.09:39
Damascenehttps://bugs.launchpad.net/bugs/25137809:40
ubottuLaunchpad bug 251378 in synaptic "Synaptic's Generate download script does not update package lists" [Wishlist,Triaged]09:40
nxvlStevenK: is it integrated with software center?09:40
Damascenehttps://bugs.launchpad.net/ubuntu/+bug/57277609:40
ubottuLaunchpad bug 572776 in ubuntu "Ubuntu should provide update packages for download and use for offline users" [Undecided,New]09:40
StevenKnxvl: No, it's like 6 years old09:40
maconxvl: whereas i was thinking that you should only need to make one trip to the internets. put a list of whats currently installed into a file, go to the internets, import that list, and have a program download the Packages.gz, figure out what is needed, and do the downloading09:40
nxvlStevenK: :D09:40
Damascenemvo, got the links?09:41
nxvlmaco: that *can* work aswell, but it will need more server-side computing09:41
mvoyes09:41
maconxvl: or a smarter gui09:41
nxvlmaco: which i don't see possible, or at least realistic09:41
nxvlmaco: or a windows app that does that for you09:41
nxvlwindows/mac/linux09:41
StevenKhttps://alioth.debian.org/projects/apt-offline/09:41
nxvlStevenK: what soren just mentioned :D09:42
Damascenemaco, your idea is really nice. but what I suggested will work with the tools we have in hand09:42
nxvlStevenK: can you please tell debian sysadmins to update the ssl certs09:42
StevenKnxvl: "Meh"09:42
nxvlStevenK: :D09:42
nxvli hate insomnia09:43
mr_pouitpitti: are calls to libhal{,storage} supposed to start hal? It doesn't seem to happen (Bug #546992 - thunar doesn't call hal directly, it only uses it through libhal for removable devices).09:46
ubottuLaunchpad bug 546992 in policykit-1 "Xubuntu 10.04 usb automount fails if autologin is on" [Medium,Confirmed] https://launchpad.net/bugs/54699209:46
nxvlwell, will try to get some sleep, read yoy tomorrow09:46
DamasceneStevenK, that need two ubutnu machines. right?09:49
StevenKDamascene: I think set up requires a Ubuntu machine that is off the net, and some Linux machine that is on the Internet09:51
DamasceneStevenK, is there a apt-offline for rpm based?09:51
StevenKDamascene: I have no idea, I've not used a RPM-based distro for 9 years09:52
Damascene:)09:52
joaopintogood morning09:55
pittimr_pouit: yes, they should start hal09:55
pittimr_pouit: that just won't work if the program explicitly checks if hal is running before trying to connect to it (e. g. lshal does that)09:56
joaopintoslangasek, shouldn't the "-d" on update-manager be dropped on the release notes ?09:57
mvojoaopinto: for 8.04 -> 10.04 we still need either -d or --proposed as its not yet enabled by default09:57
joaopintomvo, but is it safe once maverick is setup ?09:58
mr_pouitpitti: I couldn't find any check like that for thunar (quick grep), so apparently it doesn't start it (or start it too late, since it only happens on autologin?)09:58
joaopintomvo, it doesn't comply with the "-d" purpose ;)09:58
pittimr_pouit: might be a race condition, yes09:58
pittibut it still sounds weird09:59
pittimr_pouit: does it work for you?09:59
mvojoaopinto: yeah, this is why I think "--proposed" is better, less confusing09:59
joaopintoKeybuk, where do the "Checking disks" messages come from ? mountall or plymouth ?10:00
mr_pouitpitti: I don't have access to my system, and I don't use autologin (I didn't feel very confident to use autologin with an encrypted /home...)10:01
pittimr_pouit: it wouldn't work, right10:02
pittimr_pouit: I mean, does it work for you without autologin?10:02
pittimr_pouit: if it does, then it's not a principal problem with dbus activation and libhal, but rather a race condition somewhere10:03
mr_pouitpitti: ah yes, it worked10:03
joaopintothe "Checking disks for errors" should provide a reason, for the inexperienced user a random "Check disk for error" without a motive could lead to wrong interpretation10:04
mr_pouitpitti: and some comments on the bug report state that it works correctly if they log out/log in after the first autologin10:08
pittimr_pouit: seems that it doesn't deal too well with that unexpected delay of hal's answer perhaps?10:08
pitticjwatson: I'm merging debhelper, FTR10:21
mr_pouitpitti: well, I just grepped again in thunar & thunar-volman source code, and they use libhal_* functions... So I guess that somewhere libhal returns the wrong status that let think thunar that hal isn't available...10:23
pitticjwatson: (uploaded)10:31
hyperairwhat's "TREE" and "MERGE-SOURCE" in bzr rebase context?10:34
hyperairugh, bzr is such a pain to use.10:37
pittijust as with merges, remote vs. local version?10:38
hyperairpitti: which is which?10:38
pittitree -> local10:38
hyperairis TREE local, or MERGE-SOURCE local?10:38
pitti(at least with merges)10:39
hyperairwell i thought that would be so10:39
hyperairbut then i removed all the "-l" from all su invocations in my local tree, but they're in the "TREE" section10:39
hyperairugh10:39
hyperairso i thought it might be the other way round, but it doesn't seem right either10:40
hyperairhonestly, why can't bzr be more straightforward, like git?10:40
pittirotfl10:40
* hyperair kicks bzr10:41
pittihyperair: well, it should be in either tree or merge-source? there's no third party involved there10:41
hyperair@_@10:41
hyperairwell i think i got it. i think.10:47
* hyperair sighs.10:47
hyperairpitti: is there a way to force-push a branch?10:48
pittibzr push --overwrite10:48
hyperairit complains about diverged-branches since i've rebased10:48
pittihyperair: oh, you rebased a publicly pushed branch?10:49
pitti(that's why rebasing is bad :) )10:49
hyperairpitti: i was asked to rebase by crankyadmin10:49
hyperairwhoops10:49
hyperairbad tab complete10:49
hyperaircrimsun:10:49
hyperairpitti: it's a one-fix thing anyway10:49
pitti--overwrite is the equivalent of git push -f10:49
hyperairah okay, thanks10:49
crankyadmin?!?.... /me knows nothing....10:49
hyperaircrankyadmin: yeah sorry i typed cr and tabcompleted, and your name came out10:49
crankyadminnp10:50
arandIs tasksel supposed to be usable on an in-use system?11:01
joaopintoarand, I think so11:09
joaopintoarand, at least the wiki entry for LAMP advises to use tasksel11:10
arandjoaopinto: Hmm, well the old wiki doc states that it shouldn't really be used outside a liveCD, I was just thinking how valid my Bug #574287 actually was...11:11
ubottuLaunchpad bug 574287 in tasksel "tasksel uninstalled my system!" [Undecided,Confirmed] https://launchpad.net/bugs/57428711:11
sitsI can't tell if this is the right channel11:12
sitsbut using a custom compiled kernel with 10.04 seems to stop plymouth from starting (it seems to segfault)11:13
ionIf the standard kernel works, your kernel is probably configured incorrectly.11:14
sitsion: indeed - but the next question is how is it misconfigured :)11:14
sitsit has KMS in it already so that's not a problem11:14
tkamppeterpitti, hi11:14
sitsthings are muddied further because it seems really hard to get plymouth going on a "normal" kernel if I boot to single user mode first11:15
sits(e.g. boot to single user mode then boot to runlevel 5 then try and start plymouth)11:15
sitsit does the VT switching but no splash screen is displayed11:16
pittihi tkamppeter11:17
arandjoaopinto: Hmm, well the old wiki doc states that it shouldn't really be used outside a liveCD, I was just thinking how valid my Bug #574287 actually was...11:18
ubottuLaunchpad bug 574287 in tasksel "tasksel uninstalled my system!" [Undecided,Confirmed] https://launchpad.net/bugs/57428711:18
arandjoaopinto: Sorry, ignore.11:18
joaopintoarand, the bug seems valid to me, there is nothing on the manapage which refers *install only*11:19
joaopintoanyway if that was the case why should it allow to remove packages during the install :) ?11:20
narendraHI,,11:20
tkamppeterpitti, I see you have worked on the SRU of bug 564633. You have put it to "Fix Released" and then back to "Fix Committed". What did you actually do?11:20
ubottuLaunchpad bug 564633 in system-config-printer "system-config-printer: make driver installation optional" [High,Fix committed] https://launchpad.net/bugs/56463311:21
narendraI am working on custom live DVD, may you want to share your experince with me11:21
arandjoaopinto: Aha!, Think I figured it out it removes the whole of ubuntu-desktop, provided you have manaually removed the ubuntu-desktop meta pkg earlier...11:21
narendraI am creating a 2.5GB live iso of ubuntu , but i am worry about distributing and installation methods11:22
kklimondatkamppeter: an update from -proposed has been copied over to maverick and it's fixed there. it's still not fixed in lucid so a new task has been opened.11:22
narendrawe need to distribute this iso to hundreds of schools in India11:22
tkamppeterkklimonda, pitti: So for Lucid it is still in the "verification needed" mode?11:23
kklimondatkamppeter: yes11:23
tkamppeterpitti, this SRU is a fix of the problem for most users, but there is a second problem with the Kyocera FS-1800+ of the poster of the bug for which I have created a three-line patch now. How should this be handled? Replace the current SRU by a new one with this patch added or wait for this SRU to be closed and then post a second SRU?11:26
narendraDear developers, please show some lights,,,,,,,,,,11:28
joaopintonarendra, I am not sure which kind of guidance are you looking for :) do you have a specific doubt ?11:30
narendrajoaopinto, yes, basically, I am making a SchoolOS livd DVD, 2.5 GB ubuntu karmic based11:31
narendrajoaopinto, which is going to distribute in various schools of india at mass level11:31
narendrajoaopinto, but the problem, is i am unable to figure out the distribution method.11:32
joaopintokarmic is a bad option, specially for a large distribution11:32
narendrajoaopinto, then what should i choose ?11:32
joaopintonarendra, lucid, because it's a long term support release11:33
narendrajoaopinto, http://schoolos.org this is indias biggest foss implentation project11:33
narendrajoaopinto, I afraid of bugs, and it is very recently released11:33
joaopintonarendra, your question hints that you are not prepared yet for such a large task :)11:33
narendrajoaopinto, I have prepared beta version using karmic,,11:34
joaopintonarendra, right, but there are greater chances of having the bug fixes, compared to karmic11:34
narendrajoaopinto, iso size is  2.1GB big.,11:34
narendrajoaopinto, but schools hardware maynot be able to boot 2.1GB live iso image11:35
narendrajoaopinto, If i send , LiveDVD, then ram size is issue....11:36
pittitkamppeter: I copied lucid-proposed to maverick; this causes some confusing LP bug spam, sorry about that11:36
joaopintonarendra, you are starting the way around, first you have the image and the OS ready, and now you are checking the requirements :) ?11:36
narendrajoaopinto, See idea from day one is to give big DVD with all software because many schools do not nave Internet or have limited bandwidth to download updates,,11:38
pittitkamppeter: you are welcome to upload a followup SRU; if you do that, please use debuild -S -v 1.2.0+20100408-0ubuntu5 (this will make both changelogs appear in the sources.changes)11:38
narendrajoaopinto, I think live USB is a good option, live usb won't be that show as live DVD11:39
tkamppeterpitti, thanks, I will do so.11:39
tkamppeterpitti, can you approve the SRU of Ghostscript for bug 539708? Thanks.11:40
ubottuLaunchpad bug 539708 in ghostscript "/usr/lib/cups/filter/pdftoraster failed " [High,Fix committed] https://launchpad.net/bugs/53970811:40
pittitkamppeter: I'll have a look at it soon; but it's a very complicated patch, thus will take a bit11:40
tkamppeterpitti, Daniel J Blueman has reported in the upstream bug report at Ghostscript that it fixed segfaults on 2032 files for him.11:42
pittitkamppeter: ugh, on how many did he test it? :-)11:52
=== oubiwann is now known as oubiwann_
tkamppeterpitti, this I do not know, at least he had many segfaults and these segfaults have gone away for him.11:53
pittinice11:53
tkamppeterpitti, for me I also did not succeed to make Ghostscript segfaulting any more when converting PDF to CUPS raster.11:54
tkamppeterpitti, new s-c-p SRU uploaded.11:57
sianisasac: could you review this merge? https://code.launchpad.net/~sianis/desktop-webmail/hungarian-translations/+merge/2456211:59
ari-tczewseb128: could you open a task on karmic in bug 574262 ? I want to prepare a patch to fix it by sru.11:59
ubottuLaunchpad bug 574262 in gdm "Please backport to Karmic GDM fix for bug #463376" [Low,Fix released] https://launchpad.net/bugs/57426211:59
seb128ari-tczew, sorry I've to run for lunch now, will have a look later12:00
ari-tczewin total anyone could open a task in above bug, please12:00
seb128ari-tczew, why do you care about it? users who need the feature will have moved away from karmic since and new installs will be lucid12:01
seb128there is no reason to install karmic over lucid now12:01
seb128lunch, bbl12:02
patrick42hsun's coming up, time for bed12:02
ari-tczewseb128: bug-requester says that he administrate 18 computers based on karmic, so I think that would be nice help him, or not?12:02
asacsianis: yep. thanks. will look at that when i get to that merge request email ;)12:05
highvoltageis maverick open for uploads yet?12:07
ajmitchdon't think it's quite open, though there've been toolchain uploads12:11
Laneyhttps://lists.ubuntu.com/archives/maverick-changes/2010-May/thread.html12:13
james_wit's frozen, so you can upload and everything will just land in UNAPPROVED anyway12:14
james_wit will probably be a day or two before it is fully open12:14
pittigcc-4.5 FTBFSed12:14
Laneylp/ubuntu/maverick just times out :(12:14
highvoltagejames_w: ok great12:15
=== dendro-afk is now known as dendrobates
=== dendrobates is now known as dendro-afk
ari-tczewwhere maverick is looking in Debian? testing or unstable?12:45
=== dendro-afk is now known as dendrobates
neardoes anyone knows why every time i reboot my desktop settings go back to default?12:47
nearI have to set up the dual monitors configuration and the wallpaper all over again12:47
=== oubiwann_ is now known as oubiwann
looljames_w: Hmm, ISTR I could branch from lp:debian/unstable/foo as well as lp:debian/sid/foo, was it my fantasy?  I can't with lp:debian/unstable/logcheck at least14:13
looljames_w: Also, logcheck 1.3.8 was uploaded 2010-04-15, but that didn't make it to the sid branch14:14
loolHmm it's listed in the 16738 outstanding jobs at http://package-import.ubuntu.com/status/14:14
LutinKeybuk: would you have any clue as to why upstart would hang upon receiving a 'amba-device-added' event on startup ?14:24
james_wlool: no, you can't do that, there's a bug open on LP requesting that feature14:33
james_wlool: and the importer is currently paused while we open maverick14:33
james_wshould be unpaused today or tomorrow14:33
looljames_w: Ok14:33
looljames_w: It's been uploaded 15+ days ago though14:34
loolBut I guess there was some backlog for whatever reason14:34
looljames_w: thanks for checking14:34
james_wlool: hmm, let me look then14:34
james_wlool: that was because launchpad hasn't seen it yet for some reason: https://edge.launchpad.net/debian/+source/logcheck/+publishinghistory14:35
pittiKeybuk: hm, it seems ureadahead reads the queue/rotational attribute for deciding between SSD and HDD14:38
pittiKeybuk: did you ever see that lie? I'm ssh'ed to a bug reporter's machine with an SSD, and all /sys/dev/block/*/queue/rotational are 114:38
pittithat might cause ureadahead to not perform optimally?14:38
looljames_w: Ok thanks14:39
pittiKeybuk: (it's correct on the mini, though)14:42
=== thekorn_ is now known as thekorn
pittitkamppeter: ghostscript and new s-c-p accepted15:28
LutinKeybuk: and more importantly, would you know what could generate this amba-device-added event ?15:32
zulpitt: is it possible to accept the landscape-client in *-proposed...thanks15:34
pittizul: can you please reupload the karmic one with correct -v? it's two new versions in one15:35
zulpitti: sure15:35
pittizul: jaunty, too; please upload with e. g. -v1.4.4-0ubuntu0.9.04 for jaunty15:36
pitti(or merge the changelogs)15:37
zulpitti: will do15:37
pittizul: thanks15:37
dholbachhttps://wiki.ubuntu.com/UbuntuOpenWeek starting in 23m in #ubuntu-classroom15:38
tkamppeterpitti, thanks.15:38
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
ari-tczewseb128: ping16:37
seb128ari-tczew, hey16:37
ari-tczewseb128, so what about bug 574262 ?16:37
ubottuLaunchpad bug 574262 in gdm "Please backport to Karmic GDM fix for bug #463376" [Low,Fix released] https://launchpad.net/bugs/57426216:37
seb128yes?16:37
seb128sorry I'm at a sprint right now16:38
seb128i.e busy with other things16:38
ari-tczewaha ...16:39
seb128if you want to work on it get a debdiff on the bug and subscribe sponsors16:39
seb128I've no interest in the issue to be honest, as said people are dealing that for 6 months now and should be moving to lucid now16:40
seb128i.e I don't want to spend time on karmic...16:40
zulpitti: it looks like i cant nominate bugs for lucid sru's is there anything i need to do?16:49
=== mathiaz_ is now known as mathiaz
smosercjwatson, what is the bug number of the plymouth bug you were chasing on thurdsay (that resulted in your posts at http://lists.freedesktop.org/archives/plymouth/2010-April/thread.html#350)16:51
El_Presidenteim sorry if i bug in here but im not sure if that bug report i filed is enough to tell the devs what error i have https://bugs.launchpad.net/ubuntu/+bug/57214616:52
ubottuLaunchpad bug 572146 in alsa-driver "crackling sound from microphone with 2.6.32-21 kernel" [Undecided,New]16:52
crankyadminEl_Presidente, "HOW TO GET FIXEX NO SOUND IS NO FUN I NEED CRISPY SOUND SYSTEM 5.1HOW TO GET FIXEX NO SOUND IS NO FUN I NEED CRISPY SOUND SYSTEM 5.1" Wtf?!?!16:57
El_Presidentethatzs not mine :(16:58
macowow caps16:58
crankyadminmy bad! :D16:58
El_Presidenteim sorry for that nerd but thats not me16:58
El_Presidentecrankyadmin, any suggestions what information i can add?17:00
crankyadminEl_Presidente, nope. Although I would have my finger on Alsa screwing with interrupts for what ever reason.17:03
El_Presidentecrankyadmin, i did a fresh installation of 10.04 on a second harddrive to avoid upgrade bugs17:04
El_Presidentebut the error is still present17:04
El_Presidentebut thank you ;)17:04
crankyadminIt is more likely a change that has been made within the drivers. Someone will pick it up...17:05
El_Presidenteok17:05
El_Presidenteis there a way to remove the spam comment?17:06
sitshi I've been trying to debug a mountall issue17:46
sitshttps://bugs.launchpad.net/ubuntu/+source/mountall/+bug/57170717:46
ubottuLaunchpad bug 571707 in mountall "fsck progress stalls at boot, plymouthd/mountall eats CPU" [High,Confirmed]17:46
=== radoe_ is now known as radoe
arandsits: Ask the question-proper ;) (I'm also poking in that one as you might've noticed, you got anything interesting?)17:56
sitsarand: :)17:57
sitsit seems to happen with multiple themes17:58
sitsa backlog seems to be build up17:58
sitsand when I try and trace it the last thing I get is inside the plymouth library17:58
=== al-maisan is now known as almaisan-away
sitsarand: I genuinely think it could be down to plymouth being sent thousands of messages18:02
sitse.g. 80% 80% 80%... 81% 81% 81% etc.18:03
sitsarand: I'm wondering if the progress bar is being read too frequently18:07
arandsits: Hmm, well I ran it with plymouth:debug and it does spew a bunch of messages, but not overly frequently, so I*m not sure if that's it...18:07
=== dpm is now known as dpm-afk
sitsarand: duplicates apparently get weeded out18:10
sitsbut the question becomes at what stage?18:10
arandsits: Hmm, I wrote the log to file and since I pretty much left it running for an hour or two I only have 91% as the lowest thing in the log, of which there are 1658 occurences, but I've no idea if that says anything..18:12
sitswell it suggests some filtering could take place I guess18:13
sitsthe progress needs to be read and sent less frequently18:14
sitsbut that suggests you were only getting updated every second18:15
arandsits: Ah, no that's only for the messages stating  "91%" However it must've remained at that point for more than 2-10 minutes though.. I think18:16
sitshmm scratch that18:16
sitsah ok18:16
arand*5-1018:16
sitsmy question is does fsck really generate that many updates18:17
sitsor are some being synthesized?18:17
arandThe speed at which the precentage in the plymouth splash increases is definitely quicker that the interval at which the debug messages are being printed if switching out to the debug via arrowkey.18:19
sitsarand: well in my case the fsck is finished in seconds18:22
sitsbut then the updates keep being sent18:22
sitsarand: we could do with a breakpoint in fsck_reader18:23
sitsto see how often plymouth_progress is really called18:23
* arand felt something whosh over his head just there18:24
Tm_Tarand: yeah, we have some bats around here, just ignore them18:25
=== uaa is now known as Damascene
arandsits: Yea... I'm severely lacking in that gdm-fu you speak of...18:26
sitsarand: ah 'tis a pity18:27
sitsyou see to have a better setup for testing this than me18:27
arandsits: but if you want me to copy-paste commands I'm all ears :)18:27
sitsdo you have more than one system?18:27
arandsits: I'm using a virtualbox18:28
sitsah good good18:28
sitsarand: ok first you need to go to runlevel 118:29
sitser18:29
sitslet me msg you18:29
=== yofel_ is now known as yofel
=== rgreening_ is now known as rgreening
shtylmandid something happen to libactiverecord-ruby ?19:02
Picishtylman: see debian #57415419:06
ubottuDebian bug 574154 in ftp.debian.org "RM: libactiverecord-ruby, libactivesupport-ruby" [Normal,Open] http://bugs.debian.org/57415419:06
shtylmanPici: thanks19:07
robertzaccouri recently discovered a bug, reported it this morning. i did apport-collect but don't see anything added to the report19:13
robertzaccourhttps://bugs.launchpad.net/ubuntu/+bug/574406 thats the bug19:14
ubottuLaunchpad bug 574406 in ubuntu "Audio capture doesn't record anything." [Undecided,New]19:14
narendrahi19:27
robertzaccouri recently discovered a bug, reported it this morning. i did apport-collect but don't see anything added to the report19:27
robertzaccourhttps://bugs.launchpad.net/ubuntu/+bug/574406 thats the bug19:27
ubottuLaunchpad bug 574406 in ubuntu "Audio capture doesn't record anything." [Undecided,New]19:27
narendramay anybody tell me how to build ubuntu ? I mean what script ubuntu developer use to create ubuntu distribution19:27
robertzaccourthats the bug19:28
narendramay anybody tell me how to build Ubuntu ? I mean what script Ubuntu developer use to create Ubuntu distribution... I want to learn about Ubuntu build process19:29
jdongpitti: ack sorry, end of term life has been hectic. which bugs would you like me to look at?19:31
robertzaccourme?19:31
robertzaccourhttps://bugs.launchpad.net/ubuntu/+bug/574406 thats the bug19:31
ubottuLaunchpad bug 574406 in ubuntu "Audio capture doesn't record anything." [Undecided,New]19:31
cjsTwo release note issues:19:34
cjs1. The 10.04 release notes still go to http://www.ubuntu.com/testing/lucid/beta1, which claims it's still in beta.19:34
cjs2. This should really go in the release notes: https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/57466519:35
ubottuLaunchpad bug 574665 in qemu-kvm "kvm + virtio disk corrupts large volumes (>1TB)." [Undecided,New]19:35
narendraanybody give some hint19:36
narendra how to build Ubuntu ? I mean what script Ubuntu developer use to create Ubuntu distribution... I want to learn about Ubuntu build process19:36
cjsnarendra: It's not on the wiki?19:36
respiresorry to ask a support(ish) question but its to support my dev. Synaptic somehow knows that I chose to use root not sudo to execute priviliged instructions. How does it know this, more to the point how can I make my program know this19:37
Picicjs: The first should be filed as a bug in the ubuntu-website project on launchpad19:37
* respire can think of several ways to do it but wasnts to be consistent with the "ubuntu way"19:38
narendracjs : I think not19:38
robertzaccourhttps://bugs.launchpad.net/ubuntu/+bug/57440619:38
ubottuLaunchpad bug 574406 in ubuntu "Audio capture doesn't record anything." [Undecided,New]19:38
robertzaccourthats the bug i reported this morning19:38
narendracjs: wiki contains how to make custom live CD, It tell about remastering or spining distro but not about the actual build process19:39
robertzaccourok added stuff to the bug report20:00
robertzaccourhttps://bugs.launchpad.net/ubuntu/+bug/57440620:00
ubottuLaunchpad bug 574406 in linux "Audio capture doesn't record anything." [Undecided,New]20:00
robertzaccourany hope?20:00
robertzaccouri'm gonna get a plug in mic and see how my luck is20:01
=== BlackZ_ is now known as BlackZ
baptistemm_there is no way to have voice on #ubuntu-kernel20:23
JFobaptistemm_, there is, but your nick must be registered with FreeNode20:26
baptistemm_hmm, okay thanks20:26
JFonp\20:26
xaver__hi Keybuk20:45
qensehttp://popcon.ubuntu.com/ seems to have troubles with loading images on the overview page.20:48
=== dpm-afk is now known as dpm
ccheneyjcastro: did you see the 3s lucid boot?21:21
jcastrowhat?21:21
ccheneyhttp://ubuntuforums.org/showpost.php?p=9124494&postcount=40421:21
ccheneywell 3.14s to be exact21:21
ccheneylater in the thread he mentioned a couple optimizations he did to speed it up21:22
ccheneynoop scheduler and preempt kernel apparently21:22
bluefoxicywhy does firefox not render anything until it has images ~_~23:07
=== dendrobates is now known as dendro-afk
nxvlkees: can you please add me to ubuntu-sponsors23:30
keesnxvl: sure, one sec23:35
xnoxIs there a package which is ubuntu-required ubuntu-base23:35
xnoxgenerated from the seeds23:36
xnoxasking because I'm helping to bootstrap ubuntu lucid without using debootstrap23:36
keesnxvl: done23:36
xnoxor am I better off just sending the list generated by deboostrap --print-debs23:36
nxvlkees: thanks23:37

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