[00:17] slangasek, done [00:17] thanks [02:10] I'm trying to understand this http://people.canonical.com/~scott/daily-bootcharts/20100219.1-max.png [02:11] Why do so many tasks end at the same time? [02:11] ah [02:11] now I get it [02:12] they are processes that continue to run when the system is up [02:13] and I should be looking at blue stuff [02:47] Is there a trivial reason some GTK2 apps will show network places (~/.gvfs mounts) in their file dialogs and others won't? i.e. a compile-time option? === funkyHat is now known as drunkyHat [02:54] yes, gvfs isn't a kernel mount, so all apps need to be updated to use it. [02:56] how trivial is that, do you know? It's something I'd be interested in working on, because it's bothered me for years. [02:57] I figured it'd be a toolkit thing [02:58] i.e. any GTK2 app would have it inherently [03:01] tlp: You need to use a GVFS, a different VFS API, to access GVFS mounts directly. [03:04] doesn't GVFS also put FUSE mounts in ~/.gvfs anyway, so when you rm -rf a home directory it'll trash all your network shares too ^W^W^W^W^W^W^W^W^W^W^W^ legacy apps can use it [03:04] jdong: Yup. [03:04] I don't understand why even, say, GNOME Terminal doesn't let me pick "Network" in its file dialog [03:04] GNOME Terminal has a file dialo!? [03:04] +g [03:04] File->Save Contents [03:04] I'm guessing save transcript [03:05] Ah. [03:05] That menu item lies. [03:05] It doesn't not have an ellipsis. [03:05] Er, "does not". [03:05] I understand how to get to mounts via ~/.gvfs, but I highly doubt an end user would. [03:06] tlp: So, the key is that real GVFS apps don't use ~/.gvfs at all. [03:06] right. Just seems weird that core GNOME stuff isn't using it already :) [03:09] jdong: seriously? that's pretty scary. [03:09] tlp: what, recursion down network shares? [03:09] tlp: you have to replace all your file IO [03:09] tlp: there's a reason why rm is aliased to rm --one-filesystem in my zshrc right now :) [03:09] tlp: there is an argument that its basically a bad idea and that they are solving the wrong problem. [03:09] fortunately the remote system was over a low-bandwidth link AND gvfs sftp is slow :) [03:10] lifeless: .gvfs is a bad idea, or the GVFS API? [03:10] the GVFS API is kinda annoying to use IMO... [03:10] I'd rather wish they did everything over FUSE [03:10] tlp: both [03:11] (then again, the UNIXes where FUSE isn't available/common...) [03:11] fuse still has some issues last I heard [03:11] I've never used it myself. I'm a renegade from a more complex UNIX-like operating system and only recently started caring about out-of-the-box user experiences [03:12] anyway, seems like a big usability problem [03:16] jdong: so, here's a question. an rm alias seems like an easy thing to commit to prevent that; any idea why nobody has? [03:17] tlp: probably because it's confusing behavior to those not expecting rm to be crippled in such a way [03:18] I didn't even know about ~/.gvfs until last year when someone pointed it out to me. I was using smbfs or something. I'd be pretty pissed if I accidentally nuked a network share. [03:18] good thing I don't rm my home directory much [03:18] indeed. Though IMO this still files under the general category of "careful with -f / -r on rm" [03:18] it's an abnormal usecase to be zapping the home directory all that often. [03:19] that's true [03:19] fortunately, the gdm-guest-session setup does not allow any sort of usermounts thanks to apparmor policy. [03:19] I tried without success in tricking it to rm -rf a mountpoint.a [03:19] don't we prevent rm /? [03:19] upstream does. [03:19] specifically it refuses to recurse starting from /. [03:19] but you can still use /* and other cute variants to do just as much damage. [03:24] I don't see much harm in a simple y/n confirmation for that sort of thing [03:24] few people intend to do that, I think. [03:24] IMO that's something to take up with coreutils. [03:25] -f should mean -f. [03:25] would be mean to scripts to change such a primitive command like rm to include new prompts [03:25] still trying to figure out how the Ubuntu project is structured (starting by staying in here more often) [03:25] ah, true [03:25] you can turn on prompting [03:26] lol but then it seems to prompt too much :) [03:26] either either nag nag nag or *BOOM* [03:26] at least users always feel that way ;-) [03:29] I've been using rm -rf so long that I almost forgot -f is supposed to force those things :p [03:30] My university aliases rm to rm -i, which means everybody is immediately taught to always use -f, which is sort of bad... [03:31] yeah. something I learned early on and just never thought about again. [03:31] yeah, a lot of distros I've used do the same. [03:31] at least for the root user. === drunkyHat is now known as funkyHat === Hellow_ is now known as Hellow [05:25] anyone good with python give me a little help with parts of an apport hook? [05:26] here's the pastebin: http://pastebin.com/d67f035af [05:27] something to do with lines 15 to 20 is causing some issues [05:36] nigelb: missing an 'import re'? [05:36] slangasek: oh! [05:38] now something more seems to go wrong [05:40] slangasek: I dont seem to be going into line 20 [05:40] somewhere I'm doing something wrong [05:41] and the file isn't being attached either [08:25] hi [08:26] i am trying to build a initramfs which can load wireless driver [08:26] i tried modprobe iwl3945 [08:26] but its not activating wlan0 [08:26] i think due to inability of firmware loading [08:27] howto load firmware without udev ? [08:28] slynux_: which ubuntu are you using? [08:28] slynux_: and what kernel? [08:28] 9.10 [08:28] 2.6.31-14-generic [08:28] slynux_: the module is called "iwlagn" [08:29] oh [08:29] slynux_: and the way to make it be loaded in the initramfs is by adding iwlagn to the bottom of /etc/initramfs-tools/modules [08:29] okay [08:30] but i cannot find iwlagn in my lsmod o/p [08:30] why? [08:31] o/p? [08:31] output [08:31] lsmod | grep iwl [08:31] my wireless driver is working when the ubuntu boots up [08:31] when udev runs [08:31] iwl3945 77212 0 [08:31] iwlcore 112508 1 iwl3945 [08:31] mac80211 181236 2 iwl3945,iwlcore [08:31] led_class 4096 3 sdhci,iwl3945,iwlcore [08:31] cfg80211 93052 3 iwl3945,iwlcore,mac80211 [08:32] i want it to be activated during initramfs [08:32] before udev [08:33] why would you want that? [08:33] i am trying to implement a wireless LTSP [08:36] wow, is that gonna suck bandwidth wise [08:36] still i wanted to do it [08:36] there is some wltsp came with windows [08:37] but why before udev? [08:37] why not after udev in initrams? [08:38] when an apport hook is added, I only need to add one line to the debian/rules file.. correct? [08:38] i dont want to include udev in intramfs [08:38] once network is up i need to chroot to remote box [08:38] does this line work "cp debian/rhythmbox.apport debian/rhythmbox/usr/share/apport/package-hooks/source_rhythmbox.py" [08:39] why you want to _remove_ udev from initramfs [08:39] slynux_: what's wrong with having udev there anyway? [08:39] i need to add udev in initramfs [08:40] it will become heavy [08:40] i just need to keep it very small [08:40] udev is in initramfs already [08:42] initramfs is currently 9.8MB [08:42] how much of this is udev? [08:42] oh [08:42] i though udev was outside [08:42] * hyperair sighs [08:42] sorry [08:42] its both [08:42] thanks [08:45] anyway the wireless module is useless in initramfs unless you add iwconfig and ifconfig/iproute2 - not sure how big deps these would be [08:46] ok, wireless-tools would add only libiw29 so it's likely fine [08:50] hey, anyone have a rough guess at how much of Ubuntu's code is sent back upstream to Debian? [08:52] I see a wireless ltsp quite useless in most cases, just because wlans suck when you have enough traffic, but for ad-hoc setups it would be nice, maybe [08:53] of course you would still need to have local initramfs, because you can not usually tftpboot wlan adapters (there is no way to configure the wlan to use for booting etc) [08:56] oh he left. I would have first tried running for example 10 normal clients with remote X sessions over wlan and see if it's usable, before wasting time with fully diskless setup... [09:51] * ogra wonders if anyone else is seeing horrible slowness with X since the last upgrade [09:51] and also a lit of flickering in terminals [09:54] ogra: mmm, and possibly the source of my rhythmbox/xorg cpu abuse [09:54] ogra: there's definitely something odd going on in the X stack [09:55] ogra: intel? [09:55] xchat takes about 60% CPU here [09:55] yeah, intel [09:55] in xchat i see a lot of flickering on the line that separates nicks from typed text ... like there are dots jumping on the line [09:56] (i already disabled compiz, its definately not composite related) [09:58] i'm seeing a little bit of terminal flickering, not as bad as it was a few days ago, none of the other problems you mention though [09:59] * ogra downgrades the intel driver and restarts X [10:01] nope, didnt help [10:02] * ogra downgraades xorg-server too [10:05] hmm, still broken [10:11] I think I also seem some flickering in virtualbox. if this is what you mean: http://abs.getdeb.net/X.avi [10:11] hi guys [10:12] do you have some ideas to create an autocompletition texteditor? [10:12] to create post with bbcode for blog? [10:18] sigh [10:19] c_korn, my system is to slow to play a movie, both CPUs are at 100% [10:19] funnily one is completely taken by evolution-alarm-notify [10:19] the rest by X and xchat [10:20] ogra: oh, well what you would see is the consolte of the update-manager which flickers when a new message is displayed [10:20] yeah, gnome-terminal behaved similar [10:20] though that part seems to be fixed with the X downgrade i just did [10:21] i still see the flickering one the spatrator line in xchat though [10:21] *on th [10:21] e [10:22] ah, killing evo i can at least type again without delay [10:23] I've noticed that Evo's been pretty slow the last day or two. [10:23] But it's not eating my CPU. [10:23] i dont think its evos fault but something lower in the stack [10:24] nothing in any log though [10:24] Indeed. [10:25] if it's in gtk-based text widgets, maybe testing with plain xterm would be good idea... problem even there and it's likely server level issue [10:48] yeah, downgrading to libgtk2.0-0 2.19.5-1ubuntu2 fixes the world \o/ [10:49] its all sebs fault ! [10:51] the world!? [10:51] * highvoltage installs now === tkamppeter_ is now known as tkamppeter [11:46] hi [11:47] how do you create official ISOs? [12:12] hi [12:13] i am trying to activate wlan0 in initramfs [12:13] when i do modprobe iwl3945 [12:13] it shows error -2 [12:13] and says cannot probe the device [12:13] is that problem with firmware loding? [12:13] i added a hotplug script [12:13] but it has no subsystem=firmware uevent [12:13] any comments ? [12:36] StevenK: as you offered to help on the OCaml transition: findlib and facile need a rebuild (bug 522363) and camlp5 and havea need a sync (bug 522358) [12:37] Launchpad bug 522363 in xml-light "[OCaml 3.11.2 transition][round 2/6] Please rebuild packages involved in OCaml transition" [Undecided,Fix released] https://launchpad.net/bugs/522363 [12:37] Launchpad bug 522358 in hevea "[OCaml 3.11.2 transition][round 2/6] Please synchronize packages involved in OCaml transition from Debian sid to lucid" [Undecided,Confirmed] https://launchpad.net/bugs/522358 [12:57] ogra: interesting -- thanks [12:57] ogra: the ol' IZ GTK BUG [12:58] ogra: hah, that resolves my rhythmbox issue too :-) [13:04] * debian/patches/062_client_side_decoration.patch: [13:04] - upload the work from Cody Russell on client side decoration to lucid [13:05] j'accuse! [13:10] what's about this issue with having "LGPL 2.1 or later" licensed code? [13:26] azeem: I guess some people are afraid of the FSF releasing a newer version that says something they dislike? [13:28] right, but first off, the FSF actually vowes to not change the spirit of the license [13:28] and saying something like "Because seriously, everything should be this way. None of us should be saying "LGPL 2.1 or later". Ask a lawyer, even one from the FSF, how much sense it makes to license your software that way." seems unproductive [15:57] hello [16:03] hi [16:03] I came to ask it the devs know about possible improvements around screen detections in the next Ubuntu ? and about the network as well ? (network-manager still the default app ?) [16:11] melodie_: yes nm will still be default [16:11] hi asac [16:12] do you know if it will get improvements ? (last version I gave up on Ubuntu because it didn't even see my ethernet :/ ) [16:12] melodie_: did you have anything in /etc/network/interfaces configured? [16:13] what about screen detections ? no new tool coming out to get it configured the good way ? [16:13] if not, its a driver bug and has nothing to do with nm [16:14] anyway out ... getting food and weekend action and so on ... enjoyy [16:14] what I can tell you, is that it was just after installing, and that it didn't work whereas before it used to work out of the box. I could connect only with dhclient, whatever I tried. I'm not a /etc/network/interfaces specialist on the other hand [16:14] (dont know about screen detection ... usually everything happens automatically quite well nowadays) [16:14] asac, not really. well, I guess I should go back to bugzilla more often ;) [16:14] cant tell then. would need to see the logs and the files. most likely its fixed in lucid [16:15] asac, I'll give it a try again when out. thanks asac [16:15] melodie_: give it a try before its out ;) ... otherwise its hard to get things fixed [16:15] at least pop in the live cd [16:15] and check if network is good [16:16] asac, allright, I'll try it in virtualbox and will yell if the vbox drivers aren't included. ;-) [16:16] take tomorrows dailies ... http://cdimage.ubuntu.com/daily-live/ [16:16] asac, ok ! thank you ! [16:16] melodie_: virtualbox is not really important [16:16] for me it is [16:16] what is important is your main system [16:16] that's where I do many tests [16:16] virtualbox usually worked here [16:17] at a time I could not get a gui in vbox, without strange tweaks [16:17] if your problem was about virtualbox then its odd because i am sure i would have gotten more complains if it didnt work at all [16:18] now I have a Xubuntu in it, but only 800x600 and no way to know how to change it unless I get a xorg.conf from another install [16:18] which has also occurred in a real machine several times [16:18] the ethernet problem was on my laptop ibm T30 [16:19] ok I get a tomorrows' daily anyhow. [16:19] do you know if is it possible to dd it to a usb pendrive and boot it from there ? I like to save cd-r when I can [16:21] melodie_: I don't think dd will work, usb-creator works to best or unetbootin [16:22] hi LaserJock [16:22] to use usb-creator you have to have a Ubuntu installed don't you ? [16:23] yeah, or I think Windows [16:23] I don't use Windows [16:23] unetbootin is available for most distros [16:24] I tried it once, it screwed my menu.lst [16:24] never again ! [16:26] I can have it in Archlinux ! Great ! [16:27] which one there is the one asac told me about ? http://cdimage.ubuntu.com/daily-live/ [16:27] which is the latest ? 20th ? [16:46] melodie_: usually there is a link to "current" [16:46] LaserJock, ok, thank you [16:47] I'm dowloading from the "daily-live" directory now. [17:21] do one knows if is there any development done especially for the netbooks ? [17:26] please sponsor debdiffs for fakesyncs: bug 512430 ; bug 524955 ; bug 524957 [17:26] Launchpad bug 512430 in geronimo-jpa-3.0-spec "Fake sync geronimo packages (main) from Debian testing (main)" [Undecided,Confirmed] https://launchpad.net/bugs/512430 [17:26] Launchpad bug 524955 in polkit-qt "Fake sync polkit-qt 0.9.3-1 (main) from Debian testing (main)" [Undecided,Confirmed] https://launchpad.net/bugs/524955 [17:26] Launchpad bug 524957 in polkit-qt-1 "Fake sync polkit-qt-1 0.95.1-1 (main) from Debian testing (main)" [Undecided,Confirmed] https://launchpad.net/bugs/524957 [17:31] LaserJock, do you know if there are specific components at Ubuntu, that one should install when having a notebook Samsung NC10 ? Or if some special developments are led ? (kernels for notebooks, or ?) [17:58] melodie_: for netbooks you should install une [17:58] melodie_: http://cdimage.ubuntu.com/ubuntu-netbook/daily-live/ [18:06] melodie_: you can look at the "Hardware Specific Help" section of https://help.ubuntu.com/community/UbuntuNetbookEdition [18:13] james_w: when you get a moment, can you look at bug 524980? if that patch is ok, I'd like to get it in; this error drives me crazy. :) [18:13] Launchpad bug 524980 in lazr.restfulclient "does not retry temporary failures" [Undecided,New] https://launchpad.net/bugs/524980 [18:15] kees: nicely phrased :-) [18:15] I have the same code externally [18:15] heh [18:15] yeah, I have the same code externally in lots of different scripts. :) [18:15] I'd like to get leonard's opinion though [18:15] okay [18:16] yeah, I figured it should get a little review [18:17] kees: that patch is bad ;) [18:17] 1st. using while True: is lazy habit imo ;) [18:17] 2nd. you have an endless loop there, no? [18:18] asac: well, I suppose it could be better, but it's not endless, we either break or reduce retries. [18:18] run a s/retires/retries/ [18:18] ;) [18:18] asac: I wanted to avoid a needless trailing sleep(1) [18:18] kees: you dont reduce retries, but retires ;) [18:18] oop, but yeah, typo is ugly [18:18] which probably never retires ;) [18:18] cheers [18:19] * kees updates patch [18:19] kees: fwiw, you've probably discovered this already, but just that isn't really sufficient for unattended jobs [18:20] james_w: oh? it seems to work well enough for mine jobs? [18:20] s/mine/my [18:20] you often need to back of more than that [18:20] back off [18:21] if it's in cron and will try again an hour later then you are probably fine [18:21] but there are often storms that go on longer than the ~30s that will retry for [18:22] hrm, I hadn't noticed anything that got that crazy, but I think stalling for 30m in a client call isn't probably good. [18:23] no [18:23] so if it's appropriate you may want to do it at a higher level [18:23] yeah. but this should catch, at least for me, the bulk of the glitches [18:23] yeah [18:24] I just did the same thing hoping it would solve more than it ended up doing [18:24] lets see what Leonard says, I'd rather not stick it in if clients can't rely on it being there [18:26] oh, and "ValueError: No json object could be decoded" means that haproxy can't reach an appserver, a couple of bugs mean you get a silly error [18:36] kirkland: hrm, does qemu-nbd hang for you? I can't get it to connect === |sistpoty| is now known as sistpoty [19:07] hi again [19:11] I think I ought to post a bug report, but I don't know against what I should post it : Xorg ? what is the configuration tool for resolution ? I installed in vbox, and did search a simple solution within Ubuntu to fix the resolution, but found none. Ended with using a foreign xorg.conf to fix it. [19:11] http://pastebin.ubuntu.com/380521/ [19:11] what should I do to help improve this ? It's a problem I also met before, on a real install [19:12] I feel that there is really something missing... [19:50] melodie_: is this on Lucid or a stable release? [19:53] LaserJock, on Xubuntu Karmic, and last year same happened in Ubuntu Karmic on an install on one of my machines [19:53] LaserJock, maybe you are not interested in bugs about Karmic for now ? [19:58] LaserJock, one more among many ? -> https://launchpad.net/+search?field.text=xorg+resolution+Karmic&field.actions.search=Search [19:58] :o === nhandler_ is now known as nhandler [20:06] has something got decided about the fakesync version scheme? [20:13] cjwatson: ^? [20:17] I try to install Lucid in a pendrive : does it need a ext2 partition to work out ? [20:17] (with usb-creator-gtk) [20:29] nobody for the fakesync versioning scheme? [20:34] bigon: you've read the mailing list discussion? [20:34] yeah but I'm not sure something have been really decided :p [20:34] well, you have all the information that I know of then [20:36] well I will upload my pkg with -Xfakesync1 then [20:49] bigon: the tools today won't know how to handle that [20:50] :/ too late [20:50] I wouldn't recommend using new package version schemes before there's a clear consensus and a committment to implementing the tool support [20:57] slangasek: do you want me to leave bug 283217 open for the thinkpads? [20:57] Launchpad bug 283217 in pulseaudio "Volume HotKey shows OSD, doesn't change mixer level" [Medium,Incomplete] https://launchpad.net/bugs/283217 [20:57] otherwise I'm inclined to close it for the OR's hw [20:58] Is mvo the only specialist on update-manager? I'm about to update my hardy LAN server, but it picked hardy->intrepid instead of hardy-lucid. I did some initial triage together with mvo and reported a ticket in LP. Since then I've not heard back from him even though I pinged him a couple of times. I finally want to go ahead with the upgrade, but that will then be the end of further data collection. [20:59] crimsun: eh, why would you use that bug to track anything about thinkpads when there's no mention at all of thinkpads in it? [20:59] Laibsch: he is the primary author [21:00] its just python though, so anyone that can program can debug it reasonably easily [21:00] who is econdary author? [21:00] ;-) [21:00] +s [21:00] lifeless: computers are just 0s and 1s, anybody who can count can debug them ;-) [21:01] slangasek: given the title, I did not know whether you intended to track thinkpad symptoms there [21:01] slangasek: anyhow, I'll close it, thanks. [21:01] crimsun: yes, please close it [21:02] I wasn't intending to be sarcastic [21:02] I feel like you were though in your response; is that accurate? [21:03] * hyperair finds python pretty hard to debug [21:03] is there some gdb for python? [21:04] hyperair: yes; pdb and also gdb can do python [21:09] oh that's cool [21:10] lifeless: gdb doesn't appear to do python. says it's not an executable format [21:11] you either need to run 'gdb python' and then pass the script as an argument to 'run', or attach to the process by pid [21:12] is 6 Go enough for Lucid, in a Virtual machine, please ? [21:13] melodie_: it should be, yes [21:13] melodie_: you may find #ubuntu+1 is a better forum for asking such questions [21:14] thks slangasek, what is that chan dedicated to ? [21:14] hyperair: http://wiki.python.org/moin/DebuggingWithGdb [21:16] melodie_: #ubuntu+1 is for support of the current development release [21:16] slangasek, I'll also look for the dedicated bug tracker [21:21] melodie_: sorry, dedicated bug tracker for what? [21:21] for Lucid [21:22] sorry I was disconnected :- [21:22] melodie_: https://help.ubuntu.com/community/ReportingBugs [21:23] thank you ! theses pages can look like a real jungle when you don't know where the right one is [21:24] no [21:24] I meant the bug tracker dedicated to Lucid. there I have gone allready, I am logged in launchpad too [21:36] If you don't take some time to read the chances are that you will not write a good bug report. The effect of that will be that you will take up valuable developer time that could have otherwise gone into fixing a bug. [21:36] Please make sure you write the best possible bug report you can so that the devs can actually fix it [21:37] You may find out that your "bug" is not a bug after all or has a known workaround [21:37] melodie_: Launchpad *is* our bug tracker [21:37] melodie_: bugs.launchpad.net/ubuntu is the dedicated bug tracker [21:37] melodie_: its big because Ubuntu is big. [21:38] I know that [21:38] I used it at it's very beginning [21:40] melodie_: this isn't the right place for you to ask for support, please use #ubuntu+1 or #ubuntu-bugs. Please be sure to read the wiki pages pointed out to you. You will find all the answers you need. [21:41] Laibsch, I am logged there, I was just asking where the dev's are looking for bugs to fix [21:41] melodie_: bugs.launchpad.net/ubuntu [21:42] in the whole place ? [21:42] * Laibsch is getting a certain feeling [21:43] melodie_: yes [21:44] lifeless, thank you [22:25] Is there a web site / way to view the patches that have been applied to packages in Lucid. (I'm looking to see if a particular has been fixed since Karmic.) [22:27] Dunkirk: patches.ubuntu.com === Simira_ is now known as Simira [22:51] what is todays Lucid ? alpha ? alpha 1 ? [22:51] or more ? I want to report this : http://pastebin.archlinux.fr/378262 [22:55] melodie_: Lucid questions in #ubuntu+1 [22:57] allright thanks [23:34] bye === dendro-afk is now known as dendrobates