[00:32] Keybuk: ping regarding Bug #571707 [00:32] Launchpad bug 571707 in mountall "fsck at bootstrap is too slow" [High,Confirmed] https://launchpad.net/bugs/571707 [00:33] arand: what kind of ping? [00:34] Keybuk: We had a discussion regarding it in here before, and since it seemed to be very general (at least three in here reproduced esily)... If you are not already, it might be worth seeing if it's fixable before the general public starts to hit the scheduled fsck:s ... [00:35] Keybuk: If you are the right person to look at it, of course. [00:35] right now is the weekend [00:35] then it's the Canonical SomeHands [00:35] then it's UDS [00:35] you're looking at 2-3 weeks for anyone to look at this [00:36] ...hmm, ok. [00:37] Well, at least we have an easy but extremely non-obvious workaround... I guess that's at least something. [00:47] i run notify-osd with xmonad and no gnome stuff at all, but the font in the bubbles look ugly, any idea why? [03:11] Since Ubuntu has no easy way to disable services and configure runlevels (like Fedora and OpenSUSE make easy), can anyone give me the low down on how to do this? I mean I understand the sysv-rc type paradigm but "upstart [03:11] confuses me [03:12] and upstart has zero documentation [03:12] Some of us have limited RAM and this dont need 20 extra services for braille and the like. [03:13] For instance if I try to use a toold like "bum" or "sysv-rc-conf" they don't work. They list services as not running when I know they are running. And others that I uncheck still start at boot. So, obviously, upstart is screwing it all up. [03:14] FeasibilityStudy, disabling jobs is a future feature, for now you can just modify the upstart job and comment out the 'start on' line. [03:14] look in /etc/init/ [03:15] (I did mention as much in #ubuntu ;-) [03:15] cody-somerville: Right now I am doing the old "update-rc remove service" command. Will that suffice? [03:15] no, it won't. [03:15] Well there needs to be better documentation about this! It's ridiculous. [03:15] FeasibilityStudy: bug #94065 [03:15] Launchpad bug 94065 in upstart "init: add non-destructive means to disable a job" [Wishlist,Invalid] https://launchpad.net/bugs/94065 [03:16] Upstart has ZERo docs I can find. [03:16] FeasibilityStudy, have you visited http://upstart.ubuntu.com ? [03:17] Why is there still even an /etc/init.d directory in Ubuntu? [03:17] If Upstart does away with the init system? [03:17] for backwards compatibility [03:18] note, there are hundreds of packages that use initscripts; it wasn't feasible to migrate them all to upstart semantics [03:20] Grrr. I mean I applaud upstart for trying to fix some shortcomings with sysv, but I think it is not helpful to screw the whole init system up by having a hybrid upstart-rc system. It just confuses everyone. [03:20] Thats always a great excuse to do nothing. [03:21] I didnt say do nothing, but I think upstart should wait until the whole system can be converted to it. Now, we have a hybrid type of deal where the old rc.d stuff doesn't work for upstart services. [03:21] FeasibilityStudy, We can't force every piece of software out there to convert to upstart jobs [03:22] Frankly, it is a PITA and is a bigger nuissance than PulseAudio. [03:22] change can be stressful. [03:22] indeed [03:22] Humans don't like change. That'll never change. [03:22] Just like Obama's change..ahem..Ok but. [03:23] I am stressed to tha max..Going crazy! arghh [03:23] FeasibilityStudy, take a break [03:23] Getting stressed over something like this is absolutely pointless [03:24] OK Im calm. Now, where can I find a list of all the upstart jobs? And what line do I need to configure again? [03:24] FeasibilityStudy, all the upstart jobs are in /etc/init/ [03:24] the upstart jobs, as I mentioned in #ubuntu, are in /etc/init. You can also see their symlinks to /lib/init/upstart-job in /etc/init.d [03:25] ou can just modify the upstart job and comment out the 'start on' line. <--------- Where do I find this line? [03:25] FeasibilityStudy, look in the files. Its at the top. [03:26] FeasibilityStudy, sometimes the statement spans multiple lines. be sure to comment out the entire statement. [03:27] In the bluetooth service script I see a "start" [03:28] FeasibilityStudy: if you haven't wandered past http://upstart.ubuntu.com/getting-started.html, that is a nice place to start. [03:29] crimsun: I am there but it seems to be mainly a page about how to develop for upstart [03:30] FeasibilityStudy: however, it also describes (albeit briefly) semantics [03:30] Ok I am looking.. I think I am on the right page..Maybe there is some info. [03:30] e.g., it mentions that jobs are defined in /etc/init [03:31] it also describes "start on", "stop on", etc. [03:38] OK where can I find all sysv services on the machine? I assume they are not in /etc/init.d [03:50] I don't see the problem sysv-rc-conf lists everything correctly for me. I assume there's something broken in your system in case it doesn't do this for you. [05:09] hey, im trying to build a package for iscsitarget. the application just uses a makefile, no configure scripts. i ran dh_make, it built the dir out, but when I try to use debuild, it fails, because the makefile tries to install files in $DISTDIR/usr [05:09] in install: build, in rules, i did export DISTDIR=$(CURDIR)/debian/iscsitarget;$(MAKE) DESTDIR=$(CURDIR)/debian/iscsitarget install [05:09] and it now works. is that legit? or is there a better way to do it? [05:11] if you are unsure you could reference another package that uses a configure-less build [05:11] can't think of any off the top of my head, but they exist [05:11] well this one specifically ignores DESTDIR pretty much and instead use DISTDIR [05:14] df00z1: you could always add a debian/patch too that fixes the makefile to do the right thing(tm) and then send the patch to the upstream devs [05:15] yes the patch sounds better === rgreening_ is now known as rgreening [05:35] hi [05:35] i'm vavelin kevin, i'm a student in reunion island [05:36] i just want to thanks all developpers of ubuntu for their work ^^ [05:36] ubuntu 10.4 is simply, powerfull, fast ;) [05:43] What software gets invoked when I hit a volume key on a laptop? [05:43] (Have a bug there) [05:49] the keys need to be mapped by xorg, dunno where they go from there [05:52] YokoZar: linux, gnome-settings-daemon [05:53] and, possibly other bits if you want extra things [05:54] df00z1: minor detail maybe, but I don't think the usage of $(MAKE) is correct there [05:55] $(MAKE) is (I think) intended for usage of recursive make [05:56] debian/rules happens to use make too, but that's a detail; it might as well use a shell script (although I think debian policy dictates the use of a makefile for debian/rules) [05:57] anyway, export FOO=val won't work for Makefiles I think [05:57] although your usage of make FOO=val is correct [05:58] but I suspect it's just a broken Makefile which fails to take DESTDIR into account [05:59] dh_make created the $(MAKE) [05:59] export did work, but i dont think its correct [05:59] does anyone know if plymouth needs anything special in the kernel to function properly? [05:59] I doubt dh_make put the export line in there :) [05:59] no I did that [05:59] ;) [06:00] export does NOT what you think it does [06:00] what export does is [06:00] sets an environment var? [06:00] it does, but not within a makefile [06:00] odd [06:00] it does work, and doesnt without it.. [06:00] what really happens is make spawns a subshell ($SHELL), which export's the variable [06:01] ah [06:01] but since you actually don't do anything from the subshell, the export isn't actually usfull anywhere [06:02] in the makefile's install sections, they do @install -vD usr/ietd $(DISTDIR)/usr/sbin/ietd [06:02] yes [06:02] is there a proper way to set DISTDIR? [06:02] but DESTDIR refers to a *makefile* variable [06:02] not to a shell variable [06:02] ah [06:02] you need to do DESTDIR=foo [06:02] or DESTDIR:=foo [06:03] either, don't recall offhand [06:03] but the makefile doesnt reference destdir at all [06:03] and it doesnt work [06:03] it uses distdir [06:03] like I said [06:03] 06:58 < Chipzz> but I suspect it's just a broken Makefile which fails to take DESTDIR into account [06:03] yeah [06:03] -e, --environment-overrides Give variables taken from the environment precedence over vari- ables from makefiles. [06:03] DESTDIR isn't magical, it's just a convention [06:04] i think environment variables cross into makefiles? [06:04] according to the manpage [06:04] make sometimes uses variables from the environment it was started from [06:05] but [06:05] they dont set DISTDIR inside the makefile [06:05] its just blank :\ [06:05] unless you set it [06:05] "export DISTDIR=$(CURDIR)/debian/iscsitarget;$(MAKE) DESTDIR=$(CURDIR)/debian/iscsitarget install" [06:05] note the ; [06:05] those are 2 seperate statements [06:05] yeah [06:05] and iirc make will spawns *two* subshells to execute both [06:06] hm [06:06] as *seperate* commands in *seperate* subshells [06:06] is there a way to pass params to make directly? [06:06] which is why I believe the export doesn't take effect [06:06] yes, the second part [06:06] it does take effect...its just not correct... [06:06] cexport DISTDIR=$(CURDIR)/debian/iscsitarget;$(MAKE) DESTDIR=$(CURDIR)/debian/iscsitarget install [06:06] oh can i set DISTDIR=whatever? [06:07] argh [06:07] $(MAKE) DISTDIR=$(CURDIR)/debian/iscsitarget install [06:07] $(MAKE) DESTDIR=$(CURDIR)/debian/iscsitarget install [06:07] which is where the difference between = and := is significant [06:07] (iirc) [06:08] ah [06:08] but it ignores destdir and i need distdir to be set :( [06:08] unless i redo their makefile [06:08] you're ignoring sth I already said [06:09] 07:03 < Chipzz> DESTDIR isn't magical, it's just a convention [06:09] convention being the operative word [06:09] I know...I'm saying D_I_STDIR [06:09] not DESTDIR [06:10] not a typo [06:10] if your makefile chooses to ignore that convention, then it's broken [06:10] yes. it is. [06:10] (fsvo "broken") [06:13] run. Every environment variable that make sees when it starts up is transformed into a make variable with the same name and value. But an explicit assignment in the makefile, or with a command argument, overrides the environment. [06:13] So make will pull in environment variables [06:14] i sent the devs an email asking if theres a specific reason they're using DISTDIR...if they don't fix it..i can probably patch it [06:14] yuck... [06:14] and then have to fix it every time they do an update upstream.. [06:15] http://www.gnu.org/prep/standards/html_node/DESTDIR.html like everything says to use DESTDIR [07:06] Is it known there is a Grub issue with 10.04? [07:09] k776: What kind? Got a Bug #? (nb, I'm not a developer, and this channel might be quiet during the weekend now...) [07:13] The kind that doesn't boot the OS though it installed ok [07:13] Just gives me a black screen [07:19] k776: Well, support goes in #ubuntu, but my general thing is to reinstall grub [07:21] tried #ubuntu. Some helpful people but no luck with their suggestions [07:21] including reinstalling grub [08:09] https://bugs.launchpad.net/ubuntu/+bug/572776 [08:09] Launchpad bug 572776 in ubuntu "Ubuntu should provide update packages for download and use for offline users" [Undecided,New] [08:10] don't you think it would be useful to implement [08:34] TheMuso: how do you read pdfs? there's someone saying that orca + [evince|okular|adobe pdf] doesnt work and they have to copy and paste to gedit to get things going. figured youd be the person to ask for info about this [09:25] i run notify-osd with xmonad and no gnome stuff at all, but the font in the bubbles look ugly, any idea why? [13:23] Hi. I prepared a patch for LP #572151. What should I do to get it uploaded to lucid-proposed? (I'm not a Ubuntu developer.) [13:23] Launchpad bug 572151 in libanyevent-perl "Tries to overwrite files from anyevent-perl without Replacing it" [Medium,In progress] https://launchpad.net/bugs/572151 [13:55] aburch: have a look at https://wiki.ubuntu.com/StableReleaseUpdates and https://wiki.ubuntu.com/SponsorshipProcess [14:37] hi , I had a general question abt filesystem and Flash memory , is this the right place to ask ? [14:39] probably not [14:39] but you could try [14:39] do take into account that it's a weekend though, so not many ppl here [14:40] ok , How can I access a file stored on USB ( NAND ) Flash Memory at the the "block(s)" and "pages" level . [14:41] that question doesn't make sense [14:41] and shows you don't understand how linux works :) [14:42] linux has block devices, and filesystem use that as the underlying storage [14:42] linux doesn't care weither the block device is a hard disk or flash memory [14:42] so that part of your question is irrelevant [14:43] second [14:43] you want to access a file [14:43] but you fail to take into account there are several possible file-systems which could be on the block device [14:44] all of which have different layouts and storage strategies [14:45] apart from that, I fail to see why you would want to access a file at the block or pages level in the first place [14:46] there's really very few to none at all good reasons to do so [14:46] unless you're writing a fs defragmenter or sth like readahead, you really don't want to do that [14:46] hm... ok , basically as far as I knew before you told this , to access a file at that level say from a C program , I would have to use stat() to get the inode index number , look it up in the indoe table and access it using ioctl() . [14:47] but since flash memoryis electronic ioctl() is of no use , as ioctl() is primrily used to communicate with the hardware [14:47] but what are you trying to achieve in the first place? [14:48] because 9/10 you are not taking the right approach [14:48] I basically want to change any random bit in the ECC code of each block the file is stored in [14:49] ^^ [14:49] why? [14:49] * gangil someone told me to try this out [14:50] help , you can understand this as this is an assignment for me [14:50] to see how well ecc works? [14:50] yeah , maybe , he told me to do this and I want to make this work [14:51] sigh [14:51] :-/ [14:51] you realize that you are intentionally trying to fuck up your data? [14:52] no problem [14:52] I am willing to exp. [14:52] you also realize that this won't even do you any good? [14:52] since ecc is meant to correct hardware failure [14:52] I realize everything ! :) dont worry [14:52] I want to know how can I do it [14:52] if you're writing to ecc memory, the checksum bit will change too, as the change is INTENTIONAL? [14:53] this won't help you test ecc one bit... [14:53] would you tell me how can I do that ? [14:53] I have no idea [14:56] Chipzz: are you intentionally not telling me , please tell me , I want to do this . [14:59] 15:53 < Chipzz> I have no idea [15:00] ok , thanks :) for other useful info === blueyed_ is now known as blueyed === pretto_ is now known as Pretto === dendro-afk is now known as dendrobates === 40FAAUE7J is now known as dendro-afk === yofel_ is now known as yofel [17:37] Hello, how do I remove the keyboard layout from the systray? I asked in #ubuntu but they didn't know. [17:38] Also, that may be a bug, so I thought you should know. === almaisan-away is now known as al-maisan === NCommand1r is now known as NCommander === NCommander is now known as Guest43091 === Guest43091 is now known as NCommander [19:06] hi... I'm seeing problems using initramfs-tools... looks like this issue : http://www.mail-archive.com/linux-fai@uni-koeln.de/msg02954.html [19:07] I don't think busybox is being put into the initrd - so there's no grep, rm, pidof ... [19:07] I've managed to get an init=/bin/bash shell and done a manual udevd& udevadm trigger to get a minimal system up [19:08] oh, this occured during a 9.04->9.10 upgrade === al-maisan is now known as almaisan-away === radoe_ is now known as radoe [21:22] guys, WTF [21:22] why does ambiance mess with gconf settings for window buttons? [21:22] how the hell does the *GTK+* theme touch my window titlebar settings? [22:13] here is yet a bug issued about black screen (no console, no x11) switching back to first session after an user switch? === 15SAASDC2 is now known as dendrobates [23:03] what was the component you want to look at the log output for when inserting a dvd+rw disc in the drive causes it to vanish from nautilus? === 77CAALSRA is now known as dendro-afk === dendro-afk is now known as 77CAALSRA [23:16] I've found a regression in kernel currently available in -proposal; how could/should report it? [23:19] elleuca: https://wiki.ubuntu.com/StableReleaseUpdates [23:20] elleuca: ""In the event of a regression, immediately notify the Ubuntu Technical Board via email, and ask for help on #ubuntu-devel in making urgent contact with a member of the Board or the SRU team: state the problem, the bug number, and ping "(list of names)""" [23:21] arand: that's the procedure for regressions which get released into -updates, not into -proposed. For -proposed I think it should be sufficient to comment on the relevant bug [23:22] arand, mdke see latest comments from myself here https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/362875 [23:23] Launchpad bug 362875 in linux "rt2x00usb_vendor_request: Error - Vendor Request 0x07 failed for offset 0x308c with error -71." [Undecided,Incomplete] [23:23] elleuca: see the section "Verification" in that wiki page, steps 1 to 5 [23:24] mdke, thank, then I'll reboot using kernel from -proposed and I'll file a bug [23:24] cool [23:25] anyone using rhythmbox + smb:// uri's w/success? [23:35] mdke, cool, using `ubuntu-bug linux` there is a step by step path helping to address the issue === 77CAALSRA is now known as dendro-afk === dendro-afk is now known as dendrobates