[00:05] anyone aware of an open bug for xorg's outrageous cpu usage? current 96% for me === marienz_ is now known as marienz [02:13] mysql isn't running on startup... I don't know if this is a +1 related issue, but I don't remember having it before so thought I'd better ask here ⢁D [02:14] funkyHat: /etc/init/mysql.conf is there? [02:14] yofel: yes, and running start mysql works [02:15] I don't know anything about upstart really, so I don't know if there's a system for enabling or disabling services [02:15] well, what mysql depends on is [02:15] start on (net-device-up [02:15] and local-filesystems [02:15] and runlevel[2345]) [02:15] I know about update-rc.d, but I guess that's not relevant anymore? [02:16] hm, it did start here when I rebooted a few days ago [02:17] oh... "Jul 21 21:07:41 tallest init: mysql post-start process (6549) terminated with status 2" [02:20] funkyHat: correct. I think !boot in general is a little out of date now, actually. [02:21] I wonder if I can set a higher debug level and get some more info about why it's failing next time I boot [02:22] funkyHat: sudo initctl log-priority info [02:22] funkyHat: https://wiki.ubuntu.com/MeetingLogs/devweek1007/UpstartJobs is a good intro :) [02:23] rww: I need it to persist after reboot, mysql comes up fine if I run it after boot [02:23] funkyHat: ctrl-F that command on the page I linked and read the next few lines [02:25] I'm gonna forget to do this when I swtich my PC on tomorrow [02:26] agh I have 87 tabs open in firefox === yofel_ is now known as yofel [03:01] ARHGGHGH [03:01] Xorg is treating my middle button as a left button! [03:01] How do I make it not do that? [03:02] generally or just touchpad? [03:02] Just touchpad. [03:02] And yes, it does have a middle button. [03:02] I haven't tried other mice with it. [03:03] Even evtest on the event file gives left for middle. [03:04] hm, checked synclient settings? [03:05] Oh, and they say we don't need "shmconfig" anymore.... [03:05] bollocks... how else are you supposed to do synclient -m? [03:05] If it's a security vulnerability, then fine, make it require sudo. [03:06] hmm, even cat /dev/input/by-path/platform-i8042-serio-4-mouse | hd [03:06] reports same for left and middle. [03:06] Synaptics: Clickpad mode enabled [03:06] Ah, no wonder. [03:06] I don't have a ClickPad! [03:07] Time for a bug report, then. [03:18] Bug filed. [03:18] Oh wait, an error occured, and trashed my submission. [03:18] Thanks, Launchpad. [03:18] Great. [03:19] Well, screw that. [03:24] Grrr. [03:24] So, how do I fix that so I can use my middle button again? [03:25] argh [03:26] rebooting to Lucid.... that button mishandling is a showstopper. [03:32] Argh, Synaptics really irks me sometimes. [03:32] Through an accidental driver release, they've proved that ALL their touchpads can do basic multi-touch... yet they cripple that if OEMs don't pay for gesture support. [03:33] I have a P2 laptop that can do multi-finger detection... yet my Core 2 Duo can't? That's baloney. [03:40] DanaG: that's probably a patent issue [03:40] No, it's a money issue. [03:41] They sell some touchpads with full multi-touch, and some with none. [03:41] All of the old ones had very trivial multitouch, at least. [03:41] DanaG: right, because they probably have to pay for a patent license if they sell stuff with multitouch [03:41] The new ones have either none, or have some if you install the driver that patches the touchpad firmware. [03:42] funkyHat: fine, then they should at least let users pay for the feature, even if the OEM didn't! [03:42] Wait, and weren't Synaptics the ones to win multitouch patent lawsuits? [03:43] No, looks like there's a lawsuit of Elan suing Apple. [03:43] ah, Elan versus Synaptics: "Last year, that lawsuit and Synaptics' countersuit were both dismissed after the companies reached a cross-licensing agreement." [03:44] Damn patents. [03:55] the ruinous cost of all of the lawsuits will be passed on to consumers who just want the stuff to work and at a cheaper cost, something the patents prevent === Melodist is now known as Symphonist [13:24] when i use indicator-network [13:24] can i safely remove network-manager? [13:29] DrHalan, network-manager isn't the tray icon. [13:43] Well, adding --debug and --verbose to the boot options didn't make upstart log anything else about mysql. Still in the dark about why it's crashing [14:22] running into problems with "aptitude safe-upgrade" with the package libvirtodbc0 http://paste.ubuntu.com/467997/ [14:37] anybody having issues retrieving gmail with evolution? [14:38] IdleOne, I can't receive emails from my Live account, Gmail works fine. [14:39] Strange, I have two gmail accounts, both were working fine and now it seems like I only get email when it feels like it [14:39] no errors that I can see [14:42] There's something that's bothering me to hell. [14:42] gnome-system-monitor no longer shows the "Memory" thing. I have to set it in the preferences everytime I start it :| [14:43] Also I think I need to teach my sister about spam email. [15:40] Looks like flashplayer doesn't like pulseaudio anymore. === patdk-h3 is now known as PatrickDK === jdobrien is now known as jdobrien|lunch === luftikuss is now known as bullgard [17:50] xubuntu 10.10 from comp 1 to compt 2 wire to share network doestn work dorsny show any wired network available. in comp 1 settings is made config shre to other comps. in comp 2 is made automatic. [17:51] when i plug wire from comp to router to comp then all works. comp to compt doesnt. why? i used https://help.ubuntu.com/community/Internet/ConnectionSharing#GUI%20Method%20Using%20Network%20Manager%20%28Ubuntu%209.10%20and%20up%29 [18:07] charlie-tca: ? maybe u know? [18:56] I don't use network manager. All my connections are static ip managed by me [19:45] hi. short question about bootchart [19:46] disk throughput seems really low and disk utilization is almost 100 % for the whole boot process [19:46] is this normal? [19:46] or is my hdd s%#t? [19:48] also cpu is most of the time in io-wait state [19:49] http://knittl.is-a-geek.net/public/bootchart.png [19:49] to give you an idea [20:02] knittl: doesn't seem too strange. readahead is usually used to increase io efficiency, but i don't know if they did something about it in maverick. [20:02] void^: well, i was just thinking that cpu is most of the time waiting for disk [20:03] and disk is not reading at full throughput [20:03] (the 90something spike) [20:06] hm? a traditional harddisk can only achieve its "full throughput" when reading consecutive data in order [20:14] void^: i thought that was the exact goal of ureadahead? [20:14] moving data on disk so it can be read consecutively? [20:17] knittl: afaik, it just figures out the optimal order for reading the required files into memory. but as i'm not seeing any readahead process in your chart, it may not even be installed or they might be using something different in maverick. [20:17] void^: readahead != ureadahead? === C-S-B_ is now known as C-S-B [20:26] knittl: ureadahead is ubuntu's favourite readahead implementation [20:27] i see. but i can see it in my bootchart ^^ [20:27] somewhere. i'm pretty sure i did :D [20:28] and it's installed. do i have to do something to enable it? === gorgonizer_ is now known as gorgonizer [20:31] ah, right. there it is. seems to do awfully little, so there's probably something wrong with it. check the manpage, it explains how to force a retrace iirc. [20:31] hm [20:31] deleting packfiles [20:32] i'll try a reboot [20:49] another thing i notice: after logging in hdd is still reading for like 2 min [20:49] could this be related to ubuntuone? [20:51] and i get »ureadahead main process terminated with status 5« upon boot [20:51] in bootchart it's still only running for <1s [20:52] knittl: did you reboot again? maybe it was reprofiling this time [20:52] yofel: yes, i did the reboot twice [20:52] odd [20:53] immediately after hdd went silent (and blinking stopped) i hit reboot [20:53] there is like 6 min difference between chart generation time [20:53] and bootup time is the same (75s) [20:53] gdm login shows up pretty quick btw [20:54] well, thanks to upstart, some system services get started *after* gdm shows, so hdd activity after login is ok IMO (maybe use iotop to check) [20:55] but 2+ minutes? [20:55] and desktop is like frozen [20:55] if i move some window over the icons they will not show again for a long time [20:57] and mysql is not starting lately [20:57] :( [20:57] (yeah i know, maverick is not released yet …) [20:58] hm, you're not the only one with mysql issues iirc [20:58] manual start works [20:58] i.e. sudo service mysql start [20:58] funkyHat: ^ [20:58] yofel: but it could be related, that my databases are in /home/mysql [20:58] instead of /var/lib/mysql [20:59] and home is a separate partition [21:00] I now have no sound :( [21:00] (just updated) [21:00] Chipaca: use the volume control to turn it up [21:00] (or use alsamixer) [21:01] knittl: I have no alsa devices [21:01] Chipaca: hu? [21:01] knittl: in fact, no snd* modules loaded [21:02] interesting [21:02] not particularly, no [21:02] :) [21:03] so ok, i'll try commenting every line in fstab except /, /home and swap [21:03] but this shouldn't give problems … [21:10] sudo sed -i 's+^start on starting mountall+start on mounted MOUNTPOINT=/var+' /etc/init/ureadahead.conf [21:10] now this seems like a solution [21:18] yofel: so im here :P [21:18] hehe, welcome :P [21:19] * shadeslayer has come fully prepared.... [21:19] a helmet.. kevlar... riot shield... [21:19] seems im set [21:19] * yofel bombards shadeslayer with failsave-x [21:19] oh.. my katana [21:20] * shadeslayer slices failsafe-x with katana [21:20] btw [21:20] anyone with sendmail installed experience insanely slow boot times? [21:21] ok, let's reboot and have a look at iotop === shadeslayer is now known as Razac [21:25] and yeah, i was right [21:25] io nearly 100 % for: [21:25] python /usr/lib/ubuntuone-client/ubuntuone-syncdaemon === Razac is now known as shadeslayer [21:27] still syncdaemon running [21:28] * shadeslayer goes to remove sendmain [21:28] *mail [21:28] now desktop is working again [21:28] ubunut1 still 50+ % io [21:29] and ureadahead won't regenerate packfiles :( [21:29] ok, now syncdaemon is finished [21:29] 4 minutes? oO [21:34] hm. my first boot with maverick was only 33 s [21:34] now it's 80 [21:35] my last boot with lucid was 70 [21:35] i don't understand it … [21:36] on july 3rd boottimes went up to 100s (from 30s a day before) [21:45] uargs. i don't fucking know what "terminated with status 5" means. and everything that google tells me, is status 4 [21:46] stupid #$~|\*/ ureadahead [21:46] but at least io stays down after login now (thanks to disabling ubuntu1) [21:46] knittl: please keep the language clean [21:46] which is still running? wtf [21:46] IdleOne: sorry … [21:48] knittl, do you have var on separate partition ? [21:49] Dink: no [21:49] only /home [21:49] and i even disabled every other partition except /, /home and swap [21:53] you get anythign with ureadahead -v? [21:56] status 5 I think means no pack file... Anyways i am out... later and good luck [21:58] Dink: well, it should create a file [21:58] but it doesn't. it just terminates [22:02] and indeed, 5 means no packfile [22:03] but only when trace fails [22:03] i'd really prefer not reading the sourcecode for this xD [22:04] oh ok. i shouldn't delete this fake mountpoitn [22:04] raaah, rebooting [22:20] uh... [22:20] do i read the topic right? [22:20] Official Ubuntu Maverick support [22:20] Official ? [22:20] :) [22:35] since when?? [22:51] This is not an official support channel? [22:54] never knew it as such charlie-tca [22:54] devel is *totally* unsupported [22:54] so no "oficial" support can be provided [22:54] now this is strange/funny [22:54] what is ? [22:54] stock kernel won't boot anymore … [22:55] wfm [22:55] maybe it's nouveau related, i just removed nvidia-current [22:55] but nouveau has worked before [22:55] The following packages have unmet dependencies: [22:55] hplip: Depends: hplip-data (= 3.10.5-4ubuntu1) but 3.10.6-0ubuntu1 is to be installed. [22:55] libvlccore4: Depends: vlc-data (= 1.1.0-4ubuntu1) but 1.1.1-1+exp1ubuntu1 is to be installed. [22:55] this kinda worries me more [22:55] been stuck on that for a few days now [22:55] i didn't have that [22:55] and 'not booting' is not entirely corretc [22:56] but it stops before X shows up [22:56] otherwise i wouldn't be here obviously :D [22:56] and my custom kernel still refuses to mount my ext4 root partition [22:57] altough ext4 support is compiled in (not even a module) [22:57] any hints BUGabundo? [22:57] humm [22:57] .xessionserrors? [22:57] syslog? [22:57] xorg.log? [22:57] anything obvious? [22:57] no, why my custom kernel won't mount :D [22:57] i will try reinstalling nvidia-current and see if that works [22:58] hm, stock kernel does not even load alsa it seems. now this is peculiar [22:59] mocp: FATAL_ERROR: can't recieve volume from server [22:59] wth? [22:59] fsck it ? [23:00] no, my fs should be clean [23:00] i blame nouveau right now [23:00] i don't mind working in the console for a while [23:00] want to fix my custom kernel first :> [23:02] but also strange error (with custom and with stock kernel): address space collision: host bridge window conflicts with PCI mem [23:02] any idea? === Andre_Gondim_ is now known as Andre_Gondim [23:29] seems i can't boot with nouveau right now … [23:30] is installation of the "upstream" nvidia drivers supported in 10.10? Have they updated their installer? [23:30] when is ubuntu1 syncdaemon started? [23:30] or from where? [23:30] i removed it from startup applications, but it's still eating my hdd after login -.- [23:35] it's nice that exactly half of packages use -utils suffix and the other half -tools [23:36] nothing like guessing the right package name to make your day [23:38] kklimonda: [23:39] knittl: doesn't work when I write an answer to the post ;)