[12:12] hm, that looks slightly more reasonable now [12:12] mdz: successfully builded images for powerpc,amd64 and i386 [12:12] Kamion: you want urls, eh> === usual [~colin@alb-69-202-45-196.nycap.rr.com] has joined #ubuntu-devel [12:13] amu: by hand? [12:13] mdz: give http://cdimage.ubuntu.com/daily-live/20050107/hoary-live-i386.iso a try [12:13] mdz: by script [12:14] Kamion: can you make rsync:// work? [12:14] it does [12:14] @ERROR: Unknown module 'daily-live' [12:14] put /cdimage/ first [12:14] but there's a maxconn limit :( [12:14] @ERROR: max connections (15) reached - try again later [12:14] gah [12:15] you have an account on little, rsync over ssh :) [12:15] what's the local path? [12:16] mdz: /srv/cdimage.no-name-yet.com/www/full/daily-live/20050107 [12:16] ew. something bloated. [12:16] 637MB [12:16] Kamion: i386 is there [12:17] lamont: amd64? [12:17] amu: building [12:17] lamont: ok, but won't be able to do anything with it until tomorrow [12:18] lamont: the amd64 image is 100mb bigger compared to i386 [12:18] *** Error -5 compressing block 0! (compressed=0x7fbffff830, len=548682070016, uncompressed=0x526010, blocksize=65536) [12:18] ./livecd.sh: line 130: 1630 Segmentation fault create_compressed_fs $IMG 65536 >livecd.cloop [12:18] ok, I'm off for the night now ... [12:19] Kamion: n8 [12:19] total size is 547250176 speedup is 81.06 [12:19] not bad [12:19] Kamion: night [12:19] amu: I have a quest for you [12:19] amu: we need to find out what is doing so much writing during startup on the live CD [12:20] eek [12:20] Kamion: it's broke [12:20] mdz: i#ll look to it === Nicolas__ [~Nico@u81-11-166-150.adsl.scarlet.be] has joined #ubuntu-devel [12:20] amu: it uses like 15M of snapshot space just to boot [12:20] Kamion: could not find kernel image: linux [12:21] mdz: means memoryuseage is d-i + 15M ? [12:21] mdz: did you explicitly type 'linux' at the boot prompt? [12:21] amu: dmsetup status /dev/mapper/casper-snapshot [12:22] Kamion: I tried both pressing enter and typing 'linux' [12:22] bleh, ok, maybe isolinux defaults to linux or something [12:22] aum: note that the script has probably changed since you grabbed a copy... [12:22] what should the name be? [12:22] it's probably broken, let me attempt to fix/rebuild [12:23] mdz: which version we talking? Kamion's? your's? mine? [12:24] amu: they should all have the same characteristics [12:24] lamont: ok, still same place? [12:24] amu: run dmsetup status /dev/mapper/casper-snapshot [12:24] amu: not quite yet... [12:24] AARGH, MY CAPS LOCK OR SHIFT KEY OR SOMETHING IS STUCK [12:26] Kamion: if it's not trivial to fix, it can wait, don't lose sleep over it [12:26] mdz: try 20050107.1, see how it goes [12:27] Kamion: same result [12:28] erk, I was doing repeated >isolinux.cfg rather than >>isolinux.cfg [12:29] 20050107.2 will happen in a moment and should fix that problem; in the meantime bed :) [12:29] try that when it appears in a few minutes [12:30] it's up [12:31] oh! mirnyy has space now; I can round-robin cdimage too [12:31] [that'll help with the rsync limits] [12:32] rsyncing [12:32] elmo: oh, the limit is for the sake of server resources (rather than bandwidth)? [12:32] Kamion: booting now [12:33] mdz: totally resources - 25 was hitting the box so hard, colin couldn't sync a sounder image [12:34] my word [12:34] somebody needs to fix rsync === bronson [~bronson@node-40240852.sjc.onnet.us.uu.net] has joined #ubuntu-devel === bronson [~bronson@node-40240852.sjc.onnet.us.uu.net] has left #ubuntu-devel ["Client] === zenrox [~zenrox@wbar7.sea1-4-10-181-189.sea1.dsl-verizon.net] has joined #ubuntu-devel [12:38] + create_compressed_fs livecd.fsimg 65536 [12:38] Opening input: File too large [12:38] Hrmpf. [12:38] mdz: so how do you feel about just making the file system an even maxint in size? [12:39] lamont: that's fine, but we ought to fix create_compressed_fs [12:39] should be trivial [12:39] yeah [12:39] elmo about? [12:40] Kamion: it's gotten as far as starting sysvinit; looks good, thanks [12:40] yeah [12:41] mdz: this pass, I'll just hardcode the size, then I'll fix cloop [12:41] lamont: can we quash that scary postfix inode warning? [12:42] mdz: paste it to me [12:42] can't paste from qemu, bu tit goes like this: [12:42] QUEUE FILES WERE RENAMED TO MATCH INODE NUMBERS OH MY GOD [12:42] THE GODDAMN SKY IS FALLING [12:42] this on livecd? [12:42] something like that [12:42] yes [12:43] there _shouldn't_ be any queue files... === lamont goes to look [12:43] probably got some from debconf or whatever [12:43] daniels: on the X40 does s4_bios need a special partition creating? [12:43] also [12:43] lamont: sources.list still says 'echo' [12:43] oh, nm [12:43] this is the same image [12:45] 3 messages sent during the install. I'll make em go away [12:45] hmm, what would be the best way to stop cron.daily from running on the live CD, I wonder [12:45] as nice as it is to have an up-to-date locate db, it's not very convenient to have it start up soon after boot... [12:46] I suppose I should just not let it start up [12:47] lamont: hmm, /etc/init.d/dbus-1 is missing [12:48] lamont: are you keeping a list? [12:48] and mkinitrd? [12:49] and sbin/udevd [12:49] udevd is OK [12:49] is this in the latest image, or the one from lunch? [12:50] verified OK in the current image [12:50] diverts were kinda borked at lunch, like I said [12:51] DIVERTS="usr/sbin/mkinitrd usr/sbin/invoke-rc.d etc/init.d/dbus-1 sbin/udevd" [12:51] but udevd came in later, and invoke-rc.d apparently got better somehow... [12:51] any changes to /etc/kernel-img.conf? === doko_ [doko@dsl-082-082-188-062.arcor-ip.net] has joined #ubuntu-devel [12:54] lamont: in both [12:55] colin used the same fs image from lunch to build the test CD === ogra [~ogra@p508EB45C.dip.t-dialin.net] has joined #ubuntu-devel [12:58] if you want etc/cron.daily/man-db to go away, that can be arranged... [12:59] does it get ran at isntall time? [01:00] nah, if I just don't start anacron, it fixes all that === lamont throws another round of builds into the fire, with postfix queues flushed, but cron.daily/man-db intact [01:00] cron suppresses the daily stuff when it sees anacron is installed [01:01] lamont: you know what would be nice [01:01] lamont: if at the end of the build, you ran cron.daily/man-db and cron.daily/slocate [01:01] brb, reboot [01:05] night all [01:06] night elmo [01:08] night elmo [01:09] night === sdfsdfs [sdfsdfs@tkp-ip-nas-1-p56.telkom-ipnet.co.za] has joined #ubuntu-devel === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel === bronson [~bronson@node-40240852.sjc.onnet.us.uu.net] has joined #ubuntu-devel === mirak [~mirak@AAubervilliers-152-1-52-19.w82-121.abo.wanadoo.fr] has joined #ubuntu-devel [01:25] pitti around? === robertj [~robertj@66-188-77-153.cpe.ga.charter.com] has joined #ubuntu-devel [01:27] gah, my panel is fucked [01:27] I have Blank Panel Syndrome [01:27] mdz: BPS [01:27] it's kinda refreshing though [01:27] minimalism is in [01:28] does anyone know the workaround for the panel-bug-o-day? [01:28] dunno, about to find out ;) [01:29] 1m40s remaining on my download [01:29] mdz: I found a reliable sudo pkill -u mdz [01:29] seb128: awake? [01:29] mdz: and log back in [01:30] mdz: there's a race in there somewhere, but I couldn't find it === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel [01:30] Keybuk: yep === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel [01:30] seb128: http://people.ubuntu.com/~scott/nautilus-2.9.1_keybuk-places-bookmarks.patch [01:31] thanks [01:31] Keybuk: cool, thanks [01:32] seb128: my mixer applet is showing the red crossed-out icon even when it is unmuted [01:32] Keybuk: that's a bit late to read it now, but I'll try it tomorrow and let you know [01:32] mdz: do you have a way to reproduce it ? [01:32] seb128: it is happening right now; I have never seen it before [01:33] seb128: I lowered the volume to zero and brought it back up, and now the icon is normal [01:33] mdz: etc/cron.daily/man-db is _scary_ [01:33] seb128: but when I mute, it still doesn't change [01:33] how do you mute ? keybinding ? [01:34] context menu [01:34] right, same bug here [01:36] seb128: the panel startup problem, is it #4794? [01:36] it just happened to me on the first login after a reboot, so no processes running [01:36] mdz: man-db starts background processes (why???), and that causes a race condition that I'm not going near. [01:36] startup problem ? [01:36] freeze you mean ? nautilus and blank panels ? [01:36] seb128: I logged in, and got empty panels [01:36] right [01:36] yeah [01:36] that's it [01:37] lamont: background processes? where? [01:37] seems to be a deadlock in gnome-vfs-daemon [01:37] cat /etc/cron.daily/man-db [01:37] I've not had it since mataro [01:37] lamont: I did [01:37] start-stop-daemon --startas mandb [01:37] etc [01:37] start-stop-daemon doesn't fork unless you give it -b [01:37] I thought it was fixed with the hal support off in gnome-vfs, but apparently not :/ [01:37] oh. that's better then. [01:37] seb128: are you able to reproduce it? [01:37] no [01:37] is there anything I can do? [01:38] I've not had it since mataro [01:38] (if I see it again) [01:38] while we're on the subject; gnome-battery-applet blinking whilst charging is annoying and distracting [01:38] seb128: yes after a reboot [01:38] seb128: I just rebooted and it's happened [01:39] however it didn't happened5 days ago when I booted the computer [01:39] rebuilt a non-stripped libgnomevfs2-common (yeah, the package split is borked, I wonder why a -common has a server part) [01:39] and get a backtrace with symboles of the lock [01:40] to compare with the previous one in #4794 [01:40] which was with the hal support [01:40] are you planning to upload gnomevfs sometime soon? [01:40] if not, I will build it for debug now [01:40] no, not in the next days [01:40] you could upload an unstripped version to the archive so that everyone can debug it :-) [01:41] yeah, I want to fix that stupid "-common contains the server" stuff === bronson [~bronson@node-40240852.sjc.onnet.us.uu.net] has joined #ubuntu-devel [01:42] -common should be an arch all, and the server should be in a -bin or with the lib [01:42] not with the lib [01:42] unless it's versioned with it? [01:43] no [01:43] it should be in a -bin in fact [01:43] with the gnomevfs-ls/cat/info/copy/... === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel [01:57] seb? === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel [01:59] lamont: include the full nickcname to highlight :p === mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #ubuntu-devel [02:01] yeah, sorry [02:02] mdz: you disagree with the place menu not working as a target to fix for Hoary ? [02:03] (#5123) [02:03] mdz: /usr/lib/prebaseconfig.d writes a lot and takes the long time .... [02:03] mdz: pong [02:05] jdub: gpdf will probably get type3 fonts support in the next days (that works with evince and will be backported to gpdf), I think we should set it as the default pdf viewer for hoary :) [02:05] seb128: agree. [02:06] seb128: I assumed the submitter had set it [02:06] seb128: since it was also BLOCKER [02:07] jdub: status of adding upgrade-notifier to the default session? [02:07] amu: which bit? [02:07] mdz: I've set both. Blocker is not a "has to be fixed" ? [02:07] mdz: imho this one has to be fixed for hoary, having place not working sucks [02:07] seb128: blocker is WAIT DON'T RELEASE YET [02:07] seb128: I agree, it ought to be fixed [02:08] but I don't see much point in setting blocker severity on bugs when we are still far from release [02:08] is blocker appropriate or not ? just to know for the next time :) [02:08] mdz: that's one for seb [02:08] seb128: does it make it easier for you to track your bugs? [02:08] yep [02:08] if so, go ahead [02:08] ok, thanks [02:08] often people file random bugs with severity=blocker, and they show up in my global reports [02:09] yeah, I know :) [02:09] seb128: can you add upgrade-notifier to the default session? [02:09] I should have added a comment, I've thinked about that after clicking on the button :p [02:09] mdz: ok [02:10] seb128: do you add a dependency when you do that, or should I change the seed? [02:10] (for upgrade-notifier) [02:10] er [02:10] update-notifier [02:10] we have both in the archive right now, ick [02:10] change the seed [02:10] and update-notifier is in universe, blecch [02:10] session start it if available [02:10] I don't need to add a depends [02:11] jdub: update-manager ready for desktop as well? [02:11] seb128: ok, thanks [02:11] np [02:14] time to sleep, 'night guys [02:17] mdz: no, mvo and i have to talk about those before they're added to the seed [02:17] ok [02:19] GAH [02:19] arch_commit: unable to acquire revision lock (internal error in archive-pfs.c(pfs_lock_revision)) [02:19] this is getting so old === stub [~stub@dsl-246.248.240.220.dsl.comindico.com.au] has joined #ubuntu-devel [02:27] mdz: didnt found it out, i've also some problems after using some programs, the snaphot becomes invalid, stauts is 0 3965184 snapshot 16904/20480 === sdfsdfs [sdfsdfs@tkp-ip-nas-1-p56.telkom-ipnet.co.za] has left #ubuntu-devel [] === nicedreams [~nicedream@ip68-2-79-215.ph.ph.cox.net] has joined #ubuntu-devel === nicedreams [~nicedream@ip68-2-79-215.ph.ph.cox.net] has left #ubuntu-devel [] [02:30] mdz: 16904 grows up to 20480, after it reaches 20480, i cannot run other programs === amu moves to bed [02:34] amu: sleep well [02:34] amu: invalid means you filled it up [02:34] after which it won't work anymore [02:36] did't catched it, if it goes near to 20480 [02:37] azeem: hi [02:38] amu: you can boot with ramdisk_size=65536 to get more space [02:39] sure [02:39] but I think we can get it down to a more reasonable level with some tweaks [02:42] mdz/Kamion: i386: 526249079; powerpc: 567143169; amd64: waiting for kernel love on the buildd machine; ia64: broken depends [02:42] lamont: kernel love? [02:42] lacks loopback support [02:42] lamont: did elmo not compile the loop module or something? :-P [02:42] haha [02:42] modules, what modules. this is an elmo kernel [02:42] CONFIG_ELMO=y [02:42] will have loopback support tomorrow london time [02:47] lamont: could you run updatedb before clooping the chroot [02:50] amu: we run /etc/cron.daily/{man-db,slocate} [02:50] amu: aha [02:50] amu: for some reason the noatime is not taking effect [02:50] or it is being removed by the mount --move [02:50] that will cause a huge number of writes === salgado [~salgado@200-158-222-150.dsl.telesp.net.br] has joined #ubuntu-devel [02:52] or maybe busybox mount doesn't even support -o properly? [02:52] mdz: hppa: ubuntu-desktop not installable (well, has no depends); sparc: unknown [02:52] mdz: it's better, i run just a tty, guess i've to setup X :) [02:53] hmm, no, busybox mount seems ok [02:53] amu: mount -o remount,noatime /dev/mapper/casper-snapshot / [02:53] should save a lot [02:54] mdz: as in the friendly ghost? [02:54] lamont: yes [02:54] or the dopest ghost in town [02:55] I thought he turned back into a dead little boy. :-) === stub [~stub@dsl-246.248.240.220.dsl.comindico.com.au] has joined #ubuntu-devel [02:57] mdz: yeah it save a lot, .... still trying to configure X [02:58] I've got a crippled hedgehog. My background and my panels have not loaded (the panels are there, but blank). Is there a fix for this? [02:58] I don't know where it loses noatime though [02:59] I tested a mount --move on my desktop, and it preserved noatime [03:00] maybe it has to do with mtab [03:00] I've build evolution with debugging symbols (as file /usr/bin/evolution confirms me that), but when I run it in gdb, I can't get those symbols. [03:00] I've built evolution according to http://www.ubuntulinux.org/wiki/DebuggingProgramCrash. is there something I may be missing? [03:02] nope, tested with mount -n [03:02] wtf [03:02] maybe the pivot changes it? [03:02] mdz: did you want those cloop's where you could fetch them? [03:02] lamont: what's responsible for writing the entry for root into mtab? [03:03] lamont: sure, I'll download them so I can rsync the full images faster tomorrow [03:03] I think mount does it, but have'nt had to look before.. [03:03] it's one of the early init scripts [03:04] checkroot.sh [03:04] salgado: those instructions should do it [03:05] salgado: if file /usr/bin/evolution says it's not stripped, then you did the right thing [03:05] salgado: can you paste the gdb output to me in /msg ? [03:05] mdz: sure [03:06] mdz: http://people.ubuntu.com/~lamont/livecd/livecd-20050108-{i386,powerpc}.cloop [03:07] mdz: how often do you want them built, btw? [03:07] lamont: daily is fine [03:07] lamont: daily, in time for the daily CD run [03:09] 0615 is when the d-i build runs, sounds like about a good time. [03:09] mdz: i'll play little more tomorrow with it, too late now [03:09] amu: ok, good night [03:09] amu: tomorrow, see if you can find out where the noatime goes away [03:09] so we can fix that [03:10] mdz: last suggestion, what's about a swapfile on the cd, i've a lot ram, probalby other user have not too much ... [03:11] mdz: where do you put your last d-i? i'll try to build also homeeditions iso's ;) === usual [~colin@alb-69-202-45-196.nycap.rr.com] has joined #ubuntu-devel [03:12] mdz: k [03:12] amu: on the cd? [03:12] a swap file on read-only media is hard to use :-) [03:12] I do think we should detect and use swap on the hard disk [03:13] mdz: certainly [03:13] mdz: fallback to RAM drive if there's not enough space? [03:14] n8 -R * [03:14] sivang: very funny [03:14] amu: I am using the daily d-i build [03:14] no need to build it here anymore [03:15] mdz: so daily install cd, with a bit of hackification? [03:15] lamont: what I'm using right now is the CD Colin built, in fact [03:15] unmodified === lamont needs a 'so you want to build a home-edition livecd' cookbook [03:15] it's way easy [03:16] all you need to do is put the cloop image in /casper/filesystem.cloop === srbaker [~srbaker@blk-224-172-150.eastlink.ca] has joined #ubuntu-devel [03:16] and pass casper/enable=true on the kernel command line [03:16] boom [03:16] (assuming you have a full complement of udebs, including casper-udeb, on the CD) [03:17] so you could start from an install CD, remove all the .debs, drop in a filesyste image, tweak isolinux, and you're done [03:17] /casper/filesystem.cloop on the CD, yes? [03:17] or I could fetch hoary-live-i386.iso and use that, eh? [03:18] lamont: oh, we're going to want pcmcia-cs in the cloop image [03:18] if I tweak filesystem.cloop in the iso, are there any checksum files to fix? [03:18] lamont: nope === lamont sighs, adds pcmcia-cs. anything else? [03:18] if you're starting from a live CD, you can swap in any filesystem you want [03:19] lamont: actually, hmm === lamont contemplates putting nethack on the system... [03:19] the reason we don't install pcmcia-cs by default is that it breaks some machines [03:19] and frozen-bubble/kobodl, of course. [03:19] we can arrange for it to be installed if needed, of course... [03:19] but that wastes valuable COW space [03:20] I guess the udeb isn't enough? [03:20] don't worry about it just yet [03:20] I need to consult with colin on the pcmcia-cs situation [03:20] sladen: i don't know, sorry [03:21] daniels: hey, your magic xkb dvorak recipe stopped working [03:21] daniels: your xkb hero status is in jeopardy [03:21] mdz: so us(pc104)+dvorak(basic) as the symbol set is no longer any good? what happened? [03:22] daniels: dunno, did anything change xkb-wise in that last X upload? [03:22] (correct, it no longer does what I want) [03:22] mdz: yeah, a little, but nothing I would think would be relevant [03:22] (i.e., give me an Alt_R) [03:22] daniels: it should be setxkbmap -keymap 'us(pc104)+dvorak(basic)', right? [03:23] I thought it was just setxkbmap 'us...dvorak...' [03:23] but that doesn't work, so I assumed I had done it differently before [03:23] er, -symbols 'us(pc104)+dvorak(basic)', i think [03:23] mizar:[~/data/src/canonical/casper] setxkbmap 'us(pc104)+dvorak(basic)' [03:23] Error loading new keyboard description [03:23] hmm [03:23] ahh, ok, maybe that was it [03:23] daniels: what do I put in xorg.conf? [03:24] rtw c br, e.ucbcyc.nf dak. ekrpat [03:24] mdz: just put us(pc104)+dvorak(basic) in xorg.conf for XkbSymbols [03:24] mdz: and if you're going to keep breaking dvorak, you need to teach it to me [03:25] daniels: switch your keycaps around [03:25] you probably don't need to look at the keycaps for qwerty anyway [03:25] mdz: well, good point [03:25] (btw, the above random garbled text was something like, 'yeah, switching into dvorak works') [03:25] daniels: when I was learning, I would leave an image viewer on sticky with an image of the layout [03:25] mdz: heh :) [03:26] mdz: that would work nicely on my crt (21", 1600x1200), but maybe not so much on the laptop [03:26] daniels: should I delete my XkbModel/XkbLayout when I add XkbSymbols? [03:27] mdz: think it should be fine as is; what is it now? [03:27] Option "XkbRules" "xorg" [03:27] Option "XkbModel" "pc104" [03:27] Option "XkbLayout" "dvorak" [03:28] ah, I see [03:28] so change XkbLayout to "us(pc104)+dvorak(basic)" [03:28] and leave model and rules alone [03:29] i'm not entirely sure it's the most elegant, but it should work [03:32] mdz: ;) [03:35] uh. if i'm going to write a gnome bittorrent gui, as listed on the Ubuntu Bounties page, where do i find out the person to contact? [03:37] I heard the latest bittorenr GUI is a much improved GTK one, so you might want to check that first (I don't use bittorent) [03:38] daniels: but setxkbmap -layout 'us(pc104)+dvorak(basic)' doesn't work [03:38] but XkbLayout will? [03:38] srbaker: that would be me [03:39] mdz, ah. is there a list of requirements? [03:39] mdz, i'm curious what the requirements are, and what the value is. [03:39] srbaker: yes, there was a discussion on the mailing list where we worked out the requirements [03:39] mdz, i'm trying to schedule it in. [03:39] oh, okay. ML archives then. [03:39] mdz: must be XkbSymbols, then [03:39] srbaker: though, someone mentioned on the wiki that bittorrent upstream seems to be developing a pygtk gui [03:39] so there may be a starting point there [03:39] mdz: try XkbLayout "us" XkbSymbols "dvorak(basic)" [03:39] daniels: and leave everything else in place? [03:39] yeah. [03:40] mdz, well, my plan was something like azureus, only not so heft. [03:40] hefty [03:40] mdz, i'm trying to schedule as many of these bounties as i can in between client obligations [03:40] daniels: "setxkbmap us; setxkbmap -symbols 'dvorak(basic)'" breaks my alt keys [03:41] srbaker: what other projects are you interested in working on? [03:41] srbaker: feel free to email me proposals [03:43] mdz: ok, so try setting (in the config -- you want to start another X server) XkbLayout "dvorak(basic)", and XkbSymbols "us" [03:44] mdz: (if you leave XkbLayout as "dvorak" and have XkbSymbols as "us(pc104)+dvorak(basic)", it should work, but I'm looking for something slightly more elegant [03:48] I'll play with it later [03:49] cool [03:49] sorry I couldn't be of more help; my grasp of XKB is kind of handwavey [03:49]