WebHosting Paid by #1Payday.Loans
[03:19] blindcod1r (i=blindcod@gateway/gpg-tor/key-0x8997BE0F) joined #rocklinux. [03:31] brian|lfs (n=Brian@216-15-45-30.c3-0.161-ubr1.lnh-161.md.cable.rcn.com) joined #rocklinux. [03:36] kasc_ (n=kasc@dslb-084-060-109-065.pools.arcor-ip.net) joined #rocklinux. [03:39] blindcoder (i=blindcod@gateway/gpg-tor/key-0x8997BE0F) left irc: Remote closed the connection [03:43] kasc (n=kasc@dslb-084-060-117-164.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [03:43] Nick change: kasc_ -> kasc [03:57] stf^rock1inux (n=user@M1182P006.adsl.highway.telekom.at) joined #rocklinux. [04:10] stf^rocklinux (n=user@M1227P007.adsl.highway.telekom.at) left irc: Read error: 110 (Connection timed out) [04:36] blindcod1r (i=blindcod@gateway/gpg-tor/key-0x8997BE0F) left irc: Remote closed the connection [04:38] blindcoder (i=blindcod@gateway/gpg-tor/key-0x8997BE0F) joined #rocklinux. [04:54] Brian_L (n=brian@ip24-254-0-84.rn.hr.cox.net) joined #rocklinux. [05:01] brian|lfs (n=Brian@216-15-45-30.c3-0.161-ubr1.lnh-161.md.cable.rcn.com) left irc: "Leaving" [06:25] SMP (n=stefanp@vanessa.wronline.net) left irc: Read error: 113 (No route to host) [06:35] Brian_L (n=brian@ip24-254-0-84.rn.hr.cox.net) left irc: "Leaving" [06:42] <blindcoder> moin [06:53] <blindcoder> ARGH! [06:53] <blindcoder> bdb 4.5 [07:19] rocklogbot (i=blindcod@pallas.crash-override.net) left irc: Remote closed the connection [07:21] rocklogbot (i=blindcod@pallas.crash-override.net) joined #rocklinux. [08:20] rocklogbot (i=blindcod@pallas.crash-override.net) left irc: Remote closed the connection [08:20] rocklogbot (i=blindcod@pallas.crash-override.net) joined #rocklinux. [08:27] SMP (n=stefanp@vanessa.wronline.net) joined #rocklinux. [09:44] SMP (n=stefanp@vanessa.wronline.net) left irc: Read error: 113 (No route to host) [10:27] daja77_ (n=daja77@85-124-206-39.dynamic.xdsl-line.inode.at) joined #rocklinux. [10:29] daja77 (n=daja@85-124-93-158.zollergasse.xdsl-line.inode.at) left irc: "es kann nur einen geben" [10:32] Nick change: daja77_ -> daja77 [10:49] netrunner (n=andreas@anvame.net) left irc: Read error: 110 (Connection timed out) [10:50] netrunner (n=andreas@anvame.net) joined #rocklinux. [11:14] <stf^rock1inux> moin [11:55] <th> daja77: do you plan a openssh update soon? [11:56] <daja77> should i? [11:57] <mnemoc> 0.9.8d [11:57] <daja77> ssh too? [11:57] <mnemoc> CVE-2006-2937 CVE-2006-2940 [11:57] <daja77> this is ssl [11:58] <mnemoc> oh, [11:58] Action: mnemoc steps back [11:59] <daja77> i committed that patch today ;) [11:59] <mnemoc> :) [12:01] <daja77> ok i should [12:01] <daja77> th: is this a security thing or just a normal update? [12:02] <th> security [12:02] <daja77> hm they mention security fixes in the release notes [12:02] <daja77> ok i'll do it [12:03] stf^rocklinux (n=user@M1268P006.adsl.highway.telekom.at) joined #rocklinux. [12:06] <netrunner> daja77: what is the openssl-daemon? [12:06] stf^rock1inux (n=user@M1182P006.adsl.highway.telekom.at) left irc: Read error: 145 (Connection timed out) [12:06] <daja77> ? [12:09] <netrunner> daja77: you wrote one should restart the daemon after the upgrade [12:10] <daja77> sigh yeah i saw it [12:10] <daja77> the template was the bind mail [12:10] <daja77> >_< [12:10] <netrunner> :) [12:10] <netrunner> actually it is corrrect, you should restart all daemons that are using ssl [12:13] <daja77> .oO (good to know that someone is actually reading that) [12:17] <netrunner> :) [12:41] <netrunner> daja77: you forgot to change the subject line :) you should use an empty template with FILLME tags :) [12:44] <daja77> i am sleepy .. [12:44] <daja77> but right i need a real template somewhere [12:50] <stf^rocklinux> th: I tried your irfs yesterday. Works quite well :) [12:51] <stf^rocklinux> th: I've also added rootdir support, so it can build initramfs for other architectures. [12:51] <stf^rocklinux> th: maybe you want to add it as a ROCK package so I can start sending patches? :) [12:51] <blindcoder> I guess we should finish started work first [12:52] <blindcoder> like /etc/net [12:52] <stf^rocklinux> ok, so what needs to be done for etcnet? An init script? [12:53] <blindcoder> init script and a stone plugin [12:53] <blindcoder> but it brings its own init script anyway [12:53] <blindcoder> I guess we can use that [12:53] <blindcoder> how is the etcnet package in pilots repository coming along? [12:54] <stf^rocklinux> afaik it basically works (e.g. DHCP) [12:54] <blindcoder> great [12:54] <blindcoder> is it CORE flagged? [12:55] <blindcoder> because the rocknet package isn't [12:55] <stf^rocklinux> not yet [12:55] <blindcoder> I see [12:55] <stf^rocklinux> oh, and etcnet expects ifrename in /usr/sbin/ as a temporary workaround [12:56] <blindcoder> hmm [12:56] <stf^rocklinux> when you have ifrename in /sbin, you have to remove the workaround from the ROCK-specific config of etcnet [12:56] <blindcoder> is Pilot going to make it arbitrary? or de we have to create a patch? [12:57] <stf^rocklinux> etcnet-0.8.5/contrib/50-RockLinux-3: [12:57] <stf^rocklinux> # Temp hack [12:57] <stf^rocklinux> IFRENAME=/usr/sbin/ifrename [12:57] <blindcoder> oh [12:57] <blindcoder> so it's hacked to expect ifrename in /usr/sbin? [12:57] <blindcoder> which means the wireless-tools package is wrong? [12:58] <stf^rocklinux> yes, but there's a wireless-tools patch by now. [12:58] <blindcoder> great :) [12:59] <blindcoder> ah, I already have it in my journal [13:02] <blindcoder> Installing fontconfig postinstall Script ... [13:02] <blindcoder> cp: cannot stat `/ROCK/package/x11/fontconfig/x11-30-fontconfig-postinstall.sh': No such file or directory [13:03] <blindcoder> stf^rocklinux: should the file be called x11-30-postinstall.sh or x11-30-fontconfig-postinstall.sh ? [13:03] <blindcoder> I assume x11-30-fontconfig-postinstall.sh [13:03] <stf^rocklinux> blindcoder: the latter. Did I get the name wrong? :/ [13:03] <blindcoder> yes :) [13:03] <stf^rocklinux> how is the postinstall in fontconfig called? [13:04] SMP (n=stefanp@vanessa.wronline.net) joined #rocklinux. [13:04] <stf^rocklinux> I'll rediff the patch... [13:05] <blindcoder> the actual filename is x11-30-postinstall.sh [13:05] <blindcoder> the file it tries to install is x11-30-fontconfig-postinstall.sh [13:05] <blindcoder> I now did a svn mv x11-30-postinstall.sh x11-30-fontconfig-postinstall.sh in my tree [13:05] <stf^rocklinux> ok, I'll fix the patch [13:08] <stf^rocklinux> blindcoder: 2006092913071523974 [13:10] <blindcoder> stf^rocklinux: that was not the patch I was talking about [13:10] <blindcoder> err [13:10] <blindcoder> yes [13:10] <blindcoder> it was [13:10] <blindcoder> blind, you know :) [13:11] <stf^rocklinux> ^^ [13:18] <blindcoder> == 09/29/06 13:17:32 =[5]=> Finished building package fontconfig. [13:30] <th> 12:51:19 < stf^rocklinux> th: maybe you want to add it as a ROCK package so I can start sending patches? :) [13:30] <th> will do [13:31] <th> stf^rocklinux: i can give you access to irfs on my personal svn if you like [13:32] <blindcoder> th: already got a plugin infrastructure built up for it? [13:44] <th> blindcoder: since the beginning [13:44] <th> blindcoder: it is a build.d/ directory [13:49] <blindcoder> so still not a full installable version? [13:50] <blindcoder> as in /sbin/mkinitramfs /etc/initramfs and so on? [13:52] <stf^rocklinux> th: I'll wait for the package for now [13:56] <th> 13:49:24 < blindcoder> so still not a full installable version? [13:56] <th> what you mean by that? [13:57] <th> blindcoder: since the beginning it includes a mkinitramfs.sh script which should be installed to /sbin or something [13:57] <th> blindcoder: "so still not a": it _never_ hat NO such script. did you ever look at it? [14:10] <stf^rocklinux> th: the build.d location has to be changed, though [14:10] <stf^rocklinux> for an installable package, I mean [14:12] <th> stf^rocklinux: propably [14:30] <blindcoder> that's exactly what I mean [14:31] <blindcoder> when I looked at it I found no way it could be used outside of a directory structure like this: [14:31] <blindcoder> /home/root/initramfs/mkinitramfs.sh [14:31] <blindcoder> /home/root/initramfs/build.d/ [14:31] <blindcoder> etc [14:31] <blindcoder> maybe I'll see it when there actually is a ROCK package for it [14:34] <blindcoder> anyway, feierabend [14:50] <th> blindcoder: so the very big issue is to move the build.d directory to another place when installing the package rockirfs? [15:16] rocklogbot (i=blindcod@pallas.crash-override.net) left irc: Remote closed the connection [15:16] rocklogbot (i=blindcod@pallas.crash-override.net) joined #rocklinux. [15:31] <blindcoder> th: seems like it is [15:31] <blindcoder> th: and maybe call it rockinitramfs or I'll always think of infrared stuff [15:32] <blindcoder> th: I'll have a closer look when the rock package is there [15:32] <blindcoder> th: and please write some documentation how to put plugins into it [15:35] <th> blindcoder: of course the documentation exists since the beginning in build.d/README [15:37] <blindcoder> so if everything has been there "since the beginning" [15:37] <blindcoder> where's the ROCK package? [15:41] <th> it does not exist. i wanted some rock-feedback first [15:41] <th> stf's was the first today. [15:41] <blindcoder> well, hard to get any when only saying 'it exists' on IRC every now and then [15:42] <blindcoder> besides, I've been talking to SMP about the initrd-vs-initramfs thing [15:42] <blindcoder> as I understood him, changing the mkfs call to the appropriate mkinitramfs call in the rockinitrd packgae would get us a nice initramfs, too [15:48] <blindcoder> https://pallas.crash-override.net/~blindcoder/rocklinux_20060924.txt [15:48] <blindcoder> search for initramfs [15:48] <blindcoder> the third or fourth occurance should be it [15:48] <blindcoder> 9:03 <SMP> no, initramfs is just a cpio. no /dev magic [15:48] <blindcoder> 19:03 <SMP> no, initramfs is just a cpio. no /dev magic [15:48] <blindcoder> starting there [16:01] <th> blindcoder: i dont only say it exits - i always said where to get a tar to test it [16:02] <th> blindcoder: where did smp say that replacing the call in rockinitrd would give a nice irf? [16:04] <blindcoder> 19:13 <blindcod1r> so we'd still need the whole /etc/conf/initrd/initrd_* infrastructure, a custom /sbin/init for the initramfs and stuff? [16:04] <blindcoder> 19:14 <SMP> the content is more or less the same as with an initrd [16:04] <th> the content. yes. he says that the resulting content of what rockinitrd produces and what rockirfs produces is similar [16:05] <th> blindcoder: where did smp say that replacing the call in rockinitrd would give a nice irf? [16:05] <blindcoder> 19:13 <blindcod1r> so we'd still need the whole /etc/conf/initrd/initrd_* infrastructure, a custom /sbin/init for the initramfs and stuff? [16:05] <blindcoder> 19:14 <SMP> the content is more or less the same as with an initrd [16:07] <th> he is talking of the resulting content. not of anything which relates to "< blindcoder> as I understood him, changing the mkfs call to the appropriate mkinitramfs call in the rockinitrd packgae would get us a nice initramfs, too" [16:07] <th> changing mkfs-call to mkinitramfs call is wrong. cause mkfs-call is part of rockinitrd as cpio-creating is part of mkinitrfs call. [16:08] <th> and calling mkinitramfs call in rockinitrd is like calling rockinitrd in rockinitrd [16:08] <blindcoder> "changing the mkfs call to the appropriate mkinitramfs call" [16:08] <blindcoder> whatever it is that has to be called in mkinitramfs [16:09] <th> okay - this might work somehow but is inefficient. [16:10] <blindcoder> how so? [16:11] <th> you are currently doing cp-a from prepared directory on a loopmounted newly created fs. [16:11] <th> the difference is: [16:11] <blindcoder> the loop-mount won't be necessary anymore, of course [16:11] <th> the easy and suggested way to create a cpio is to just pipe a list of files to a gen_cpio program [16:12] <th> the gen_cpio program creates the cpio file [16:12] <th> no need for creating any filesystem [16:12] <th> no need for doing any losetup [16:12] <th> no need for root privs [16:12] <blindcoder> of course not [16:12] <th> no need for preparing a directory [16:12] <blindcoder> well, you still need root privileges [16:12] <blindcoder> or you won't get the stuff into /boot [16:12] <blindcoder> I'd be worried about every user being able to manipulate my boot environment [16:13] <th> the package creates the cpio - you need no privs for this creation [16:13] <th> as you need no privs to build a kernel [16:13] <th> of course if you want to install a kernel or a cpio to make your system boot... then you need privs [16:13] <th> you would need no privs if your system is just a qemu or something [16:13] <blindcoder> well, that's the point (usually) [16:14] <blindcoder> so we should create a possibility to create the initramfs where it belongs with a single command [16:14] <blindcoder> possibly check for userid in the mkinitramfs.sh script [16:14] <blindcoder> then stuff the result where it belongs [16:14] <blindcoder> so it would work like this if I understand you right: [16:15] <blindcoder> cat /etc/conf/initrd/initrd_* | gen_cpio [16:15] <blindcoder> with correct layout of the /etc/conf/initrd/initrd_* files? [16:16] <th> blindcoder: yes. [16:16] <th> blindcoder: the order of the files is importat [16:16] <blindcoder> I see [16:16] <blindcoder> why is the order important? [16:16] <th> blindcoder: that's why there are numbers in the build.d/ dir [16:17] <th> blindcoder: it's important because at extraction time a parent directory for a file must exist prior to it's (the files) extraction [16:17] <blindcoder> sounds reasonable [16:17] <th> it does. (since the beginning) ;-] [16:18] <blindcoder> what if a directory/file is presented twice? [16:18] <blindcoder> well, you never explained it to me like this. or I never got it. [16:18] <th> then an error occurs at extraction time because the second file cant be extracted [16:18] <blindcoder> I guess getting the order right won't be a problem [16:19] <blindcoder> cat | sort | uniq | (shell magic that sorts all the directories at the beginning and the files at the end) [16:19] <th> blindcoder: no - i just created a rockmkirfs package-pilot [16:19] <th> blindcoder: everyone interested could inspect it. like SMP and stf did [16:19] <blindcoder> I did have a look at it back then [16:19] <blindcoder> I just didn't understand it [16:20] <th> for x in build.d/[0-9][0-9]* [16:20] <th> do [16:20] <th> echo [16:20] <th> . $x [16:20] <th> done | sort -u [16:20] <th> that's where your "cat | sort | uniq" is implemented [16:20] <blindcoder> I wouldn't assume correctness of the files [16:20] <blindcoder> but rather prepare for a few problems [16:21] <blindcoder> let me just hack a shellscript up [16:21] <th> "(shell magic that sorts all the directories at the beginning and the files at the end)" is not necessary when build.d/* contains numbers and correct order [16:22] <blindcoder> exactly [16:22] <blindcoder> never rely on correct order [16:22] <blindcoder> there _will_ be people doing it wrong [16:22] <th> blindcoder: hack away - but be aware that the infrastructure you're going to create already exists [16:22] <blindcoder> and I really don't have the nerve to correct them again and again [16:23] <blindcoder> so I'd rather have it work automatically [16:23] <th> blindcoder: if you start to write "intelligent" software which helps dump people to fix some of their problems.. then you end up like "karl klammer" [16:23] <th> i propose to not correct but fail. [16:23] <blindcoder> well, ROCK is intelligent software, too [16:23] <th> i propose to not correct but check and fail. [16:23] <blindcoder> you could always do the work manually... [16:23] <th> automation is no intelligence [16:24] <th> intelligence comes in if software tries to guess user behaviour and fix it. [16:24] <th> "the user propably meant it that way" [16:24] <blindcoder> well, if a user adds a file to the list one can be fairly certain he wants it to work, too [16:24] <blindcoder> and not have the extract mechanism fail [16:25] <th> perhaps she added the directory prior to the file but had a typo in it... with your mechanism he would have two directories now. because the software's intelligence did miss that it was a typo and not a missing dir. [16:27] <th> it would be user friendly to issue a warning "the parent directory for the following files is missing - prepare for extraction-error" [16:27] <blindcoder> which wouldn't be fatal either [16:28] <th> existence of directory might be fatal sometimes [16:28] <blindcoder> or make it "the parent directory for the following files is missing - adding it automatically" [16:29] <blindcoder> lack of directory _is_ fatal [16:29] <th> that's only "verbose intelligence" which i still dont like because of the "intelligence" in it. [16:30] <blindcoder> and I like it for it [16:30] <blindcoder> it's not like it's getting in the way or something [16:30] <th> blindcoder: imagine a user who creates a /foobar/ directory and a fooBar/file after it. now she will have /foobar and /fooBar/file. but she needed /foobar/file - this lack of existence will be fatal for her [16:31] <th> but she will not notice [16:31] <th> because extraction does not fail. [16:31] <blindcoder> of course ,because the creation tells her [16:31] <th> and perhaps this fatalness does not show up as visible as extraction-failure [16:31] <stf^rocklinux> the order of files in mkinitramfs is always right, since the 'sort -u' part places dirs first in the list [16:31] <blindcoder> and it'd be more dangerous if extraction - and thus boot - fails [16:31] <th> stf^rocklinux: a dir might be missing - that's the discussion [16:31] <blindcoder> stf^rocklinux: true, but only if the directory is explicitly mentioned [16:32] <blindcoder> th: I'd be more worried if booting fails than if initramfs creation tells me how to fix it [16:32] <th> blindcoder: i prefer a not running system to a wrongly running system when it comes to early-stage-problems. [16:32] <blindcoder> because if the boot process fails (as opposed to extracting the initramfs) I can still do root=/bin/sh and do the steps manually [16:33] <blindcoder> th: compromise [16:33] <blindcoder> th: how about we make default behaviour fail on the extraction process [16:33] <blindcoder> th: and create a -auto-dir parameter to create directories automagically? [16:33] <blindcoder> so you have your behaviour and I can have mine [16:34] <th> i've no objections regarding a --auto-dir option. [16:34] <th> as long as it's not default [16:34] <blindcoder> I just want to help people not fucking up their system as I have done more than once already [16:34] <blindcoder> with a borked early-root filesystem [16:35] <th> one can include a test-extraction in the unit which copies the cpio to /boot [16:35] <blindcoder> that's a very good idea [16:38] <blindcoder> https://scavenger.homeip.net/~blindcoder/dir_files.sh [16:38] <blindcoder> just because it was already done now :) [16:39] <blindcoder> the very first "sort -u" won't work, but that' meaningless [16:42] <blindcoder> th: btw, how about a ROCK4-pre1 when /etc/net/ is integrated? [16:42] <th> can we aim to fix the missing /dev/dsp issue? [16:42] <th> is there some standard for /dev layour which we could obey? [16:42] <th> s/layour/layout/ [16:43] <blindcoder> th: udev ships a few example configurations [16:43] <blindcoder> sadly, /dev is just as standardised as network configuration :) [16:43] <th> blindcoder: have you read about the syntax the list must have? [16:44] <blindcoder> th: I remember something about [fcb] [major] [minor] <file> [16:44] <th> blindcoder: see usage of gen_init_cpio.c which is part of kernel source. [16:44] <stf^rocklinux> th: why not always create dirs automatically? Having to create them manually is an additional source of errors. [16:44] <th> stf^rocklinux: no it is not. [16:45] <blindcoder> ah, okay [16:45] <th> stf^rocklinux: at least that's my opinion [16:45] <th> stf^rocklinux: as stated above [16:46] <th> stf^rocklinux: in any case you HAVE to offer creation of directories. [16:46] <th> stf^rocklinux: to allow empty directories [16:46] <stf^rocklinux> th: if I want a file /some/where on the initramfs, I have to add the dir, and the file to the correct dir, so I have to specify the dir twice [16:46] <th> stf^rocklinux: and with what permissions will the auto-creation happen? [16:46] <th> stf^rocklinux: how to specify that? [16:46] <th> stf^rocklinux: which ownership? [16:46] <blindcoder> hmm [16:47] <blindcoder> I wasn't aware you need to supply uid/gid/mode [16:47] <th> stf^rocklinux: yes - you are right about specifying it twice. because the file-spec contains the dir. [16:47] <stf^rocklinux> th: that's a point, it would have to be some default value. [16:47] <blindcoder> I never have files that are not root:root 0755/0644 on my initrd [16:47] <th> blindcoder: that's the coolness of it. you would need privs for chown/chgrp - but dont need it this way [16:47] <th> blindcoder: no tmp dir? [16:48] <blindcoder> th: I don't see a use for this in standard 'set up the system to boot' mode [16:48] <blindcoder> th: mounted by /sbin/init [16:48] <th> blindcoder: well - of course you can do more than the "just boot". and people start to want this. [16:49] <th> with encryption, raid, lvm [16:49] <th> soon they want an interactive dialog for authtification for decryption of something [16:49] <blindcoder> th: encryption, raid and lvm _all_ need root :) [16:49] <th> blindcoder: yes in early stage we do have root [16:50] <blindcoder> th: and only lvm needs to write to the filesystem, for reasons not yet known to me [16:50] <th> blindcoder: it does not _need_ to write there in an early stage i think. [16:50] <blindcoder> th: yes, I already thought about that dialog thing. Reading the GPG Key for decryption of the harddisk encryption key from a Cardreader :) [16:51] <blindcoder> crypting crypto-keys with crypted crypto-keys :) [16:51] <blindcoder> that'd probably make bruce schneier proud :) [16:51] <th> bootup stage is not trivial much longer - so i think it is a good idea to offer modes. [16:51] <blindcoder> yeah [16:51] <blindcoder> th: anyway, I'd be very happy if we could stick to the interfaces currently provided by initrd as much as possible [16:52] <blindcoder> as in [16:52] <blindcoder> plugins are sourced by a bash [16:52] <blindcoder> using ". /etc/conf/<plugins>" [16:52] <blindcoder> that kind of thing [16:54] <blindcoder> the current initrd's /sbin/init implementation does have some problems [16:54] <blindcoder> but fixing those needs to wait for the new installer [16:54] <blindcoder> which will greatly influence the way filesystems are mounted (/etc/fstab and stuff) [16:55] <blindcoder> th: that's why I'd like to create a ROCK4-pre1 after /etc/net and /dev/dsp stuff [16:55] <blindcoder> th: because once the new installer is started, stuff is going to break [16:56] <th> i've no objections regarding "a ROCK4-pre1 after /etc/net and /dev/dsp stuff" [16:56] <blindcoder> (OT: Isn't Roquefort a brand of cheese?) [16:57] <blindcoder> th: and we should change the version name for this release then :) [16:57] <blindcoder> th: because ROCK 3 still has TRUNK in it [19:16] stf^rocklinux (n=user@M1268P006.adsl.highway.telekom.at) left irc: Read error: 145 (Connection timed out) [19:56] Pilot (i=pilot@81.200.16.172) joined #rocklinux. [19:57] <Pilot> hello [20:01] <th> hi Pilot [20:05] <daja77> these damn telecom suckers [21:13] daja77_ (n=daja77@85-124-206-39.dynamic.xdsl-line.inode.at) joined #rocklinux. [21:26] daja77 (n=daja77@85-124-206-39.dynamic.xdsl-line.inode.at) left irc: Read error: 113 (No route to host) [22:01] Pilot (i=pilot@81.200.16.172) left #rocklinux ("left"). [22:29] stf^rocklinux (n=user@M1353P031.adsl.highway.telekom.at) joined #rocklinux. [23:12] rocklogbot (i=blindcod@pallas.crash-override.net) left irc: Remote closed the connection [23:12] rocklogbot (i=blindcod@pallas.crash-override.net) joined #rocklinux. [00:00] --- Sat Sep 30 2006