WebHosting Paid by #1Payday.Loans
[00:40] daja77_ (n=daja77@dslb-088-072-037-159.pools.arcor-ip.net) joined #rocklinux. [00:51] daja77 (n=daja77@dslb-088-072-032-104.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [02:36] kasc_ (n=kasc@dslb-084-060-096-206.pools.arcor-ip.net) joined #rocklinux. [02:45] kasc (n=kasc@dslb-084-060-109-068.pools.arcor-ip.net) left irc: Read error: 113 (No route to host) [02:45] Nick change: kasc_ -> kasc [08:16] <blindcoder> moin [10:39] blindcoder (i=blindcod@tor/session/direct/x-dd86a6fdcff322b5) left irc: Remote closed the connection [10:43] blindcoder (i=blindcod@tor/session/direct/x-f2d23daa384ac2de) joined #rocklinux. [10:48] <esden> moin [11:12] <blindcoder> moin esden [11:52] <fake> moin [11:57] <blindcoder> moin fake [11:58] <blindcoder> thinkgeek-thermo-pots with caffeeine molecule printed on them rule :D [11:58] <blindcoder> forgot my coffee an hourl ago [11:59] <blindcoder> it's still warm enough to be enjoyable [11:59] <blindcoder> and hot enough to burn my tongue *ouch* [12:04] <daja77_> hi fake [12:05] <fake> hi daja77_ ! [12:21] <blindcoder> moin daja77_ [12:21] <blindcoder> daja77_: did you find out how the live bookmark or whatever works? [12:24] treo (n=xfman@L11fa.l.pppool.de) joined #rocklinux. [13:09] Action: blindcoder feels ignored :( [13:14] Action: mnemoc hugs blindcoder [13:16] <blindcoder> thanks, I needed that [13:16] <blindcoder> now I'm ready to go back to work :) [13:16] <blindcoder> bbl [13:16] <mnemoc> cu :) [13:44] <daja77_> blindcoder: no [15:08] Pilot (n=pilot@212.48.133.193) joined #rocklinux. [15:13] blindcoder (i=blindcod@tor/session/direct/x-f2d23daa384ac2de) left irc: Remote closed the connection [15:15] blindcoder (i=blindcod@tor/session/direct/x-0e074085c1bb4b29) joined #rocklinux. [15:28] treo (n=xfman@L11fa.l.pppool.de) left irc: "safe the Planet, kill yourself" [16:20] Pilot (n=pilot@212.48.133.193) left irc: "quitting" [16:34] mnemoc (n=amery@216.241.24.10) left irc: Read error: 104 (Connection reset by peer) [16:34] mnemoc (n=amery@216.241.24.10) joined #rocklinux. [17:01] <blindcoder> *grunt* [17:01] <blindcoder> it seems rockinitrd isn't broken after all [17:02] <blindcoder> rather my installed udev version is too old [17:05] <th> well - i say initrd is obsolete ;_ [17:07] <blindcoder> and initramfs is inferior to initrd [17:07] <blindcoder> oh great [17:08] <blindcoder> now udev is too new for rockinitrd [17:10] <blindcoder> maybe it works anyway *crossfingers* [17:15] <blindcoder> indeed works [17:18] <th> blindcoder: want to see my work on initramfs? [17:18] <th> blindcoder: https://www.rocklinux.net/people/teha/mkinitramfs.txt [17:20] <blindcoder> daja77_: Manage Bookmarks -> File -> New LiveBookmark [17:20] <daja77_> ah [17:21] <blindcoder> th: can you give me the instructions forr the crazy and lazy (ie: me)? [17:22] <blindcoder> daja77_: no offense, but it took me about thirty seconds to figure out [17:23] <daja77_> well i haven't really focused .. [17:24] <blindcoder> I see [17:24] <th> blindcoder: sure that's in the txt file [17:24] <th> * apply 2006041801092016294 (updates udev to 090) and rebuild your udev package [17:24] <th> * copy https://www.rocklinux.net/people/teha/hotplug.rules to /etc/udev/rules.d/ [17:24] <th> * extract mkinitramfs.tar.bz2 [17:24] <th> * run ./mkinitramfs/mkinitramfs.sh [17:25] <daja77_> livemark failed .. [17:26] <blindcoder> th: that's all? no kernel source or anything needed? [17:26] <th> blindcoder: you need no kernel source for it [17:26] <th> blindcoder: you need a "$something/lib/modules/`uname -r`/" [17:27] <th> only IF you want to have module sin it [17:27] <th> all modules are copied. [17:27] <th> and loaded automatically [17:27] <th> depending on e.g. pci-id's [17:27] <blindcoder> th: so is linux/Documentation/filesystems/initramfs.txt wrong? [17:27] <th> why? [17:28] <th> i dont have that file [17:29] <blindcoder> filesystems/ramfs-rootfs-initramfs.txt [17:29] <th> which part? [17:29] <th> the CONFIG_INITRAMFS_SOURCE thing? [17:29] <th> that's for linking the cpio directly into the kernel image [17:30] <th> we dont need this. [17:30] <th> we can pass it just as initrd to the bootloader [17:30] <blindcoder> - The old initrd was a separate file, while the initramfs archive is linked [17:30] <blindcoder> into the linux kernel image. (The directory linux-*/usr is devoted to [17:30] <blindcoder> generating this archive during the build.) [17:30] <th> "initramfs archive is linked" is wrong. it should read "initramfs archive can be linked" [17:30] <th> alternativ is to pass it through the initrd mechanism through bootloader [17:31] <blindcoder> ah, okay [17:31] <blindcoder> so it really is just mkfs vs. cpio? [17:31] <daja77_> blindcoder: you should add an encoding parameter [17:31] <daja77_> perhaps that's why it doesn't work [17:31] <th> blindcoder: you should take a minute and look at that tar.bz2 [17:31] <blindcoder> daja77_: yes, it seems that FF doesn't understand 0.91 well [17:32] <th> btw - mkinitramfs works as unprivileged user. [17:32] <blindcoder> th: I don't really have time to dig through that amount of code right onw [17:32] <th> blindcoder: that amount? [17:33] <th> let me just paste the complete udev initialization: [17:33] <th> mount -t tmpfs none /dev [17:33] <th> udevd --daemon [17:33] <th> udevtrigger [17:33] <th> udevsettle [17:33] <th> EOF [17:33] <blindcoder> I see three directories, 10kB of C code of several more shell code [17:33] <th> 10kb of c code is the files stolen from linux kernel to CREATE the cpio archive. [17:33] <th> you dont need to look at this at all [17:34] <th> files/init.d is worth a look [17:34] <th> that's 446 bytes in total [17:34] <th> perhaps another 204 bytes for linuxrc.sh [17:34] <blindcoder> I sure hope that's just a proof of concept [17:34] <th> why? [17:35] <th> i even use it for my LVM boots already [17:35] <blindcoder> eval `cat /proc/cmdline` [17:35] <th> well we could remove that line - we dont really need the real_root= option [17:36] <th> but anyhow - do you think attackers could exploit that? [17:36] <blindcoder> for getting kernel paramaters of the name=value form, read /target/lvp/x86/initrd/keymap [17:36] <th> when you are able to change the kernel-command line then you wont insert malicious code [17:36] <th> i liked the eval `cat /proc/cmdline` [17:36] <th> for x in `cat /proc/cmdline` ; do [17:37] <th> haha [17:37] <blindcoder> no, but what about this: [17:37] <blindcoder> linux devfs=nocompat 5 [17:37] <blindcoder> so? [17:37] <blindcoder> with the above line, your code will spit an error message [17:37] <blindcoder> with mine, it won't [17:38] <blindcoder> blindcoder@crazyhorse:~$ eval 'a=b 5' [17:38] <blindcoder> bash: 5: command not found [17:38] <blindcoder> or consider this: [17:38] <blindcoder> linux devfs=nocompat debug [17:38] <blindcoder> same thing [17:39] <th> ok. point taken. i'll remove the line. [17:40] <blindcoder> :) [17:40] <blindcoder> so, for the user side it's really just the same if I say [17:40] <blindcoder> # mkinitrd [17:40] <blindcoder> or [17:40] <th> yes [17:40] <blindcoder> # mkinitramfs [17:40] <blindcoder> correct? [17:40] <th> mkinitramfs.sh for now. [17:41] <blindcoder> then where is my mkinitramfs package? :) [17:41] <blindcoder> rockramfs :) [17:41] <th> * run ./mkinitramfs/mkinitramfs.sh [17:41] <th> not yet created. [17:41] <th> this is only a p-o-c ;) [17:41] <blindcoder> btw, our initrd does not return to the kernel [17:41] <blindcoder> hehe [17:42] <th> "does not return to the kernel"? [17:43] <blindcoder> wel, according to the text file above (rootfs-*) the initrd should return control to the kernel [17:43] <blindcoder> so as I see it we would already have everything we need in our initrd [17:43] <blindcoder> and alll that's left to do is practically change the mkfs call to the gen_cpio call [17:44] <th> they are not interchangable. [17:44] <blindcoder> what's the big difference? [17:45] <th> gen_init_cpio reads a list of inodes. [17:47] <blindcoder> yes? [17:48] <th> yes. [17:48] <th> and it must contain all directories. [17:49] <th> basically the mkinitramfs is much more readable then current rockinitrd [17:50] ringo (n=info@ip565668e1.adsl-surfen.hetnet.nl) joined #rocklinux. [17:50] Action: blindcoder .oO( it used to be very readable 0 [17:50] <th> i missed that stage of it [17:52] <blindcoder> th: it was before everyone and their dog started messing with it [17:52] <blindcoder> around the time it used to be in package/blindcoder [17:53] <blindcoder> daja77_: this is strange [17:54] <blindcoder> daja77_: the blog.crash-override.net rss feed works fine [17:54] <blindcoder> daja77_: the makemyrss feed doesn't [17:54] <th> tfile=${targetdir}/${target}${f#${file}} [17:54] <blindcoder> daja77_: even though they are syntacticaly identical [17:54] <th> that's not readable [17:54] <blindcoder> th: as I said [17:54] <blindcoder> before everyone started messing with it [17:54] <th> blindcoder: that's your code [17:54] <blindcoder> I can't remember writing that code [17:54] <th> svn blame does. [17:55] <blindcoder> hmm [17:55] <blindcoder> maybe it is [17:55] <blindcoder> still it's not hard to read :P [17:55] <th> well. it's not only about readability it's also about how the readability changes when other people contribute to it. [17:56] Action: ringo *gives wifi another shot tonight did not work yesterday.* [17:56] <daja77_> blindcoder: you don't have an encoding specified [17:56] <blindcoder> th: noe [17:56] <blindcoder> th: that's only the part I cleaned up already [17:56] <blindcoder> mine looked like this [17:56] <blindcoder> mkdir -p $targetdir/${target%/*} [17:56] <blindcoder> cp $file $targetdir/$target [17:56] <blindcoder> daja77_: so? neither do i have in /rss.php [17:57] <daja77_> hm [17:57] ringo (n=info@ip565668e1.adsl-surfen.hetnet.nl) left irc: "munchtime." [17:57] <daja77_> blindcoder: but perhaps you don't have utf8 chars in it [17:58] <blindcoder> that might very well be [17:58] <blindcoder> hmm [17:58] <blindcoder> akregator reads them just fine ._< [18:00] <daja77_> so perhaps they break it unless you specify encoding=utf8, like the original rss does [18:00] <blindcoder> let's try [18:07] <blindcoder> cool, now akregator doesn't read it anymore, either [18:10] <daja77_> <?xml version="1.0" encoding="UTF-8"?> [18:10] <daja77_> ah ok [18:10] <daja77_> hm [18:16] <blindcoder> I don't get it [18:17] <daja77_> blindcoder: what if you just rename it to sth with .xml in the end= [18:18] <blindcoder> the same [18:18] <blindcoder> the funny thing is [18:18] <blindcoder> it doesn't validate in feedvalidator.org [18:18] <blindcoder> and I don't understand the error [18:19] <daja77_> link? [18:20] <blindcoder> https://feedvalidator.org [18:24] <daja77_> looks like you can't use html markup in <description> tag, + there are some weird conversions in it [18:24] <daja77_> < is converted e.g. [18:25] <daja77_> look at the areas that are marked yellow [18:26] <blindcoder> so should it be &gt;? [18:26] <daja77_> guess it should be < [18:28] <daja77_> line 57, column 198 [18:28] <blindcoder> seems to be the case, strange enough [18:29] <daja77_> ah onlyone error left [18:29] <blindcoder> that's fefe's broken html [18:30] <daja77_> guess it should be <a href=".. [18:30] <blindcoder> yes [18:33] <daja77_> one regex for all, so far I can see ;) [18:39] <daja77_> have to go now [18:42] <blindcoder> line 12, column 29: Undefined description element: a (13 occurrences) [help] [18:42] <blindcoder> yay... [19:06] <blindcoder> okay, I don't get why FF doesn't like them [19:06] <blindcoder> it doesn't say why [19:06] <blindcoder> akregator reads them [19:06] <blindcoder> I'm content with that [22:35] blindcoder (i=blindcod@tor/session/direct/x-0e074085c1bb4b29) left irc: Remote closed the connection [22:36] blindcoder (i=blindcod@tor/session/direct/x-946c61c0af4ebfa7) joined #rocklinux. [23:59] ringo (n=info@ip51cf58c9.direct-adsl.nl) joined #rocklinux. [00:00] --- Fri Apr 21 2006