| View previous topic :: View next topic |
| Author |
Message |
Matteo Azzali Retired Dev


Joined: 23 Sep 2004 Posts: 1133
|
Posted: Fri Oct 05, 2007 8:51 am Post subject: |
|
|
I think these tests need some discussion to be understood:
1- Is the line finishing before end time meaning that it was faster achieving the same operation?
(I think so, and this would change completely the figure as MB/sec and Throughtput/sec become the overhead
needed to do the same operation, more is worse, this would be a big advantage for btrfs....)
2- Cinebench "read" and "read compiled" results are not so good and that is a little scaring. Is this gonna improving
or it's already optimized for speed? (I swear if this isn't for checksumming rechecked at every read) _________________ Every day a new distro comes to birth. Every day a distro "eats" another.
If you're born distro, no matter what, start to run.
---- http://www.linuxprinting.org/ ---- http://tuxmobil.org/ |
|
| Back to top |
|
 |
arnvidr Guru


Joined: 19 Aug 2004 Posts: 441 Location: Oslo, Norway
|
Posted: Fri Oct 05, 2007 2:03 pm Post subject: |
|
|
| steveL wrote: | | arnvidr wrote: | | AforAlexander wrote: | I do hope people realise that I was being sarcastic, because your signature makes me look like an idiot  |
Well, I was away from this forum for two years, you would know about the general intelligence of the people here more than me I guess. If you want I'll add something indicating it was sarcastic
Or did you mean that people could connect it with the other quote in my signature? I could swtich them around perhaps... |
| AforAlexander wrote: | | Good thing data corruption is only a minor issue in filesystems. |
| AidanJT wrote: | | Do you make a habit of being completely and utterly wrong? |
Ouch! That's bad ;P I'd definitely switch em around (and maybe add a <sarcasm> thing.) |
How do you like it now? _________________
|
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Fri Oct 05, 2007 2:50 pm Post subject: |
|
|
| Matteo Azzali wrote: |
I think these tests need some discussion to be understood:
1- Is the line finishing before end time meaning that it was faster achieving the same operation?
(I think so, and this would change completely the figure as MB/sec and Throughtput/sec become the overhead
needed to do the same operation, more is worse, this would be a big advantage for btrfs....)
2- Cinebench "read" and "read compiled" results are not so good and that is a little scaring. Is this gonna improving
or it's already optimized for speed? (I swear if this isn't for checksumming rechecked at every read) | I had the same questions when reading that pdf as well. Both ext3 and btrfs had lines that didn't go as far as xfs, I just took it to mean that they completed early, because they then showed a graph that was only ext3 and btrfs of the same data. So, I am thinking that they finished earlier than xfs. If that is case, you are right in that it is a big advantage. I would like to see tests compared to recent zfs developments. |
|
| Back to top |
|
 |
steveL Veteran

Joined: 13 Sep 2006 Posts: 1441 Location: The Peanut Gallery
|
Posted: Fri Oct 05, 2007 8:28 pm Post subject: |
|
|
| arnvidr wrote: | | How do you like it now? |
Much better :-) |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Fri Oct 05, 2007 9:40 pm Post subject: |
|
|
Since frenkel disappeared, can someone post ebuilds for all of btrfs: The kernel module and utils?
Thanks |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Fri Oct 05, 2007 9:53 pm Post subject: |
|
|
sure. Here are the current versions from our overlay...
sys-fs/btrfs-0.8.ebuild:
| Quote: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
DESCRIPTION="The btrfs filesystem"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/";
SRC_URI="http://oss.oracle.com/projects/btrfs/dist/files/${P}.tar.bz2";
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
inherit linux-mod
DEPEND=""
RDEPEND="=sys-fs/btrfs-progs-0.8"
pkg_setup() {
BUILD_PARAMS=""
BUILD_TARGETS="all"
MODULE_NAMES="btrfs(fs:${S}/"
linux-mod_pkg_setup
if ! kernel_is 2 6; then
eerror "Need a 2.6 kernel to compile against!"
die "Need a 2.6 kernel to compile against!"
fi
if ! linux_chkconfig_present LIBCRC32C; then
eerror "You need to enable LIBCRC32C in your kernel!"
die "You need to enable LIBCRC32C in your kernel!"
fi
}
src_install() {
linux-mod_src_install
}
pkg_postinst() {
ewarn "WARNING: Btrfs is still highly experimental software. The diskformat may "
ewarn "still change during developement. These ebuilds are only provided to"
ewarn "test btrfs!"
} |
sys-fs/btrfs-progs-0.8.ebuild:
| Quote: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
DESCRIPTION="The btrfs filesystem programs"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/";
SRC_URI="http://oss.oracle.com/projects/btrfs/dist/files/${P}.tar.bz2";
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
DEPEND=""
RDEPEND=""
src_install() {
dobin mkfs.btrfs
dobin btrfsctl
dobin btrfsck
dobin debug-tree
} |
_________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Fri Oct 05, 2007 10:29 pm Post subject: |
|
|
Great. Thanks. These will go to good use frying my root partition. Nah. I think I will put portage on it and maybe /var/tmp/portage. That folder gets used a lot, so it would be a great test, and /var/repos is large since that's where distfiles is and I never clean that out. hmmm. I could delete all distfiles and then do emerge -d world. That would test the partition out. Oh yeah.
*runs off to screw up system* |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Fri Oct 05, 2007 10:45 pm Post subject: |
|
|
| Sorry to double post, but I can I give subpartitons labels. I have /dev/hda6 formatted as ext3 for /var/repos and it's called "Repo". My fstab is all setup to mount by label, so can I format /dev/hda6 as btrfs, then divide that up into Repo and /var/tmp/portage? Currently /var/tmp/portage is mounted in ram, so I have no other physical location except under the main / (/dev/hda5). All the other partitions are all used for important stuff, so I don't want to resize or move them. |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Fri Oct 05, 2007 11:54 pm Post subject: |
|
|
I try to make a new subvol and it says that ioctl returns 0 and I don't see anything when I do a ls. I don't even see the default subvol. :S
Thanks |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Sat Oct 06, 2007 8:32 am Post subject: |
|
|
You might want to try mounting them with "-o subvol=name" as this works now. I am using btrfs on /usr/portage and /var/tmp/ccache and it works great
Rei _________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Sat Oct 06, 2007 2:58 pm Post subject: |
|
|
I can mount the subvols directly now, but I thought the btrfs site said that we were able to see the subvols, if we have the main partition mounted and we did a "ls" in that mount point. :S
Also, do we need to have the main partition mounted to be able to mount and use the subvols? So I have the btrfs on /dev/hda6, do I need to mount hda6 to some place then mount the subvols to where I want them? Or can I not have the hda6 mounted, and just mount the subvols?
I am thinking this might be cool to have a btrfs partition and make /var /home /usr /... all the stuff under root and mount those as subvols. Then you could snapshot each subvol as needed. Like /etc changes the most with new configs and whatnot, so that would be great to snapshot more often in case you botch an update.
Thanks for all your help. |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Sat Oct 06, 2007 5:18 pm Post subject: |
|
|
You do not need to mount the main partition now. You are just mounting the subvolumes to where you need them.
Concerning the main partition, I am not sure if this is a bug or a new feature that you do not see the subvolumes anymore. Anyways - stay tuned for new releases  _________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Sat Oct 06, 2007 5:29 pm Post subject: |
|
|
Cool! Thanks! I guess I need to post on the ML about seeing subvols by using "ls".
Anyways, I swear that my repos on btrfs is much faster than when the repos were on reiserfs 3.x (whatever the latest 3.x version is). Searching portage is extremely quick compared to what it used to be. I am sold on this filesystem. |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Wed Oct 10, 2007 8:22 am Post subject: |
|
|
I did Live-Ebuilds for fetching the latest development snapshots of btrfs and btrfs-progs from the mercurial repositories as I discovered a bug (or the bug discovered me) that causes a kernel oops which was fixed soon after the 0.8 release.
Feel free to use/improve them...
sys-fs/btrfs-progs-9999.ebuild:
| Code: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
inherit mercurial
DESCRIPTION="Live ebuilds of the btrfs filesystem programs"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/"
SRC_URI=""
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
DEPEND=""
RDEPEND=""
S="${WORKDIR}/btrfs-progs"
src_unpack() {
mercurial_fetch http://oss.oracle.com/mercurial/mason/btrfs-progs
cd "${S}"
}
src_install() {
dobin mkfs.btrfs
dobin btrfsctl
dobin btrfsck
dobin debug-tree
}
|
sys-fs/btrfs-9999.ebuild:
| Code: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
inherit mercurial linux-mod
DESCRIPTION="live development version of the btrfs filesystem"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/"
SRC_URI=""
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
DEPEND=""
RDEPEND="=sys-fs/btrfs-progs-9999
dev-util/mercurial"
S="${WORKDIR}/btrfs"
pkg_setup() {
BUILD_PARAMS=""
BUILD_TARGETS="all"
MODULE_NAMES="btrfs(fs:${S}/"
linux-mod_pkg_setup
if ! kernel_is 2 6; then
eerror "Need a 2.6 kernel to compile against!"
die "Need a 2.6 kernel to compile against!"
fi
if ! linux_chkconfig_present LIBCRC32C; then
eerror "You need to enable LIBCRC32C in your kernel!"
die "You need to enable LIBCRC32C in your kernel!"
fi
}
src_unpack() {
mercurial_fetch http://oss.oracle.com/mercurial/mason/btrfs
cd "${S}"
}
src_install() {
linux-mod_src_install
}
pkg_postinst() {
ewarn "WARNING: Btrfs is still highly experimental software. The diskformat may "
ewarn "still change during developement. These ebuilds are only provided to"
ewarn "test btrfs!"
} |
_________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Wed Oct 10, 2007 1:54 pm Post subject: |
|
|
| StifflerStealth wrote: | | I can mount the subvols directly now, but I thought the btrfs site said that we were able to see the subvols, if we have the main partition mounted and we did a "ls" in that mount point. :S |
As far as I can see, the volume itself got something like a default subvolume. You should be able to use it just like a common filesystem by now.
Rei _________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Wed Oct 10, 2007 2:49 pm Post subject: |
|
|
Just a few tidbits about the ebuilds.
On btrfs-progs, you do not need the src_unpacked() function.
For the btrfs module ebuild, you do not need both src_unpack() and src_install() functions. In pkg_postinst(), you need to include the function call to linux-mod_pkg_postinst before the three ewarn statements. I think that function call is missing in the 0.8 ebuilds, and it is important because it calls depmod, which explains why I had to manually run depmod to get btrfs to load. That function should go in the 0.8 ebuild as well if it's missing.
Thanks for posting these. I think I might check out the live sources. ^^ Hmmm ... what would be cool is to have both installed and running, the "stable" 0.8 release on /var/repos and the "live" module on some test partition. I don't think that is possible though.
Cheers. |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Mon Oct 15, 2007 7:16 pm Post subject: |
|
|
I fixed up the live ebuilds to include the functions and the correct src_uri.
sys-fs/btrfs-progs-9999.ebuild
| Code: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
inherit mercurial
DESCRIPTION="Live ebuilds of the btrfs filesystem programs"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/"
EHG_REPO_URI="http://oss.oracle.com/mercurial/mason/btrfs-progs"
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
DEPEND=""
RDEPEND=""
S="${WORKDIR}/btrfs-progs"
src_install() {
dobin mkfs.btrfs
dobin btrfsctl
dobin btrfsck
dobin debug-tree
} |
sys-fs/btrfs-9999.ebuild
| Code: | # Copyright 1999-2007 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
inherit mercurial linux-mod
DESCRIPTION="live development version of the btrfs filesystem"
HOMEPAGE="http://oss.oracle.com/projects/btrfs/"
EHG_REPO_URI="http://oss.oracle.com/mercurial/mason/btrfs"
LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~x86"
IUSE=""
DEPEND=""
RDEPEND="=sys-fs/btrfs-progs-9999"
S="${WORKDIR}/btrfs"
pkg_setup() {
BUILD_PARAMS=""
BUILD_TARGETS="all"
MODULE_NAMES="btrfs(fs:${S}/"
linux-mod_pkg_setup
if ! kernel_is 2 6; then
eerror "Need a 2.6 kernel to compile against!"
die "Need a 2.6 kernel to compile against!"
fi
if ! linux_chkconfig_present LIBCRC32C; then
eerror "You need to enable LIBCRC32C in your kernel!"
die "You need to enable LIBCRC32C in your kernel!"
fi
}
pkg_postinst() {
linux-mod_pkg_postinst
ewarn "WARNING: Btrfs is still highly experimental software. The diskformat may "
ewarn "still change during developement. These ebuilds are only provided to"
ewarn "test btrfs!"
} |
Happy compiling. |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Tue Oct 16, 2007 6:46 am Post subject: |
|
|
Will test them
But there are some things on my mind we should take care of:
- Is there a way to add the modules to the list of modules to be rebuilt by module-rebuild
- The btrfs Makefile should be patched to care for /usr/src/linux instead of the current kernel (gentoo way)
Rei _________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Tue Oct 16, 2007 3:32 pm Post subject: |
|
|
In this function:
| Code: | pkg_postinst() {
linux-mod_pkg_postinst
ewarn "WARNING: Btrfs is still highly experimental software. The diskformat may "
ewarn "still change during developement. These ebuilds are only provided to"
ewarn "test btrfs!"
} | I added the linux-mod_pkg_postinst function. This function adds the mod to the module db list. So if I run: | Code: | ~ # module-rebuild list
** Packages which I will emerge are:
=sys-fs/btrfs-9999
=x11-drivers/nvidia-drivers-100.14.11 | The other ebuilds: both 0.8 and live ebuild lacked that function. I read through a lot of eclasses to figure out exactly what was going on with that function does. Most module ebuilds skip the pkg_postinst() function in the ebuild, so that function gets called automatically because the linux-mod eclass defines it, so if you define your own function in the ebuild, you need to call the function that the eclass would call.
I hope this answers your question. |
|
| Back to top |
|
 |
ToeiRei Veteran


Joined: 03 Jan 2005 Posts: 1113 Location: Austria
|
Posted: Wed Oct 17, 2007 6:45 am Post subject: |
|
|
Thanks a lot, StifflerStealth.
I usually do not do that much ebuilds - in fact, the btrfs ebuild was my first one and I was happy that it worked.
Rei _________________ Blog | btrfs | Please stand by - The mailer daemon is busy burning your messages in hell... |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Wed Oct 17, 2007 4:53 pm Post subject: |
|
|
You're welcome. I work a lot with ebuilds. I know a lot of tricks with them.
Although, now that Waninkoko included BTRFS patch in 2.6.23-kamikaze3: http://forums.gentoo.org/viewtopic-p-4384529.html the btrfs module ebuild might become obsolete. Maybe not. But, it's patched into that kernel if you want to look at the patch and add it your own kernel, you can do that. I converted several people to BTRFS in the 2.6.23-kamikaze2 thread.
Personally, my thoughts are that BTRFS and ext4 are really fast filesytems. When ext4 is final, it will be my new root FS. BTRFS is really fast for portage, and it may become my root fs some day. I'm glad I tried them. Reiser4 gave me some troubles awhile back and I haven't been brave enough to try it again. I basically lost my whole system back in the 2.6.8 days. R4 may be more stable now, but I think these two fs's are fast enough and definately stable with features I like, like subvolumes for BTRFS and stuff.
My two cents. |
|
| Back to top |
|
 |
wrc1944 Advocate

Joined: 15 Aug 2002 Posts: 2637 Location: Gainesville, Florida
|
Posted: Wed Oct 17, 2007 9:10 pm Post subject: |
|
|
StifflerStealth,
So, if I understand correctly, your btrfs-progs ebuild above will work with kamikaze3, or is that incorrect- or we don't know yet and I just need to try it?
I just compiled kamikaze3, and under File Systems enabled btrfs as a module (wouldn't let me compile it in the kernel). Can we now use it like we do any other FS, and just do mkfs.btrfs /dev/sdax, for say /usr/portage? _________________ Main box- ASRock 880GM-LE AM3
Phenom II x6 1090T, 3.2 GHz, 8GB GSkill DDR3 1333mhz
Samsung SATA 500GB, Radeon HD 4670 1GB DDR3
Gentoo ~x86, ~amd64, glibc-2.15-r1, gcc-4.7.0, kernels 3.3.7, 3.4.0 w/auto-cgroups |
|
| Back to top |
|
 |
StifflerStealth l33t


Joined: 03 Jul 2002 Posts: 954
|
Posted: Wed Oct 17, 2007 9:32 pm Post subject: |
|
|
You should be able to. I asked Waninkoko what version he used to make the patch but he seemed to skip over that question in that post. I think he used the svn copy because I mentioned that was the most stable, so the above ebuild should work. I don't remember the difference between v0.8 and svn of the progs. But, I do think it will work. I still need to update my kernel to try it out. I just can't reboot my computer at the moment.
Still, I think it's great that Waninkoko was able to patch in BTRFS even though it's only available as a module. It really helps on rebuild and reboot.
You can also make subvolumes with BTRFS, so you can make two subvols: one for portage and one for /var/tmp/ccache. That's explained somewhere in this thread. I like them. It keeps my data separate and allows snapshoting of different subvols, so you don't need to snapshot the entire partition. In theory, one could make each revision of the entire portage repository (minus the distfiles) a snapshot, so if a package gets removed, you just mount that snapshot and point to it in the overlay directory. So, it's like built in versioning, but more compatible with emerge. That way, something like winamp would never be gone, per se and all it's depends are still available as well. Interesting thoughts ... Hmmm .... |
|
| Back to top |
|
 |
wrc1944 Advocate

Joined: 15 Aug 2002 Posts: 2637 Location: Gainesville, Florida
|
Posted: Thu Oct 18, 2007 5:12 am Post subject: |
|
|
I'm having problems with the btrfs-progs ebuild. I made the /usr/local/portage/sys-fs/btrfs overlay directory, cd'd in and copied your ebuild there, and tried to digest it (I've done that lots of time). First I had problems like this:
| Code: | gentoo btrfs # ebuild btrfs-progs-9999.ebuild digest
Appending /usr/local/portage to PORTDIR_OVERLAY...
!!! /usr/local/portage does not seem to have a valid PORTDIR structure.
gentoo btrfs # | My make.conf file is correct, so I didn't understand this message.
Then I checked the ebuild again, and made an /usr/local/portage/sys-fs/btrfs/btrfs-progs directory, and tried again, and got this: | Code: | gentoo btrfs # ebuild /usr/local/overlays/sys-fs/btrfs/btrfs-progs/btrfs-progs-9999.ebuild digest
Appending /usr/local/portage/sys-fs to PORTDIR_OVERLAY...
!!! 'btrfs/btrfs-progs' has a category that is not listed in /etc/portage/categories
>>> Creating Manifest for /usr/local/overlays/sys-fs/btrfs/btrfs-progs
!!! 'btrfs/btrfs-progs' has a category that is not listed in /etc/portage/categories
gentoo btrfs # emerge -p btrfs-progs
These are the packages that would be merged, in order:
Calculating dependencies /
emerge: there are no ebuilds to satisfy "btrfs-progs". |
I created /etc/portage categories, but don't know what to put in there (syntax)
googled for 2 hours, with no clues as to what i needed to use for the categories line. Read man portage, but it doesn't help much- nothing I tried works, like sys-fs, which from man portage should have worked AFAIK.
What should I do? _________________ Main box- ASRock 880GM-LE AM3
Phenom II x6 1090T, 3.2 GHz, 8GB GSkill DDR3 1333mhz
Samsung SATA 500GB, Radeon HD 4670 1GB DDR3
Gentoo ~x86, ~amd64, glibc-2.15-r1, gcc-4.7.0, kernels 3.3.7, 3.4.0 w/auto-cgroups |
|
| Back to top |
|
 |
Matteo Azzali Retired Dev


Joined: 23 Sep 2004 Posts: 1133
|
Posted: Thu Oct 18, 2007 7:23 am Post subject: |
|
|
@wrc1944:
| Code: |
cd /usr/local/overlays/sys-fs
mv btrfs/btrfs-progs ./
|
@StifflerStealth:
IMHO it's too early for either Btrfs and Ext4. Beside all my complaints about inode based ext4 (it's not impossible that
they'll find a way to skip those lenghty fsck without removing most of the data integrity advantage of extX filesystem, I just want to
see that.) Btrfs is still under heavy development, and both fs have to be tested and benchmarked in a more serious
way (i.e.: with recent processors AND with the main target in mind: writing or finding a million 16Kb files may be useful,
but you'll do that just once in a while, and on desktop you'll definitely need speed for reading of mixed size files and
short fsck). _________________ Every day a new distro comes to birth. Every day a distro "eats" another.
If you're born distro, no matter what, start to run.
---- http://www.linuxprinting.org/ ---- http://tuxmobil.org/ |
|
| Back to top |
|
 |
|