Re: changed ownership of / by mistake...
Gilles wrote: dpkg --get-selections | perl -ane 'print $F[0], if $F[1] =~ /^install/' SELECT [...] and press enter... [Note/WARNING: I did not test this!] I've done something very similar to repair a system after filesytem corruption. It worked quite well. But if ownership is screwy, dpkg may have problems running. I'm also not sure that this will fix /etc since dpkg tends to want to keep user changes there. Can you boot into Knoppix (or similar) and start chowning files back to root? Note that not all files should be root:root; on my laptop: # find /etc -not -user root -or -not -group root [I've annotated the output to include the users and groups of things] /etc/ppp/peers /etc/ppp/peers/anu /etc/ppp/peers/snap [... more ppp peers - these are all root:dip ...] /etc/ppp/ppp_on_boot.dsl /etc/bind /etc/bind/ [more files here, mostly root:bind] /etc/cups /etc/cups/ppd /etc/cups/ppd/* [all files here are root:lpadmin] /etc/cups/classes.conf /etc/cups/classes.conf.O /etc/cups/classes.conf.dpkg-dist /etc/cups/cupsd.conf /etc/cups/printers.conf /etc/cups/printers.conf.O /etc/cups/printers.conf.dpkg-dist /etc/exim4/passwd.client [root:Debian-exim] /etc/chatscripts /etc/chatscripts/provider [root:dip] /etc/shadow [root:shadow] /etc/gshadow [root:shadow] /etc/at.deny [root:daemon] There are a bunch of files in /var with non-root:root ownership too. Did you interrupt chown before it got to /var or would a list there also be useful? Obviously the list of what things are not root:root depends on what packages you have installed but hopefully this should be a fairly typical desktopish system. Cameron signature.asc Description: Digital signature
Re: X86_64 on Shuttle-XPC ST20G5
Andrew Sharp wrote: what a PITA. I've got an ATI X300SE PCIe on my system, came right up with X when I installed etch, it works fast and furious, I didn't have to do anything except answer a couple of questions from debconf (or whatever) when it was installing. ... To get 3d working, it's a bit of work regardless of which you choose. For for fast and easy as pie 2d, ATI just works. FWIW this has also been my experience. With an X300 (and Radeon = 9250) you should have 3D acceleration too - thanks to the capital-F Free drivers written by ATI and included as part of the kernel/Xfree/Xorg. The low-end Radeons and Intel i8xx/i9xx are my graphics chipsets of choice for this reason. Cameron. signature.asc Description: Digital signature
Re: OO 2.0 released
Greg Madden wrote: It is not a 64 bit version of OOo.org, it is 32 bit, and they (Ubuntu) have included and configured ia32libs to make it work. It's an ugly ugly hack. It does appear to work, though. $ cat /usr/lib/openoffice2/program/soffice.bin #! /bin/sh GTK_PATH=/usr/lib32/gtk-2.0 LD_PRELOAD=/usr/lib32/libpangohack.so.0.0 exec $0.real $@ Cameron signature.asc Description: Digital signature
Re: Checkpoint Firewall Client on AMD64
Lennart Sorensen wrote: On Wed, Jun 29, 2005 at 01:20:05PM +0100, Clive Menzies wrote: I just looked at openswan but on trying to install it got: WARNING: untrusted versions of the following packages will be installed! Never seen that message before. Are you running some silly signature checking feature in apt or debsigs? If so expect to see a lot of that until some day when all debian packages are actually signed. Debian packages _are_ signed, and have been for a while. The amd64 packages are signed with a different key to the standard Debian ones, though, so maybe you need to tell apt where to get the signature from? curl http://amd64.debian.net/archive.key | apt-key add - (Untested but should do the trick.) Cameron. signature.asc Description: Digital signature
Re: verification of packages with gnupg/apt-key
Lennart Sorensen wrote: I was under the impression the majority of packages in debian were not signed, since no one has come up with a way for the buildd to sign a package using a package maintainers key (and I imagine no one should try either). All packages are signed. Ones uploaded by the buildd's are signed by the buildd owner instead (semi-manually after the build completes, not as an automated process AIUI). Cameron. signature.asc Description: Digital signature
Re: notebook choice advice...
Lennart Sorensen wrote: Certainly based on my past experience, if you run linux and want it to work, you always pick nvidia over ati for graphics. My experience has been that Intel graphics is much better supported than either. A pity you can't get an AMD machine with Intel graphics and wireless cards :-( The older Radeons are much better supported than the newer ones, though: up to the Radeon 9200 / X300 you get 3D acceleration with the Free drivers. Suspend/resume is supported too, without any special kernel patches. I've never had much luck with Nvidia's proprietary drivers, but all of my Nvidia cards have been old (TNT2, Geforce 2MX) and it sounds like they've improved a heck of a lot since then. Cameron. (Not a gamer :-) signature.asc Description: Digital signature
Re: SATA Silicon Image 3114 support for A64 images ?
eternalnewbee wrote: This one is at least supported in 2.6.10 and above, in plain SATA mode. Using the proprietary software raid crap in the Sil chips isn't recomended. Why not? The lowly Windows has been doing just fine with SiI; why is it not recommended under Linux? Linux's md RAID subsystem has been around for a long time, can do more than the SiI's drivers (e.g. if you wanted RAID 5 or RAID 6; or some of your RAID components were on a different controller), makes it easier to move the array to a different controller in the future, and often gives much higher performance than other software RAID implementations. The dmraid driver under Linux (which I've never used) should support the on-disk format used by your SiI controller's firmware if you need it for some reason, e.g. compatibility with another operating system on the same array. However it isn't supported by the Debian installer yet, so you'd have to install the system temporarily onto another (non-RAID) drive, and then move it across later. Note that it's still software RAID, the same as the RAID functionality provided by SiI's Windows drivers is. Cameron. signature.asc Description: Digital signature
Re: Bootsector not writable?
Andre Timmermann wrote: The system works without any problem if I boot it from my grub-floppy. You could try installing grub from that boot floppy. It'd be something like: grub root (hd0,0) grub setup (hd0) Replacing (hd0,0) with the partition that contains grub (e.g. /boot or if you don't have one, /). Cameron. signature.asc Description: Digital signature
Re: Bootsector not writable?
Andre Timmermann wrote: Serial ATA Raid-Controller: Adaptec AAR-2410SA 3 x 80GB SATA harddiscs But when I try to boot the system, the bios shows Operating system not found This may be a stupid question - but it's something that worth checking, since everyone suffers from moments of stupidity. It sounds like you've got three identical drives in there. Are you sure you're telling the BIOS to boot from the one that Linux thinks is sda? Cameron signature.asc Description: Digital signature
Re: SATA Silicon Image 3114 support for A64 images ?
Tudiatya wrote: No, I didn't try sata_sil .. I don't know, how. :) The Debian Installer modules list (which modules to load) doesn't contain sata_sil, only sata_nv. :( Hi, While the SiI controllers should normally be fairly well supported, but what you say suggests that the installer hasn't picked the the SiI controller at all. If you send the output of `lspci` and `lspci -n`, we can look into why this might be the case. Cameron. (Who, erm, doesn't own an AMD64 or any SATA equipment, but has a vague understanding of how Debian's hardware detection works.) signature.asc Description: Digital signature
Re: SATA Silicon Image 3114 support for A64 images ?
Tudiatya wrote: Okay, I went from the installer to Execute a shell, but in the shell, lspci was not recognized as a command, so this won't take me further. :( (I don't have Linux, WinXP only ! Would it help you if I let a Suse Live CD run and lspci from there?) Yes, that would be great. It might also be possible to get PCI IDs from within Windows XP, but I'm not sure how to go about doing that. BTW, are you sure that it's a SiI 3114 controller that you have? What interesting is, that there's a module called siimage in the installer's list, but I think that's for another type of silicon image controller. Correct. It's for the SiI 680 ATA133 controller (I have one of these in a machine at home). I believe that the autodetection method that Debian uses for IDE controllers is load all the drivers and see which ones work, so that might be why you see that one there :-) Would it perhaps help, if the installer would include sata_sil in it's list of selectable modules ? The installer does include the sata_sil module (it's in the same sata-modules package that the sata_nv module is), it's just a matter of convincing it that it's the right module for your hardware. A quick look at the source code shows that the device IDs that sata_sil recognises should be the same as the list that Debian uses for working out which module to load, so it's quite likely that even if you did convince the installer to load sata_sil, it wouldn't pick up your controller. Out of curiosity, have you tried a Debian i386 install CD to see if that picks up your SATA controller? Cheers, Cameron. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian AMD64 Archive Move
Ed Cogburn wrote: Note: non-free is NOT provided yet. We need to decide what we do with it, as we may be forbidden to distribute some of the software in it (we aren't Debian). Wait a second, if you *aren't* Debian, it should be *easier* for you to provide non-free, not harder. Nope. It is guaranteed that all packages in the main archive are distributable by anybody, whether they're the Debian project or not (DFSG#8). This is not necessarily the case for non-free packages, hence they'd have to be examined individually to determine whether the licence was acceptable. Cameron. signature.asc Description: Digital signature
Re: Recommended Graphics card for OpenGL (was Matrox P650 / Athlon 64)
Gaius Mulley wrote: I wonder whether anyone knows (if any) graphics board can be bought which also has complete source code for reasonable, even average, performance OpenGL (Mesa) which also works on pure64? I've had good luck with the Radeon 9200 in a number of 32-bit machines (though I've no idea how well they'd work on AMD64). The driver is completely open source and supports 2D and 3D acceleration. The higher end Radeons have no 3D support under Linux except for ATI's proprietary 32-bit-only drivers. Cheers, Cameron. signature.asc Description: Digital signature
Re: Diskless AMD64; mknbi
Jeroen Coumans wrote:
I have a server running with pure64 and would like to follow the
method on http://wiki.debian.net/index.cgi?DiskLess. I'm wondering
what the status is of porting mknbi to AMD64, since it's the only
package which doesn't compile.
It's a perl script, so you /might/ be able to get away with just
extracting the i386 version somewhere.
Cameron.
% head =mknbi-linux
#!/usr/bin/perl -w
# Program to create a netboot image for ROM/FreeDOS/DOS/Linux
# Placed under GNU Public License by Ken Yap, December 2000
# 2003.04.28 R. Main
# Tweaks to work with new first-dos.S for large disk images
BEGIN {
push(@INC, '/usr/lib/mknbi');
signature.asc
Description: Digital signature
Re: amd64 and video card experiences?
Raul Miller wrote: Not exactly. Current generation X (for example X.Org) seems to run fine on current generation ATI. [Well... it runs fine on 9600 and 9800 -- I'm only presuming it runs fine on X800.] It's the 3d acceleration which is not yet supported on amd64 for ATI. If you don't want 3D acceleration, why buy a 9600 instead of a 9200? Cameron. signature.asc Description: Digital signature
