Re: exploring debian's users and groups
On Tue, Aug 07, 2001 at 01:35:48AM -0400, Joey Hess wrote: sudo: HELP: Nothing uses it here, and I have sudo installed.. Maybe there's a way to only let users in this group use sudo? sudo uses this group internally. Members of this group do not need to type their password. see /usr/share/doc/sudo/OPTIONS Rainer -- KeyID=759975BD fingerprint=887A 4BE3 6AB7 EE3C 4AE0 B0E1 0556 E25A 7599 75BD
autmagically install -dev and -doc packages
Hi! Does anybody know an easy way to install -doc and -dev packages automatically when installing a certain package? Say I select libgtk1.2 in dselect and it suggests libgtk1.2-dev, libgtk1.2 and libgtk1.2-dbg, too. Adding Suggests: *-dev, *-dbg to each Package is a) much work b) annoying if you don't want -dev installed by default. In this example it could be done using the Source: field. All Packages have Source: gtk+1.2, but I suppose thats not The Right Thing. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
ftp.de.debian: Archive_in_progress - still?
Hi! Hmm, according to /debian/Archive_Maintenance_In_Progress ftp.de.debian.org update is starting at 13:52 US Central time, and ending at about 15:30. That's 19:52-21:30 CET, isn't it? Currently it's 0:50 CET and this file still exists. Is Archive_Maintenance_In_Progress incorrect, is the update still in progress or did ftp.de.debian.org miss to delete this file? please keep me CC:'ed Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: Synchronizing Windows NT to Debian.
Hi! Jay Barbee ([EMAIL PROTECTED]), Re: Synchronizing Windows NT to Debian.: There is a program called Tardis that has client and server time services. Seems pretty powerfull: Yes, I know, but Greg Frye ([EMAIL PROTECTED]) already mentioned it. Does NT's net time /set use ntp's port 123? No, it uses SMB. I assume the linux box runs samba, too. It doesn't that many bells and wistles, is lesss accurate, but kind of sufficient for basic requirements. (and saves some Meg of RAM?) Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpemRt0dKxx4.pgp Description: PGP signature
Re: nfs and 2.2.x
Hi! Shaleh ([EMAIL PROTECTED]), nfs and 2.2.x: I am prepping to setup NFS on a 2.2.x machine. I would like to use the kernel based one as the docs point to it being faster/better. Is there support in Debian for this? The package is in project/experimental. (I never had the time to actually try it out) Any reasons why I should NOT use the kernel based NFS solution? IIRC knfsd's exports don't recurse mountpoints: If host1's /usr is a different volume than /, mounting host1:/ on another box won't give you access to host1:/usr. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpMzDBSEFxDk.pgp Description: PGP signature
Re: Synchronizing Windows NT to Debian.
Hi! Greg Frye ([EMAIL PROTECTED]), Re: Synchronizing Windows NT to Debian.: There is a precompiled binary of xntp3 for NT if you want the NT box to be a time server that sync's to the Debian box. It you just want the NT box to be a client, I like Tardis a lot. I have it on NT workstations set up to set the time from ntp broadcast. ... and if you don't need that much precission use net time /set at the cmd prompt (maybe as sheduled job?) ... Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp38Z1simIYD.pgp Description: PGP signature
Re: NT and Debian...O.S. loader 4.01 sux!
Hi! Martin Schulze ([EMAIL PROTECTED]), Re: NT and Debian...O.S. loader 4.01 sux!: [EMAIL PROTECTED] wrote: what up! scenario: i had Linux on my laptop running efficiently well...however, i just installed NT4.0 without preparing LiLo for the image of NT...now, my laptop will only boot to NT (o.s. loader 4.01 takes over) and I dont know of any way to get back into Linux Specs: my hard drive is split up to where there is an NTFS partition and an ext. partition in which the Linux swap and native are stored. If you installed LILO into the beginning of your root partition (default for debian), you need to move the active/bootable flag back to your root-partition. Check Diskmanager wether it can do this - or use a plain DOS-fdisk. Of course you can uuse linux' fdisk, too. If LILO used to reside in the MBR, you need to boot from a rescue disc and reinstall LILO. You may want to check some of those linux+winnt-HOWTOs for more details. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: aic7xxx kernel
Hi! Richard Hall ([EMAIL PROTECTED]), aic7xxx kernel: Can anyone tell me which parameters to choose during 'make config' to fix this? There are three questions that come up when I choose the AIC7xxx driver. I've answered yes and no to overriding driver defaults, yes and no to verbose error reporting, and tried 5 and 15 second waits after resetting SCSI bus. What am I missing? Appearantly you're missing a working driver :-) Either Get the newest aic7xxx driver (5.1.6) from ftp://ftp.redhat.com/pub/aic/ or ftp://ftp.dialnet.net/pub/linux/aic7xxx/, patch your kernel source and make your kernel as usual or grab the sources for 2.0.36 - it ships with a more recent aic7xxx driver which should work (but IIRC it's not the most recent one). Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpGuXxg5j6w7.pgp Description: PGP signature
Re: Can't access RAID - why?
Hi! Martin Schulze ([EMAIL PROTECTED]), Can't access RAID - why?: After that I added this MD device to the system: # mdadd /dev/md2 # cat /proc/mdstat Personalities : [2 raid0] read_ahead 8 sectors md0 : active raid0 hda2 hdb2 819312 blocks 4k chunks md1 : active raid0 hda4 hdb4 7405952 blocks 8k chunks md2 : inactive sda2 sdb2 1984818 blocks It isn't running. Either supply -r to your mdadd command to add + run the device or explicitly run it with mdrun /dev/md2 . Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgptlEutVzZWl.pgp Description: PGP signature
micro-HOWTO (was Serial consoles)
Hi! I've seen FreeBSD being installed over a serial console and thought Hmm, how can I do this with debian?. I was aware of the sercons stuff in recent 2.1 kernels. Initiated by a question from Shane Wegner on debian-user I tried it out (up to partition a harddisk - I had no spare disk to continue). Cleto Pescia [EMAIL PROTECTED] finished an installation and sent me all those missing hints. I'm summing this up for all those of you intereseted in it. It's based on Debian 2.0 (hamm), but should work with other versions, too. Get a recent 2.1 (or 2.2) kernel or patch your 2.0 sources with Miquel van Smoorenburg's patch to backport the sercons stuff: ftp://ftp.cistron.nl/pub/people/miquels/kernel/ Build a replacement kernel for the installation disk. Enable sercons, initrd, ext2, msdosfs and all needed drivers. Do not use modules (or prepare to hack the driver disk, too). Check the install manual for more details (in case I missed something) Put the kernel on the rescue disk: mount /tmp/resc1440.bin /mnt -o loop cp the new kernel to /mnt/linux and run sh /mnt/rdev.sh Edit /mnt/syslinux.cfg to automatically boot with sercons: set TIMEOUT to 1 and add console=ttyS0,9600n8 to all APPEND lines (for COM1, 9600 Baud, no parity, 8bits, software handshake) (possibly unneded if you have no graphics card and keyboard attached) Add the needed /dev/console to the rescue disk's root-fs: zcat /mnt/root.bin /tmp/root.bin mount /tmp/root.bin /floppy mknod -m 622 console c 5 1 as root umount /floppy gzip -c9 /tmp/root.bin /mnt/root.bin The new rescue disk image is done. umount /mnt Make a boot floppy cp /tmp/resc144.bin /dev/fd0 boot from the rescue disk, select monocolor and your keyboard (used for virtual console), select open a shell from the menu, manually run fdisk (cfdisk segfaults), exit the shell and continue installation as usual (alternatively symlink cfdisk - fdisk in root.bin) before you reboot your new system, reopen a shell. make the new-style console device rm /target/dev/console cp /dev/console /target/dev/ (alternatively modify base*.tgz) Instruct lilo to use the sercons, too: echo 'append = console=ttyS0,9600n8' /target/etc/lilo.conf echo 'serial=0,9600n8' /target/etc/lilo.conf lilo Make root logins via the sercons possible: echo ttyS0 /target/etc/securetty and put a getty on the serial line: Edit /target/etc/inittab and uncomment the line starting with T0. (alternatively modify /etc/init-tab in root.bin) reboot Special Thanks to: Miquel van Smoorenburg for the sercons patch Cleto Pescia [EMAIL PROTECTED] for finishing what I started. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpgNlVVEgd91.pgp Description: PGP signature
Re: Diald
Hi! [EMAIL PROTECTED] ([EMAIL PROTECTED]): I've been able to configure diald to restrict dialing out at certain times (like after I go to bed). However, if I have the connection up when the restriction time comes, it will kill the connection. I'm wondering if it would be possible for diald to prevent dialing out, but not to kill a live connection? remove the restrict lines from your config, copy the standard-filter and replace all accept with keepup, add a cronjob which replaces the filter file used by diald and write reset to the control fifo Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp04Iz6pGiuC.pgp Description: PGP signature
Re: Serial consoles
Hi! Rainer Clasen (bj): Shane Wegner ([EMAIL PROTECTED]): I posted here a few months ago about getting the Debian installation process to work on a serial console over a null modem connection to a vt100 terminal. You need to modify root.bin! gunzip it, mount it via loop and replace /dev/console. ohh - wait - /dev/console isn't used by the installation program. maybe you need to kludge around and make /dev/tty1 the same as /dev/console. Check /usr/src/linux/Documentation/serial-console.txt for details. I became curios and actually tried it. I added console=ttyS1,9600n8 to syslinux.cfg's APPEND lines and set DEFAULT to 1. Furthermore root.bin was modified with mknod c 5 1 /dev/console. I rdev'ed the kernel and booted the box from this floppy. E'voila - worked like a charme. I stopped at partiton a hard disk since there was no disk attached. Now I'm thinking about removing the floppy and booting of the net. :-) This is really kewl - I've expected more problems. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpLfCM2g3I1C.pgp Description: PGP signature
Re: Serial consoles
Hi! Shane Wegner ([EMAIL PROTECTED]): I posted here a few months ago about getting the Debian installation process to work on a serial console over a null modem connection to a vt100 terminal. ... in a dos directory. My question is will simply rebuilding that linux file as a developement kernel and editing install.bat appropriately be enough or do I need to modify anything on root.bin? Also, do I have to rebuild You need to modify root.bin! gunzip it, mount it via loop and replace /dev/console. ohh - wait - /dev/console isn't used by the installation program. maybe you need to kludge around and make /dev/tty1 the same as /dev/console. Check /usr/src/linux/Documentation/serial-console.txt for details. drv1440.bin with the 2.1 modules and what modules and kernel compile options should I enable for the Debian install program to go smoothly? You need to change /dev/console in base*.tgz, too. I suppose you needn't change anything else if you don't use modules - compile the needed stuff into the kernel. Of course you can switch back to a modularized kernel when installation is finished. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpkWvurOsFxg.pgp Description: PGP signature
Re: how can i find out a netbios name from an IP address.
Hi! Joe Emenaker ([EMAIL PROTECTED]): are there any other tools that can do it? The only thing I could find is to look in /var/samba/browse.dat and then use nmblookup on each one. Problem is, browse.dat doesn't hold all of the machines. What you need is a way to list all of the machines in a domain... but I couldn't figure out how to make samba do that. if you are a master browser and browse.dat doesn't show the whole domain, your setup is broken (or your samba is not yet running the magic 45 minutes it can take till a master browser is updated). Usually you can do smbclient -L netbios name of any masterbrowser to get a list of all hosts in your domain. Usually the domain controller has the most recent list (and delivers it to each subnets' masterbrowser). Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp6VT9tRy7EO.pgp Description: PGP signature
restore(8) barfs on tape 5/5
Hi! I'm currently trying to restore my stripe from tape with restore -rf /dev/st0. I am stuck at the prompt for the last tape. It refuses to continue. Yes I made a mistake while dump(8)ing to tape: The drive was still busy when I told dump to continue. dump got EOF imediately, wrote nothing, but increased it's tape counter. Are there hints on how to continue? What happens when I cancel restore (with tape: none) and resume restore with -R? Do I need to do the full restore again and interactively? Does anybody know a way to write a dummy tape5? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: strange non-trivial routing problem
Hallo erstmal! Jeff Katcher ([EMAIL PROTECTED]): Rainer Clasen wrote: Hi! This may be offtopic - its not debian-specific. But I'm not sure where to ask at all (linux-net is announced as development list ...). I appreciate pointers to the correct forum. Ok, first some ASCII-art to confuse the reader ;-) 10base2 | V NET_A |- ---| | || |NET_B|| BOX_A --- BOX_B BOX_C | |NET_C ^ || | -- 100baseTX crosslink | BOX_D As you can see BOX_A is my 100mbit router running 2.1.125. BOX_B is 2.0.35 with masquerading stuff turned on. BOX_B has a route to NET_C via BOX_A's IP in NET_B. BOX_C's default route points to BOX_B. But: BOX_D can't connect/ping whatever to any IP on NET_A!! I traced a ping to BOX_C by adding logged ACCEPT rules to BOX_A's and BOX_B's firewall: Ping gets in to BOX_A, BOX_A sends it out to NET_A, BOX_C replies, BOX_B gets it, and sends it out through NET_B - BUT BOX_A doesn't see it If I telnet from BOX_D to BOX_B, I get those unknown syslog entries known from half-port scanning. Rainer Have you set up Firewalling through ipfwadm for BOX_B (or is it packet forewarding)? No. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
strange non-trivial routing problem
Hi! This may be offtopic - its not debian-specific. But I'm not sure where to ask at all (linux-net is announced as development list ...). I appreciate pointers to the correct forum. Ok, first some ASCII-art to confuse the reader ;-) 10base2 | V NET_A |- ---| | || |NET_B|| BOX_A --- BOX_B BOX_C | |NET_C ^ || | -- 100baseTX crosslink | BOX_D As you can see BOX_A is my 100mbit router running 2.1.125. BOX_B is 2.0.35 with masquerading stuff turned on. BOX_B has a route to NET_C via BOX_A's IP in NET_B. BOX_C's default route points to BOX_B. But: BOX_D can't connect/ping whatever to any IP on NET_A!! I traced a ping to BOX_C by adding logged ACCEPT rules to BOX_A's and BOX_B's firewall: Ping gets in to BOX_A, BOX_A sends it out to NET_A, BOX_C replies, BOX_B gets it, and sends it out through NET_B - BUT BOX_A doesn't see it If I telnet from BOX_D to BOX_B, I get those unknown syslog entries known from half-port scanning. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: XFCom_Matrox and Backspace
Hi! Alexander Kushnirenko ([EMAIL PROTECTED]): I got Matrox Millenium G200 Card and installed XFCom_Matrox from SUSE. For some reason Backspace is not working (there is a little beeping sound). If I Any suggestions how to fix that? I use XKEYBOARD extensions SUSE's server use a different directory in /var to compile the keyboard map. follow the installation instructions shipped with the server and make this dir (or make it a symlink to /var/lib/xkb) Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpcHWcd2IVjF.pgp Description: PGP signature
Re: Frame Buffer
Hi! Jeff Beley ([EMAIL PROTECTED]): I've just upgraded to 2.1.130 and am expirementing the frame buffer that's built into the kernel. I've been able to use the fbset utility to set the resolution and suchhowever X is very fuzzy(for lack of a better term)...I read in the documentation that there is a fbdev server for X, however I have not been able to locate that X server. I have an ATI Mach64 card recognized at bootup. IIRC debian has no FBDEV server (at least for i386). Either do your own or get a precompiled one. You might want to check out the framebuffer-HOWTO at http://www.tahallah.demon.co.uk/programming/prog.html it has instructions/ links for both. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpkNfs7aRSL4.pgp Description: PGP signature
Re: Optical Jukebox support in Linux
Hi! Jay Barbee ([EMAIL PROTECTED]): I was wondering if Linux (kernel and tools) had the ability to control an optical jukebox. The box in question is an HP OEM that has 144 slots to ... Anybody have any experience with this or could point me in a direction to get some help? there is a driver fir SCSI-Changers at http://user.cs.tu-berlin.de/~kraxel/linux/ Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpZfzEgBLSUd.pgp Description: PGP signature
Re: Getting Rid of Debian/Linux
Hi! David McDonald ([EMAIL PROTECTED]): On Fri, 27 Nov 1998, Austanners Wet Blue Pty Ltd wrote: I have a requirement to remove Debian from one of our workstations and re-install DOS/Windows (Horror!) Just boot from your DOS installation disks, run the FDISK program (I can't recall if it will do this automatically or not - you may have to exit to the DOS prompt with the F3 key (again, that's the way I remember it, not necessarily the way it is). Note: DoS won't delete logical non-DoS partitions. You need to use something else to delete them. For example you could use dd if=/dev/zero of=/dev/hda count=1 to nuke ALL partitions on /dev/hda or simply use Linux's fdisk to delete those parts (or change their type to a DoS one) Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp6lWLVReORx.pgp Description: PGP signature
Re: Can't remote X
Hi! Kent West ([EMAIL PROTECTED]): Now, for the Thanksgiving weekend, I've traveled 150 miles east (from Abilene to Ft Worth) to my folks' house, and am using my Debian laptop on my Mom's ISP and trying to telnet back to my Abilene Debian box. However, I'm getting an error to the effect that the client is not authorized to connect to the Xserver. Is this some sort of default setting in Debian that I need to turn off somewhere, or is it some sort of barrier in Mom's ISP service, or what? either do an ssh to your box at home - it will set up everything correctly (and encrypted!!) or do something like rsh box.at.homexauth add `xauth list $DISPLAY` . Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpHkx687SVH2.pgp Description: PGP signature
Re: diald time restriction
Hi! Peter Bartosch ([EMAIL PROTECTED]): is it possible to restrict diald only to work at e.g. one hour per day? It's easy to say diald is allowed to bring up the line from 3:00 till 4:00 either by restrict options in the configuration file or by sending block/unblock commands to the control fifo. But I suppose you want to say the total time the line is allowed to be up is 1 hour - I could imagin a cron job using a log generated from pppd's ip-{up,down} scripts to decide wether the hour has passed and to block diald. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpl4vVL8OtNm.pgp Description: PGP signature
Re: Problem with X...Windowmaker...locale?
Hi! Ed Cogburn ([EMAIL PROTECTED]): Rainer Clasen wrote: Do you mean reinstalling xbase_3.3.2.3a-7? apt and IIRC the other access methods to dselect won't install a package if the same or newer version is already installed. a) If we are talking about the missing 'C locale' problem, its the latest 'xlib6g' X11 package that solves this (it has the missing dir in it). Ooops, ... Sorry, somehow my brain got mixed up. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgphlTWSfOhTZ.pgp Description: PGP signature
Re: libc and staroffice 5
Hi! Peter Bartosch ([EMAIL PROTECTED]): after upgrading libc on my hamm-box to slink´s 2.0.7u-4 my apt won´t install anything further! IIRC you are currently running apt_0.1.6, aren't you? i´ve installed the new libc´s for staroffice 5. would installing slink´s apt (0.1.7 iirc) fix such problems? maybe. BTW, currently apt is at 0.1.9. But I'd say wait for libc6-2.0.7u-5, and install this one. There are currently some problems related to -4 - check debian-devel for details. Updating package status cache...done Checking system integrity...dependency error You might want to run apt-get -f install' to correct these. Sorry, but the following packages are broken - this means they have unmet dependencies: apt: Depends:libc6 but libc is installed: ... This might be a versioned depends/conflicts (sorry, no apt_0.1.6 avail. to verify). Check this either in dselect or with the help of dpkg --status apt Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp5VDJYq4beq.pgp Description: PGP signature
Re: smbmount via /etc/fstab?
Hi! Marc Haber ([EMAIL PROTECTED]): How can I do smbmounts via /etc/fstab? I'm not 100% sure if this is the correct link, but you might take a look at http://www-scf.usc.edu/~vibber/linux/smbfs.html . IIRC there are at least some links to patches against mount to do this. I can't verify the link - I'm currently in a no-web situation ... Please don't be disappointed if this only is the link to a wrapper script for 2.1.x's smbmount. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpqpwVEVZxx1.pgp Description: PGP signature
Re: Problem starting X for first time
Hi! Amanda Shuler ([EMAIL PROTECTED]): did you run xf86config? maybe he has ;-) IIRC there are situations where some parameters aren't set as chosen. edit /etc/X11/XF86Config (it has a manpage :-) and go to Section monitor. set HorizSync and VertRefresh according to your monitors manual. On Thu, 19 Nov 1998, Bret Craw wrote: ... Fatal server error: No valid modes found _X11TransSocketUNIXConnect: Can't connect:errno=111 Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpIjLzBKtl6Q.pgp Description: PGP signature
Re: Debian2.0/XFree86 3.3.2.3 on TOSHIBA 4400: X does not start
Hi! Marc Fleureck ([EMAIL PROTECTED]): Installed all relevant packages (xbase, xbase-clients, xlib6, libc6, and xserver-vga) without errors. When we do: startx It complains: X: exec of /usr/bin/X11 failed. put your desired Xserver in the first line of /etc/X11/Xserver - replace the wrong /usr/bin/X11 for example with /usr/X11R6/bin/XF86_VGA16 Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpEu8GqUaQDT.pgp Description: PGP signature
Re: Updated installation disks/images?
Hi! Amanda Shuler ([EMAIL PROTECTED]): On Thu, 19 Nov 1998, Rainer Clasen wrote: Hi! Amanda Shuler ([EMAIL PROTECTED]): Is there anyone out ther who has the installation images for Debian 2.0 with a kernel version 2.0.36 or greater? I have 2.0.34 and I am having trouble with booting... it freaks out when it detects my SCSI card. It's an Adaptec AIC-7xxx; I was told that the newer kernels have better support for Adaptec cards. I heard there are debian images at the aic7xxx distribution sites. IIRC they are not yet 2.0.36, but with an up to date aic7xxx driver. ftp://ftp.dialnet.net/pub/Linux/aic7xxx ftp://ftp.redhat.com/pub/aic Uhhm - and some time before that I heard there are images at http://www.debian.org/~doko but I'm not shure about this one. You may check the archive of debian-user. next thing I'd try is using slink bootdisks. Enrique Zanardi said he'll upload new bootdisks (fitting on 1440 disks) in the next days. IIRC 2.0.35 in slink already has the updated aic7xxx driver or is going to get it. You might take a look at one of the incoming mirrors - eg. ftp.de.debian.org/pub/debian-incoming Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp7MHsNijpd0.pgp Description: PGP signature
Re: Problem with X...Windowmaker...locale?
Hi! Ben Jorgensen ([EMAIL PROTECTED]): Ok, lets see. I assume you've installed 3.3.2.3a-7, haven't you? Either update to 3.3.2.3a-8 or reinstall xbase. According to debian-user this fixes the local problem. Have you installed those new packages: xmodmap, xfonts-*, xterm? Is there some problem with apt/dselect as well because nothing happened when I did Update in dselect. I had to get it and install it manually.. Do you mean reinstalling xbase_3.3.2.3a-7? apt and IIRC the other access methods to dselect won't install a package if the same or newer version is already installed. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpDEqfnKu28n.pgp Description: PGP signature
Re: cd-rom problems
Hi! Mike Fetherston ([EMAIL PROTECTED]): I've been having troubles installing debian on my system. If I run boot.bat from \boot it starts the installation just fine, but then when it reboots for the first time to install the packages using dselect it can't find my cdrom. My cdrom is an LG Electronics/Goldstar 8160B 16X IDE on the secondary IDE channel as master and it is all alone on the cable. If I create boot diskettes using resc1440.bin, I can use my cd-rom. From chatting on #debian on irc.debian.org it seems to be a kernel problem. Is there any way of fixing this so that when i boot from my hard drive i can gain access to my cd-rom to complete the installation? By loooking at /var/log/messages it seems that boot isn't finding my secondary pci ide channel. what kind of motherboard / chipset / IDE-controller do you have / does linux show in in /var/log/messages? which kernel are you running? the one from the rescue disc? Or did you somehow install another kernel (-package)? sis you try booting with the rescue disc and specifying the root-fs at the prompt like this: rescue root=/dev/YourRootPartition Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpDAYMS6QuRO.pgp Description: PGP signature
Re: where'd my cd go?
Hi! Richard E. Hawkins Esq. ([EMAIL PROTECTED]): does this happen with any CD? Or only with this newly purchased ones? Do you have another drive handy for testing the readability of the discs? All of the CD's, even the one's i've already installed. And it doesn't cooperate under drdos, either, though I have no idea whether my drdos installationis in good shape or not. I think i tampered with the silly config files several times. Uhhm, neither DRDOS nor Linux read any CD? Maybe the drive got broken? Nov 18 19:00:10 eyry kernel: hdc : tray open Nov 18 19:00:10 eyry kernel: end_request: I/O error, dev 16:00, sector 0 Nov 18 19:00:10 eyry kernel: FAT bread failed Nov 18 19:00:10 eyry kernel: hdc : tray open or drive not ready Nov 18 19:00:15 eyry last message repeated 7 times Nov 18 19:00:15 eyry kernel: hdc : tray open Nov 18 19:00:15 eyry kernel: end_request: I/O error, dev 16:00, sector 64 Nov 18 19:00:15 eyry kernel: isofs_read_super: bread failed, dev 16:00 iso_blknu m 16 block 32 Nov 18 19:00:15 eyry kernel: hdc : tray open or drive not ready looks familiar. One of my dying drives spits this out, too. daemon.log:Nov 17 18:05:03 eyry modprobe: can't locate module block-major-11 block major 11 is for SCSI SCSI cdroms. Are trying to mount /dev/scdX ? daemon.log:Nov 18 12:26:12 eyry modprobe: can't locate module net-pf-5 That's appletalk - unrelated. see /etc/conf.modules on how to avoid this message. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpnGGUiss5Gq.pgp Description: PGP signature
Re: Problem with X
Hi! Pierre-Antoine CHAMPIN ([EMAIL PROTECTED]): Since yesterday, I can't run X anymore : the graphical screen and mouse appear for one second, then X stops with the message : waiting for X server to shut down FreeFontPath: FPE /usr/X11R6/lib/X11/fonts/misc refcount is 2, should be 1; fixing. Uhhm, I don't think thats your problem. I have this message from time to tome, too. I suppose the problem is hiding in the previous output. Run startx 21 | less to get the full output. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpUmcZiUi9EK.pgp Description: PGP signature
Re: Xconsole
Hi! [EMAIL PROTECTED] ([EMAIL PROTECTED]): After I upgraded to slink, my xconsole doesn't start automatically any more (and doesn't appear on the xdm login either) The only trace I found is the line run-xconsole in /etc/X11/config /etc/X11/config ist split into several files. you need to put run-xconsole into /etc/X11/xdm/xdm.options Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpnJaN1vChwV.pgp Description: PGP signature
Re: Updated installation disks/images?
Hi! Amanda Shuler ([EMAIL PROTECTED]): Is there anyone out ther who has the installation images for Debian 2.0 with a kernel version 2.0.36 or greater? I have 2.0.34 and I am having trouble with booting... it freaks out when it detects my SCSI card. It's an Adaptec AIC-7xxx; I was told that the newer kernels have better support for Adaptec cards. Sorry for my ignorance, but have you tried installing with linux aic7xxc=no_reset entered at installation disk's boot prompt? If there is someplace online where I can download the images, that would be great as well. I have already tried to download kernel sources and replace the kernel image on the installation diskette with an updraded kernel, but it is always telling me that the new kernel is corrupted or invalid. what did you copy to the disk? arch/i386/boot/(b)zImage? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpXBC5GgxAzB.pgp Description: PGP signature
Re: Moving computer without losing things
Hi! Jose L. Gomez Dans ([EMAIL PROTECTED]): Hi! I have been using a debian system for a couple of months now, and I am extremely happy with it. However, since this isn't my computer, I have been told that I'll have to leave this computer, and take a newer one. I was wondering if there is a way to actually send all the files I've got here (I've installed loads of packages, compiled loads of sources) to the new computer (I could install debian on that one without much hassle) over the ethernet connection we've got. At the moment, the only alternative is to copy everything to floppies or zip drives :DDD. There MUST be a better option :))) idea 1: - enable rshd for root on your old machine (kind of dangerous) - get a rescuedisk with rsh and cpio on it (Toms root + boot disks??) and boot your new computer with it. - partition your new computers disk and mount all partitions the intended way under eg. /mnt - run as root cd /mnt rsh oldbox find / -xdev -depth | cpio -o | cpio -i --sparse expand / to the whole list of fs'es you want to move. - run lilo -r /mnt idea 2: - make a cpio/tar file of your old machine's disk find / -xdev -depth | cpio -o /var/tmp/foo.cpio - export /var/tmp (or wherever you placed the .cpio) via nfs - use debian's rescue disk to boot the new machine, start setup, load your NIC's modules, load nfs module, configure the network, act as if you're install to a bogus partition (you're going to delete later) mount your desired partition setup eg. under /foobar open vt2 and do mount oldbox:/var/tmp /mnt -orsize=4096 cd /foobar star /mnt/foo.cpio Uhhm, please take 2. only as idea - possibly you need to use tar insted of cpio - and there may be other issues I missed. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgppgNatlcUQn.pgp Description: PGP signature
Re: pt_BR or br_BR locales in Xlib
Hi! Mário Olímpio de Menezes ([EMAIL PROTECTED]): I have my locale set to pt_BR in profile. After upgrading to the new xlib6g, netscape is giving me this message: netscape: locale `pt_BR' not supported by Xlib; trying `C' according to another post, you are running 3.3.2.3a-7, aren't you? There is a bug with its locales. Either update to -8 or reinstall (the same) xbase_3.3.2.3a-7.deb Maybe this kills your Staroffice, too? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgptntJbXaPtR.pgp Description: PGP signature
Re: startkde cannot connect to X server
Hi! Stephan Böni ([EMAIL PROTECTED]): i have installed kde on debian linux 2.0. commands like startx or startkde return the error cannot connect to x server. what's wrong? If running startx says cannot connect to xserver this usually indicates that your Xserver isn't configured properly. Install doc-linux-text and read /usr/doc/HOWTO/XFree86-HOWTO.gz and /usr/doc/X11/* . *THEN* run either XF86Setup or xf86config to configure your xserver. startkde is run within X. see /usr/doc/kdebase/README.Debian. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpdjXQTFqSmX.pgp Description: PGP signature
Re: Permissions for vfat mounted drive
Hi! John Stevenson ([EMAIL PROTECTED]): Any ideas how I can get no root access to my win98 partion, or is it a Fat32 problem. I use this in my fstab: /dev/sda1 /dos/c vfat umask=0002,gid=110,uid=0,showexec 0 2 gid 110 is called dosfs, but you may use dos's gid. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgppoRzBkiJOG.pgp Description: PGP signature
Re: IPFW works a little to well...
Hi! Anthony Landreneau ([EMAIL PROTECTED]): below. The problem, when I execute the script nothing comes in, nothing goes out, the perfect firewall. The bad news is I need some traffic to pass. The network behind the firewall is a subnet of a class B network with Uhhm, it seems you mixed some source/destination ports. I'm not shure about -b - I've removed it (by mistake??) I'd try the following modifications: # By Default DENY ALL services first ipfwadm -F -p deny # # Flush all Commands ipfwadm -F -f ipfwadm -I -f ipfwadm -O -f # # Allow email to NCTAMS01 ipfwadm -F -a accept -b -P tcp -S 0.0.0.0/0 1024:65535 -D 111.229.13.13 ipfwadm -F -a accept -P tcp -S 0.0.0.0/0 -D 111.229.13.13 25 # Allow email to NS1 Relay host ipfwadm -F -a accept -b -P tcp -S 0.0.0.0/0 1024:65535 -D 111.229.13.2 ipfwadm -F -a accept -P tcp -S 0.0.0.0/0 -D 111.229.13.2 25 # Allow email to outside mail severs from NCTAMS01 ipfwadm -F -a accept -b -P tcp -S 111.229.13.13 25 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -P tcp -S 111.229.13.13 -D 0.0.0.0/0 25 # Allow email to ouside mail servers from NS1 ipfwadm -F -a accept -b -P tcp -S 111.229.13.2 25 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -P tcp -S 111.229.13.2 -D 0.0.0.0/0 25 # Allow DNS traffic to NS1 ipfwadm -F -a accept -b -P udp -S 0.0.0.0/0 53 -D 111.229.13.2 ipfwadm -F -a accept -b -P tcp -S 0.0.0.0/0 53 -D 111.229.13.2 ipfwadm -F -a accept -b -P tcp -S 111.229.232.0/24 82 -D 111.229.13.2 # what is port 82? I'm skipping this one ipfwadm -F -a accept -P udp -S 0.0.0.0/0 -D 111.229.13.2 53 ipfwadm -F -a accept -P tcp -S 0.0.0.0/0 -D 111.229.13.2 53 # Allow Web connections to outside Web Servers ipfwadm -F -a accept -b -P tcp -S 111.229.13.0/24 80 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -b -P tcp -S 111.229.13.0/24 -D 0.0.0.0/0 80 # Allow FTP connection to outside Servers ipfwadm -F -a accept -b -P tcp -S 111.229.13.0/24 20 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -b -P tcp -S 111.229.13.0/24 21 -D 0.0.0.0/0 1024:65535 # not touching FTP - trying to avoid emitting bogons. # Allow Telnet connections to outside Servers ipfwadm -F -a accept -b -P tcp -S 111.229.13.0/24 23 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -P tcp -S 111.229.13.0/24 -D 0.0.0.0/0 23 # Allow NTP time to NS1 ipfwadm -F -a accept -b -P tcp -S 111.229.13.2 123 -D 0.0.0.0/0 1024:65535 ipfwadm -F -a accept -P tcp -S 0.0.0.0/0 -D 111.229.13.2 123 Ok, since my packet filter is less restrictive (and based on an allow-policy) I'm definatly not shure if this is correct. At least the first packet to initiate a connection should find it's way :-) Don't know if the adressed host is able to reply. Try re-adding -b's if it doesn't work. I wonder why you're (only) using a forward rule. I'd set the Incoming rules(, too). Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp5LiEgTt4SO.pgp Description: PGP signature
Re: Problem with X...Windowmaker...locale?
Hi! Ehren Wilson ([EMAIL PROTECTED]): Hello I just recently started running slink and I have run into a a problem with X or Windowmaker. XDM starts up fine, but when I log in, the screen goes black and then returns me to the login screen. I have included my .xsession-error file to see if anyone has the same problem or knows how to fix it. Ok, lets see. I assume you've installed 3.3.2.3a-7, haven't you? Either update to 3.3.2.3a-8 or reinstall xbase. According to debian-user this fixes the local problem. Have you installed those new packages: xmodmap, xfonts-*, xterm? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgppZSMnUUu0E.pgp Description: PGP signature
Re: where'd my cd go?
Hi! Richard E. Hawkins Esq. ([EMAIL PROTECTED]): Unless my memory has faded even more, a possibility I won't rule out, I used to be able to mount cd's with mount /dev/hdc /cdrom -t iso9660 I now get, mount: block device /dev/hdc is write-protected, mounting read-only mount: wrong fs type, bad option, bad superblock on /dev/hdc, or too many mounted file systems does this happen with any CD? Or only with this newly purchased ones? Do you have another drive handy for testing the readability of the discs? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgps6xX9Rfvq1.pgp Description: PGP signature
Re: Weird behavior with Tcsh
Hi! Ossama Othman ([EMAIL PROTECTED]): The prompt in my xterm-1 still shows ~/test/me. When I do an ls in xterm-1 I get no error. Shouldn't I get an error since the directory isn't supposed to exist anymore. Is this some behavior in tcsh that I wasn't aware of? In the past I used to get errors from ls of a non-existent directory if the directory was removed when I was in that directory, at least from what I recall. Am I missing something here? I think so. You didn't remove the directory - It still exists. Its Inode is still the same. speculative mode on If ls uses the . entry of the current directory, everything is absolutely fine. if you you do ls ~/test/me you do get an error. speculative mode off Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpRxnhmwSfSS.pgp Description: PGP signature
Re: Problem!
Hi! Andreas Rasmussen ([EMAIL PROTECTED]): ...some people still use modems, I do! :-( ie. you can't install directly from ftp/http? A friend of mine downloaded and created the 3 Debian ISO-CD-IMAGES from you're mirrors... Everything works fine untill I come to the dSelect program.. Some packages is on the first cd (BINARY.RAW) and som are on the second cd (CONTRIB.RAW).. Ahh, I see, the multi-CD problem. There are some workarounds: a) ignore the CONTRIB CD for now. b) see a), but use dselect only with main and add contrib/non-US/non-free packages manually. (HINT: dpkg --info foo.deb showes you the dependencies, dpkg --install foo.deb installs it.) c) get a second CD-drive and mount both CDs d) copy ony CDs to harddisk e) copy the Packages.gz files and one .deb for each section (contrib/ non-free/ non-US) to eg /debian/CONTRIB (same path as on the CD) and when dselect wants to install packages from this CD mount the CONTRIB cd as /debian/CONTRIB. This is kind of advanced - and untested ... Contact me to make me work out more details. Maybe some other guys on this list have easier ideas. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpp9nvgmYV72.pgp Description: PGP signature
Re: How do you use su under X?
Hi! Sorry, for hiding this to you - my direct reply to Chip Grandits should have gone to the list ... Helge Hafting ([EMAIL PROTECTED]): I want the cookie security so other users can't do anything wrong. I also want root to be able to use the X display no matter who is running x. What is the simplest way to achieve this? sudo xterm ssh -l root localhost xterm su enter pass export XAUTHORITY=/home/your login/.Xauthirity Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpxJMO7DQYYs.pgp Description: PGP signature
Re: Console mode with 80x24 possible?
Hi! Eric House ([EMAIL PROTECTED]): Can I rebuild the kernal to use my entire display? My hamm-equipped laptop has an 800x600 screen. When in console mode it uses only the middle of the screen for an 80x24 display -- even though there's room on the screen for at least 120x32. try adding vga=extended or vga=ask to your boot-prompt Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpCT8zFDPbNP.pgp Description: PGP signature
Re: keyboard switchers
Hi! Chris Evans ([EMAIL PROTECTED]): Anyone know if Linux/Debian (Hamm) gets irritated if it finds no keyboard where one was a few minutes/hours before? Uhhm, I don't think so. I'm playing the dangerous game of hot-plugging keyboards for some years. I use one keyboard with 7 boxes. One of them used to be NT. There was only *one* occurance I lost keyboard input with a Dell desktop running Debian. I suppose it was the BIOS - not debian ignoring the keyboard. Telnetting in and rebooting fixed it. The only annoyance of hot plugging keyboards is the lossage of my faster typematic rate :-( Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgp2lkBZ5FUhD.pgp Description: PGP signature
Re: lsof problem
Hi! Groumph ([EMAIL PROTECTED]): lsof does not work on my system: groumph:~# lsof /dev/audio lsof: WARNING: compiled for Linux release 2.0.35; this is 2.1.125. ^^^ Of course it doesnt't! Take a look at /usr/doc/lsof*/00FAQ.gz You need to recompile lsof to use 2.1's proc-based interface. get the source (.dsc .orig.tar.gz .diff.gz) of your debian CD or favorite mirror, extract it with dpkg-source -x lsof*.dsc, (?modify debian/control,?) and run debian/rules binary Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpInmYWHFFVX.pgp Description: PGP signature
Re: problems after upgrading libc6
Hi! [EMAIL PROTECTED] ([EMAIL PROTECTED]): Checking system integrity...dependency error You might want to run apt-get -f install' to correct these. Sorry, but the following packages are broken - this means they have unmet dependencies: apt: Depends:libc6 get apt 0.1.9 from slink and hope nothing else is brocken. Currently I suspect slink's libc breaking rwhod. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpvBf4ZJ8Pd9.pgp Description: PGP signature
Re: SMP
Hi! [EMAIL PROTECTED] ([EMAIL PROTECTED]): The motherboard is an AMI Titan 3 - w/ two 166MHz Pentums - 64MB RAM. Here's what I know: Linux version 2.1.123 ([EMAIL PROTECTED]) (gcc version 2.7.2.3) #3 SMP Fri Oct 16 10:22:12 EDT 1998 oSo I know I've got SMP support compiled in... Hmm, IIRC a friend of mine had an option in his Gigabyte BIOS to switch of SMP support in a way. Maybe this board has this, too? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpjZBt9aHxv6.pgp Description: PGP signature
Re: Diamond Viper 330 AGP card.
Hallo erstmal! Alexander Kushnirenko ([EMAIL PROTECTED]): Is X running on Diamond Viper 330 AGP card? Could someone comment is it a good card? A friend ran it with hamm's SVGA server. I switched his setup to nvidia's peliminary server, because there was noise on moving windows. Now it works satisfying, but the server soaks up 20MB RAM. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpYBDuvJTYUm.pgp Description: PGP signature
Re: how do I extract a 2.6 gigabyte .tar.gz file ?
Hi! Darxus ([EMAIL PROTECTED]): | Any more ideas ? :) Only thing I can think of is getting access to a 64-bit machine, decompressing the file there, tarring the contents off to tape and then restoring them on your machine. Or at least putting them into sub-2GB chunks before taking them back. I don't know how to get it there, but you might be able to read it from a raw device (/dev/hdd /dev/nst0 ...) Maybe this works: Since you have it on fat32: defragment this file, find where it is located and use dd if=/dev/hd?? skip=XX count=YY | zcat | tar... Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: SNMPD - MRTG- PPP stats iregularity
Hi! Steve Lamb ([EMAIL PROTECTED]): On Fri, 23 Oct 1998 19:48:19 -0400, Michael Stone wrote: Linux doesn't ordinarily have any facility for saying _how much_ data goes through an if. The snmpd's base their throughput figures on _number of packets_, which isn't exactly the same thing. IIRC, you can run ip accounting and hack the snmpd's to get their numbers via that mechanism. Well, that blows. Packets don't mean diddly compared to bytes when it comes to bandwidth utilization. Do you know if that will be fixed any time soon? 2.1.* maybe? Yes. 2.1 counts bytes in /proc/net/dev. But IIRC there are still some NIC-drivers which don't supply these numbers. tulip does. I've no clue about SNMPD, but I suppose you'll have to teach it to use the byte counters, too. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: two things
Hi! Shane Wegner ([EMAIL PROTECTED]): 2) When I installed slackware, I did it over a null-modem serial connection to an ms-dos machine running a program called telix. I do this because I use screen reading software to read the screen and as of yet, there is nothing available under Linux. Therefore, in slackware I log in as root and go. /sbin/agetty 9600 ttyS0 vt100 and do the whole setup remotely by running setup from the remote machine. Is this possible to do while using Debian? I understand that it goes into the setup program as soon as you boot the bootdisk or cdrom so you'd have to break out somehow and spawn the agetty and run it from the remote machine. Anyone have any hints on how to do it? Hmm, I have thought about this, too. I run several boxes with a serial console, but actually I never tested to install one over serial line. It's not as easy as you describe with slackware - there is no getty on the rescue disc. dinstall (the setup program on the rescue disc) doesn't need one. Maybe It's still doable with a modified rescue disk: patch the kernel with sercons support (to see boot messages on your serial line and to hopefully fake your real tty) or use a recent 2.1 kernel make /dev/tty0 what usually becomes /dev/console for sercons kernel (dinstall seems to open /dev/tty0 instead of /dev/console). Alternatively one may patch dinstall to open the required /dev/console. replace syslinux with lilo (to get your bootprompt on serial line). Alternatively you may hardcode the boot parameters in syslinux.cfg. hope dinstall doesn't fail eg. when loading the keymap... I'll try this if you're interested. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpMMFbvMgCiY.pgp Description: PGP signature
Re: permission denied in root.
Hi! Brent McMillan ([EMAIL PROTECTED]): I'm trying to get the wordPerfect demo up and running, but when I try to run the setup executable (./Runme) I get a permission denied. the file is executable, I own it, I own the directory, I tried as root and non-root alike. The directory is not mounted on my main partition (if that has anything to do with it. No executable files will run in this directory. What's going on? Uhhm, ./Runme is on a cd? You have a line in your fstab for your cdrom with option user? This means you mounted the disc noexec - your not allowed to run anything from it. try mount /dev/cdrom /cdrom -tiso9660 -oro as root Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpMZ8bXoxolD.pgp Description: PGP signature
Re: mountd/nfsd
Hi! David Boisvert ([EMAIL PROTECTED]): Ok, first, my mountd server always says that I can't mount the server from the client. cat /etc/exports: / 192.168.1.1(rw) #client side this is server's /etc/export, isn't it? Here is the error in syslog: Oct 18 09:37:32 server mountd[148]: Unauthorized access by NFS client 192.168.1.1 Oct 18 09:37:32 server mountd[148]: Blocked attempt of 192.168.1.1 to mount / check this on your server: Did nfsd and mountd reread /etc/exports after editing? hamm: /etc/init.d/netstd_nfs reload slink: /etc/init.d/nfs-server relaod other linuxes kill -HUP pids of mountd and nfsd Do /etc/hosts.{allow,deny} allow access to portmap from 192.168.1.1? Consult man 5 hosts_access and /usr/doc/netbase/portmap.txt.gz for more details. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgproWZI0Znxv.pgp Description: PGP signature
Re: dhcp and/or nis questions?
Hi! Lee Bradshaw ([EMAIL PROTECTED]): On Wed, Oct 14, 1998 at 11:18:13AM +0200, Miquel van Smoorenburg wrote: In article [EMAIL PROTECTED], Lee Bradshaw [EMAIL PROTECTED] wrote: nis specific problems: When I run dhclient, /etc/resolv.conf is written with the new dns info, but /etc/defaultdomain doesn't change. I guess the file doesn't need to change, but the value returned by (nis)domainname isn't modified either. I'm using the dhcp-beta_2.0b1pl6-0.1 and dhcp-client-beta_2.0b1pl6-0.1. I set the domainname to a garbage value, ran dhclient, and then expected to see the updated nis domainname from dhcp. But on my system the nis information doesn't appear to be updated by dhcp. Am I missing something? Here is the dhcp server nis information again: option nis-domain alantro.com; option nis-servers 192.168.50.1; Do I need to do anything on the client to enable dhclient to override a previous value for the nis domain? theoretically this should work: a) put apropriate request / require lines in your dhclient.conf. dhclient only asks for dhcp-options it needs / you told it to ask. cat /var/db/dhclient.leases to see what options it got. b) You need to extend your /etc/dhclient-script. By default only the bare minimum of all those nice dhcp-options is used: IP adress(es), ip-alias(es), default route(s), netmask?, broadcast, DNS domain name, nameserver(s). Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpI6gof9pBlA.pgp Description: PGP signature
Re: Hiding a linux computer
Hi! Jeremy Blonde ([EMAIL PROTECTED]): However a few problems arise. They are using dhcp, which I can work with (I get the ip address and can hit all the servers, etc.), but this also leaves a record in the dhcp ip address listing in this format: ipaddress hostname mac address The Linux machine shows up like this: X.X.X.X (null) XX:XX:XX:XX:XX While NT machines show up with an actual hostname not null. try putting a send host-name statement in your dhclient.conf (see man 5 dhclient.conf for details) I'd like to not even use the dhcp server but I think that would mean I'd have to setup the Linux machine to be a DNS server wouldn't it? No! Not using dhcp means guessing a free IP - or better getting a real staic one. I don't know what that would do to the rest of the machines on our network, I can't be messing them up or else I'll be in hot water. You can mess at least one box on your net - if you use the same IP as it. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgprr6744PUsN.pgp Description: PGP signature
Re: DebianNet.pm
Hi! Mario Olimpio de Menezes ([EMAIL PROTECTED]): I upgrade some packages yesterday and found a problem with the above module. Some *.postint script tried to include it from /usr/lib/perl5/i386-linux/5.005 but it's installed at /usr/lib/perl5. This broke my upgrade and was solved only with a ln -s to the right location. Is there some other way to do this? How to inform perl where are the .pm modules located? Any more had this problem? I'm using slink. IIRC this is already discussed on debian-devel and/or already filed as bug. This dir seems to be missing from @INC. I symlinked /usr/local/lib/site_perl - /usr/lib/perl5 as workaround (no local perl stuff installed). Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: chmod help
Hi! E.L. Meijer Eric ([EMAIL PROTECTED]): [EMAIL PROTECTED] (Default Debian Reader) | | How can i change the perms on a directory so that a file that is copied into that directory is automatically owned by another user and group? Don't know if anyone mentioned it already, but you can actually do this for the group ownership. If you do a `chmod g+s dir', new files created in this directory will get the same group. Hmm, do you have an idea how to deal with an anonFTP like incoming directory, too? This dir is direct accessible and (the big problem) the guy to sort all this stuff will (hopefully) never get root. I made a diaradm group and added this guy to it. Then made the dir sgid diradm. guy has to cp things to its new location to change owner. drawbacks - luser xyz does chgrp xyz subdir and guy is locked out :-( - cp may have some problems when space becomes limited and its SLOW compared to a mv (on the same partition) Do I need to write somethind suid root which guy can run (like chown -R guy dir )? Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB
Re: using debian passwords in NT.
Pere Camps ([EMAIL PROTECTED]): Does anybody know of a way of making winNT use the usernames and the passwords of a debian machine? Maybe you can try NISGINA. It appears to turn your NT box into a NIS client. Can't tell more - I've never got to use it. Try this URL, I don't know if its still valid http://www.dcs.qmw.ac.uk/~williams/ Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpOcAyzWYeUO.pgp Description: PGP signature
Re: Reading CD image from CD: slow!
Hi! Zini Enrico ([EMAIL PROTECTED]): When I try to read the ISO image from a CD, either with cat or dd, the reading speed (I read with a SCSI 4x/2x Philips CDD2600 cd-writer) seems to be around 70Kb per second (half 1x !). If I mount the CD I can read at full 4x, no problem: what could it be? Do I need some utility to set cd reading speed or such? I suppose dd's default blocksize of 512 Byte kills performance on several CD-Drives. I use dd if=/dev/scd0 of=tmp.iso bs=2k count=$(( `isosize /dev/scd0` / 2048 )) which is reasonable fast (and avoids reading the unreadable sectors at the end) Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpbiHYB9zjSP.pgp Description: PGP signature
Re: Hard drive sriping
Hi! john mcpeek ([EMAIL PROTECTED]): know how to set them up to work as one drive. I saw something about drive striping and I installed the leaner drive module. I need to know how I should format the drives and how to make it think two drives are really one. It's explained in the Software-RAID mini-HowTo. Maybe you'd also want to take a look at the Root-RAID-HOWTO. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpcsAgewE0K6.pgp Description: PGP signature
Re: apt - priorities?
Hi! B. Bell ([EMAIL PROTECTED]): 2) Track everything in unstable, but only try to upgrade a package if a) the user requests to have the package upgraded b) another package depends on the new version I use this workaround: Usually I have unstable commented out in sources.list. If I want to install a single package from slink re-add it to sourcec.list, run apt-get update apt-get install package remove it from sources.list and rerun apt-get update. It's not perfect, since new packages from slink are listed as orphaned, but I can live with it. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpKwd0fVpFLd.pgp Description: PGP signature
Re: Adaptec 2940UW failure with current rescue disks; Buslogic BT-958 succeeds.
Hi! Jameson Burt ([EMAIL PROTECTED]): Debian rescue disks have supported my Adaptec 2940UW disk for 1 1/2 years. Now, none of the rescue disks in .../dists/hamm/disks-i386 work with my Adaptec 2940UW. I am not concerned about this problem; I do not seek a solution since I have a very old rescue disk. Still, I welcome comments in the interest of others and my understanding. I imagine these solutions / workarounds: - try to enter linux aic7xxx=no_reset at the boot prompt. *Sometimes* this helps :-) - use a kernel older than 2.0.34 and put it on the boot-disk ( I suppose there is a readme how to do this.) Maybe you can simply use the kernel of a bo boot disk. This should be possible without linux, too. - compile your own kernel with current 5.1.x-pre patches and put it on the boot disk. Don't forget to read the aic7xxx list :-) This is the only way to deal with adaptec's new U2W controllers. - ask someone to compile a kernel with 5.1.x(-pre) drivers for you. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgptMYeIN27MK.pgp Description: PGP signature
Re: passwd vs. yppasswd under nis setup question
Hi! Stuart Marshall ([EMAIL PROTECTED]): the password file but not update the nis map. I seem to recall long ago that the nis installation made passwd call yppasswd? I am confused as to how it should be set up. currently /bin/passwd isn't linked to yppasswd on installation and I think this is the Right Thing (TM). passwd is still needed for local accounts. You can link it yourself: either put a link in /usr/local/bin (and hope everybody has this in its PATH before /usr/bin) or run these commands # dpkg-divert /usr/bin/passwd # mv /usr/bin/passwd /usr/bin/passwd.distrib # ln -s yppasswd /usr/bin/passwd But beware: Last time I tried, rpc.yppasswd (the daemon running on the NIS master) dumped core if I used yppaswd. If I use the +:: entry in the /etc/passwd file on the master, then the nis maps are used for logins, etc. But if you put +:: in /etc/passwd of the NIS master you - must put your passwords for the NIS map in another file - and tell rpc.yppasswd where to find this file Don't put the plus entry in NIS master's /etc/passwd if you use it as source for the NIS map. since the maps are unchanged after using passwd the passwd program reports success and the password appears to be unchanged to the user. passwd fails for accounts not in /etc/passwd. therefore passwd won't work on NIS clients. You can run passwd on the NIS master *if* you make your NIS maps from /etc/passwd. And as you observed correctly the NIS map won't be updated. The next time root runs make in /var/yp the maps are updated and the password is changed. This is a source of confusion naturally. so I suggest you either - tell all users to only use yppasswd ( *lol* ) - link passwd to yppasswd - move NIS accounts to another file than /etc/passwd (to make ordinary passwd fail) - patch NIS master's passwd to invoke make -C /var/yp after successfully changing a passwd - let cron execute make -C /var/yp Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB pgpZFicoqfkOn.pgp Description: PGP signature
Re: Several questions before I run Debian
Hi! Matthew A. Reklau ([EMAIL PROTECTED]): Do the X servers include support for the Diamond Viper V330? yes, they do. Some month ago I installed debian pre-hamm on a box with one. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] /dev/null
Re: I burned hamm on NT and now all the files end with ~1.
Kevin Traas ([EMAIL PROTECTED]): You probably burned it as a Joliet filesystem... not supported under 2.0.x kernels yes, but at least the current debian 2.0.33 kernel is already patched for joliet. To burn it for use under Linux, you have to burn it as iso9660 with RockRidge Extensions However, I don't know of any software on Win95/NT that supports that. (Adaptec CD Creator Deluxe does not.) I suppose mkhybrid does :-) I don't know where I found them, but somewhere I found a port to NT. IIRC it's using Cygwin.dll. I never used it, but it IS available. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Convert RedHat install to Debian ..
Nathan E Norman ([EMAIL PROTECTED]): I've got a Sun Netra I5, with serial console only. The current Debian boot disks don't like the serial console, so I used RedHat's instead. Now I've got this Netra Linux box :) I don't know sun at all, but how about modifying the bootdisks to spawn a getty and the installprog on ttyS? I tried to take a look at disks-sparc/root.bin, but (after gunzipping it) I wasn't able to mount it - Are they somehow special? So I refer to i386's root.bin. Indeed I never used anything else than i386, but maybe the basic Idea applies. root.bin doesn't contain a getty and it's init is the installation-prog. So you need to rename init to something else, copy a real init and getty to the disks, make an inittab. Probably this won't fit on the original image, so you need to make a bigger one. # pwd /var/tmp # zcat /pub/debian/hamm/main/disks-i386/current/root.bin root.img # ls -l root.img /sbin/init /sbin/getty -rwxr-xr-x 1 root root12468 Jan 24 22:48 /sbin/getty* -rwxr-xr-x 1 root root24188 Apr 15 17:09 /sbin/init* -rw-r--r-- 1 root root 2662400 May 12 11:47 root.img # set newsize=$[ ( 2662400 + 24188 + 12468 ) / 512 +1 ] # mount root.img /oldroot -oloop # dd if=/dev/zero of=newroot.img count=$newsize # mkfs.blah newroot.img # mount newroot.img /newroot -oloop # cd /oldroot; find -depth | cpio -pmd /newroot # cd /newroot/sbin; mv init install_prog # cd /sbin; cp init getty /newroot/sbin # vi /newroot/etc/inittab and add a line like T1:12345:respawn:/sbin/getty -m -l /sbin/install_prog -L ttyS1 9600 vt100 # umount /newroot [insert an install disk in fd0] # mount /dev/fd0 /floppy # cp /var/tmp/newroot.bin /floppy/root.bin # umount /floppy This is completely untested (I hate to reboot) and as mentioned above I have no clue regarding sun. Please let me know if this works or if it is nonsense. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to build Debian Linux cluster?
Pierre Blanchet ([EMAIL PROTECTED]): Here, i have 85 PC to install. And the method i will try will be slighty different (note: it will only work in a network, with a local mirror): 3) Write a shell script, dpkg-record, something like : Just out of curiosity: What speaks against using rdist? IIRC somebody suggested this in a previose discussion on this topic. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to build Debian Linux cluster?
Pierre Blanchet ([EMAIL PROTECTED]): RC == Rainer Clasen [EMAIL PROTECTED] writes: RC Just out of curiosity: What speaks against using rdist? IIRC somebody RC suggested this in a previose discussion on this topic. The idea behind dpkg-record/dpkg-replay is flexibility and simplicity. You can install whatever package wherever you want automatically (if you have installed this package already). Ahhh! Appearantly my assumption is false, that you want to keep all boxes as equal as possible ... Naturally thats no (easy) job for rdist. I like your idea to keep a sitewide default for package configuration. This just goes one step ahead detached installation (which curently is impossible with dselect) - wow! I Hope there will be a time Apt can do this by default. Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
lilo on serial line (was: Re: how to set up headless machine?)
Hi! Jack Kern ([EMAIL PROTECTED]): I'm not sure I understand what is required but the lilo doc, Manual.txt.gz, in the Global options section (/usr/doc/lilo/Manual.txt.gz) seems to have a relevant passage: SERIAL=parameters enables control from a serial line. The specified serial port is initialized and LILO is accepting input from it and from ... Did anybody get this running propperly? I added serial=1,9600n8 to my lilo.conf and on reboot I got lilo´s prompt, but wasn´t able to enter anything. When getty starts I can login. Or was this the fault of minicom running on the other end? BTW: I used bo´s lilo_19-2 Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Convert RedHat install to Debian ..
Hi! Nathan E Norman ([EMAIL PROTECTED]): I've got a Sun Netra I5, with serial console only. The current Debian boot disks don't like the serial console, so I used RedHat's instead. Now I've got this Netra Linux box :) AFAIK there is enhanced serial console support in 2.1.*. But don't forget, 2.1 is unstable! I have no experience with Suns at all, but Maybe you can try this: make a standard debian bootdisk use your RedHat system to compile a 2.1 kernel with console support and either replace the bootdisk's kernel with this one make the disks bootloader pass the necessary parameters for the serial console or install the kernel on your system copy the disks root image to your system make your bootloader (SILO?) load this image as initrd as the bootdisks do and pass the parameters for the serial console This is only an idea - maybe this works... Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
hamm's last soaks up 9MB RAM
Hi! I recently updated an 386 with 8MB to hamm and after fixing utmp/wtmp I ran last too check if everything is ok. It seemed to do nothing, neither ^C nor ^Z worked. I opened another ssh-connection and ps showed me root 210 6.9 49.5 9952 3400 p0 D13:39 0:02 last After a while last completed sucessfully. I copied wtmp to a faster hamm-box, which apearantly had no problems, but also used 9MB RAM. bo's last only takes 800k - so what's wrong?? BTW: its from sysvinit_2.74-4 Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Zen and the art of working with NT
Hallo erstmal! Stephen Carpenter ([EMAIL PROTECTED]): create a second extended filesystem in a file on the NT home dir and then mount that filesystem (I need a filesystem that will hole my some time ago I tried the same, and somebody told me loop support doesn't work with any remote FS. According to that I wasn't successful with smbfs and nfs. Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: supressing return mail with at
Hi! Richard E. Hawkins Esq. ([EMAIL PROTECTED]): I tossed together a 2 line script to run with at, to give me a ^Q every minute: echo -n bbb /dev/ttyp8 at -f testfile now + 1 minutes avoid output to stdout/err! at only sends mail if there is anything to send. see at(1) at -f testfile now + 1 minutes /dev/null 21 but why don't you use cron? see crontab(1,5) and run crontab -e to add a line like this: 1 * * * * echo -n bbb /dev/ttyp8 Gruß Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror
Hi! John Boggon ([EMAIL PROTECTED]): How can I get mirror to leave the debian/contrib and debian/non-free directories but still retrieve the hamm | slink/contrib, hamm | slink/non-free directories ? exclude_patt+|^contrib/ exclude_patt+|^non-free/ and when you dont't want mirror to create them as empty dir replace '/' with '(/|$)' eg: exclude_patt+|^contrib(/|$) Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
how do I compile with libc6?
Hi! ok, I tried to compile the little hello.cc example from libgtkmm-dev. I ran g++ hello.cc -lgtkmm -o hello -I/usr/lib/glib/include and got: /usr/bin/ld: warning: libc.so.5, needed by /usr/lib/libc5-compat/libXi.so.6, may conflict with libc.so.6 /lib/libc.so.5: the getwd' function is dangerous and should not be used. /lib/libc.so.5: the gets' function is dangerous and should not be used. /lib/libc.so.5: warning: siggetmask' is obsolete; sigprocmask' is best shouldn't g++ try to link with libc6? I recently copied and updated the copy from bo to hamm using autoup.sh. maybe I broke something? BTW: updating with dselects mount method (and a slinghly imperfect mirror) is a real hassle! Why didn't I use apt-get from the beginning? At least I had to run ldconfig by hand, since an ldd on a freshly compiled classic hello_world (stabdard C) said it was neither a.out nor ELF :-( I have no libc5-*dev installed - so what did I do wrong? some versions: gcc:2.7.2.3-4 g++:2.90.27-0.3 binutils: 2.8.1.0.23-1 libc6: 2.0.7pre1-4 libgtk1:0.99.9-1 libgtkmm: 0.7.18-1 Regards Rainer -- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
patch for 2-step mirror-ing
Hi! Phone fees prevent me to mirror over my modem. But since I have access to a fast connection at my university, I practice a 2-step method: Make a list of files to get on my local machine runing mirror -n other options | sed -n 's/^get file \(.*\) as .*/\1/' and use the fast connection to put them on a tape. Then I restore the tape at home and rerun mirror (still working on an offline sollution). Unpatched mirror needs to be connected during the list generation - even when used with a local_ls_lR_file (uhhh - expensive !!) I modified mirror (2.8-16bo) to avoid staying/getting online when run with -n and a local_ls_lR_file. I *think* I didn't break mirror when called without -n and/or local_ls_lR_file. Since someone asked for a similar solution some time ago, I think others may also be interested in this patch. I'd appreciate, if someone could test it. Since it is 2K gzipped I don't want to bother the list with it. Please ask and I'll send it to you. Maybe this could be include in the debian-package and/or forwarded upstream if it stands the test? Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: embarassing X questions
Hi! Christopher J. McNicholas ([EMAIL PROTECTED]): 1) Is there somewhere I can read up on the different types of windows managers? http://www.plig.org/xwinman/ 2) I loaded X without a windows manager installed. I can move my mouse around, but that's about it, no menus or anything. I switched VCs just to make sure I wasn't locked up :-).. The question is..how do I properly close out of X without a windows manager? CTRL-ALT-BACKSPACE (if not disabled) terminate the process exec'ed from .xsession or /etc/X11/Xsession eg: type exit in an xterm. or switch to a VC an kill the Xserver or the process that was exec'ed from xsession Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Running X11 over a network
Hi! Uhmm, I need to correct myself: I think removing xbase is only recommended if you exactly know what you are doing. It contains all manpages, documentation, the X wrapper, all config-files ... And since xbase depends on xlib6 you cannot remove it, too. Maybe someone else has a step-by-step instruction to cope without xbase... You may still want to use the laptop's fonts (via xfs) to avoid duplicate installation. The basic fonts do not need much space, but maybe you are curiose enought to try to get along without xbase - and then you must get them from a font-server. ok, I'll try to cleanup a little: 1) set up your network (see Ben Pfaff's mail, Net-3-HOWTO) I assume you gave your desktop the hostname desktop and your laptop the hostname laptop 2) set up the laptop: - install xbase, xlib6, xlib6g (if you're running hamm), xfntbase, xfnt75 - make it run xdm and xfs: Put start-xdm and start-xfs into /etx/X11/config to start them automatically on reboot. See /usr/doc/X11/debian.README for further information about this file. If you want to avoid a reboot, you can start xdm / xfs manually by executing /etc/init.d/{xfs,xdm} start 3) set up your desktop: The absolute minimum package to *run* your desktop as X-terminal is: xserver-*the Xserver you need for your graphic card And since it simplifies the setup, provides doc, ... xlib6if you are running hamm, you need xlib6g instead xbase During comfortable setup you additionally need Xserver-vga16it contains XF86Setup xfntbase - install all above mentioned packages - run XF86Setup to configure the xserver - run startx to test the configuration (maybe tweak it with help of xvidtune) - run dselect and purge xserver-vga16 (if you don't intend to run it) xfntbase HINT - this would be the point to remove, not purge xbase. Removing keeps the config-files (XF86Config, ...). And if you use /usr/X11R6/bin/XF86_* as your xserver (instead of the wrapper /usr/X11R6/bin/X) and start it as root you may be able to run without xbase and xlib6. /HINT - tweak /etc/X11/XF86Config to use your laptop as fontserver: go to Section Files and add a line FontPath tcp/laptop:7100 to this Section. Maybe you need to comment out other FontPath entries. I don't know if they mustn't exist and if they are removed automatically. - check, that /etc/X11/config doesn't contain start-xdm to avoid it starting up atomatically - now you can start X (runnig on the laptop, displaying on the desktop) with X -query laptop - since you didn't remove xbase, you should still be able to run a local X with startx, but your laptop always needs to be reachable, since it's fonts are required. - you may write your own /etc/init.d/Xserver or so (take a look at /etc/init.d/skeleton) and make the appropriate inks in /etc/rc*.d with update-rc.d to start the Xserver automatically. Incidently, do you know what kind of performance hit I will get running X over 20m of BNC cable (there are 2 other (Win'95) machines on the network)? I agree to Ben Pfaff that this shouldn't be a bottleneck. I was using X on a *heavy* loaded network and it was never much too slow. BTW, I wasthinking about the IP addresses and would it make sense for me to use 192.168.xxx.1 and 192.168.xxx.2 (xxx=random number) in case I added someone elses Linux system to my network for some reason in the future? as long as xxx is constant: yes ;-) I would reserve at least *.1 for a possible router. I think it is very common to use this for a router Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Multi-Boot
Hi! Dale Smith ([EMAIL PROTECTED]): On Wed, 25 Mar 1998 12:47:56 -0600, you wrote: So, you may indeed be able to do this. In any case you should be able to use the NT Boot Loader to start linux. This is in fact what I do. Set up LILO in your linux partition and have it write the boot block there. Eg., if your linux partition were on the third scsi disk in the first partition, a command to do this would be 'dd if=/dev/sdc1 bs=512 count=1 of=linuxboo.bin'. Then copy that boot block to a file and stick that file in your main NT partition. Now edit your c:\boot.ini file and add a line for Linux, ala 'C:\linuxboo.bin=Linux' (assuming you named the file linuxboo.bin). As long as your BIOS supports reading from the third drive this should work. another automatic solution I'm currently using: # mount your c: drive RW. mount c: /mnt-tvfat # this file must exist when lilo is run. otherwise lilo complains. dd of=/dev/zero of=/mnt/linux.bin bs=512 count=1 and put boot=/mnt/linux.bin to your lilo.conf I have read about this in one of the HOWTO's. This is a real pain for me because my NT partition is ntfs, and I can't use linux tools to copy the boot sector. You can use bootpart: http://ourworld.compuserve.com/homepages/gvollant/bootpart.htm It generates such chain loaders. Then the boot-sequence looks like: nt thingie bootpart lilo bootsector on your linux disk Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Running X11 over a network
Hi, Ian! So I skip the export, startx and xhost steps replacing them with X -q... on the desktop. correct! Packages installed as above, uses window manager on the laptop? yes, this way all X-applications (including WMs) run on the laptop. Does XDM require you to log in? If so, can I make it automatically send the info I used to log in to the console (both will have identical user details). I think XDM requires you to log in. But you may start the desktop's X-server on startup to avoid you to log in twice. BTW, does the laptop need xbase and xlib6g? yes, the laptop needs them and of course every application you want to run. You needn't bother with setting up a Xserver on your Laptop if you don't want to. According to my knowledge there are two basic setups for your desktop: - with locally installed fonts glancing at package dependencies it needs Xserver, xfntbase, xlib6, xbase - using the laptop as fontserver now it only needs Xserver and an appropriate configured XF86Config :-) I suggest to start with the first setup + xserver-vga16. Then you can use the graphic XF86Setup to configure your desktops xserver. When you become more experienced (and diskspace is an issue) you can switch to the second setup and purge xbase, xfntbase, xlib6 (, xserver-vga16). Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Running X11 over a network
Hi! Ben Pfaff ([EMAIL PROTECTED]): Try this, which is slightly different: * Give 192.168.1.1 to the laptop. * Give 192.168.1.2 to the desktop. * Install any X software you want, but no xserver package, on the laptop. * Install only xserver-* (whichever you need), xbase, xlib6g on the server. * On the laptop do `export DISPLAY=192.168.1.2:0' in bash. * Run `startx', then `xhost +192.168.1.1' on the 486. * Run your xclients from the laptop command line. How about running 'X -query 192.168.1.1' on the desktop? This should give you the XDM-login of your laptop. IIRC every host can connect to an XDM by default. Otherwise you need to add a line with a * to /etc/X11/Xaccess. Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: New handling of scripts external to PPPD (ie ip-up, ip-down)
Hi! Adam Heath ([EMAIL PROTECTED]): I am thinking of more places to add runlevels. Currently, I am trying to eliminate the connect option from pppd's option files. Ppp-functions would be called like this: /etc/ppp/ppp-functions dial hostname username ip-up parms, with stdin, stdout, and stderr redirected to the serial port. Ppp-functions could then use whatever means nescessary to dial. This would allow for dialing several numbers in round-robin fashion, better redial capabilities, enhanced logging, etc. hmm, somehow I feel a little uncomfortable with stderr /dev/ttyS?? a) errors *MAY* confuse a modem b) warnings *WILL* confise a modem c) nobody except of myself needs to know when I wasn't able to write a working script I'd prefer stderr /dev/null And put the following to the connect script: # you may uncomment the following to debug this script #exec 2/tmp And since you mentioned improved logging: How do you think about teaching chat to use syslog for REPORT, SAY and (optionaly) specify a filename|stderr to override this. This way you needn't fill up your syslogs with the verbose chat output and still keep track of your connect-rate. I modified chat from 2.3.2 to (only) sylog REPORTs - it wasn't very hard. I think it's great to add something like hostname and username. But I'm a little curious where from you intend to get the hostname? :-( Hmm, I suppose using the argument from pppd's call option isn't appropriate, isn#t it? Did you already elaborate how handle inbound setup? Or diald? How Do you intend to tell pppd wether it needs to run ppp-functions dial. If a package wants to be run when pppd is on, then include a file /etc/ppp/scripts.d/packagename. In the postinst, run the command ppp-update.d. And what happens when a package is installed before pppd is? Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: OK I'm stupid. How do I get to my fl
Hi! Richardson,Anthony ([EMAIL PROTECTED]) wrote: If the NT partition is NTFS you have quite a bit more work to do. NTFS read-only support for Linux is available, but requires kernel patches and a kernel rebuild. You need't rebuild your kernel! There is a debian package with an ntfs-module. you only need to install this pkg. The last version I saw (not the debian-package) had an option to mount NTFS RW! But I didn't check it out. I think its pre-alpha ;-) Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- E-mail the word unsubscribe to [EMAIL PROTECTED] TO UNSUBSCRIBE FROM THIS MAILING LIST. Trouble? E-mail to [EMAIL PROTECTED]
Re: Smail/Sendmail Question
Hi! mfrattola localhost mail deliver locally don't rewrite header mfrattola LAN mail deliver thru eth0 don't rewrite header mfrattola internet mail deliver thru ppp0 rewrite header This is possible by a) defining mailers, and modifying sendmail.cf to add a inexpensive smtp mailer (copy the smtp mailers and call them lsmtp et akk b) look at ruleset 1 and prevent rewrites for the LAN, while allowing rewrites for external addresses I wrote such m4s and Marco asked me to post em to the list. I hope nobody is angry because of their size (6k gz) in addition they are capable of rewriting usernames restrict internet-mail to listed users. deliver mail to non-local adresses which are known to be reachable localy They aren't perfect, but they work. At least they are a good start to continue tweaking. Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB sm_hack.tar.gz Description: GNU Zip compressed data
Re: Smail/Sendmail Question
Hi! mfrattola localhost mail deliver locally don't rewrite header mfrattola LAN mail deliver thru eth0 don't rewrite header mfrattola internet mail deliver thru ppp0 rewrite header This is possible by a) defining mailers, and modifying sendmail.cf to add a inexpensive smtp mailer (copy the smtp mailers and call them lsmtp et akk b) look at ruleset 1 and prevent rewrites for the LAN, while allowing rewrites for external addresses I wrote such m4s. I don't think they're perfect, but they work !-) I offer to share em. shall I send em to the list? Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- E-mail the word unsubscribe to [EMAIL PROTECTED] TO UNSUBSCRIBE FROM THIS MAILING LIST. Trouble? E-mail to [EMAIL PROTECTED] .
kernel 2.0.33 smbfs bug
Hi! there is a bug in kernel 2.0.33 which woes files' timestamp on mounted NT4 shares. There is already a bugfix patch, which seems to slip into 2.0.34. kernel-source-2.0.33-2 didn't include this patch. My questions: -shall I report this as bug or wait till kernel-*-2.0.34? -against what shall I file this bug? Regards Rainer --- KeyID=58341901 fingerprint=A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- E-mail the word unsubscribe to [EMAIL PROTECTED] TO UNSUBSCRIBE FROM THIS MAILING LIST. Trouble? E-mail to [EMAIL PROTECTED] .
Re: NAT solution for Linux?
Hi! You wrote: Is there a kernel module and/or software that provides Network Address Translation (something similar to what Cisco's PIX product does)? I don´t know Cisco´s PIX, but there is domething called NAT at http://www.csn.tu-chemnitz.de/HyperNews/get/linux-ip-nat.html Regards Rainer --- PGP fingerprint = A5 57 04 B3 69 88 A1 FB 78 1D B5 64 E0 BF 72 EB -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .