Re: Explorer-type file manager
hey bill, for folks who are really wanting a windows-like environment, i'd recommend going with a gnome or kde desktop environment. both come with built in file browsers, and i believe that they both also provide the click and go functionality for desktop icons as well. i don't know too much about installing either on your desktop, because i'm rather happy with fvwm2 personally, but i've sat down at debian stations running both kde and gnome, and found that both were quite featureful and user-friendly (though i don't remember what the default underlying window manager was in either case). sean On Mon, Feb 17, 2003 at 03:54:38PM -0800, Bill Moseley wrote: > Trying to ease the transition to Linux for the family, I'd like a > MS Windows-like Explorer file manager, both in look and usage. > > Not that icon view, just the plain old "Explore" with directories > on the left and directory contents on the right. And where I can > configure what mouse buttons do for each file type (e.g. right click on > .mp3 and select from menu xmms play, xmms queue, or mpg123, or scp to > some set machine). Also click on column titles to sort by that column. > > Window manager is icewm. > > I like Rox and FileRunner and a few others, but this is a case where I'm > trying to make it feel a lot like Windows. > > Thanks, > > -- > Bill Moseley [EMAIL PROTECTED] > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg31378/pgp0.pgp Description: PGP signature
Re: Explorer-type file manager
hey bill, if you're reluctant to try and install gnome or kde, i'd recommend you download and burn yourself a knoppix CD. if you're not familiar with what knoppix is, it's a bootable linux on a cd, based off of debian, and comes with gnome, kde, and iirc even icewm as options from the boot prompt, as well as a nice collectoin of g* and k* software. it's a good way to get a feel for the different desktop environments, and from my own experience comes in very handy as a rescue cd as well, so it's good to have anyway. sean msg31538/pgp0.pgp Description: PGP signature
Re: jack: coding failed, error#32512
On Tue, Feb 18, 2003 at 02:50:38PM -0500, Thomas H. George,,, wrote: > I want to try mp3. I used apt-get to install jack. From the > description I expected to work with its standard .jackrc file. It > started smoothly using cdparanoia but, when the first cd track was > extracted the error#32512 appeared. > > The man page for jack suggested other man pages including lame(1). I > have no such man page and apt-cache search lame does not find a page > with this name although there is a glame. lame is your mp3 encoder, and isn't available in debian because it's technically (though debatebly) illegal to not pay frauenhoeffer a royalty for using an mp3 encoder. if you happen to find the source code for lame online, however, you'll notice it ships with a ./debian directory, meaning you can do something like fakeroot ./debian/rules binary and presto have your own .deb for lame. anyway, if it's not installed, i'd guess that's why you can't make mp3's unless you have another one installed. _However_, i'd suggest you consider migrating to ogg vorbis. smaller file size and better sound quality at the same time i've found, and it's supported by the latest versions of mp3 players for all os's. i don't know about jack, but abcde (another friendly frontend) handles both ogg and mp3 quite well. hth sean msg31584/pgp0.pgp Description: PGP signature
Re: USB mass storage and kernel 2.4.18
heya, i have it running fine, and i'm balthasar[~]00:13:09$ uname -a Linux balthasar 2.4.19 #1 Sat Nov 16 15:49:07 EST 2002 i686 unknown unknown GNU/Linux but i believe i've had it working on 2.4.18 as well. i believe you just need more modules installed. here's an edited output of my lsmod, maybe you need some of these other modules loaded too, like usbcore or usb-uhci? balthasar[~]00:13:14$ /sbin/lsmod Module Size Used byNot tainted usb-storage97744 0 (unused) usb-uhci 20908 0 (unused) usbcore54112 0 [hid usb-uhci] vfat9276 0 (autoclean) fat29080 0 (autoclean) [vfat] sd_mod 9932 0 (autoclean) scsi_mod 80296 1 [sd_mod] if this isn't the case, try doing a tail -f /var/log/messages before you plug it in, then plug it in, and take a look at whatever it says for clues hth sean On Thu, Feb 20, 2003 at 12:46:44PM +0800, Dai Yuwen wrote: > Hi, All > > I've a 64M USB flash disk. I searched the internet, and did the following > steps: > > 1. plug USB disk into my PC > 2. modprobe sg usb-storage > > Then I tried to mount it: >mount /dev/sda1 /mnt > > But a messages said `/dev/sda1 is not a valid block device'. > > Then I checked this file: > > 3. cat /proc/scsi/scsi > Attached device: none > > I use the stock kernel image of debian: > uname -r > 2.4.18-686 > > I also found some info that said kernel 2.4.18 had some problem in USB. Is > that true? If yes, I'll have to compile another version of Linux kernel. > Any idea? Thanks in advance. > > best regards, > Dai Yuwen > > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg31810/pgp0.pgp Description: PGP signature
Re: USB mass storage and kernel 2.4.18
On Thu, Feb 20, 2003 at 04:15:26PM +0800, Dai Yuwen wrote: > > excellent :) question: if you plug yours in and unplug it, and then plug it > > in again, is it still /dev/sda? every time i do that on mine it changes to > > sdb then sdc, etc... > > > I've no this problem. But I umount /dev/sda1 before I unplug the USB > device. Have you umounted it? yeah. i think it's just a symptom of my reader not being fully supported. but as long as i can get stuff off of it, i'm happy, really. i'm glad i don't have any other scsi devices though, otherwise i'd probably get a little confused. sean msg31891/pgp0.pgp Description: PGP signature
Re: OT: mod_ssl (apache) log entries -- wtf?
On Thu, Feb 20, 2003 at 01:56:40AM -0600, Will Trillich wrote: > what does this mean? are there black hats involved? (maybe even > a gray fedora?) i'd guess one of two things: a - someone else has messed up a dns entry or href on a webpage b - this computer is trying to test for exploitable weakness in your web server. you could find out for sure by running the packet sniffer of your choice and dumping the whole conversation to a log, and then look at what kind of data the client was sending. oh wait... https... nevermind. there's probably a way to turn up verbosity on apache then :) sean msg31892/pgp0.pgp Description: PGP signature
Re: USB mass storage and kernel 2.4.18
On Fri, Feb 21, 2003 at 11:37:06AM +0800, Dai Yuwen wrote: > which device should I use in /dev (sda? or sdb? ... ) whichever one mounts :) if you want to take a sneak peak, do file -s /dev/sda1 or /dev/sdb1, and if it says something about a vfat parition or an x86 boot sector, that's the one. sean msg32051/pgp0.pgp Description: PGP signature
Re: ssh problem ssh_exchange_identification
On Fri, Feb 21, 2003 at 01:23:45PM +0100, Rudy Gevaert wrote: > ssh_exchange_identification: Connection closed by remote host this looks like a hosts.allow or hosts.deny issue on the remote box... sean msg32175/pgp0.pgp Description: PGP signature
Re: USB mass storage and kernel 2.4.18
On Fri, Feb 21, 2003 at 07:34:22AM -0500, Kevin Coyner wrote: > A nice utility to use for figuring out which scsi the device is on is > > apt-get install sg3-utils hey, yeah, that is pretty nice. maybe i can even use that to get my reader to work as a hotplug device... sean msg32176/pgp0.pgp Description: PGP signature
Re: nvidia driver for debian testing
On Fri, Feb 21, 2003 at 11:25:57AM -0800, Dennis Krinke wrote: > Hi, I am trying to install nvidia 3D driver for riva tnt on debian testing, > kernel 2.4.20-k6. Originally, I tried from nvidia source, > now I am trying from debian packages. I can compile the package > nvidia-kernel-2.4.20-k6_1.0.3123-3+_i386.deb, > but it does not load. > If I try to force insert the module I get many unresolved symbol errors, > just like when I tried to compile from nvidia source. are you sure you're building against the kernel headers for the kernel that you're running? if you're running a stock debian kernel, move any kernel source you have in /usr/src out of the way, install the kernel-headers package for your kernel (apt-get install kernel-headers-2.4.20-k6 maybe?), and then re-compile your module against that like you did previously. hth sean pgp0.pgp Description: PGP signature
Re: (Newbie) Functioning In Debian
On Fri, Feb 21, 2003 at 09:32:34PM -0500, M. Kirchhoff wrote: > Another site I frequent uses streaming Windows Media. Am I totally out > of luck there? I know there's this Crossover package that will run WM, > but it's definitely non-free. I haven't run into needing RealPlayer > support yet, but I wonder if there's a free clone of that that works > under debian-mozilla. no! i highly reccommend mplayer. it's a good, free movie player and it can play a good number of non free formats such as MS asf. the folks who made it are even kind enough to ship it with a ./debian directory, which means that you can: $ fakeroot ./debian/rules binary and then you'll have a mplayer .deb that you can install with dpkg. give it a shot. (note you might need a couple other packages to do this, namely fakeroot). sean pgp0.pgp Description: PGP signature
Re: kernel produces unresolved symbols
On Sat, Feb 22, 2003 at 10:50:20AM +0100, Roman Joost wrote: > After the compiliation process, i copy my kernel into /boot and reboot. Last > time i get some unresolved symbols after a depmod: > depmod: *** Unresolved symbols in /lib/modules/2.4.19/kernel/drivers/char/drm/gamma.o > depmod: *** Unresolved symbols in /lib/modules/2.4.19/kernel/net/ipsec/ipsec.o > depmod: *** Unresolved symbols in /lib/modules/2.4.19/kernel/net/irda/irda.o > > This kernel is made by: > make menuconfig > .. > make-kpkg kernel_image > make_kpkg modules_image > > I read that a good kernel is made by: > make dep > make bzImage etc. wait, which are you doing, the make-kpkg way or the manual way? make-kpkg makes a binary .deb in /usr/src (or wherever you're keeping the source), that you install with dpkg -i. there's no copying of files needed. if you just copy the files you're likely to get symbol version mismatches... > I would ask, what im doing wrong in these steps? Is something wrong with my > debian way of kernel compilation? it looks like the compilation is ok, but that it's the kernel installation that's getting you... sean pgp0.pgp Description: PGP signature
Re: Compilation of a single kernel module.
On Sat, Feb 22, 2003 at 08:24:31PM +0200, Shaul Karl wrote: > I had to compile a single kernel module, namely isa-pnp, and did not > want or had to compile the whole kernel. > I wonder whether the following method is good and will work for other > modules as well. i think you can get by just by replacing the kernel headers with what was used to build your kernel (the kernel_headers package if you're using a stock debian kernel or make-kpkg'd your own), and then compiling it as you did in the end of your example, without having to do any make oldconfig etc... sean pgp0.pgp Description: PGP signature
Re: command-line biff?
hey nori, On Sat, Feb 22, 2003 at 05:29:36PM -0500, Nori Heikkinen wrote: > does anyone know of a command-line version of some biff or buffy or > whatever (mail notification program)? i just wrote a little shell take a look at mailstat (comes w/procmail), which might do something along the lines of what you're interested in. sean pgp0.pgp Description: PGP signature
Re: command-line biff?
On Sun, Feb 23, 2003 at 11:38:29AM -0500, Nori Heikkinen wrote: > > take a look at mailstat (comes w/procmail), which might do > > something along the lines of what you're interested in. > > this only uses procmail logs, though, right? and i can't see how to > make it tell me just what's *new* in the mailbox? right. the idea is that you use it without -k, such that it truncates the log every time you call it, and if there's mail in the log, it's new. of course if you've already read it it doesn't matter, so this program does have its deficiencies. plus, if you want to actually keep your logs around for posterity... how does your script do this? sean pgp0.pgp Description: PGP signature
Re: debian kernel configuration
hi andreas, On Sun, Feb 23, 2003 at 09:29:14AM -0700, Andreas J Guelzow wrote: > Hi, > > does anybody know where I can find the kernel configuration for the > debian kernels (the 2.4.18-smp to be exact)? I have to recompile the > kernelafter changing a few settings and would like to start with the > settings in the default kernel (since I know that they are working, > except that they don't support all the memory I would like to use). the first place i'd recommend looking would be in /boot if you have the kernel already installed. there should be a config-`uname -r` with your current kernel's config if you've installed a debian stock kernel or one you've made yourself with make-kpkg. if you can't install it, i'd recommend downloading the package and extracting the config manually. hth sean pgp0.pgp Description: PGP signature
Re: dpkg package configuration level(?)
hi corey, On Sun, Feb 23, 2003 at 04:40:33PM -0800, Corey Hickey wrote: > Does anyone know what I'm talking about, or am I just massively > confused? try # dpkg-reconfigure debconf hth sean pgp0.pgp Description: PGP signature
Re: Upgrading from Stable to Testing
On Sun, Feb 23, 2003 at 10:44:16PM -0600, Hanasaki JiJi wrote: > IMO, yes. > > Although, I use "sarge" not "testing" to be sure that I dont > inadvertantly upgrade to the next version of testing by accident. right, but beware, there's lots of folks who say that testing is worse off than unstable right now, because of the whole gcc/g++ transition holding up all the upgrades from making it into testing... sean pgp0.pgp Description: PGP signature
Re: OS X Games
On Mon, Feb 24, 2003 at 02:32:21PM +, Colin Watson wrote: > They're not sufficiently similar for that to work, no. Apart from > anything else, unless you're running Linux on PowerPC then it's a > different processor architecture. Also, while MacOS X under the hood is > similar to Linux in that it's a Unix-like operating system, it's a > different Unix-like operating system, essentially BSD. that leads me to wonder if it would work on a NetBSD or FreeBSD box sean pgp0.pgp Description: PGP signature
Re: your mail
salut, in the future, please make a point to put a Subject: in your emails, as lots of folks often don't read emails without subjects. i would give the installation a try with the 'bf24' kernel. i'm not sure which .iso it's on, but it's definitely on one of them, and you can also get the bootfloppies images off of ftp.debian.org as well. i think that kernel has support for lots more stuff than the standard 2.2 kernels do. good luck, sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
heya, On Tue, Feb 25, 2003 at 07:25:20PM +0700, Brian Durant wrote: > /dev/hda7reiserfs9.29 GB(mount point) / > > /dev/hda5ext215.1 MB/boot > > shmfs shm 57.9 MB/dev/shm from my own experience, if you have the time i'd recommend splitting that up a bit more. debian likes having big /var directories, so i'd recommend shaving half a gig off of / and making a /var, also maybe have /home separate too, so over ambitious downloading doesn't bring the system to it's knees. sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
hi brian, On Tue, Feb 25, 2003 at 09:54:55PM +0700, Brian Durant wrote: > So where would I do this in the install process, at the prompt for hda > disk partition? Skip over the swap prompt? What comes after, a dialog > listing of all the hard drive partitions? /home I understand, but what > is a "/var"? How would all of these partitions look together in > schematic form and what values (sizes) do you suggest? Which of these > should be ReiserFS partitions "/" and "/home", if I want to use ReiserFS? at some point you'll be asked if you want to initialize/format your partitions. assuming you have nothing important on your system that can't be backed up beforehand (which one really always should do before installing any os), i'd recommend completely starting over from scratch and doing a setup like* /boot 25-50MB swap(amt. RAM x 1.5) / (most of your hard drive) /var512 MB /home 100-500 MB, depending on taste note that this is > 4 partitions, so you'll need some logical/extended entries. i think cfdisk does this somewhat transparently for you, but it's been a while... the reason i mentioned /var is because it's really easy for the subdirectory /var to fill up in many situations. for example, a runaway process or infinitely looping bouncing mail (anyone who's learned the hard way about what not to do with a .forward...) might put a couple thousand entries in a logfile in /var/log in a matter of minutes, and if you don't have /var in its own partition, it won't be long before your disk is full and things start breaking. also, debian stores downloaded .deb packages in /var/cache, another reason to limit it. having a separate partition is kind of like putting fireproof walls and doors in a building, it does wonders for damage control. if you're really paranoid, or just enjoy putting up blast doors in your home, you might want to do it. but also, like some folks mentioned in this thread, this isn't exactly the control center for a nuclear reactor, so if you don't feel like spending the time, you can always just have a big / partition and be done with it. wrt to ReiserFS vs. ext2 vs. ext3 vs. ???, i think that's more a matter of personal taste than anything else. also, i'm not sure what kernels support what fs's in the boot floppy series, so you might need to check to make sure that it's supported by whatever you're using to install. hth sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Tue, Feb 25, 2003 at 10:43:02AM -0600, Kent West wrote: > If you're a newbie, I'd recommend taking the first scheme, use Debian > for a day or two. Then do the install all over again, wiping the drive, > using the second scheme. Then do it again in a week or so, wiping the > drive, and making a choice then as to which scheme you feel comfortable > using. The first two installs are not for the purpose of figuring out > which scheme you want to use; they're just to get you familiar with the > install process, kind of like how rebuilding your car's engine will give > you a huge education, but rebuilding it thrice will give you an even > better education. fwiw, i agree with this, but it also depends on how much time you want to put into learning this stuff. i think if you do it this way, you'll learn a lot more about how your computer and linux (and os's on your computer in general) work, but of course it's more of a time commitment as well (especially if you're doing a net install...). sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Wed, Feb 26, 2003 at 03:10:08PM +0700, Brian Durant wrote: > DHCP failed. I booted the install with the command "bf24", so support > for the built into motherboard SiS 900 Fast Ethernet driver should be > loaded. What do I do now??? well first things first see if the networking works :) try $ /sbin/ifconfig eth0 to see if it has an ip address. if you're still having issues, it may be that you need to insert a module--so try modprobe sis900 (i believe that's the name of the module, though you might want to double check that). to restart your networking (and give another jab at dhcp), # /etc/init.d/networking restart anyway, if you installed with a 2.2 kernel, then that's probably what got installed on your hard drive. at this point to get a newer kernel on it, i'm not sure what would be the easiest method. normally i'd say apt-get, but you kind of need either networking or the cd set to do that try booting into the installer with the bf24 again, and see if you can skip all the partitioning/mkfs'ing, then mount your partitions, and then re-install the kernel with the installer. sean pgp0.pgp Description: PGP signature
Re: Ethernet startup problems
On Wed, Feb 26, 2003 at 07:20:39AM -0800, Michael Rudmin wrote: >Problem is, after that point, my ethernet doesn't > work [RTL8139]. Looking at the syslog, I now no > longer see ethernet even mentioned during the startup. > > >Going over to modprobe -c, I again don't see it. > So [I'm sure this will seem to be a stupid question] > how do I reinstall that module? how about # modprobe rtl8139 or # modprobe 8139too if either of those work, put the module name in /etc/modules, and you should be set. if not, i would guess that you inadvertantly deleted the module or mucked your kernel install. also, hopefully this isn't the case, but from my own personal experience rtl8139 based chipsets are really, really, flaky... >Also, how do I know what else has been blown to > pieces by this apparent crash? generally, try to use it, if it doesn't work it's probably broken :) sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Thu, Feb 27, 2003 at 10:30:26AM +0700, Brian Durant wrote: > OK, I started the install process from the beginning again, remounted > partitions, etc. Not only did I boot the install using "bf24", but I > also chose the long install, to make sure that I was installing a 2.4 > kernel from the Deb 3 rev. 1 full install CDs. I tried /sbin/ifconfig > eth0 and got a lot of numbers in response that I don't really > understand, but what seems to be more interesting, is when I did a > "modprobe sis900", the response was "can't locate module sis900". I then > looked under the kernel/net modules for the SiS900 driver and it is not > listed .-( well looking at the kernel config for SIS900 in the bf24 kernel config on debian's ftp server: CONFIG_SIS900=y which means it's hard-compiled into the kernel, and you don't need to do a modprobe. can you give us the output of ifconfig? you're most interested in the "inet addr" field". for example: oil[~]22:59:16$ /sbin/ifconfig eth0 eth0 Link encap:Ethernet HWaddr 00:06:5B:DC:32:B2 inet addr:130.58.82.95 Bcast:130.58.82.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:7692088 errors:0 dropped:0 overruns:91 frame:0 TX packets:8289238 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:1962609489 (1871.6 Mb) TX bytes:341377353 (325.5 Mb) Interrupt:11 Base address:0xec80 the important thing to clean is that addr exists. if you have that line, that means you should be ok. try pinging another ip address, and if that works, try pinging a hostname like google.com > According to a posting from Klaus Imgrund, there is a module for SiS900 > on the "testing CD", but I am unsure what "testing CD" this is and how I > can load the module into the kernel I am installing. or it could be that. can you post us the output of dmesg (or better, put it online somewhere and give us a link)? if it really is recognized by your kernel, there should be something mentioning it. sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Thu, Feb 27, 2003 at 11:19:25AM +0700, Brian Durant wrote: > Giving the output is not an easy task, as the computer I am installing > on has a USB floppy (the Celvin is a kind of iMac for the PC world) and > hopefully will work when I get Debian installed. As the network isn't > working either, I can't send it as a text file to another box on my heh... > Link encaps:Ethernet HWaddr 00:30:67:06:4F:86 > Broadcast Multicast MTU 1500 Metric:1 but you don't have an ip address. this means your kernel see the device and configured it, but dhcp isn't coming through. do you have lines in /etc/network/interfaces that say something like: auto eth0 iface eth0 inet dhcp ? also, if this is on a private lan, how about statically assigning an ip address to your card[1]? this problem might be signs of a completely different problem (like faulty cable or connecter or card), or it might just be dhcp being flaky for some reason. hth sean [1] for example: # ifconfig eth0 192.168.0.99 pgp0.pgp Description: PGP signature
serial line laptop console for headless setup
hey -users, i'm thinking that this computer sitting right by my bed, which is on 24/7 and producing profuse amounts of noise, really ought to be moved into my closet, far away from the monitor on my desktop (which would still be used for another workstation that goes on and off) and coincidentally, far away from my bed. so i've got this extra laptop with a serial port, and i'm thinking it'd make a nice console for this machine, such that i could open my closet door, flip up the lcd, and look at stuff if something ever went wrong, and at the same time be able to close the closet door and get a silent night's sleep. so, without even googling or apt-cache'ing, i found /usr/share/doc/HOWTO/en-txt/Serial-Laplink-HOWTO.gz, already on my system, which looks pretty promising. however i'd still like to field some opinions/experience from anyone on the list who's done something similar, so i might be able to avoid pitfalls, find shortcuts, et c... thanks sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Thu, Feb 27, 2003 at 02:16:43PM +0700, Brian Durant wrote: > The response is "not found". Remember, I am still in the install. I the response to what? ifconfig? lspci? also, were you able to install the kernel onto the hard disk? if you can finish the install process and boot off the hard disk (even if it's only a base system and you can't install anything else yet), that's one less variable to worry about. > don't have any problems dhcp on any of the other connected boxes, all > running a version of Win at this time. All receive an IP address > automatically through dhcp. I tried "ifconfig eth0" again (had to do a > " " to get out of little problem with grep in the ctrl alt delete to what? if you need to kill your way out of a program, try ctrl-c > shell as per my last posting and start the install process one more > time) and noticed a line with: "Interupt: 11 Base address: 0xe400", but > everything else is all zeros. I hope this helps, at this point I am > utterly clueless. after you boot up from your hard drive, how about # grep -i eth0 /var/log/dmesg and # grep eth0 /etc/network/interfaces good luck... sean pgp0.pgp Description: PGP signature
Re: autologin in console mode
On Thu, Feb 27, 2003 at 07:17:33PM -0500, Levi Waldron wrote: > Is it possible to do an autologin into console mode? ie, when turning on the > machine a particular user gets logged in every time without entering a > username or password? > > It's for a visually-impaired user, so having to type that stuff in before the > voice prompts are activated is a barrier even with a simple > username/password. Then with an entry in .bashrc starting "emacspeak -o" the > computer will go straight into voice mode every time it boots up. i just got something kind of like that to work. install the rungetty package, and then open up /etc/inittab. change the line that says: 1:2345:respawn:/sbin/getty 38400 tty1 to 1:2345:respawn:/sbin/rungetty -u username --autologin username tty1 -- bash -i where you obviously replace username with the appropriate user. hth sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Fri, Feb 28, 2003 at 12:34:22PM +0700, Brian Durant wrote: > "SIS 900 Internal MII PHY Transceiver found. at address 1 > Using transceiver found. at address 1 as default. > SIS 900 PCI Fast Ethernet at 0xe400, IRQ 11, 00:30:67:06:4f:86" okay, that's Good, it means your kernel sees and happily loaded the driver for your card. > By the way, thanks for the tip on the ctrl-c. I seem to not be writing > the grep variables too closely first time around and then grep just > churns away for hours. actually, it was probably just waiting for input. if you don't specify the file to grep through on the cmdline, grep assumes stdin. try just "grep foo", and give it some input by entering lines with and without the string foo in them to see what i mean. ctrl-c kill(1)s your current process. > The output of "# grep eth0 /etc/network/interfaces" was nothing. i was using the "#" to indicate you needed to do the command as root (as opposed to "$", which means as anyone)... you didn't type that part in, did you? if you didn't and just nothing showed up, put these two lines in /etc/network/interfaces: auto eth0 iface eth0 inet dhcp then do # /etc/init.d/networking restart > Even after the base install, the "ifconfig" command returns "command not > found. are you doing this as root? it's possible that it's not in your path for some reason. try again but with /sbin/ifconfig instead of just ifconfig. do this after making sure that the above is set up, and see if you get an ip address. if you don't, you can set it manually with the same program. hth sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
hi brian On Fri, Feb 28, 2003 at 04:19:38PM +0700, Brian Durant wrote: > OK, I didn't get the difference between "#" and "$". Tried again, in > root (#), same result. Added the two lines in /etc/network/interfaces > restarted the network and tried again. Same result - nada - zip -zilch interfaces didn't exist? strange... > response. No response with /sbin/ifconfig either. Thanks for all your what do you mean by "no response" was there output, but just not the address? was there an error? > time so far. Is there any place that we can go from here? I did a sort > of default full install with tasksel and dselect. Got a lot of stuff I > didn't want, but even the stuff that I do want will not do any good if I > don't get this @£5#+* eth0 working. right. okay, at this point a few things posted would be very helpful. i know it's hard to get it to use without a network, but the following might shed some light: $ dpkg -l dhc* | grep ^ii (this will tell use what kind of dhcp client you have installed, hopefully) after you do /etc/init.d/networking restart, show us anything that looks relevant from: # tail /var/log/messages /var/log/syslog (those two files are log files for various parts of the system) $ /sbin/ifconfig eth0 namely, for the above, we're interested in the first few lines. also, what are you running that's giving out the dhcp addresses? do you have anyway of looking at it's logs? okay, and if none of this works, try this: # apt-get --reinstall install netbase ifupdown pump (pump is a dhcp client that i think is a little less picky than the dhclient -- i don't really like it all that much, but it works) after doing this, try just typing "pump" in a root prompt and see if magic happens. also, everything else failing, istr someone said something about the sis900 driver maybe not supporting dhcp, so we can try setting it manually. so when you have a chance, go to another machine and find out its ip address, netmask, and dns information. i know on the older windows machines this was by running "winipcfg", but i don't know off the top of my head what does it in XP these days. good luck, sean ps - since there are a lot of instructions and questions in here, when you reply, please interleave your responses with the text of this mail, like what's done at the top of this mail, to avoid as much confusion as possible pgp0.pgp Description: PGP signature
Re: Using one mailbox at ISP for many people
On Fri, Feb 28, 2003 at 03:45:35PM +, Jonathan Matthews wrote: > What combination of programs do I use to get one mailbox successfully > segregated when it hits my machine? I'm using fetchmail -> exim -> > procmail at the moment, but I've lurked on the procmail list long enough > not to ask about using it as a MDA (or is it an MTA that it shouldn't > be treated as?). i would think about doing something like the following: - cron fetchmail as your username - have the following two rules at the top of your .procmailrc :0: * [EMAIL PROTECTED] /path/to/her/mbox :0: * [EMAIL PROTECTED] /path/to/your/mbox if you have other procmail rules for your own filtering, you can just drop the last rule. also, you'll need write permission to her mbox in order to put stuff there obviously. i'd recommend creating a group that only the two of you are in, and make the file group-writable by that group. i'm sure there's a more graceful way to do this using procmail as an MDA, but that's what i could come up with off the top of my head hth sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Sat, Mar 01, 2003 at 12:07:55PM +0700, Brian Durant wrote: > If it is OK with you, I would rather not interleave my responses. I just > got over some serious eye problems and find that spending too much time > editing and working with the dim Debian text output on my daughter's > computer, makes my eyes complain a lot. I am trying to keep the eye > strain at a tolerable level. I hope you understand. okay, well then just be extra clear what you're talking about.. > The output of "# tail /var/log/messages" was as follows: > > sda: status=1, message=00, host=0, driver=08 > sda: I/O error: dev 08:00, sector0 > unable to read partition table that looks like unrelated errors... > The output of "# tail /var/log/syslog" was as follows: > > named[170]: exiting (due to assertion failure) > lpd[195]: restarted > kdm[229]: server /usr/X11R6/bin/ X cannot be executed. > kdm[219]: Server unexpectedly dies. Server for display : 0 cannot be > started, session disabled. again, unrelated... > 1 Ethernet adapter : > > IP Address. . . . . . . . . : 192.168.1.100 > Subnet Mask . . . . . . . . : 255.255.255.0 > Default Gateway . . . . . . : 192.168.1.1 okay, and since your other box is 192.168.1.101, i think it's safe to assume that your dhcp server is giving out addresses on the 192.168.1.0 network. how about trying to manually configure your device then, with: # /sbin/ifconfig eth0 192.168.1.111 and then see if you can ping some ip addresses, like: $ ping 216.239.57.100 (that's www.google.com). if that works, you just need to copy the dns information over too. sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Sat, Mar 01, 2003 at 06:17:41PM +0700, Brian Durant wrote: > dhcp works in Knoppix. Should I run the same tests that have been posted > to me for Deb? What would the conclusions be for my Deb install if > Knoppix does work? if knoppix works, it will probably automagically work, because that's what knoppix does. you put it in your cd rom drive and boot off of it, and it figures everything out on its own. so when you get dropped to a desktop, try opening up a browser and go to some website and see if it just works... sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Sat, Mar 01, 2003 at 03:10:14PM +0700, Brian Durant wrote: > OK Sean, the output of "# /sbin/ifconfig eth0 192.168.1.111" was "eth0 > Media Link Off". The result of "$ ping 216.239.57.100" was "Network is > unreachable" :-( i've never seen that error out of ifconfig before, and can't even find it in the source code for the program. okay, while you're waiting for the knoppix cd, how about you try this: - in /etc/interfaces, take out all lines mentioning eth0, then put the following in: auto eth0 iface eth0 inet static address 192.168.1.111 netmask 255.255.255.0 and then just reboot the whole machine, and see if you get an address out of ifconfig when you get back to a prompt. also, i just have to ask--have you tried this with another cable, and have you tried this in one of the ports on the hub that you know to work? sean pgp0.pgp Description: PGP signature
Re: Newbie bull brings own china shop.
On Sun, Mar 02, 2003 at 05:29:56PM +0700, Brian Durant wrote: > Knoppix got finished earlier than expected. I first tried it on > another machine on my WAN/LAN to see if the CD would work with another > configuration. This went off without a hitch, though it is necessary to > prompt Knoppix to search for a dhcp assignment of an IP address by using > "netcartconfig". I gave the command "knoppix noapic" at the boot prompt > on my daughter's computer, and then used the "netcartconfig" tool to get > an IP address assigned by DHCP. Unfortunately, this didn't work. that's odd... for everything in my lan, i don't have to do anything, knoppix figures it all out and automatically does dhcp. what do you have giving out dhcp addresses, anyway? > I added the lines that you suggested to /etc/interfaces (there were no > contents to start with in the file). I then rebooted. There was no > output to the command "# /sbin/ifconfig" or "ifconfig". However, a new > line prompt did appear. that's really odd that this file wasn't there. are you sure you have the ifupdown package installed? try $ dpkg -L ifupdown if it outputs a list of files, then it's probably installed. if it isn't, try installing it. also, just to clarify, you aren't typing those # and $ characters, right? sean pgp0.pgp Description: PGP signature
Re: Working with WAV files
On Tue, Mar 04, 2003 at 04:07:07PM -0500, David Turetsky wrote: > What packages are out there which facilitate editing? a good one that you can run from the command line is sox. it has a bunch of basic effects and filters you can pass the sound through, i've used it in the past and been happy with it sean pgp0.pgp Description: PGP signature
Re: kernel menuconfig
On Wed, Mar 05, 2003 at 11:12:09AM -0800, Curtis Vaughan wrote: > Has anyone created a file that shows all the options available in > menuconfig when doing a kernel compile? how about /usr/src/linux/.config ? or maybe i misunderstood you? you can also see what kernel options are compiled into your current running kernel by looking at /boot/config-`uname -r` hth sean pgp0.pgp Description: PGP signature
where'd my mouse go?
hey -user, so i just upgraded my laptop's kernel, threw in devfs, apt-got devfsd, and booted back up with minimal hassle[1]. everything is completely functional and happy, with the exception of my touchpad mouse. the device /dev/psaux no longer exists. what i do have is /dev/mouse -> /dev/gpmdata /dev/usbmouse -> /dev/input/mouse i messed around with various things (/etc/devfsd/*, update-devfsd) but to no avail. this is very strange because on my desktop which has an almost identical setup (unstable, devfs, 2.4.20), i do have /dev/psaux, which is a symlink to /dev/misc/psaux. has anyone else had this problem, and more importantly, does anyone haev a fix? thanks sean [1] actually, i forgot to install devfsd, and didn't realize until i booted. fortunately i had a knoppix cd handy... pgp0.pgp Description: PGP signature
Re: where'd my mouse go?
On Fri, Mar 07, 2003 at 06:26:21PM +1100, Rob Weir wrote: > Stupid question: did you include PS/2 mouse support in your kernel? > Have you loaded it? No related device nodes will show up until you do. i have: CONFIG_PSMOUSE=m but no module that looks like it ought to be the psaux module. very odd. also, Configure.help in my linux src doesn't say anything about this being able to be a module. i think i'll just recompile the kernel with it built in and report back... sean pgp0.pgp Description: PGP signature
Re: Some kernel compile questions
On Fri, Mar 07, 2003 at 01:14:36PM +0530, Sukrit wrote: > 1. How do i decide which modules to load at boot time, which file is > to be edited? (i am thinking that i'll compile support for lots of > devices - cd-writer, different network cards - as modules that way i > won't have to recompile kernel when i add those devices. So i don't > want all my modules to be loaded at boot time. Also, i don't want to > dynamically and automatically load/unload modules) /etc/modules > 2. What is the function of system.map file, after compiling kernel > (the non-debian way) is it necessary to copy the file to /boot and edit > lilo.conf to reflect changes. i think it maps addresses in the kernel to symbols > 3. i compiled the kernel, now do i need to get kernel-headers? What is > the function of kernel-headers? kernel-headers are the headers you used to compile your kernel. they come in really handy for compiling add-in modules later, such as the nvidia drivers, and take up much less space than the kernel source. the best way to make them is to make them with make-kpkg (same goes for your kernel). to do this, do something like: # cp yourconfig /usr/src/linux/.config # cd /usr/src/linux # make-kpkg --config=menuconfig --revision=custom.1.0 binary and then you should have kernel, kernel-source, kernel-header, and kernel-documentation .debs in /usr/src that you can install with dpkg at your leisure. > 4. Right now i am copying kernel bzImage into /boot renaming it, and > editing lilo. Also for modules i > #make modules > #make modules_install > > If i already have the same version of kernel running does old module > tree get over-written or not? probably does! note you don't have to do any of this with the above method. (er, it's done for you :) regards sean pgp0.pgp Description: PGP signature
Re: where'd my mouse go?
On Fri, Mar 07, 2003 at 10:23:05AM -0500, sean finney wrote: > but no module that looks like it ought to be the psaux module. very > odd. also, Configure.help in my linux src doesn't say anything about > this being able to be a module. i think i'll just recompile the > kernel with it built in and report back... gah, still no beans. this is bizarre, it worked in 2.4.18 just fine. i guess i'll try compiling another kernel, without devfs support and make sure *that* works... sean pgp0.pgp Description: PGP signature
[solved] Re: where'd my mouse go?
so about a dozen kernels later... i reverted to the debian stock kernel for 2.4.20, and things worked on that. so i examined the relevant differences between my config and debian's: 8,11c8,11 < CONFIG_MOUSE=m < CONFIG_PSMOUSE=y <--- debian --- > CONFIG_MOUSE=y > CONFIG_PSMOUSE=m <--- mine the funny thing, according to menuconfig, it should be the way my config has it listed. there's no option to set CONFIG_PSMOUSE to m, it's on or off! so what did i do? i opened up .config in vi, made the change manually, recompiled the kernel, and what do you know, i have a mouse again. the *really* strange thing is that neither CONFIG_MOUSE nor CONFIG_PSMOUSE generate insertable modules when set to m... On Fri, Mar 07, 2003 at 04:10:02PM -0500, Narins, Josh wrote: > Your new mouse (i am pretty sure) is at /dev/input/mouse, which you knew. /dev/psaux (->/dev/misc/psaux in devfs) actually, because it's a laptop touchpad mouse not usb. sean pgp0.pgp Description: PGP signature
Re: SIS900 DCHP Problem (was: Newbie bull ...)
On Thu, Mar 06, 2003 at 05:06:06PM +0100, Michael Bona wrote: > Sorry, I am confused: Detection with DHCP fails - I can confirm that ;-( > Then what do you do to make the detection "work fine? do you have anything in your kernel logs that looks like: ... Error EERPOM read ... ? i remember a fellow on another list i'm on had a problem with his sis900 card too, it was a bad eeprom that returned bogus values. i sent him a patch that i thought might fix it, but he never got back to me if it worked... sean pgp0.pgp Description: PGP signature
Re: about matrix text editing
On Mon, Mar 10, 2003 at 04:02:47AM +0900, Youichi Mano wrote: > > a 1957111 > b 1902222 > c 2001111 > > > i.e. the output will be > > a 1957111 > c 2001111 > grep -E '^[a-z]+[[:space:]]+[0-9]+[[:space:]]+111$' < yourfile ? i think what might be slowing it down is the -P you're passing to grep. hth, sean pgp0.pgp Description: PGP signature
Re: Newbie questions
On Sun, Mar 09, 2003 at 08:43:24PM +0100, Inge Thorin Eidsaether wrote: > 1 - I keep getting console messages about > 'eth0: link up, 100Mbps, full duplex, lpa 0x41E1' and > 'eth0: link down'. These two messages alternate regularly. > When the link is down, of course I cannot connect to anything. > Also, how do I avoid getting these annoying messages? are you sure this isn't a problem with your connection? do you get similar messages on your freebsd box? i think the easiest way to keep those messages off your console is to redirect them in /etc/syslog.conf(5) > 2 - Doing a 'man-k ' (or man -f ) does not work. > Is there a misconfiguration somewhere? man -k searches for keywords, not commands... or do you mean that it just doesn't work at all? i know that man -k was segfaulting on my unstable box for a while, but it's since been fixed and you said you're running stable. if it doesn't work at all, what version of man-db do you have installed? (you can find this out with dpkg --status man-db) > 3 - How do I go backwards in a man page reading? Looks like 'more' is used to > page the ouput to screen, but 'b' or ^B does not work here. > Silly question, maybe... it might be that you only have more installed. try apt-get install less and see if that fixes your problem. hth sean pgp0.pgp Description: PGP signature
Re: Newbie questions
On Sun, Mar 09, 2003 at 05:26:00PM -0500, [EMAIL PROTECTED] wrote: > In order to force the NIC to use a particular mode, you need to pass an option to > the pcnet32 module when it loads. I looked a while for the syntax but didn't find > it. Perhaps someone else here knows. instead of passing it as an option to insmod, how about you install the ethtool package? you can use it to force a link down into 10mbps mode among other things. i just checked, and according to README.Debian, your card is supported. hth sean pgp0.pgp Description: PGP signature
Re: Automatically log in
On Sun, Mar 09, 2003 at 03:46:56PM -0800, Leo Spalteholz wrote: > Hi, > I'm setting up a debian box to use as an mp3 player for a car.. What > I need it to do is automatically login when you power it up. I found > some resources on how to accomplish that by patching mingetty but I > was wondering if theres an easier way. > > Any suggestions? i suggested this a week or two ago for someone who needed the same thing. take a look at the rungetty package, and try replacing the entry for tty1 with rungetty. if you want an example, look at my previous post, i don't have it off hand but it's really, really simple. hth, sean pgp0.pgp Description: PGP signature
Re: Mirroring apt repositories?
On Sun, Mar 09, 2003 at 06:30:30PM -0700, Pete Ashdown wrote: > Has anyone know of a method for mirroring through apt? I'm aware of > apt-move, but I can't seem to get it to do a complete mirror that grabs > everything and not just the distribution that I have in the sources.list. > It also creates empty directories for all the architectures, regardless of > whether they are actually used or not. have you looked at the debmirror package? it does this pretty well, and has examples at the bottom of the manpage. however, i think it only does rsync and ftp. wonder why it doesn't do http... sean pgp0.pgp Description: PGP signature
Re: Mirroring apt repositories?
On Mon, Mar 10, 2003 at 09:30:35AM +0100, Sven Hoexter wrote: > Well http is not that easy to parse and handle when you try to mirror a > whole dir recursiv especially when directory index is forbidden. but i think you miss something... with debmirror, it doesn't ever do a recursive listing. it gets the Packages/Sources/Release files, and then gets stuff as it should be gotten according to that. i just hacked in http into the debmirror script, and it seems to (almost) work. wow, the things i do with my time... i'll see about cleaning it up and submitting a patch :) > If you intend to use debmirror I suggest to use it with debmirror-wrapper > http://people.debian.org/~absurd/debmirror-wrapper i haven't heard of that before, i'll check it out... regards, sean pgp0.pgp Description: PGP signature
Re: Mirroring apt repositories?
On Mon, Mar 10, 2003 at 09:14:20AM +, Colin Watson wrote: > Sure, but debmirror doesn't need to do directory indexing because it > uses Packages files to figure out what to download, so that's > irrelevant. Bug #154364 contains a patch which adds HTTP support (I > haven't tried it myself). aww.. serves me right for not checking bts first... i just got it working :( sean pgp0.pgp Description: PGP signature
Re: wtf? (long and frustrated)
(breaking the in-reply-to header, since this is really a different topic) On Mon, Mar 10, 2003 at 11:49:03AM -0700, Glenn English wrote: > When the installer says, "Have fun," and reboots, the screen blinks a > couple times, and a curses dialog box comes up saying it can't run X, > telling me why, and offering to run the X configuration program - that's > cool. I say, "Yes," and a program starts - IN X!!! > > This is not funny, folks; it's sadistic. heh, oops, sorry, wasn't supposed to laugh :) okay, let's do a little triage here. mouse: in a console, run gpmconfig. if you can get the mouse going in console mode, you're set. i don't know enough about your setup to know what kind of mouse you have, but i'd guess ps2. when it asks you about "repeating", tell it you want to repeat "raw" (w/out quotes). X: log in on a console and try dpkg-reconfigure xserver-xfree86 and let debconf take you through the configuration. i find it's the best way if you don't want to mess around with config files yourself. to get your mouse working, if you got it working on the console, give it /dev/gpmdata for your mouse device. PCMCIA: assuming you have a default debian kernel, you should install the pcmcia-modules package for your kernel version, and then the pcmcia-cs package. works pretty well in my experience. sean pgp0.pgp Description: PGP signature
Re: Package system vs. source vs. both
hey radek, i used to be real gung-ho on rolling my own everything, since i much of what i first learned about adminning was on a solaris box. after a while though, you'll find it gets really, really annoying when you want to upgrade (or lord forbid, remove) software. configuring X in debian is one of the weak points of the distribution at present imho (though that actually says a lot of good about debian, if you think about it), but i think it's worth your time to try and get it configured the Debian Way, and installed via the package management system. if you haven't already, try to get it working via dpkg-reconfigure xserver-xfree86 iirc it even asks you what level of expertise you want it to assume that you have wrt configuring X. otherwise, if compiling from source magically makes it work, perhaps you should try compiling the xserver-xfree86 package from source. you can do this with apt-get -b source xserver-xfree86 though you may still have to install its dependencies, this will build you the xserver package from source. this way you still the benefits of a .deb that you can install on your system. hth sean On Tue, Mar 11, 2003 at 10:36:04AM -0500, Radek Zajkowski wrote: > Here is a problem, I could not get the X to run from the packages on my > machine. It either would not configure, hang during configuration or refuse > to accept my settings. I am not an advanced user, therefore I often have no > facility to trace these kind of issues. > > As a result I downloaded the binaries of Xfree and it runs as a charm. I > compiled Emacs and my crappy pentium200 is now a bit more friendly, or at > least it offers the alternative to terminal. > > Although initially this was a matter of curiosity, X runs very smoothly and > I would like to keep it, and enrich it with more software. > > The problem I have created here is rather obvious, the package manager > doesn't know I have X libraries on my system and therefore, anything > requiring Xlibs will not install, since it forces the dependecies to be > configured as well. You probably get the rest of the story. > > I would like to find out from you some of the experiences and tactics you > might be employing when dealing with the package systems. Is it packages or > sources all the way, or are hybribds a common thing? > > Thanks, > > R> > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > pgp0.pgp Description: PGP signature
Re: [off topic] Learning Shell from an old UNIX book
On Tue, Mar 11, 2003 at 05:41:46PM +0200, Aryan Ameri wrote: > Hi There: > 1 ) Can this book be beneficial for me? or is it so obsolete that it is not > usefull anymore? i would say this makes it even *more* useful. personally, i do all my scripting in /bin/sh (that's the Bourne shell), because that's really the only thing you're guaranteed to have on a UNIX system. this is, of course, if portability is at all a concern of yours. imho, there is a zen to sh scripting, and staying with such an approach will gain you much. > The book shows examples for all of these tree shells. Therefore I wonder > 2 ) Bash is more similar to which one of these Shells? Korn Bourne or C ? Bash is the "Bourne Again" shell, if that's at all a subtle hint :). > 3) What things shall I keep in mind when reading example programs. Do commads > on Korn, Bourne and C, usually work on Bash? Or is Bash using a completely > diffrent syntax? ksh and bash will likely be the same for your needs and uses, but csh is horribly, horribly different and you shouldn't be coding in it anyways, you should be avoiding it like the plague... sean pgp0.pgp Description: PGP signature
Re: Problem with perl & locale
On Wed, Mar 12, 2003 at 03:21:03PM -, [EMAIL PROTECTED] wrote: > I've tried to: 1) regenerate locale via locale-gen; 2) re-install locales package. > To no avail! I can't make head or tail of it. > > Please help have you tried: dpkg-reconfigure locales ? it prompts you for a list of the locales you want to generate. sean pgp0.pgp Description: PGP signature
Re: SIS900 DCHP Problem
On Wed, Mar 12, 2003 at 03:47:51PM +0100, Robert Epprecht wrote: > I am not the guy you talk about, but with kernel 2.4.18 dmesg said: > sis900.c: v1.08.02 11/30/2001 > eth0: Error EERPOM read > > I was told, that the reason was the driver not able to read the mac address > on the card. I had to update to kernel 2.4.20 and switch ACPI off to get > sis900 running. (The ACPI stuff seems to be related to my motherboard). yeah, that's the error i was talking about. you can tell because of the typho (should be EEPROM, not EERPROM). do you still get that message? i wrote a patch for it that i believe works, but i don't have a card to test it on :) sean pgp0.pgp Description: PGP signature
Re: your mail
(don't forget your Subject: line in the future, most people aren't reatding this email because there isn't one) On Wed, Mar 12, 2003 at 10:31:53AM -0600, linux stuff wrote: > ssl question: i have a debian box w/apache, mysql, php, etc, running just > fine ... now a client requires secure sockets ... one place i read > (aboutdebian.com) made it sound impossibly difficult to set up ... is that > true? or is there a HOWTO that could help an beginning/intermediate admin > succeed? or should i give up? all help appreciated > apt-get remove apache apt-get install apache-ssl or have i missed something? i did this on my box for my own ssl based needs, and it's really just a drop-in replacement. if you have complicated configs, you might have to spend some time merging them into your new config directory (/etc/apache-ssl) but otherwise it's a piece of cake. hth sean pgp0.pgp Description: PGP signature
Re: Re: Problem with perl & locale
On Wed, Mar 12, 2003 at 05:16:08PM -, [EMAIL PROTECTED] wrote: > >dpkg-reconfigure locales > > > YES, the same result as locale-gen. It seems to finish correctly but it simply > doesn't work... and you selected your language, and it generated it? that's odd. what's in: /etc/locale.gen and what's the output of 'env'? sean pgp0.pgp Description: PGP signature
Re: Kernel panic: No init found.
On Wed, Mar 12, 2003 at 01:09:50PM -0500, Reid Mumford wrote: > EXT2-fs warning: checktime reached, running e2fsck is recommended > VFS: Mounted root (ext2 filesystem). > VFS: Mounted root (ext2 filesystem) readonly. > change_root: old root has d_count=1 > Trying to unmount old root ... okay > Freeing unused kernel memory: 156k freed > Kernel panic: No init found. Try passing init= option to kernel > > Any ideas on how to save this system without having to reinstall? try booting up with init=/bin/sh and see if you can get yourself a root prompt. if so, you're golden. you might need to reinstall some packages (like sysvinit), and i don't know what the easiest way is to find packages that might have corrupted files though. hth sean pgp0.pgp Description: PGP signature
Re: Setting Static IP Adreess, done...now setup Apt-Get
On Wed, Mar 12, 2003 at 02:45:02PM -0800, CM Miller wrote: > > Thanks to everyone who helped me setup the static > ipaddress. I can now ping and do loopback. The only > problem is that now I ran /usr/sbin/apt-setup and it > cannot resolve any of the www addresses that I select. > > any ideas why? can you ping ip addresses? for example, can you ping 216.239.33.101 ? (that's www.google.com) if so, you're only missing entries for your DNS name servers from your ISP. once you have those, put them in /etc/resolv.conf like: nameserver 192.168.0.1 where 192.168.0.1 is replaced with your name server, naturally. hth, sean pgp0.pgp Description: PGP signature
Re: considered harmful (was [off topic] Learning Shell from an old UNIX book)
On Thu, Mar 13, 2003 at 06:45:15AM +0800, [EMAIL PROTECTED] wrote: > I'm not convinced. As the article itself points out, there are > workarounds for the perceived weaknesses of csh. And why pick on > the granddaddy, when the grandson is very much alive and kickin? > I've been using tcsh for both my scripts and its more > user-friendly shell. Admittedly I don't unleash them "unto the > world". But I know of at least one big project that uses tcsh > scripting, OpenOffice.org. i think one of the points of shell scripting is to be as portable as possible, and nothing is more portable in the world of shell scripting than /bin/sh... afaik *csh is horribly inconsistant wrt POSIX compliance, and there are even parts that can be optionally compiled POSIX-compliant and non-compliant... that suggests one could get some nasty headaches on different vendor's implementations of *csh. sean pgp0.pgp Description: PGP signature
Re: SIS900 DCHP Problem
On Wed, Mar 12, 2003 at 10:49:21PM +0100, Robert Epprecht wrote: > > yeah, that's the error i was talking about. you can tell because of the > > typho (should be EEPROM, not EERPROM). do you still get that message? > > no good, maybe it's already fixed in upstream then... > I could try starting the old kernel which would probably give us back > the error ;-) (Currently I'm on another machine, but I can try tomorrow) > > If I can help in testing your patch please tell me. eh. unless someone else is still actively having that problem, i say let sleeping patches lie. you'd have to recompile a kernel or module, and it looks like it's already been fixed upstream... sean pgp0.pgp Description: PGP signature
Re: Debian Boot Process
On Thu, Mar 13, 2003 at 10:01:02AM -0800, Kris wrote: > Ok I am trying to do some fancy stuff and need to know how the debian 2.4.18 > boot process works. Ok ouch I will compare. If for example I wanted to > know the step by step process of an msdos system I would say something like okay, so - boot sector gets read from hard drive, instructs bios to load in kernel from specified location. - kernel load ups, uses initrd if configured (and loads up modules found in initrd if needed) - kernel mounts real root filesystem (ro), reads in /etd/modules, and loads those modules. - kernel spawns off init - the rest is in /usr/share/doc/sysvinit/README.runlevels.gz hth sean pgp0.pgp Description: PGP signature
Re: using white noise for cryptography
hey martin, On Thu, Mar 13, 2003 at 11:43:54PM +0100, martin f krafft wrote: > 1. is there already a package that enables this? not that i know of. > 2. is there a way to inject bytes into the entropy pool of the > linux kernel? yeah, at least as a module. don't know about from userland. basically you declare a device as being able to contribute to the random entropy pool, and give a means for doing so (i don't recall if you have to actually patch the kernel or not, i can look into this tomorrow when i have my hands on a linux kernel book) > 3. how can i read the data from the microphone? it's being > amplified by my soundcard, but cat'ing /dev/dsp gives nothing > really (well, the same byte repeatadly). you can read from it with rec (in the sox package). i used it a couple days ago to record audio from a cassette being pumped into the mic jack. the trouble is, i don't know how to read from it and have it not come out the speakers (i turned themic up, and the master volume down when recording). then again, you hardly need the mic on at all for this, because if you're talking about what i think you're talking about, you get your random entropy from normalizing the amplified white noise, right? sean pgp0.pgp Description: PGP signature
Re: wget usage help, please
hey stan, On Fri, Mar 14, 2003 at 09:48:28AM -0500, stan wrote: > I;ve tried thinhs likke --mirrot and --convert-links, but I wind up having > wget chase loinks all over the web. How can I restrict it to jsut follow > links on this site itseelf? i usually use wget -m for this, which is the same as --mirror. so i'm not sure where your problem is coming from. if you are still getting links from other hosts, i guess you can explicitly say which domains to follow with the -D option (man wget for the desc.). hth, sean pgp0.pgp Description: PGP signature
Re: Convincing someone to switch to Linux
hey roberto, like the other post said, really, i don't know how much of a chance you have for convincing them. that doesn't mean it isn't worth a shot though. i think your best bet is downloading and burning a knoppix iso, and when they get their new shiny computer, put it in and see what they think. it's effectively an entire debian system running off of cd (so they you don't endanger their dear, dear MS stuff), and if they like it, it's a rather painless install off of cd onto the hard drive. sean pgp0.pgp Description: PGP signature
Re: Samba, rsync, home network w/ XP, backups?
hi barry, On Sat, Mar 15, 2003 at 01:33:43PM -0500, Barry Mathieu wrote: > I think rsync is the correct application for such a task; is this > correct? I don't want try to redesign the wheel and I think many others > have passed this way before. I'm a bit confused by the rysnc > documentation - it doesn't appear to be used for interfacing using SMB. i can think of a couple ways of doing this, depending on your desired level of security, and whether or not you want to regularly automate it or just manually run it every now and then. a ] share your entire xp drive (ew...), smbmount it onto the woody box, and then just rsync -a between two directories (i wouldn't recommend that as an automated solution, leaving your entire drive shared is bad...) b ] install cygwin on your xp box, and through cygwin install rsync, cron, and ssh. then, rsync the directories when you want, just like you would on another linux box (this assumes you know how crontabs work if you want to automate it). (probably the better solution, but if you want to automate it you'll need to learn about pubkey authentication) hth sean pgp0.pgp Description: PGP signature
NFS hangs??
hi -users, i'm running with two unstable boxes, one as an nfs client and the other as a server. when i try and connect from a client, it hangs indefinitely, and not only that, doesn't respond to backgrounding with ^Z or die with ^C. not surprisingly, it also does not respond to any signal delivered via kill. at some point later (maybe 5, 10 minutes), it looks like it finally recieves one of the 5000 kill signals i sent it, and dies. now the REALLY strange part... after it dies, the directory is mounted! does anyone know what's going on here? umount does not seem to have these problems afterwards, either. i had this problem about a year ago too, and thought it had to do with mismatched nfs kernel implementations, but now it seems like it's not. here's the relevant info: exports on the server: /usr/local/music192.168.0.0/24(ro,sync) kernel settings on the server: CONFIG_NFS_FS=m CONFIG_NFS_V3=y # CONFIG_ROOT_NFS is not set CONFIG_NFSD=m CONFIG_NFSD_V3=y # CONFIG_NFSD_TCP is not set # CONFIG_NCPFS_NFS_NS is not set from the server's log: Mar 15 20:25:53 balthasar rpc.mountd: authenticated mount request from 192.168.0 .99:802 for /usr/local/music (/usr/local/music) kernel settings on the client: CONFIG_NFS_FS=m CONFIG_NFS_V3=y # CONFIG_ROOT_NFS is not set CONFIG_NFSD=m CONFIG_NFSD_V3=y # CONFIG_NFSD_TCP is not set # CONFIG_NCPFS_NFS_NS is not set any help would be greatly appreciated. sean pgp0.pgp Description: PGP signature
Re: NFS hangs??
On Sat, Mar 15, 2003 at 10:26:58PM -0700, Bob Proulx wrote: > As to the original trouble, check out the errors that are undoubtedly > occurring in /var/log/syslog. Try running tcpdump at the same time > and see if there is a clue as to why you are hanging. Run 'showmount > -e servername' and see that you are allowed access. hmm... okay, _after_ the thing "fails" (and the stuff is mounted) i have: in client's messages: Mar 15 20:09:04 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:14:19 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:19:34 sativa kernel: portmap: server localhost not responding, timed o ut in client's syslog: Mar 15 20:03:21 sativa kernel: lockd_down: no lockd running. Mar 15 20:09:04 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:14:19 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:19:34 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:29:15 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:34:30 sativa kernel: portmap: server localhost not responding, timed o ut ooh, and this looks interesting, kern.log: === Mar 15 19:13:39 sativa kernel: lockd_up: makesock failed, error=-5 Mar 15 19:18:54 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:03:21 sativa kernel: lockd_down: no lockd running. Mar 15 20:09:04 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:14:19 sativa kernel: portmap: server localhost not responding, timed o ut Mar 15 20:19:34 sativa kernel: portmap: server localhost not responding, timed o ut error -5... i can find the spot in the kernel code where it's happening: /usr/src/kernel-source-2.4.20/fs/lockd/svc.c:253 failing to create a socket? but why... anyway, maybe i'll get lucky messing around with some mount options, and otherwise i'll look into tcpdump, and showmount, and report back in a bit... (in the meantime, using smbfs for my needs...) thanks sean pgp0.pgp Description: PGP signature
Re: Auto apt-get upgrade
On Sun, Mar 16, 2003 at 09:55:38AM -0500, [EMAIL PROTECTED] wrote: > On Sun, 16 Mar 2003 10:24:33 + (GMT) > Rus Foster <[EMAIL PROTECTED]> wrote: > I suppose you really meant "apt-get update && apt-get upgrade"? > > There's nothing at all wrong with doing "update" in a cron job, since > that just fetches the newest packages lists. > > Doing "upgrade" by cron? i like what cron-apt does by default though. it runs update, and then runs the download-only portion of upgrade. that way, when you decide it's time to upgrade, you already have the packages. sean pgp0.pgp Description: PGP signature
Re: making nvidia drivers
hi john, On Sun, Mar 16, 2003 at 09:39:59AM -0800, John F. wrote: > > I don't really think you need to have a .deb pkg to do that since > > you're 'patching' the kernel. > > However enter Nvidia_kernel dir and type 'make install', then go to > > Nvidia_glx and do the same. Then you have to edit the XFREE config > > file as stated in the nividia installation guide on nvidia site. i don't think this is correct. at the bottom of this email are some instructions that i posted a few months ago on this list. all else failing, don't forget to look in /usr/share/doc/, there are good instructions there. sean - roll your own kernel with make-kpkg, or at least install the kernel-headers package for your kernel version (after extracting the kernel source/headers to /usr/src/linux) # cd /usr/src/linux && make-kpkg binary (there will be .debs in /usr/src after that. OR...) # apt-get install kernel-headers-2.x.x - install the following packages: # apt-get install nvidia-glx-src nvidia-kernel-src - extract the nvidia module source # cd /usr/src && tar xvfz nvidia-kernel-src.tar.gz - build the kernel module for the nvidia card # cd /usr/src/linux && make-kpkg modules_image - then in /usr/src you should have a .deb package for nvidia kernel # cd /usr/src && dpkg -i nvidia-kernel-2.x.x_rev_i386.deb - then install the glx stuff # cd /usr/src/nvidia-glx_version && dpkg-buildpackage -us -uc - go back to /usr/src, and install everything that starts with nvidia-glx and ends with .deb - finally, reconfigure X, which i strongly recommend that you do with debconf # dpkg-reconfigure xserver-xfree86 there should be a new driver entry when it asks you for your video card driver, like nvidia or nvdriver or something (not nv), choose that, and you should be on your way. pgp0.pgp Description: PGP signature
Re: making nvidia drivers
On Sun, Mar 16, 2003 at 11:13:35AM -0800, John F. wrote: > I looked at that, but the part with "make-kpkg modules_image" seems like > it is starting to compile a new kernel. I don't know, but most of the > questions I was being asked made no sense to me, I'm still kinda a > newbie after a year and a half in Linux. I don't know what I'm supposed > to do. take a look at my instructions, skip the step about making the kernel and instead install the kernel-headers package that matches your installed kernel (uname -r). you may need to do ln -s kernel-headers-`uname -r` linux in /usr/src too, i haven't used kernel-headers in a while... and this is assuming you haven't compiled your own kernel. the rest of the steps should work after that sean pgp0.pgp Description: PGP signature
Re: How do I get the install disks to recognize my network?
hiya, On Sun, Mar 16, 2003 at 06:13:52PM -0500, John Fisher wrote: > Good afternoon. I am having a problem installing Debian (my first install of any > Linux, actually) and was hoping someone might be able to render assistance. congrats, and good luck, hope you enjoy the ride :) > Due to lack of funds and lack of access to CD-burner, I was hoping to use the > Floppies / net install option to install to my new comp ( Intel D845GBV moterboard > with onboard LAN ). do you know what chipset it is? i would guess that it's an ether express card of some kind because the mobo is intel, so try ctrl-alt-d2 to get a root prompt during install, and running modprobe eepro100 then ifconfig eth0 if eth0 shows up, you should be good. if not, you might need to try other modules. they should be located in /lib/modules/`uname -r`/kernel/drivers/net/ > Seems to be working up until the installation of base system, where there is no > 'Network' option. No prob, I figure, so I go to configure modules and find whatever > one I'll need so it can use my onboard ethernet info. > > But nothing there seems to be right, and when I guess, they all seem to fail. > > Any thoughts? Please? also, you said that this is your first install of linux. i'd really recommend taking the time to download and burn yourself a knoppix cd (also based off of debian). it's sort of like a sneak peak at how sweet your system can be once you've got it up and running, and you can use it to more robustly determine what devices etc. you have installed on your system (lspci is really helpful for this). hth sean pgp0.pgp Description: PGP signature
Re: make menuconfig fails
On Mon, Mar 17, 2003 at 02:14:47PM +1100, Russell Shaw wrote: > /usr/bin/ld: cannot find -lncurses > collect2: ld returned 1 exit status apt-get install libncurses5-dev (iirc) and while you're at it, apt-get install kernel-package man make-kpkg (this is a really, really cool utility for making kernel packages) hth sean pgp0.pgp Description: PGP signature
Re: nvidia follow-up and new question
heya, On Sun, Mar 16, 2003 at 06:49:21PM -0800, John F. wrote: > Okay, I got the driver to work through some secret methods (downloading > a tar file and compiling), but now I need to know how to change the PAM > settings, because when I try to run something like "tuxracer", I get the > following: you did it the nvidia way as opposed to the Debian Way, you mean? did the directions i posted not work? that's too bad... anyway, your solution is to add yourself to the video group. you can do this with the vigr program (if you know how to use vi), or perhaps some invocation of the adduser command, or maybe just editing the /etc/group file, adding your username after the colon for the video group. if you really don't care, you can just chmod 666 /dev/nvidia* (evil...) and never have this problem, but the best way is to add yourself to the video group (and add yourself to the audio and cdrom groups while you're at it), and then log out and log back in to your x session. regards sean pgp0.pgp Description: PGP signature
Re: Download accelerator
On Sun, Mar 16, 2003 at 08:44:07PM -0800, nate wrote: > Trey Sizemore said: > > Looking for a good download accelerator (similar to...Download > > Accelerator...). What are some of the favorites out there. > > is download accelerator one of those tools that tricks users into > thinking their connection is faster? no, it actually does make stuff go faster. it's a pretty nice utility, if you find yourself stuck on a win32 box. it does multiple gets in parallel, and it also has support for downloading the same file from multiple mirrors as well to help eliminate bottlenecks. > I use wget for most of my downloading needs, runs on the console > so I can fire it within screen so it's in the background, and it > supports auto resume, mirroring, and tons more features. wget is a gold standard as far as that goes. however, some sites do alter their content if they see that you're getting stuff from them with wget. of course, you can always get by that by specifying your own user-agent string (cmdline option, naturally). also, a program i've recently started using is puf (apt-get install puf), which does something similar with parallel fetching, though i don't think there's any browser plugin support for it. sean pgp0.pgp Description: PGP signature
Re: Removing CRONTAB jobs
On Mon, Mar 17, 2003 at 06:33:50AM +0100, n/a wrote: > I've just set up a firewall for some students, now it's late and i'm still awake and > > > this machine is making a lot of noise. Are there any crontab jobs i surely should > NOT remove ? in general, i think most the crontabs are there for a reason. for example, the find job (probably the noisiest of them all), updates the database that the "locate" command uses for finding stuff. if the noise keeps you up at night, i guess you could disable it, but then don't expect locate to be too helpful in a couple weeks. also, you can change the frequency of any of these jobs by moving them from /etc/cron.daily to /etc/cron.weekly if you really want. hth sean pgp0.pgp Description: PGP signature
Re: libvorbis0a? in unstable
hiya, On Mon, Mar 17, 2003 at 04:35:33AM -0500, Shawn Lamson wrote: > > It sounds like it should. If it does, you should send your solution > > to the libvorbis maintainer, since the current situation is less than > > ideal. > > Did anyone ever file a bug about this? #184764: libvorbis0a: Sould not conflicts with libvorbis0 has a discussion it looks like. anyone have an idea when a fix is going to go in? there's all kinds of packages i can't install from my fresh unstable install because of this. sean pgp0.pgp Description: PGP signature
Re: Samba, rsync, home network w/ XP, backups?
On Mon, Mar 17, 2003 at 03:33:01PM -0500, Barry Mathieu wrote: > Any idea on how to set-up XP Pro to accept smbmount? > > The XP Pro machine is named, 'red_hat'. > > Here's what happens when I try to list the shares on red_hat from my > debian machine, 'debian': > > [EMAIL PROTECTED]:~$ smbclient -L red_hat > added interface ip=192.168.0.3 bcast=192.168.0.255 nmask=255.255.255.0 > error connecting to 192.168.0.2:139 (Connection refused) > Error connecting to 192.168.0.2 (Connection refused) > Connection to red_hat failed try: smbclient -L //red_hat -Wyourworkgroupname -Uusername hth sean pgp0.pgp Description: PGP signature
Re: libvorbis0a? in unstable
On Tue, Mar 18, 2003 at 12:46:13PM -0500, Shawn Lamson wrote: > > On Mon, Mar 17, 2003 at 11:04:45AM -0500, sean finney wrote: > > > has a discussion it looks like. anyone have an idea when a fix is > > > going to go in? there's all kinds of packages i can't install from > > > my fresh unstable install because of this. > > > > Look upthread for a URL pointing you to the packages likely to go into > > sid to solve this. They work for me. > > They work for me too, however, won't updated packages overwrite > (replace) the supplemental ones I installed from that site: > http://people.debian.org/~pyro/libvorbis > ? > if this fix is as trivial as it seems, why hasn't it been uploaded yet? i don't mean to be getting cranky, but i can't install at least half a dozen packages because of this... so... is this happening any time soon, or should i put in some testing lines (the package is still in testing, right?) to fix this on my box? sean pgp0.pgp Description: PGP signature
Re: Bug#184764: libvorbis0a? in unstable
On Tue, Mar 18, 2003 at 03:37:21PM -0800, Brian Nelson wrote: > sean finney <[EMAIL PROTECTED]> writes: > It won't be uploaded soon because, at the moment, the maintainer doesn't > have access to a computer. uhoh. none at all? like, could we at least get in touch to get an ok for an nmu? sounds like the new packages work for folks, i built mine from the version in deb-src lists of testing and will hold off upgrading till the fix rolls out :) > Also, the fix unfortunately isn't so trivial. Previously, the ew, right. versioned depends sure would be nice right around now :) sean pgp0.pgp Description: PGP signature
Re: Bug#184764: libvorbis0a? in unstable
On Wed, Mar 19, 2003 at 02:59:29AM -0500, sean finney wrote: > uhoh. none at all? like, could we at least get in touch to get an > ok for an nmu? sounds like the new packages work for folks, i built > mine from the version in deb-src lists of testing and will hold off > upgrading till the fix rolls out :) turns out, hindsight comes on quick at 3am... all the software i want to install want libvorbis0a, not libvorbis0. wah. i'll go check out the new versions :) sean pgp0.pgp Description: PGP signature
Re: xv (graphics package)
On Wed, Mar 19, 2003 at 01:57:56PM -0500, [EMAIL PROTECTED] wrote: > Hi, > > I seem to recall that the xv package (graphics viewer / editor) by John > Bradley was formerly included in an earlier distribution (potato, maybe?). > It doesn't seem to be in the current distributions. Does anyone know if > it is available any longer as a debian package? Maybe I'm not looking in > the right place. i think the upstream author forbade binary redistributions of his software... sean pgp0.pgp Description: PGP signature
Re: possible IDE hard disk problem
On Mon, Mar 24, 2003 at 10:29:04AM -0500, David Roundy wrote: > My kernel is 2.4.20, which I compiled just recently, so another > Prior to this kernel I was on 2.4.17, I believe. For the moment I've out of curiosity, has the value for CONFIG_IDEDISK_MULTI_MODE changed between these two kernels? also, have you tried messing around with hdparm at all (perhaps finishing that backup first is in order:)? sean pgp0.pgp Description: PGP signature
Re: Howto redirect output from a terminal
heya, i'd suggest either nohup or screen. the former is dirt simple but not very featureful, and the second is the opposite. hth sean On Wed, Jan 29, 2003 at 10:48:04AM -0600, Michael Kahle wrote: > I am not sure how to ask for what I want here. But, here goes. I was using > dump to make a backup of a directory on my computer. This was started from > a terminal in X. As it was running I could see in the terminal window all > the status output from the program. Good. This backup job took a long time > to complete and so I went home while the backup was still running. When at > home I ssh'd into the box to see if the dump process was still running, it > was. Then I got to thinking, how can I see the output of this program in my > current ssh session? Is there a program that exists to allow me to redirect > the output of a program running in a terminal to another terminal? I hope I > am explaining this clearly. > > Thanks, > > Michael > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg27246/pgp0.pgp Description: PGP signature
Re: OT: how to convert apple/mac .dmg to .iso images?
On Fri, Jan 31, 2003 at 03:08:29PM +0100, Nicos Gollan wrote: > Chances are that those images are the same thing as iso's with another > extension. To verify this, you could try to mount such an image locally with > type iso9660. If this works and you get a readable result, it's all fine. or to take even less time, try $ file foo.dmg if it says something like iso9660 in the output, chances are you can burn right away. if it's not, i'd ask the mac people about the format, and see if you could find specifications on it. i don't image that it could be that different from an iso image, as you'd need to convert it to iso before burning it to a cd. in a worst case situation i'd guess that it's a file with a bunch of meta-data followed by an iso image. hth sean msg27641/pgp0.pgp Description: PGP signature
Re: ssh keys from two behind-the-firewall boxes?
On Fri, Jan 31, 2003 at 05:39:22PM -0800, nate wrote: > SSH v2 is not *too* much different. though it's been a while > since I tried it with DSA authentication, last time I tried it, > it was a real bitch to get working(this was about a year ago) erm, iirc it's the same thing, only if you're using an older version of sshd you need to put the public key in ~/.ssh/authorized_keys2 instead. here's a really simple example: oil[~]20:58:57$ ssh-keygen -t dsa Generating public/private dsa key pair. Enter file in which to save the key (/home/finney/.ssh/id_dsa): Created directory '/home/finney/.ssh'. Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /home/finney/.ssh/id_dsa. Your public key has been saved in /home/finney/.ssh/id_dsa.pub. The key fingerprint is: e1:a2:6a:bc:e3:c9:b8:19:57:bc:a7:6c:b6:82:87:ac [EMAIL PROTECTED] oil[~]20:59:34$ cat .ssh/id_dsa.pub >> .ssh/authorized_keys2 oil[~]21:00:18$ ssh oil # <== look ma, no passwd! Kickstart-installed Red Hat Linux Thu Jan 16 10:36:39 EST 2003 # ^===(disclaimer: not my machine) :) oil[~]21:00:26$ logout Connection to oil closed. oil[~]21:00:27$ however, the above example uses a passphraseless key, which isn't the best wrt security (same goes if you're using sshv1 rsa keys too). what's better is to use ssh-agent and a passphrase on your key. sean msg2/pgp0.pgp Description: PGP signature
Re: List of users from a certain group
On Sat, Feb 01, 2003 at 05:38:02PM +0100, Calber Chainy wrote: > I have a group of users in my system just for mail, the name of that > group is "correo". > > My question is, how do I obtain a list of users from this group? how about grep ^correo /etc/group hth sean msg27919/pgp0.pgp Description: PGP signature
Re: ssh keys from two behind-the-firewall boxes?
On Sat, Feb 01, 2003 at 12:57:08PM -0800, nate wrote: > will trillich said: > > > at what point are the passphrases required? if passwordless > > login/scp is the objective, where are the passphrases used? > > ssh-agent is designed to prompt you for your passphrase, then > it stores it in memory, and automatically 'inputs' it when you > connect. That is until you logout or reboot or something. I > have never used ssh-agent myself. for my personal account I > use SSH w/passphrase and just input it every time. I use > passphrase-less keys for mostly non interactive stuff. afaik ssh-agent stores the key in memory, not the passphrase (you never give ssh-agent the passphrase, that's from ssh-add). ssh-agent outputs some environment variables that can be inherited by child processes, and then you load in your key with ssh-add or ssh-askpass. all the child processes that get spawned from the process that launched ssh-agent inheret certain env variables that let them know how to communicate with the agent, that then provides the key-based authentication for connections. this process can further be forwarded onto another machine, and the real beauty of it is that on the remote machine nothing is stored other than a socket to talk back to the agent on the home machine. for example, this is in my .xsession, and lets me ssh without a password even though i have a passphrase on my key in any child process of my xsession (xterms, et c.): eval `ssh-agent` ssh-add .ssh/id_dsa running w/o a passphrase is still probably the most common > way to perform automated tasks. that is, stuff from cron etc. true. you can however limit what commands can be executed from authorizing with a specific key. it'd be neat to see some way someone could spawn off cron using ssh-agent, but it'd make boot-up require someone be at the console. > if the system is properly secured the chance of a key getting > compromised is not that great. that's no attitude to take towards security. > on my more secure systems I lock them down to key logins only, > so even if they have my root password or account password they > have no opportunity to input them. right, but if they're storing a passphraseless key on another machine to which someone else has root, that someone else now has access to your machine too. if that's your root key... sean msg27985/pgp0.pgp Description: PGP signature
Re: shuttle disaster
On Sun, Feb 02, 2003 at 07:56:04AM +1100, David Pastern wrote: > My deepest commiserations to the US, NASA and all families involved over the > space shuttle Columbia tragedy. I believe that they were using Debian GNU > Linux for the first time onboard the shuttle :-( i've been watching the live feed online most of the day... tragic for so many reasons, really. are you serious about debian on the shuttle? where could i find out more about that? sean msg27987/pgp0.pgp Description: PGP signature
Re: OT: adduser vs manual file manipulation [WAS: Re: List of users from a certain group]
On Sun, Feb 02, 2003 at 01:03:20PM +0100, Johannes Zarl wrote: > I always edited the group, passwd and shadow files per hand to > add/remove/alter groups/users and I never had any problem. > > Is there any drawback I missed in using this procedure? Or is this simply > a matter of convenience? it's more of a convenience for the admin really, and ensures that everything is done the right way and that nothing's forgotten. it also creates home directories, copies over skeleton files, and makes sure that uids/gids are compliant with debian policy, et al. it's a pretty nice utility actually, and it's pretty configurable to your specific tastes (see adduser.conf(5)). sean msg28120/pgp0.pgp Description: PGP signature
Re: Configuration file for a running kernel
heya, usually, the config is in /boot like you said, but if you think it might be incorrect for some reason, and since you're running the bf24 kernel, you can read the kernel config for it here: ftp://ftp.debian.org/debian/dists/woody/main/disks-i386/3.0.23-2002-05-21/bf2.4/kernel-config i find it endlessly helpful to have the kernel config for the boot floppy in a file right next to the boot images. there have been more than a few times that people have said "it doesn't support this hardware", and then i could do a simple grep through those files and say "ok, try this boot floppy set". hth sean On Sun, Feb 02, 2003 at 04:48:06PM +0100, Andrej Prsa wrote: > Dear everybody, > > Is it possible to retrieve .config file for a running kernel somehow? I > guess they are usually acompanying the kernel in the /boot directory, but > if I try to compile the kernel with that config, I get significantly > different kernel size and behaviour is slightly different (e.g. PCMCIA > modules not being installed). Is this due only to different compiler or > does this mean that the config isn't the right one? I am talking about the > generic bf24 kernel that comes with Woody 3.0r1. I have found an option in > kernel configuration that kernel binary could include its config file, but > I guess it isn't a common practice to do? > > Thanks, > > Andrej > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg28121/pgp0.pgp Description: PGP signature
Re: Recursively export NFS file systems
On Sun, Feb 02, 2003 at 05:28:55PM +0100, Aaron Isotton wrote: > and then mount all of them by hand on osiris, but I'd like osiris not > to know about the file systems of the other machines. What I want is > to specify all the file systems which should be mounted by osiris in > the individual /etc/exports, WITHOUT osiris knowing anything about > that. > > Can that be done? How? not too sure, but i would guess there isn't an "easy way" of doing it. my first suggestion would be to write a sed/awk script that writes what you'd have to manually add to /etc/fstab and /etc/exports on the respective machines, then all you have to do is a little cat'ing. also, something that might work to do this would be to export the root fs as smbfs. i'm not sure if this will work as your / fs, but i don't think it pays attention to partitions at all. then again, it's a crappy, shady implementation to have for your root fs, so maybe i wouldn't... :) hth sean msg28122/pgp0.pgp Description: PGP signature
[OT] Re: shuttle disaster
On Sun, Feb 02, 2003 at 07:13:02AM +0800, [EMAIL PROTECTED] wrote: > The slashdot post was about Redhat in space. I doubt if it was > actually on board the shuttle in question. But that is neither > here nor there. I feel sad. The tragic loss aside, it also means > that the manned space program will be set back by another couple > of years. We're now farther away from 2001 than Arthur C. Clarke > and Stanley Kubrick was in 1968. well, i wouldn't be as worried about it now as i would have in 1986 with the challenger disaster, for a couple reasons: - the american public is more accepting of this tragedy and willing to move on - we have a couple astronauts in a space station who will need food sometime in june at the very latest sean msg28124/pgp0.pgp Description: PGP signature
Re: best way to disable journaling on ext3?
heya, disclaimer: i've never actually done this afaik ext3 is just ext2 with a journal tacked on. so that given, it should be really easy. replace all entries that say ext3 with ext2 in /etc/fstab, and you should be ready to go. just make sure you shutdown cleanly after doing so. i don't believe you have to modify the filesystem at all. hth sean On Sun, Feb 02, 2003 at 12:45:39PM -0900, Britton wrote: > > Short story: > > I suspect ext3 of possible causing crashes and would like to know the best > way to disable as much of it as possible without reformatting my disk. > > Longer story: > > I recently got a new 80G ATA disk, and cloned my debian install onto an > ext3 journaled file system on it. Now some programs that move lots of > data seg fault sometimes, causing subsequent ls to hang or my machine to > lock up completely. I think the kernel may be needing to do special work > to compensate for my bios not knowing about really big drives, because > older kernels didn't see the whole of even my 40G drive, and my bios > giives me an update ESCD successful (or something like that) on every > bootup. > > > Britton Kerin > __ > GNU GPL: "The Source will be with you... always." > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg28180/pgp0.pgp Description: PGP signature
Re: Disk formatting
heya, On Sun, Feb 02, 2003 at 08:30:51PM -0500, Jonathan Brandmeyer wrote: > Now, I found that /home has taken up almost all of /'s available space. > So, I created a new logical partion from some free space, as /dev/hdb5 > (~2GB) just wait, /var will probably be your next culprit :) > My question is: How can I format the new partition as ext3, and move > /home to it safely? mount /dev/hdb5 /mnt rsync -a /home/* /mnt umount /mnt mv /home /home.old mount /dev/hdb5 /home and you're running. then put /dev/hdb5 in /etc/fstab as /home, and you're set across reboots. note that you don't have to use rsync (there's cp,tar,pax...), but i use it so much these days it's the one for which i could remember the preserving options easiest. hth sean msg28213/pgp0.pgp Description: PGP signature
Re: Disk formatting
On Sun, Feb 02, 2003 at 09:35:28PM -0500, Mike Dresser wrote: > I like your method a lot more than my own, but one minor note. Doesn't > /home have to exist to mount on it? just throw a mkdir /home in the > middle of there, and he should be all set. right you are! you need to have a point on which to mount :) sean msg28239/pgp0.pgp Description: PGP signature
Re: Disk Corruption (was: Disk formatting)
heya, just an idea, if you can see all the directories and mount all your partitions from an emergency boot disk, try re-running lilo and booting again. hth sean On Sun, Feb 02, 2003 at 11:45:13PM -0500, Jonathan Brandmeyer wrote: > > Something VERY BAD has happened. When I attempted to format the disk, > with > > mkfs.ext3, it reported that the partition table for that partition said 0 > > size, and I should reboot the computer to re-read the partion table. When > I > > rebooted, instead of LILO, I saw an endless stream of "01 " repeating > > without end! I can boot into BIOS, and probably an emergency disk, but > that > > is all. > > > > I can definitely boot into the system with an emergency disk. > I should point out that my emergency disk is the stock 2.2.20 vanilla > kernel, while the current system kernel is 2.4.18-686, also stock. All of > the partions are ext2 or swap on hdb (since hdb5 was never formatted). This > kernel does not have support for my video driver, so I am in console-only > mode. > > dmesg shows this interesting entry: > Partition check: > hda: hda1 > hdb: hdb1 hdb2 hdb3 hdb4 > > and the view in cfdisk is: > Size: 60040544256 bytes > hdb1BootPrimaryLinux ext2509.97 > hdb2PrimaryLinux swap1019.94 > hdb3PrimaryLinux ext230721.43 > hdb5LogicalLinux1998.75 > LogicalFree Space25786.26 > > What if hdb4 is actually a primary partition that I cannot see? This would > violate the partitioning rules, wouldn't it? > > Thanks for any help, > Jonathan > > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > msg28240/pgp0.pgp Description: PGP signature
Re: give me the command line
heya, On Mon, Feb 03, 2003 at 02:17:33PM +0100, Hans Christian Andersen wrote: > I recently installed woody (vanilla) and KDE on a 400Mhz box > on my home LAN. I want it to go to KDE only after my commanding > startx. As it is now it starts KDE without prompting me. > How do I change that without reinstallation? > Thank you in advance. just an idea: # apt-get remove xdm (or kdm, or whatever login manager you have installed) sean msg28274/pgp0.pgp Description: PGP signature
