Re: Kernel question
Walter Williams <[EMAIL PROTECTED]> wrote: > I have subscribed to this list server to find > out more about non-Red Hat derivatives. > Is the Debian distribution of the type that I > can, when I want to update the kernel, > download a complete kernel tar ball or a patch > file from what ever web site I choose, install it, > and have things function properly? In my experience, it is. The laptop I'm using has a 2.4.0-test3 kernel and the machine at home, dual Pentium 133MHz also runs a custom compiled kernel. I usually get the kernel tar balls from ftp.funet.fi. The Debian package system does not force me to use a precompiled kernel-image from a Debian package. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Bootdisk for ThinkPad
> Can someone point me to a rescue disk for the ThinkPad? I can install Red > Hat 6.1 and Caldera 2.3, but not Debian! Try the tecra disks, resc1440tecra.bin and resc1440tecra-safe.bin. I am using a ThinkPad 770Z, and I had to use the tecra rescue disk to get it up and running. > Dwarf // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Flow chart drawing program
Micha Feigin <[EMAIL PROTECTED]> > I am looking for a program to make flow chart, mainly for drawing > dependencies for program design. > Any good programs available for the job? You could try the old and trusty xfig or the new and flashy dia. Xfig is already available for slink and potato in every mirror. The latest version of dia is 0.80, but you have to get it from an Incoming mirror or compile it yourself. The home page of dia is http://www.lysator.liu.se/~alla/dia/ and if you want to see a newer screenshot than the one on dia's home page, check http://www.cs.tut.fi/~hessu/dia.jpg (140KB). The screenshot was created just a few minutes ago. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: du and df conflict; disk space missing
Paul Reavis <[EMAIL PROTECTED]> > I'm very confused. I have a samba server running print services, and the > only real problem I have with it is that my users keep getting "out of > disk space" errors when printing. My /var partition is totally full: > > Filesystem 1024-blocks Used Available Capacity Mounted on > /dev/sda6 253807 2407110100% /var > > But I can't see why with `du`: > > padmount:/# du -s /var > 32567 /var > > What could cause this? There are probably processes that have open files in /var. Even if you remove a file, the disk space will not be freed before the last process keeping the file open exits. One possible process that has open files in /var is syslogd. Others might include daemons taking care of print services. The tool which can help you to find these mysterious prcesses is lsof (list open files). Below is an example where I use tail to keep a file open to demonstrate how du and df see things differently. I also use lsof to show how the xemacs binary is still using the disk space even if it has been removed earlier. Notice that disk space will not be really freed until I kill tail with Control-C. If you find something like syslogd keeping a file open, something like /etc/init.d/sysklogd restart should release the file. % du -k . 315 . % df -k . Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdc4 1169519 1083042 26048 98% / % cp /usr/bin/xemacs . % tail -f xemacs > /dev/null& [1] 9862 % du -k . 3690. % df -k . Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdc4 1169519 1086417 22673 98% / % rm xemacs % du -k . 315 . % df -k . Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdc4 1169519 1086417 22673 98% / % lsof +D . 19:19:rae COMMAND PID USER FD TYPE DEVICESIZE NODE NAME zsh 530 hessu cwdDIR 22,41024 22571 . tail9870 hessu cwdDIR 22,41024 22571 . tail9870 hessu3r REG 22,4 3440528 22654 ./xemacs % fg [1] + runningtail -f xemacs > /dev/null ^C zsh: exit 130 % df -k . Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdc4 1169519 1083042 26048 98% / // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: How to Remove kernel-image-2.2.9 with Buggy Script?
Art Lemasters <[EMAIL PROTECTED]> wrote: > When I try to remove an old kernel-image, the following error > occurs. > > Can't return outside a subroutine at > /var/lib/dpkg/info/kernel-image-2.2.9.postrm line 111 > > How can I efficiently remove it? ...tried dpkg --force- and that > didn't do it. If you open the file in your editor and search for return, you will find only one instance. Comment out the offending line by adding a # in front of the return statement. After that the script will complete succesfully and the package will be removed. > Art // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: filenames beginning with "-"
Rob Mahurin <[EMAIL PROTECTED]> wrote: > Apparently about an hour ago something made a few files in my homedir > that began with "-", and now I can't read them and can't remove them: > > 15:30 ~ $ ls > total 80 > -rw--- 1 alphengl alphengl 226 Oct 10 14:30 -011405 > -rw--- 1 alphengl alphengl 226 Oct 10 14:33 -011485 > -rw--- 1 alphengl alphengl 226 Oct 10 14:42 -011611 > [...] > 15:31 ~ $ cat \-011405 > cat: invalid option -- 0 Try something like $ cat ./-011405 That is, always prepend ./ to the filename if you have files that start with - > What are they and how do I make them go away? I have no idea. Maybe doing cat will help to find out where they came from. > Rob // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: KDE
dyer <[EMAIL PROTECTED]> wrote: > deb http://kde.tdyc.com slink kde > > This question has been asked many, many times. Please check the archives > before > posting. There is also a new mirror site, which probably is not yet in the list archives. Jean-Yves, you might want to try deb http://sunsite.tut.fi/ftp/Mirror/debian/ruins.tdyc.com/pub slink kde to see if it is faster for you than kde.tdyc.com. You might also want to try deb http://sunsite.tut.fi/ftp/Mirror/debian/ruins.tdyc.com/pub slink kde contrib rkrusty for some more packages. See http://kde.tdyc.com/Debian/ for more info about what is in contrib and rkrusty. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Need "ä" in a label, in a plot.
Patrik Magnusson <[EMAIL PROTECTED]> wrote:
> When I run latex on this file I get a heap of error messages:
> ! LaTeX Error: Command \Diamond not provided in base LaTeX2e.
My LaTeX Companion says on page 184 that \Diamond was defined in
LaTeX 2.09, but it is no longer defined in the base setup of NFSS
(New Font Selection Scheme). You should put a \usepackage{latexsym}
to make \Diamond and its friends available in LaTeX2e. The book
also says that if you use amsfonts or the amssymb package, you can
access the Diamond symbol that way.
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
Re: rsync from a cron script
Michael Stenner <[EMAIL PROTECTED]> > On Wed, Aug 18, 1999 at 04:24:06PM +0300, Heikki Vatiainen wrote: > > Michael Stenner <[EMAIL PROTECTED]> wrote: > > > RSYNC_RSH set to /usr/bin/rsync > > ^ > > Shouldn't this be ssh, not rsync? > > Doh... Yes, it should, and unfortunately, that IS how it appears in > the script. I mistyped in the post :) Ok, that would have been too easy :) Anyway, since you gave same doubts about environment variables, have you tried adding "env ; exit 0" and then comparing the output of the script when executed "by hand" or by cron. Maybe that could give you hints about what is different. // Heikki
Re: rsync from a cron script
Michael Stenner <[EMAIL PROTECTED]> wrote: > I have included the command "ssh remotehost date" in order to test > that ssh is working from the script (it is) -- yes, I have > RSYNC_RSH set to /usr/bin/rsync ^ Shouldn't this be ssh, not rsync? Heikki, hoping to find an easy solution :-) -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: rsync for potato
Marcus Johansson <[EMAIL PROTECTED]> wrote:
> Is there any way of using rsync to download, and keep up to date, the
> potato/unstable dist of Debian?
I use rsync to mirror the unstable binaries and source packages.
The rsync scripts I use flatten the symlinks that point to slink
so that all the packages that are the same for slink and potato
get picked up too.
The script for main archive ("anonftpsync") has a couple of exclude
lines that I need to keep the archive size down. I am pulling the
files to the laptop I am using, so it makes sense to me to not to
get everything from the archives.
Since you are from Sweden, the closest rsync-enabled mirror is
probably the sunsite we have here.
You can get the scripts from
http://sunsite.tut.fi/~hessu/deb-rsync/
Be sure to modify at least the TO variable in the script files.
> /Marcus
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
Re: Kernel Compilation Error?
Mark Wagnon <[EMAIL PROTECTED]> wrote: > I'm trying to compile a 2.0.36 kernel on my potato system, but I > keep getting this error: > > init/main.c: In function `get_options': > init/main.c:272: warning: subscript has type `char' > make: *** [init/main.o] Error 1 My guess is that your gcc is too new. In my potato system I have these too compilers installed: % gcc --version 2.95.1 % gcc272 --version 2.7.2.3 When I just tried compiling 2.0.36 kernel, I had to edit the toplevel kernel Makefile and replace all the instances of string "gcc" with "gcc272" to get the compile going. You might want to do the same thing and install the gcc272 package. Read /usr/doc/gcc272/README.Debian for more info about the old and new gcc. > I've downloaded the sources from several places, but it still > bails. Your source is probably correct, you just need the correct compiler. > I am able to compile a 2.2.* kernel (at least I was a couple > weeks ago). Same here. 2.2.* compiles with the newer gcc, but the older kernels need older gcc. > Any ideas? > > TIA I hope this helps, // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 3com509c not recognized
westk <[EMAIL PROTECTED]> wrote: > I was planning on upgrading to the 2.2 kernel as soon as I got the box on the > network, but now it sounds like I might be better off to download a 2.2 > kernel > now in order to get the NIC working. In which case, maybe I should just > forego > slink and go straight to potato. Or am I just asking for trouble doing that? Even if you did download 2.2.11 or 2.2.12 (which should be out really soon now), your 3c905C would not work. 2.2.1[12] comes with a little bit older version of the driver that does not recognize the new 3c905C card. The reason I know that even 2.2.12 will not work with your card, is that Alan Cox wrote that he will not take the risk of putting a new version of the driver into 2.2.12. He did wrote that 2.2.13pre1 should have support for 3c905C. Also, potato has problems when upgraded from slink (bash gets removed) so you probably do not want to do that unless you have read about the workarounds to that problem. If you are willing to trust me, you can download the 3c59x.o files I compiled for 2.0.36 and 2.0.37 kernels. The driver modules have been compiled with gcc 2.7.2.3 against the correct kernel headers. If you decide to use the drivers I have compiled, you can put them on e.g. floppy and install them from there. After installing the module fails with the debian installer, mount the floppy and install the module from the floppy drive. Later you can copy the driver over the old driver on the hard disk. The modules can be found under /lib/modules/ You can get the precompiled drivers from http://sunsite.tut.fi/~hessu/3c905C/ // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 3com509c not recognized
I wrote: > If this seems to be the case, then you have the 3c905C card, which > is not recognized by the kernel (yet) but is known to work. See > http://kernelnotes.org/lnxlists/linux-kernel/lk_9908_03/ > msg00191.html for more info. I have added a patch against 2.0.36 > and 2.0.37 in the bottom of this message for you convenience. It > looks like you have to patch and recompile at least the 3Com driver > in your kernel :-/ I'm following up myself since I just noticed that Donald Becker, the driver author, already has support for 3c905C and 2.0.x kernels in his latest driver. If you go to http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html you can find the link to the latest driver. If you get the driver from there, you do not need to patch the kernel sources. Just check the compile-command from the end of the source file and compile a new module for your kernel. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 3com509c not recognized
[EMAIL PROTECTED] wrote:
> I've just done a clean install of Slink (2.0.36 kernel) on a Gateway PII-400,
> but the installer doesn't install the 3c509c driver ("Installation
> Failed...").
You mean 3c905C, the twisted pair only 10/100Mbit/s Ethernet card
and not 3c509, the 10Mbit/s card?
Also, if you say "dmesg" after you try to install the module, is
there something like "Unknown 3Com PCI ethernet adapter type 9200
detected..." near the end of dmesg output?
If this seems to be the case, then you have the 3c905C card, which
is not recognized by the kernel (yet) but is known to work. See
http://kernelnotes.org/lnxlists/linux-kernel/lk_9908_03/
msg00191.html for more info. I have added a patch against 2.0.36
and 2.0.37 in the bottom of this message for you convenience. It
looks like you have to patch and recompile at least the 3Com driver
in your kernel :-/
> I've searched the archives and the web in general, but can't find much on the
> "c" card, except that it's a "Tornado" (as opposed to "Vornado" for the "b"
> card, IIRC). I've gotten the "b" card to work before, but this box has a "c"
> card, and I'm forbidden from swapping it out.
I think the b model (3c905B) is called Cyclone and this new model
(3c905C) is called Tornado.
> Anyone have any suggestions? (BTW, I'm still pretty green in LinuxLand, so
> talk
> down to me.)
What you should do is to decide if you want to upgrade to 2.0.37
which is the latest of the 2.0.x series or if you want to stay with
2.0.36. One of the differences between .36 and .37 kernels is that
among other changes, .37 seems to have a more recent 3c59x driver
that .36.
You can patch the driver by first unpacking the kernel source and
then going in the directory drivers/net/ . You do not need to be a
root to unpack and compile. You can unpack the sources e.g. in
your own home directory and then do the compiling there.
Save this mail in a file and edit out all the text except the
actual patch. The first lines in the patch file you are now
creating should start with "---" and "+++". Apply the patch with
command "patch -p0 < the_patch.txt" while still in the drivers/net/
directory.
Now when you have patched the file, see the bottom of the file. If
you do not want to recompile the whole kernel, there is
compile-command line that tells you how you can compile just that
module. Once you have compiled the module, try "insmod ./3c59x.o"
as the root.
Here is the patch against 2.0.36 and after that comes the patch
against 2.0.37. Warning: since I do not have 3c905C card, I have
not been able to test these patches. All I know is someone has been
able to get 3c905C running with these patches.
--- linux-2.0.36/drivers/net/3c59x.cSun Nov 15 20:33:02 1998
+++ 3c59x.c Mon Aug 16 21:22:27 1999
@@ -166,7 +166,7 @@
/* Caution! These entries must be consistent. */
static const int product_ids[] = {
0x5900, 0x5920, 0x5970, 0x5950, 0x5951, 0x5952, 0x9000, 0x9001,
- 0x9050, 0x9051, 0x9055, 0x5057, 0 };
+ 0x9050, 0x9051, 0x9055, 0x9200, 0x5057, 0 };
static const char *product_names[] = {
"3c590 Vortex 10Mbps",
"3c592 EISA 10mbps Demon/Vortex",
@@ -179,6 +179,7 @@
"3c905 Boomerang 100baseTx",
"3c905 Boomerang 100baseT4",
"3c905B Cyclone 100baseTx",
+ "3c905C Tornado 100baseTx",
"3c575",/* Cardbus
Boomerang */
};
--- linux-2.0.37/drivers/net/3c59x.cSun Jun 13 20:21:01 1999
+++ 3c59x.c Mon Aug 16 20:54:30 1999
@@ -267,6 +267,8 @@
PCI_USES_IO|PCI_USES_MASTER, IS_CYCLONE|HAS_NWAY, 128, vortex_probe1},
{"3c905B-FX Cyclone 100baseFx", 0x10B7, 0x905A, 0x,
PCI_USES_IO|PCI_USES_MASTER, IS_CYCLONE, 128, vortex_probe1},
+ {"3c905C Tornado", 0x10B7, 0x9200, 0x,
+ PCI_USES_IO|PCI_USES_MASTER, IS_CYCLONE, 128, vortex_probe1},
{"3c980 Cyclone", 0x10B7, 0x9800, 0xfff0,
PCI_USES_IO|PCI_USES_MASTER, IS_CYCLONE, 128, vortex_probe1},
{"3c575 Boomerang CardBus", 0x10B7, 0x5057, 0x,
> Th
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
Re: My terminal is in a weird state
Mark Wright wrote: > I just managed to completely hose my virtual terminal display. It seems > that non-alpha-numeric characters have been replaced with ascii > 127 > characters. I.e. the '-' becomes the upside down '!'. I've tried 'reset', > but that doesn't help. All of my virtual terminals are hosed. For now I > can telnet in, but eventually I'd like to be able to work at the machine > itself, and I'd like to avoid rebooting if I can. The problem started after > I droped out of X, after installing Blackbox 5. Any ideas on how to fix the > problem? Try logging in the host from the network and run "kbd_mode -a" as root. This should fix the case when the keyboard was not properly restored when switching to virtual terminal from the X server. More exactly, when the X server is not controlling the console (X is not on, or you are on a virtual console), kbd_mode without any arguments should report that the keyboard is in the default (ASCII) mode. When you are running X, the keyboard should be in the raw (scancode) mode. This is what you should get when trying kbd_mode from another host. > Mark. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Re[2]: Dselect odd behaviour
Phillip Deackes wrote: > Dan Nguyen <[EMAIL PROTECTED]> wrote: > > In article <[EMAIL PROTECTED]> you wrote: > > : I'm using dselect under potato using apt to ftp://sunsite.org.uk/ > > : and something strange seems to have happened. I did an 'update' and > > : now all the descriptions have disappeared. Also most of the packages > > : I don't have installed have gone. I've tryed updating again but to > > : no avail. Any hints on how to deal with this? > > Strange things happening here too. I am using the following > sources.list: > > deb ftp://unix.hensa.ac.uk:/mirrors/debian unstable main contrib > non-free > deb http://ftp1.us.debian.org/debian unstable main contrib > non-free > > When I tried to install mutt, I got this: > > Reading Package Lists... Done > Building Dependency Tree... Done > Package mutt has no available version, but exists in the database. > This typically means that the package was mentioned in a dependency and > never uploaded, or that it is an obsolete package. > E: Package mutt has no installation candidate > > Whatever package I try, I get the same response. If I access the hensa > site manually via Netscape, all the packages are there and accessible. > > Any ideas? I just checked the debian-devel list archives, and the problem seems to be that the Packages file which lists the package information is empty. See subject "archive package file broken?" e.g. at http://www.educ.umu.se/~bjorn/mhonarc-files/debian-devel-small/ Something had broken at the master archive and an empty Packages file was copied to all the mirror sites. As a result, the archive contents and list of archive contents do not match each other. > Phillip Deackes > [EMAIL PROTECTED] > Debian Linux (Potato) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: checking promiscuous mode
Eugene Sevinian <[EMAIL PROTECTED]> wrote: > Hi, ppl, > Does anybody know how to test whether the network card > set to promiscuous mode. It is supposed to run it from > cron. Just last week I tried a little program called "ifstatus" on Solaris. The program is designed to run from cron so that only if an interface is in promiscuous mode, it will print out a warning message. Here is a snippet from its manual page: DESCRIPTION Ifstatus checks all network interfaces on the system, and reports any that are in debug or promiscuous mode, which may be a sign of unauthorized access to the system. You can get it from ftp://coast.cs.purdue.edu /pub/tools/unix/ifstatus/ When you compile it for Linux you need to edit Makefile like this: -- cut here --- Makefile.orig Mon Sep 21 14:38:43 1998 +++ MakefileMon May 3 00:33:03 1999 @@ -43,8 +43,8 @@ # SUNOS55 -lkvm -lelf -lnsl -lsocket # SUNOS56 -lkvm -lelf -lnsl -lsocket # -OSNAME=SUNOS55 -LIBS= -lkvm -lelf -lnsl -lsocket +OSNAME=BSD +LIBS=# -lkvm -lelf -lnsl -lsocket # # Change BINDIR, MANDIR, and MANSUF if you want. They are used with the -- cut here That is, change SUNOS55 to BSD and comment out any libraries. > Thanks, No problem. > Eugene Sevinian // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: ok, this is annying the hell outta me...
Steve Lamb wrote: > >> Does *ANYONE* know of a way to turn off the default X server > >> permanently? > > Geez, I am upset. I forgot a work. Do you mean "a word" :=) > Does anyone know of a way to turn off the default X screen saver > permenently? Does "xset s off" do what you want. You also might want to try "xset -dpms". // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: NIC trouble
Ben Messinger wrote: > I just picked up some cheap used NE2K pci ethernet cards to use for a > home lan. I installed two in this Linux box since it may become a > proxy-gateway/firewall for a DSL connection soon. [cut] > I made entries in /etc/init.d/network so that it now reads: > > #! /bin/sh > ifconfig lo 127.0.0.1 > route add -net 127.0.0.0 > ifconfig eth0 netmask 255.255.255.0 broadcast 192.168.1.255 192.168.1.1 Just a thought. Does it work if you change the order of ifconfig argument? I have noticed that ifconfig can be picky about its arguments. Try this: ifconfig eth0 192.168.1.1 netmask 255.255.255.0 broadcast 192.168.1.255 > route add -net 192.168.1.0 netmask 255.255.255.0 dev eth0 > > But when the above is executed at boot time it returns the following > errors: > > SIOCSIFNETMASK: Cannot assign requested address > SIOCSIFBRDADDR: Cannot assign requested address > > The card (eth0) is now listed in the routing table, as well as being > identified by ifconfig, but I can not pass packets to my other computers > although I see the light blinking on the card when I try to ping another > box. > > Any help will be appreciated. Thank you. If reordering arguments did not work, what does ifconfig -a output? > -Ben // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: network
Mans Joling wrote: > Hi, > I have installed a 3c905b network card using modconf. > After linux starting up I see that he found him. > I have edit the host file and added the 2 ip adresses and look like > this > 127.0.0.1 localhost > 192.168.1.9 myhost > 192.168.1.8 other > reboot > When I do a ping command "ping other" he still says network > unreachable. > Under Bill's stuff he works > I use a pentium1 64 mb ram linux 2.0.36 Did you remember to say "route add -net 192.168.1.0" after you assigned an IP address to your network card? If you think you have the IP address and route set up correctly, you should say "ifconfig -a" once more and check what it says in the HWaddr field. My guess is that the card's hardware address is FF:FF:FF:FF:FF:FF which is not correct. In fact FF:FF:FF:FF:FF:FF is the Ethernet broadcast address which the older drivers for 3c590b detect incorrectly sometimes. If this is your case, see http://cesdis.gsfc.nasa.gov/linux/misc/modules.html# which gives you alternatives about what to do. The first thing to try is to power off your machine before booting to Linux from Windows. This should take care of the FF:FF:FF:FF:FF:FF problem. The other options are to upgrade to latest driver or disable "PnP OS" from BIOS if it has such setting. > Mans Joling // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Hitting web site consistently dropped ppp connection?
Ugh, I'm following up myself since I think I got it wrong about whose modem was closing the connection. I wrote: > Nathan O. Siemers wrote: > > I've got a weird one, folks. > > Not really, it's just a feature in your modem :) What I should have said is "your ISP's modem" since you are getting the page from your ISP and their modem probably thinks the triple + is a special sequence. > > Is there a ghost in the machine? How could that happen? Is this a > > freak coincidence or the tickling/exploit of a pppd bug? It there an > > escape sequence embedded in the transfer that could reset the modem? > > Escape sequence is probably the correct quess. I just > downloaded the page and one of comments had this signature: > > -- + ++ATH0 > ^ > Space added by me to protect modem owners. My guess is your > modem saw the three plus signs and hung up. ^ Again, "ISP's modem". // Heikki
Re: Hitting web site consistently dropped ppp connection?
Nathan O. Siemers wrote: > I've got a weird one, folks. Not really, it's just a feature in your modem :) > I've got a debian slink box serving as router for a network connected > to the internet by a dialup ppp connection. It does some masquerading > for the other systems on the local network, which include another > slinkbox and a redhat 5.2 system. > > > Yesterday, following the link: > > http://slashdot.org/comments.pl?sid=99/03/20/1158230&pid=0#35 , > > reading with netscape 4.07 on the redhat 5.2 system, > > killed my ppp connection to the internet from my slink router. > > This happened reproducibly (5 times) when I hit that particular page > (not a couple other pages at slashdot), at around 64K of download. I > sent lots of other test stuff over the ppp connection with no > problems. > > > Is there a ghost in the machine? How could that happen? Is this a > freak coincidence or the tickling/exploit of a pppd bug? It there an > escape sequence embedded in the transfer that could reset the modem? Escape sequence is probably the correct quess. I just downloaded the page and one of comments had this signature: -- + ++ATH0 ^ Space added by me to protect modem owners. My guess is your modem saw the three plus signs and hung up. Thus bug was discussed on linux-kernel last month. If you go to http://linuxhq.com/lnxlists/linux-kernel/ and check the linux-kernel mailing list archives for Feb 1999 - Week 1 and Feb 1999 - Week 2, you'll find the whole discussion under subjects: "PPP bug", "Re: plus plus plus [was: PPP bug]" and "Modems and the + + + bug" Here is a quote from a message by [EMAIL PROTECTED]: Like has been mentioned here recently (last week, I recall), if you have this problem, turn off the local modem escape by issuing AT S2=255 command before doing dialing; or dial by: ATS2=255DTnnn that is, prefix the dial command (DT) with escape disable command. Check your modem manual for the S-register number, and disabling value. Document I used said 'escape is turned off with any value above 127'. If the problem is at your ISP's side, then you need to inform them. > Perhaps I am just a victim of a freak coincidence. > > Thanks for any ideas, > > nathan // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: DNS--> h2n program
Ian Setford wrote: > I have been reading the O'Reilly DNS/Bind book and it mentions a > program/utility (h2n) to convert a hosts file to the db.* files needed for > DNS. I searched the debian packages in both stable/unstable and couldn't > find any mention of it. Anyone know where I can find this program or one > like it? Since you have the book (3rd edition I presume), see page xv or go directly to ftp://ftp.ora.com/published/oreilly/nutshell/dnsbind/ and get the file dns.3ed.tar.Z h2n is in the .tar.Z file and it even has a manual page. I have not used it myself though, so I do not know how well it works. > TIA. No problem. > -Ian // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 96% packet loss on my network
Paul Nathan Puri wrote: > The Port status section of the leds remains light on the 100 > section for both cards. > > The Network Utilization section of the hub leds consistently > blinks, on the 1% led of the 100 section; after every 13th > blink, both the 100 and the 10 blink together. > > Does this tell you anything? If all the lights are green and no alert leds are on, or none of the port status leds are yellow (port partitioned), or the collision led is not flashing wildly, then my guess is bad or conflicting hardware on your PCs. > Thanks. Hows the weather in Finland? In Sacramento, > California it wet and cloudy. C-ya. Finland is cold and snowy :) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: [Fwd: Re: Network problem - 3Com card]
Kent West wrote: [cut] > And to you, Heikki, thanks for the input. I stuck with the 2.0.34 kernel > though, cause it was over my head to download the 2.0.35 kernel from > another machine and copy it to my Linux box and etc, etc. But now that I > have a working network, maybe I'll upgrade to 2.0.35 (or maybe even go to > Slink -- whooo-o-o). Good to hear you got your network going! If you decide to to upgrade to 2.0.35 or 2.0.36 by recompiling your kernel, get the driver from the page I mentioned in my previous email and copy it over the existing linux/drivers/net/3c59x.c If I remember correctly, the driver in 2.0.3x is 0.99E which does not work correctly if you warm boot from Win9x. With the old driver the card gets assigned FF:FF:FF:FF:FF:FF Ethernet address which is not valid. The new driver does not have this problem. > Have a good one, ya'll! Thanks, I will! // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 96% packet loss on my network
Paul Nathan Puri wrote: > I just set up my network (laptop to desktop). I'm running linux on > both. > > I pinged my desktop and it seems there is 96% packet loss there. Why > would this be? > > On both machines there is a fast ethernet NIC, and I have a fast > ethernet 3Com hub. > > When I telnet my desktop from my laptop, a prompt comes up, but very > slowly. It won't let me type in my login, and then it times out. Is your hub one of those listed at http://www.3com.com/products/dsheets/400317a.html and do you have access to it so you can check the port LEDs? Since the cards work at least a little, they are probably running correctly at 100Mbs. However, if your hub is autosensing 10/100 Mbits it might have problems doing the speed conversion if one of your cards is in 10Mbits and the other is 100Mbits mode. I'm only guessing here since I have no experience about how well hubs do the speed conversion. One thing I know that has caused problems is NIC having different view on duplexness than the hub or switch. If the NIC runs in full-duplex mode and the hub is only half-duplex, the hub may partition (shut down) the port. The reason for this is the NIC causing too many collisions if it does not listen the receive pair before it starts to transmit. The port LEDs on the hub should indicate if the port has been partitioned, so you might want to check the hub before and after the connection dies. > Any help in this matter would be greatly appreciated. I hope this helps // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: 3com 3C905B
> I apologize for asking the question when it's been addressed so often, but > the answers I've found in the mail archives haven't been simplistic enough > for me. > > I have a hamm box freshly installed from CD. It has a 3COM 3c905B-TX NIC in > it that doesn't seem to be recognized. I've about decided from the mail > archives that the 2.034 kernel doesn't work with it. I do not know about 2.0.34 but it works with 2.0.35. I have a 3c905B and I have tested slink boot disks which came with 2.0.35. The current disk set, version 2.1.7, comes with 2.0.36, but the earlier disks worked fine with kernel 2.0.35 and 3c905B Ethernet card. > I also have a Win95 box I can use to download packages/files. > > I assume I need to download the 2.035 kernel and install it. Can anyone > give me detail instructions about what to get, where to get it, and how to > install it so my NIC will start working (assuming this is the problem)? Do you mean instructions with compiling the kernel or getting the 3c905B working with 2.0.35 kernel? If you need help only with 3c905B I suggest you take a look at http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html which is the home page for various 3c* drivers, including 3c905B. >From the page you will find the latest driver, the one I have been using, and also precompiled binaries for 2.0.x kernels. If you trust binaries compiled by a stranger, you can get the 3c59x.o module I have been using from ftp://sunsite.tut.fi/pub/Local/linux-atm/tmp/3c59x.o MD5 checksum 56d45abc68e436e81cfab1637262b8d6 3c59x.o The module was compiled from 3c59x.c, version 0.99H which is the same version that you can find from the web page above. The module may work with 2.0.34 but you have to use "insmod -f 3c59x.o" since the module was compiled with 2.0.35 and you are using 2.0.34. > TIA, Kent No problem // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Adding foreign keys
Dave Swegen wrote:
> I will have to start writing the occasional document in swedish in the near
> future (with a bit of luck), but since I have a UK keyboard I don't really
> want to switch to a swedish layout. I would prefer to use or
> somesuch to assign the extra 6 chars needed to the current layout. Does
> anyone know how to do this for both X and VCs? Any info would be much
> appreciated.
I have changed my keyboard mapping a bit so maybe I can help.
What I did was I remapped keys asdfer so that when
I press Atl-Gr with them I get {}[]()\ respectively. As a
result, when I write code I can keep my left hand at its normal
position and hit Alt-Gr with my righ thumb to get all the
different parentheses plus backslash.
This is what my .Xmodmap looks like:
keysym a = a A braceleft
keysym s = s S braceright
keysym d = d D bracketleft
keysym f = f F bracketright
keysym e = e E parenleft
keysym r = r R parenright
keysym aring = slash
keysym space = space space backslash
The Finnish/Swedish keyboard also has aring (å) which I mapped
to slash (/).
For VCs, here's a snippet from my /etc/kbd/default.map.gz
[cut]
keycode 30 = a
altgr keycode 30 = braceleft
[cut]
The names for characters with diacritics in X seem to correspond
to HTML entities, so you might want to check out
http://strindberg.ling.uu.se/call/chars.html to find out the
keysym names to put in .Xmodmap.
Another way to remap your keyboard is using the xkeycaps program
which is available as a Debian package.
I hope this helps.
> Cheers
> Dave
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
Re: X crashes after upgrade to slink
Marko Loparic wrote:
> Apparently X can't start my window manager (fvwm2) using startx. It quits
> after showing the X screen for a second. Using xinit another window
> manager is activated (I strange one, I can't tell what it is...)
>
> The X messages shows no fatal errors, but the .xsession.errors says
> something which seems to be causing the problem:
>
> /usr/X11R6/lib/X11/xinit/xinitrc: /etc/X11/Xsession: line 47: syntax error
> near unexpected token `default)'
> /usr/X11R6/lib/X11/xinit/xinitrc: /etc/X11/Xsession: line 47: ` default)'
>
> Any suggestion??? Please email me directly. Thanks a lot,
This is a known bug in xfree86-common-3.3.2.3a-9. A bug report
has already been filed against the package, so it should be fixed
soon. Here is an explanation and instructions how to fix it from
the bug report:
http://www.debian.org/Bugs/db/32/32713.html
Package: xfree86-common
Version: 3.3.2.3a-9
startx fails with a syntax error in /etc/X11/Xsession.
The following modification fixed the problem for me:
--- /etc/X11/Xsession~ Sun Java 31 19:58:00 1999
+++ /etc/X11/Xsession Sun Jan 31 19:59:00 1999
@@ -44,6 +44,7 @@
exit
fi
fi
+;;
default)
;;
*)
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
Re: Break Key
Daniel Marquez-Klaka wrote: > No, sorry but that does not work. When i start an traceroute > on a cisco and she get stuck, the send brk does nothing. My suggestion to this problem is to reconfigure the Cisco box a little. What I have done is I reconfigured the box a little so that it accepts Ctrl-C as an escape (break) character. Just add the escape-character commands to your Cisco's configuration and if it gets stuck, hit Ctrl-C fast a couple of times. I have learnt also to ping first and traceroute only if ping says the target is alive :) Here's an example, Ctrl-A = 1, Ctrl-B = 2, Ctrl-C = 3, etc. line con 0 escape-character 3 line vty 0 4 escape-character 3 > Daniel // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Printing problems
Patrick Colbeck wrote: > I haven't made any changes to my printing setup but all of a sudden I have > started getting this error when trying to print > > no connect permissions > > I am using Debian 2.0 upgraded to Slink and I do an update by APT every > week. My last update was yesterday perhaps this may have casued the problem ? You probably have lprng_3.5.2-1.2 installed which has a bug with permissions. lprng_3.5.2-1.3.deb fixes the bug and it should available at the mirrors. If that does not help, check the mailing list archives and see what Chuck Stickelman <[EMAIL PROTECTED]> wrote under subject "LPRng - Solved!!" // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Audio CD & MP3
Gregory Vandenbrouck wrote: > Is there utilities under debian to dump Audio CD and create MP3 files > ? I have found cdda2wav which create wav files, but nothing for mp3 :o( > > If there is no .deb, is there an utility I can compile ? My current favourite is L.A.M.E., which stands for L.A.M.E Ain't an Mp3 Encoder. It is fast and the quality is good enough, at least for me. LAME needs the ISO example source code to work so you have to do some patching and compiling to get it working. All you need is at http://student.uq.edu.au/~s302585/ PS. I'm using cdparanoia to create .wav files. The .deb is in slink and potato. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: ipchains and /dev/audio
Christopher R. Barry wrote: > I just installed the 2.2.1 kernel over 2.0.36. I couldn't find > `/dev/audio' and `/dev/mixer' in the configuration options so sound is > completely broken right now. What do I need to do to get `cat foo.au > /dev/audio' to work again? [stuff about ipchains cut] I am using 2.2.0-pre9 and I have the following sound related compile time options: # # Sound # CONFIG_SOUND=m CONFIG_SOUND_OSS=m CONFIG_SOUND_SB=m CONFIG_SOUND_ADLIB=m CONFIG_SOUND_YM3812=m My sound card is SB32PnP and I load the sound modules with this command: "modprobe sb irq=5 io=0x220 dma=1". To make sure you have /dev/audio, /dev/mixer and other /dev/ files setup correctly try "cd /dev ; ./MAKEDEV audio" // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Quicktime player for Linux?
Daniel Elenius wrote: > Does anyone know of a Quicktime .MOV player for Linux? xanim might do play them. However, I have found out that it can be picky about the color depth of your X server. If the movie does not play correctly, you might want to try starting your X in 32 or 8 bit modes with something like "startx -- -bpp 32" or "startx -- -bpp8" // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: ethernet collisions
Bob Nielsen wrote: > I have a system running potato which suddenly started experiencing a large > number of collisions on its ethernet port. It seems to work fine for a > while and suddenly I get collisions and everything grinds to a halt. This > just started and I was wondering if anything had changed in potato which > would cause this, or if I might have developed a hardware problem. I have a third alternative. Maybe one of your network cards is running in full duplex mode and the other one is only half duplex. However, sometimes the cards come up in matching duplex mode and things work fine. I have heard about cases where this particular "feature" has caused problems like the ones you are describing. I myself have two machines connected with a cross over cable and the link between them is running at 100Mbits/Full duplex. The both machines have a 3c905b network card and I ensure the correct mode of operation by loading the driver modules with "options=12". > Bob // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
More about lprng and /etc/lpd.perms
The discussion below is about lprng 3.5.2-1.2. I'm using potato but the same version seems to be in slink too. I have not yet filed the bug report I promised, so any suggestions to make a more intelligent bug report are welcome. Otherwise I'll send in what I have here withing the next 24 hours. I did some more research about lprng and /etc/lpd.perms. I found two ways to enable printing from local host. The first one takes 'disable all, enable what needed' approach and the other one is 'enable all, disable what not needed'. The both patches include one common restriction (ACCEPT SERVICE=S SERVER) which I found when checking out existing bug reports against lprng. In the first patch, which keeps "DEFAULT REJECT", I had to add both localhost and the fully qualified domain name to get things working. I found this out by running lpd with debuggin on. It looks like lpd likes to do a lot of DNS lookups. Adding the both names did the trick. --- lpd.perms.orig Sat Jan 30 21:11:29 1999 +++ lpd.perms Sat Jan 30 22:29:41 1999 @@ -163,8 +163,8 @@ ## # allow root on server to control jobs ACCEPT SERVICE=C SERVER REMOTEUSER=root -# allow anybody to get status -ACCEPT SERVICE=S +# allow anybody on server to get status +ACCEPT SERVICE=S SERVER # reject all others, including lpc commands permitted by user_lpc REJECT SERVICE=CSU # @@ -173,7 +173,9 @@ # allow root on server to remove a job ACCEPT SERVICE=M SERVER REMOTEUSER=root REJECT SERVICE=M -# all other operations denied +# all other operations allowed from local host, none from a remote host +ACCEPT SERVICE=XPRQ REMOTEHOST=localhost +ACCEPT SERVICE=XPRQ REMOTEHOST=rae.ton.tut.fi DEFAULT REJECT # all other operations allowed # DEFAULT ACCEPT The second patch does what [EMAIL PROTECTED] suggested and rejects everyone but localhost. --- lpd.perms.orig Sat Jan 30 21:11:29 1999 +++ lpd.perms Sat Jan 30 22:37:22 1999 @@ -163,8 +163,8 @@ ## # allow root on server to control jobs ACCEPT SERVICE=C SERVER REMOTEUSER=root -# allow anybody to get status -ACCEPT SERVICE=S +# allow anybody on server to get status +ACCEPT SERVICE=S SERVER # reject all others, including lpc commands permitted by user_lpc REJECT SERVICE=CSU # @@ -174,7 +174,9 @@ ACCEPT SERVICE=M SERVER REMOTEUSER=root REJECT SERVICE=M # all other operations denied -DEFAULT REJECT +# DEFAULT REJECT +# reject request not from local host +REJECT SERVICE=XRPQ NOT SERVER # all other operations allowed -# DEFAULT ACCEPT +DEFAULT ACCEPT // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: No connect permission for the lp printer command (It's workings again!)
Ok, good to hear it works again. I will submit a bug report against lprng. You might want to keep an eye on your printer services since I think DEFAULT REJECT was there for a good reason and substituting it with DEFAULT ACCEPT sounds like opening a big security hole... Mario Bertrand wrote: > On 30-Jan-99 Heikki Vatiainen wrote: > :-> However, editing /etc/lpd.perms and commenting out > :-> DEFAULT REJECT > :-> from the bottom of the file and substituting it with > :-> DEAULT ACCEPT > :-> gave back all the necessary permissions and now printing works > :-> again. If you do the same thing does it work for you? > > Ya, it works. > > Thanks No problem -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: No connect permission for the lp printer command
Mario Bertrand <[EMAIL PROTECTED]> > I just upgraded a few packages from slink and there was a update of the lpr > package that seems to have broken some permissions on my system. When i try to > print with any printing command i receive that error message: > > a2ps whatever.txt > [whatever.txt (plain): 1 page on 1 sheet] > no connect permissions > job 'cfA791logos' transfer to [EMAIL PROTECTED] failed > [Total: 1 page on 1 sheet] sent to the default printer > > (logos is the name of the machine) > > Any idea? Well I had a "Me too" situation here about 10 minutes ago. I'm running potato and just noticed that printing does not work anymore. It was my first printing try after upgrading from slink to potato so I was a bit puzzled about it. However, editing /etc/lpd.perms and commenting out DEFAULT REJECT from the bottom of the file and substituting it with DEAULT ACCEPT gave back all the necessary permissions and now printing works again. If you do the same thing does it work for you? // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Debian GNU/Linux a TCP Workhorse?
Art Lemasters wrote: > I would like to find references to examples of Debian GNU/Linux > systems handling heavy Internet or other network traffic. Can any of > you give me any URLs or anecdotes? How many simultaneous accesses > have your servers handled? I would like to present the info to my > local users' group. Thanks! > > Art The others already wrote about Internet servers, I hope my experiments qualify as "other network traffic". We are using Debian systems for IP-over-ATM development here at TUT. Yesterday I decided to see how much traffic our boxes would push through, so I set up netperf and let if blast for 12 hours. The experiment was between three machines, one Sun Ultra1 and two Debian intel boxes, a PPro 200 and a PII 350. Ultra1 and PPro were generating traffic towards PII and PII was generating traffic towards PPro. Here's a picture with arrow heads showing which way traffic was flowing. - -- | PPro (urku) | <===> | PII (jaarli) | <--- | Ultra1 (basso) | - -- Here are the results from the 12 hour netperf run. The test had been running about an hour when a stream from jaarli to urku was added. That's why it lasted only 11 hours. basso% src/netperf-2.1pl3/netperf -l 43200 -H jaarli -- \ -s 131070 -S 131070 -m 1450 TCP STREAM TEST to jaarli Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^6bits/sec 131070 131070 145043202.89 20.20 urku:~# netperf -l 43200 -H jaarli TCP STREAM TEST to jaarli Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^6bits/sec 65535 65535 6553543208.72 31.72 jaarli:~# netperf -l39600 -H urku TCP STREAM TEST to urku Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^6bits/sec 65535 65535 6553539599.84 88.87 Ifconfig from jaarli shows this: lec0 Link encap:Ethernet HWaddr 00:20:48:0E:67:B7 inet addr:10.10.54.208 Bcast:10.10.54.255 Mask:255.255.255.192 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:486241122 errors:0 dropped:0 overruns:0 frame:0 TX packets:492656611 errors:0 dropped:103 overruns:0 carrier:0 Collisions:0 cat /proc/interrupts shows this: jaarli:~# cat /proc/interrupts CPU0 0:8280457 XT-PIC timer 1: 22535 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 17410 XT-PIC soundblaster 9: 299899 XT-PIC eth0 10: 912585790 XT-PIC nicstar 13: 1 XT-PIC fpu 14: 73753 XT-PIC ide0 15: 19369 XT-PIC ide1 NMI: 0 The uptime of the box is a little over day and currently the interrupt count from nicstar (Fore LE155 ATM card actually) is well over one billion. If I calculated correctly, the traffic handled by jaarli was over 670Gbytes during the test (add everything together, divide by 8 and then divide by 1024^3 to gigabytes). The test is a little silly, since it only measures raw TCP throughput and nothing else, but maybe it usable for presentations :) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Why is LILO failing?
Björn Elwhagen wrote: > I've just installed a new version of lilo and i have a problem getting > it to work. It says there is an error in my configuration. Here's the > info: > > - > [ROOT: /]# cat /etc/lilo.conf > boot = /dev/hda > root = /dev/hda2 > vga = normal > block ^ Block? I did not find any references to block keyword in lilo.conf manual page. If you take it out, does it work then? > image = /vmlinuz > label = linux > read-only > image = /vmlinuz.old > label = old > read-only // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: what happened to xvidtune?
Ian Eure wrote: > I'm just wondering where xvidtune has got to- I don't see it on any of my > slink systems (all current) and I have no hamm boxes left. I seem to recall > that it was available in bo when I first started using Debian, and I see my > old kvidtune package on one of my systems, but there is no xvidtune to be > found. Can someone help me out here? Looks like xvidtune in slink is part of xf86setup package. Had to look for it myself a while too. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: biff and comsat don't work
I suspect the reason for this is that you have a daemon which listens to netbios-ns port already running when you (re)start your (x)inetd. The culprit in question is probably nmbd which is part of samba distribution. Does the problem go away if you do these things? 1. first shut down your (x)inetd 2. execute /etc/init.d/samba stop 3. start (x)inetd Doing things in this order should kill nmbd and make it release the netbios-ns port. Max <[EMAIL PROTECTED]> wrote: > The only error in the logfile that I can find is: > > Nov 2 12:35:43 chinook xinetd[27596]: bind failed (Address already in use). > service = netbios-ns > > But netbios-ns is on a different port than comsat, so I don't see how > this should make a difference. > // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: SSH for Debian/NT
Hi, Here are the ones that I know. There might be some more too. http://bmrc.berkeley.edu/people/chaffee/ http://www.geocities.com/SiliconValley/Bay/1692/ssh-index.html http://www.zip.com.au/~roca/ttssh.html // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: xterm-debian
Hi, I solved the same problem I had with TERM=rxvt by copying the rxvt entry from Linux to /usr/share/lib/terminfo/r directory in Solaris. Another way to do this is to use TERMINFO variable and set it to e.g. $HOME/.terminfo and put your xterm-debian entry in that directory. However, I have not tried the latter method. The both methods above are second hand information. The original discussion was on debian-user last May by (at least) Will Love, Hamish Moffat and George Bonser under a subject "terminal settings from linux-> Solaris". What I have written above should summarize it, but you might want to check the list archives for more info. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland
Re: Can less read its input from the stderr ? How ?
Maybe running your program as % program 2>&1 |less works. This redirects stderr (file descriptor 2) to where stdout goes. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
Re: Looking for a PDF reader
Hello again, Correcting myself ... I just noticed that I was talking about packages in the unstable distribution. The same readers are available for the stable distribution but in different locations. For the stable (1.3.x) distribution, see text/xpdf-0.6 for xpdf, non-free/gs-aladdin-4.03 for Aladdin Ghostscript and contrib/acroread-3.0 for Adobe's acrobat reader. I hope this helps :) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Looking for a PDF reader
Hi, There are several PDF readers available. I have Aladdin Ghostscript 5.10, xpdf 0.7a and acroread (Adobe Acrobat Reader) 3.01 installed. They are all available as Debian packages in non-free/text section at the Debian mirror sites. If you try to view an encrypted PDF file with Ghostscript (using gv or ghostview as a front end) and get an error message, see http://www.ozemail.com.au/%7Egeoffk/pdfencrypt/. >From there you can download a file called pdf_sec.ps which makes reading encrypted PDF files possible. The link above has the instructions what you should do with the pdf_sec.ps file. Note: Acrobat Reader does not work with 24bps color. However, you can always use it as a command line tool with something like 'acroread -toPostScript file.pdf'. Try running it with 'acroread -help' if you'd like to learn more. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: a MANPATH problem
I have two possible solutions for the X problem. Lately I have been using only the ssh version: 1. Su and $XAUTHORITY environment variable == % echo $DISPLAY :0 % su - root # export DISPLAY=:0 # export XAUTHORITY=~hessu/.Xauthority # xterm & 2. Using ssh not su === % ssh -l root localhost # xterm & As you can see, ssh takes care of DISPLAY variable and setting the magic cookie for xauth based authorization. You can get ssh from the non-US debian mirrors. See README.non-US in any of the debian mirrors for more info about the non-US mirror site locations. I have also had problems with xman not seeing all the man pages. Here is a small wrapper script I made to get rid of the problem: #!/bin/sh MANPATH=`manpath` export MANPATH exec xman $* // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Random loss of key bindings in fvwm2
That can happen if you happen to press NumLock key and it gets set. // Heikki -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: patch for man command
The syntax you are using for setting your PAGER looks a little odd. Does it work if you try it like this? # PAGER=less # export PAGER // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: PS/2 mouse setup; /dev/mouse?
Since it is a PS/2 mouse, the Pointer section in XF86Config should probably look something like this: Section "Pointer" Protocol"PS/2" Device "/dev/psaux" EndSection // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: inconsistency between server and config file?
For some reason your X setup is different from mine. Below is a short
description about how my X works. I am using the unstable distribution but it
should be the same for the stable too.
X in Debian is only a small wrapper for the real X servers like XF86_SVGA.
/etc/X11/Xserver stores configuration information about who can run X
and which server to use. The X binary reads this information and then starts
the real X server.
There is also Xserver manual page (available with 'man Xserver') but it has
nothing to do with the /etc/X11/Xserver file. The manual page describes the X
server in general and it is probably only a coincidence that there is also a
configuration file with the same name.
I ran 'strace X' as root (needed to be root since X is a setuid binary) and
it's output shows how /etc/X11/Xserver is used. Here is a snippet from the
output:
[cut]
open("/etc/X11/Xserver", O_RDONLY) = 3
fstat(3, {st_mode=0, st_size=0, ...}) = 0
mmap(0, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x400ae000
read(3, "/usr/X11R6/bin/XF86_SVGA\nConsol"..., 4096) = 245
getuid()= 0
execve("/usr/X11R6/bin/XF86_SVGA", ["/usr/X11R6/bin/X"], [/* 16 vars */]) = 0
[cut]
As you can see, X opens /etc/X11/Xserver and reads it's contents. According
to the documentation it should then verify if I was console user and start up
the real X server. Since I was running X from console the test passed and it
went on to fire up the real X server, XF86_SVGA.
See also the additional documentation in /usr/doc/xbase.
Here is what ls shows on my system:
% ls -l /usr/X11R6/bin/X
-rwsr-xr-x 1 root root 4728 Nov 2 03:27 /usr/X11R6/bin/X
% ls -l /usr/X11R6/bin/XF86_SVGA
-rwxr-xr-x 1 root root 2703004 Nov 2 03:16
/usr/X11R6/bin/XF86_SVGA
%
I hope this answers your question :)
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: compiling gnome-0.11?
I have not compiled Gnome myself but your Makefile seems to lack -lXpm even though it's needed. Brian K Servis writes: [cut] > Now I am getting this error: > > /bin/sh ../libtool --mode=link gcc -g -O2 -o htmltest test.o > ../gtk-xmhtml/libgtkxmhtml.la -lgtk -lgdk -lglib -lm -L/usr/X11R6/lib > -lX11 -lXext -lSM -lICE -lz -ljpeg > gcc -g -O2 -o _libs/htmltest test.o -Wl,-rpath -Wl,/usr/local/lib > ../gtk-xmhtml/_libs/libgtkxmhtml.so -lgtk -lgdk -lglib -lm > -L/usr/X11R6/lib -lX11 -lXext -lSM -lICE -lz -ljpeg It looks like there are references to the Xpm library but you do not have -lXpm among the libraries to use. Try fixing the Makefile and adding -lXpm after -ljpeg. Also, if the Xpm library is in /usr/local/lib you need -L/usr/local/lib before -lXpm. You probably knew it but I sometimes manage to forget the -L flag :) > ../gtk-xmhtml/_libs/libgtkxmhtml.so: undefined reference to > `XpmCreateXpmImageFromBuffer' > ../gtk-xmhtml/_libs/libgtkxmhtml.so: undefined reference to > `XpmFreeXpmInfo' > ../gtk-xmhtml/_libs/libgtkxmhtml.so: undefined reference to > `XpmGetErrorString' > ../gtk-xmhtml/_libs/libgtkxmhtml.so: undefined reference to > `XpmCreateXpmImageFromData' > ../gtk-xmhtml/_libs/libgtkxmhtml.so: undefined reference to > `XpmFreeXpmImage' > make[2]: *** [htmltest] Error 1 > > > I have gettext, guile, gtk, slib, xlockmore, libXpm, libjpeg, libpng, > zlib all installed either with deb packages or under /usr/local as > required by the README for gnome. ldconfig -v shows libXpm in it's list > and I have run ldconfig after I installed each of the above packages. > What am I missing? I am really interested in checking out gnome. > Thanks for any help or pointers. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: exec of /usr/X11R6/bin/XF86_NONE failed
If you have your XF86Config set up ok, please check /etc/X11/Xserver. Read the comments in that file and replace XF86_NONE with your correct X server. For example I use XF86_SVGA since I have a Matrox card. The correct location of XF86Config is /etc/X11/XF86Config. Any other reference to it should be a symlink if I understand correctly. I hope this helps :) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XFree86 setup problems
The XF86Setup utility only comes with the xserver-vga16 package. The VGA16 server in the xserver-vga16 package should work with any video card, so that is why the X based configuration utility comes only with it. Since I want to use higher resolutions and more colors than plain VGA can offer, I use the SVGA server from the xserver-svga package with my Matrox card. If you want to try XF86Setup, install both xserver-vga16 and xserver-S3v. The S3v server will be your main X server and the VGA16 server will be used when configuring the main X server. Here is a small snippet from the XF86Setup manual page: DESCRIPTION XF86Setup is normally used to either perform the initial setup of the XFree86 X servers or to make adjustments to the existing configuration. When used to initially configure the XFree86 servers, it will start the VGA 16 server and then allow configuration settings to be entered. Once that is completed the appro priate server for the hardware will be started and the configuration settings will be saved. Gregory Guthrie <[EMAIL PROTECTED]> > Thanks, > > I installed the xserver-S3v server, since that is the chipset that my (S3 > Virge) graphics card has; does it also have these features? > > Greg > > At 09:48 PM 1/8/98 +0200, Heikki Vatiainen wrote: > >Hi, > > > >If you have not already installed xserver-vga16 package, please do so. With > >the xserver-vga16 package comes a utility named XF86Setup which will make > >configuring X much more easier. You probably do not want to use the VGA16 > >server so answer 'no' when the installation script asks you if you want to > >use VGA16 as your default server. > > > >The XF86Setup program also lets you to configure your mouse. > > -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: problem with man- it hangs on Debian-1.3
Are you using bash as your shell and have 'set -a' in /etc/profile or in any of your bash startup files? If so, try removing 'set -a' or running man under some other shell. Bash hanging with man when 'set -a' was in use was reported as bug #8390 during last summer. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XFree86 setup problems
Hi, If you have not already installed xserver-vga16 package, please do so. With the xserver-vga16 package comes a utility named XF86Setup which will make configuring X much more easier. You probably do not want to use the VGA16 server so answer 'no' when the installation script asks you if you want to use VGA16 as your default server. The XF86Setup program also lets you to configure your mouse. I hope this helps, // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Xwindow
Hi, How did you create the X server configuration file /etc/X11/XF86Config? The error message indicates there is something wrong with the configuration file and that is why the X server can not start. I have had good success with the XF86Setup utility which comes with the xserver-vga16 package. If you try using it to create the XF86Config file does the problem go away? For more information about XFree86 and the Tseng Chipsets see http://www.xfree86.org/3.3.1/tseng.html. Section '1. Supported chipsets' says that the SVGA server should work but there are some exceptions. David Bosque <[EMAIL PROTECTED]> wrote: > Hello, > > I'm installing the Linux Debian GNU 1.3.1 system in my PC and my graphic > card is the VGA MENTOR with one MB of memory. This is based on Tseng's > ET4000. > > Also, I'm installing the Xwindow with xsever-svga and xserver-vga16 of > the packeges in the CD, but I have an a fatal server error when I make > the "startx", this error said: > > Fatal server error: > No valid modes found > > _X11TransSocketUNIXConnect: Can't connect: errno=111 > xinit: Connection refused (errno 111) unable to connect to X > server > xinit: No such process (errno 3): Server error. > > I think, that's possible do I need another server? > > How and where can I get the ET4000 driver for this X window? // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Is it possible to start X Windows in multiple V/C .
Ralph Winslow <[EMAIL PROTECTED]> wrote: > Yep, I'm absolutely certain - I entered: > > startx -- :1 > > the only spaces were between x and -- and between -- and :1 > three tokens, and I just re-tried to be absolutely certain. Oops, > this time I ran as root and voila, it worked! So I tried it again as > myself, and once again it failed, go figure! That's, hmm, interesting :) About the only thing I can think of now is the /tmp directory permissions. Since it works as root and not as an ordinary user maybe the /tmp permissions are messed up. It's also strange that the error message mentions display 0 even though you tell it to use display 1. Maybe for some strange reason you have startx in two different locations and the root user sees different startx than the non-root user. Command 'which startx' should tell where startx lives. Also to correct myself; it's minus not slash :). // Heikki -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Is it possible to start X Windows in multiple V/C .
Are you really, really sure you did not forget the space after those two slashes. I use the command almost every day and it works just fine. However, if if type startx --:1 ^^^-- note, no space after the second slash I get the same error as you about X already being active for display 0. The two slashes mean that there are no more arguments to the startx script and all the following arguments should be passed to the X server. In this case the X server is started as X :1 and it starts on VC8. // Heikki -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Tip: how to speed up XEmacs start
Hi, If you are in a hurry do touch /tmp/.sockets/audio0 otherwise read on :) I was qurious about what makes XEmacs so slow at startup. With the help of the strace(1) utility I found the following interesting behaviour of XEmacs19. The XEmacs I am using is version 19.16-1 which comes with the unstable distribution. When XEmacs starts up it creates a UNIX domain socket and tries to connect it to /tmp/.sockets/audio0. If this special file does not exist (ENOENT is returned) it does nanosleep(2) five times and sleeps about 1 second each time. Only after nanosleeping it will continue loading. If a plain file named /tmp/.sockets/audio0 is created XEmacs gets ECONNREFUSED as return value and goes on without nanosleeping or retrying. This speeds up the startup time for about five seconds on my machine. /usr/bin/time reports the time going from about 8.7 to 3.7 seconds. I hope you find this tip useful, // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XF86_NONE failed
If you know that the SVGA server was installed ok and the XFree86 configuration file is ok, see /etc/X11/Xserver file. The first line of the file tells which X server to run. For example my /etc/X11/Xserver starts with: /usr/X11R6/bin/XF86_SVGA If the XFree86 config file is not ok, try running /usr/sbin/xbase-configure as root to re-create the file. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: pdf to text converter
Hi again, I'm replying to your mail via debian-user list since there is something wrong with your nameserver right now and the mailer fails due to it. Here is the original message: Well, the only explanation I have is that we have different versions of xpdf and my verson has pdftotext and your version does not. I'm using the unstable distribution and you perhaps have stable installed. See http://www.aimnet.com/¯derekn/xpdf/ for the authors xpdf pages. There he has 'getting xpdf' page which has a link to a binary executable (xpdf-0.7-linux2.0-elf.tar.gz) for the 0.7 version. The binary version is linked against libc5 and seems to have pdftotext in it. Anyways, here is the info about the xpdf package I have installed. % dpkg --status xpdf Package: xpdf Status: install ok installed Priority: optional Section: non-free/text Installed-Size: 590 Maintainer: Dirk Eddelbuettel <[EMAIL PROTECTED]> Version: 0.7-3 Provides: pdf-viewer, postscript-preview Depends: libc6, xlib6g (>= 3.3-5), xpm4g (>= 3.4j-0), mime-support (>= 2.01-1) Description: Portable Document Format viewer for X11 xpdf is a viewer for Portable Document Format (PDF) files. (These are sometimes called 'Acrobat' files after the name of Adobe's PDF software.) xpdf is designed to be small and efficient. It does not use the Motif or Xt libraries. It uses standard X fonts. A program for conversion from pdf into ps, pdftops, is also included. % dpkg --listfiles xpdf /. /usr /usr/doc /usr/doc/xpdf /usr/doc/xpdf/changelog.Debian.gz /usr/doc/xpdf/copyright /usr/doc/xpdf/README.gz /usr/doc/xpdf/changelog.gz /usr/X11R6 /usr/X11R6/bin /usr/X11R6/bin/xpdf /usr/X11R6/man /usr/X11R6/man/man1 /usr/X11R6/man/man1/xpdf.1x.gz /usr/bin /usr/bin/pdftops /usr/bin/pdftotext /usr/man /usr/man/man1 /usr/man/man1/pdftops.1.gz /usr/man/man1/pdftotext.1.gz /usr/lib /usr/lib/menu /usr/lib/menu/xpdf I hope this helps, // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: pdf to text converter
I have pdftotext installed, it comes with the xpdf package. If you are using the unstable distribution see non-free/text section and for the stable distribution see the text section. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: SSH and X forwarding?
Have you tried running ssh with -v (verbose) option? Also, are you sure xauth is available at the remote end when ssh processes try to set up the $HOME/.Xauthority keys. Can't comment on va.debian.org though. Jason Gunthorpe <[EMAIL PROTECTED]> wrote > Has anyone managed to make ssh and sshd forward X connections? > > It seems to set the display environment variable but it never allocates a > socket in /tmp/.X11-unix. Very odd, this is on 'va.debian.org' btw. ^^ Which end is this, remote or local? I have a non-X host that has some X-utilities in it e.g. rxvt but no X server or console X usage. I have never seen /tmp/.X11-unix directories there. Hope this helps, // Heikki -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: X Server Client Refusal?
Here are my two solutions for the xauth problem. Both of these methods rely on xauth (user based authorization) as opposed to xhost (host based authorization). Let's say I'm running X as hessu and I need to su and run some X programs. hessu% echo $DISPLAY :0.0 hessu% su - # export DISPLAY=:0 # export XAUTHORITY=/home/hessu/.Xauthority # xterm & Using ssh works also and is simpler too: hessu% ssh -l root localhost # xterm & As shown above ssh sets DISPLAY and xauth cookies automagically. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: xman problems
Manuel Sickert wrote: > I posted this question to the german user list too, no answers. > I'll try it again here. > The problem is, that xman doesn't find the X11 and related > man pages. All the other manpages are displayed correctly. > The X11R6 man directory is in the MANPATh: > ~$ manpath > /usr/local/qt/man:/usr/man:/usr/local/man:/usr/X11R6/man This doesn't mean the man directories are in $MANPATH. The manpath manual page states that manpath command will print out a warning if the user has MANPATH defined. However, xman command looks for the MANPATH environment variable to find out where to look for manual pages. If no MANPATH is set it will default to /usr/man and the X etc. man pages are not shown. After you do something like export MANPATH=`manpath` xman will find all the manual pages the usual man command finds. A better way would probably be to make an alias or shell wrapper for xman I just tried this alias and it seems to work. It will set MANPATH for xman but leave it unset for man and other commands. alias xman='(export MANPATH=`manpath`; xman)' How I understand this mess is that man(1) is smart and does not need MANPATH (it will use it if the user sets it). xman(1x) is not so smart and needs MANPATH set or else only /usr/man is seen. See the relevant man pages man(1), manpath(1) and xman(1) for more details. I hope this helps. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: K Desktop Environment
If you still need msgfmt, install the xview-dev package. Apart from the usual libc5 and X packages xview-dev also needs the main xview package. The xview packages are in the X11 section. Ben wrote: > Hi. I have tried the latest KDE .deb packages there and have found that > the kfm file manager locks up and core-dumps on a regular basis. I have > had more luck compiling from scratch, but currently I can't get kpanel to > compile because I seem to be missing something called 'msgfmt'. > Apparently my friends who have openwindows have it, but I haven't been > able to locate it in Debian. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: xrsh?
That's hmm. interesting :) Have you tried running ssh with the verbose flag (-v) on? That way you can see which version of ssh and ssh protocol the local and remote end are using. Here's an excerpt from the ssh faq: 5.21 Ssh suddenly drops connections! This is a problem which has been reported by several people for SunOS 4, Solaris 2, Linux, and HP-UX 9 and 10, with 1.2.16 and 1.2.17. It happens with scp, when transferring large amounts of data via ssh's stdin, or when forwarding an X connection which receives a large amount of graphics data (such as a MPEG movie). [rest of the answer including the patch cut] The current released version is 1.2.20, the location of faq is http://www.uni-karlsruhe.de/~ig25/ssh-faq/ Gergely Madarasz wrote: > > Unfortunatelly I had trouble with ssh... the X connections are > forwarded through the ssh channel, encrypted... and sometimes > they get broken. > For example I was unable to run an xv through an ssh connection, > xv exited with "server kill" or something like that > immediatelly :( > > Greg > // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: xrsh?
How about using ssh? Since I switched to ssh I have not had any need to do those 'xauth extract - $DISPLAY | rsh otherhost xauth merge -' tricks anymore. Ssh crypts the connection and as an extra bonus automagically sets DISPLAY and xauth authorization records during the login process. Ssh is in the non-US part of the distribution, see the README.non-US file in the ftp mirrors. Sorry I couldn't help with xrsh. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Colors in the prompt like in slackware ?
See /usr/doc/fileutils/color-ls.gz which describes how to get colors in the directory listings. The quick answer is: ls --color or alias ls='ls --color=auto' in sh type shells. I hope this helps. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: .Xdefaults vs. .Xresources
My .Xresources is a symlink to .Xdefaults and I've had no problems either. I just checked /usr/lib/X11/xinit/xinitrc script and it thinks .Xresources is the right file. Here's a small snippet from the file. The $userresources variable is defined as 'usrresources=$HOME/.Xresources' in the beginning of the file. if grep -q ^allow-user-resources /etc/X11/config then if [ -f $usrresources ] then xrdb -merge $usrresources fi fi // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Xemacs wont run under SU, is it normal?
I'll add one more way how to su and get those X apps working. Lets say I'm running X as user 'hessu' and just did 'su -' so I could run xemacs as root. # export DISPLAY=:0 # export XAUTHORITY=~hessu/.Xauthority # xemacs & // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: GNUS ??
Gnus is a mailreader that comes with Emacs and XEmacs. I usually start my Gnus session with 'emacs -f gnus' or 'xemacs -f gnus'. See http://www.gnus.org/ for more info about Gnus. Basically if you get either one of the emacsen you get everything that's needed for Gnus too. Gnus has lot's of features and options but the news reading with it can be really simple too. I usually use space, q and the arrow keys and get along just fine. I hope this helps. PS. The current version of gnus is spelled as Gnus. An older version of gnus that came with older emacsen was called GNUS. The name change happened when a new maintainer started his work making GNUS better. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Modem Test?
Hi, For the starters you could try /dev/ttyS0 instead of /dev/tty0. If /dev/ttyS0 doesn't work try ttyS1-ttyS3. The /dev/ttyS0 through /dev/ttyS3 are the com devices in Linux. For testing you could try minicom (comm/minicom_1.75-2.deb in stable) and issue some AT commands (ATI4 might output some info about the modem) to see if the modem responds. I hope this helps. PATRICK DAHIROC wrote: > > Hi > > I've been trying to connect through my schools dial up system for sometime > now with no luck. I'm starting to think that my modem is not configured > correctly. I think that my modem is on /dev/tty0. I deleted the device > /dev/modem that present initially and made a new symbolic from /dev/tty0 > to /dev/modem. Is there any diagnostic test I can run to see if /dev/tty0 > is indeed the correct device driver? > > Thanks > Patrick // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Virtual Console is garbaged
You can try reset, tput reset and echo ^V^O. I have had most success with echo ^V^O I Hope this helps. // Heikki Shaleh wrote: > > I ran grep and it ended up reading a binary file. Now my console is > gibberish -- control chars, escape sequences, and non-printable chars. > I tried kill -HUP on getty, I tried kill on getty -- no dice. Any > ideas? -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: is there tools to....?
Try 'adduser --help' or 'addgroup --help'. Addgroup is a link to adduser. dada <[EMAIL PROTECTED]> wrote: > > is there some tool to add an group in Debian? (I'm locking for something > like "adduser") // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Can't delete a file
The universal answer comes from comp.unix.faq. Try rm ./-test This should work no matter which ls you are using. Mike Miller wrote: > > I have a file on my machine that I created that begins with a hyphen. > How can I delete it. rm -test says can't delete est using option -t (or > whatever). How do I tell it not to use the hyphen as a switch? // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: 1.3.1 Upgrading Probs
You should have your gateway in your network startup file (/etc/init.d/network ?). Now you have only GATEWAY="" and it does not look right. After running the modified network startup script the route command should output something like this: # route -n Kernel IP routing table Destination Gateway Genmask Flags Metric RefUse Iface 193.96.166.00.0.0.0 255.255.255.0 U 0 01 eth0 127.0.0.0 0.0.0.0 255.0.0.0 U 0 02 lo 0.0.0.0 193.96.166.1 0.0.0.0 UG1 01 eth0 GATEWAY in the network startup file Otherwise the ifconfig and route output look ok. Peter Weiss wrote: > [cut] > # route -n > Kernel IP routing table > Destination Gateway Genmask Flags Metric RefUse Iface > 193.96.166.00.0.0.0 255.255.255.0 U 0 01 eth0 > 127.0.0.0 0.0.0.0 255.0.0.0 U 0 02 lo > [cut] > > I'm starting the network services with a slightly modified version of > the standard network script: [cut] > # Configure the ethernet device or start SLIP/PPP below. > IPADDR="193.96.166.51" > NETMASK="255.255.255.0" > NETWORK="193.96.166.0" > BROADCAST="193.96.166.255" > GATEWAY="" [cut] // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: forms.h -- what package?
forms.h is in xforms-dev, section contrib in stable and contrib/x11 in
unstable.
You can search for a package owning a file by:
o grep filename Contents.{stable,unstable}, where Contents is a
file listing all the files and packages in the distribution.
You can get the Contents file from ftp.debian.org and mirrors
o Using my search tool at http://sunsite.tut.fi/finder/
o Using the search tool at http://www.debian.org/
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: One last thing before I pass out
Here's what I do after 'su -': % whoami hessu % echo $DISPLAY :0 % su - # export DISPLAY=:0 # export XAUTHORITY=/home/hessu/.Xauthority # xterm& The X server wants a magic cookie (password string) by the client program to make sure that the client is allowed to connect to it. The magic cookies reside in users ~/.Xauthority file which MUST be readable only by the user. Since root can read any file she/he wants, root can use anyones .Xauthority as shown above. Another solution is to copy the user's .Xauthority file to root's .Xauthority file, but this is less secure since the user and root now share the same magic cookies. Hope this helps. // Heikki Daniel Laffin wrote: > > After su'ing in an xterm and trying to run any x app i get this. I tried > "xhosts -", but it still gives me this. *scratches his head* Sleep > now--thanks. =) > > apollyon:/# xterm > Xlib: connection to ":0.0" refused by server > Xlib: Invalid MIT-MAGIC-COOKIE-1 key > Error: Can't open display: :0.0 > Exit 1 > apollyon:/# -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: mtools problem
Correction: I wrote: > > Try using mcopy and mdel like this: > > mcopy a:'*' . > mdel a:'*' . This should be: mdel a:'*' Sorry, it's Monday and it sure feels like it... // Heikki -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: mtools problem
Try using mcopy and mdel like this: mcopy a:'*' . mdel a:'*' . It looks like mtools need both the source and destination, a simple "mcopy a:" is not enough. The single quotes let * pass all the way to mtools programs which then do the file name expansion. Johnny Stevenson wrote: > > Hello, > > After upgrading to Debian 1.3.1, I am now having a problem using some > of the mtools operations. The command 'mdir' works but 'mcopy a:' and > 'mdel a:' produce the following error: > > This command cannot operate on "." or ".." > > Does anyone have any suggestions, appart from mounting/unmounting the > drive for each floppy (which is a bit of a pain as I still use alot of > floppies to move info around). // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Can anyone recomend a Debian package finder web page?
There are three package finders that I know: o The one by me, http://sunsite.tut.fi/finder/ o Larry and Joe's Debian Package Finder (has mirrors too) http://larry.earthlight.co.nz/debian/finder.cgi o http://www.debian.org/packages.html // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Where's zImage
Hi, Actually the package is called bin86 and it can be found from the devel section in both stable and unstable. Here's the description: Description: Assembler and loader for kernel compilation. This is the as86 and ld86 distribution written by Bruce Evans. It's a complete 8086 assembler and loader which can make 32-bit code for the 386+ processors (under linux it's used only to create the 16-bit bootsector and setup binaries). Paul Miller wrote: > > David Densmore wrote: > > > > make[1]: Entering directory > > `/usr/src/kernel-source-2.0.30/arch/i386/boot' > > as86 -0 -a -o bootsect.o bootsect.s > > make[1]: as86: Command not found > > make[1]: *** [bootsect.o] Error 127 > > make[1]: Leaving directory > > `/usr/src/kernel-source-2.0.30/arch/i386/boot' > > make: *** [zImage] Error 2 > > > > Do I really need to recompile the kernel to get my sound card to > > work? > > it may be helpful to install the package containing "as86" .. sorry I > can't help more.. it's probably a default package because mine works > fine. > > -Paul // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: How do I recompile the kernel?
Hi, I have attached a mail by Ben Gertzfield below. The original question was about kernel modules but the 'kernel-package' handles all the things related to building a custom kernel, not only the modules. This is the way I built my kernel and it worked very well. The 'kernel-package' can be found from the misc section. After you have made your new custom kernel you can install it with 'dpkg -i /usr/src/your_new_kernel.deb'. Start of quote: >>>>> "Will" == Will Lowe <[EMAIL PROTECTED]> writes: Will> This might be going to the wrong list, but I don't think I Will> subscribe to debian-devel. Maybe it actually goes to the Will> linux kernel people. Anyway: Will> Shouldn't 1) automatically purge the Will> /lib/modules/* tree before installing? 2) update Will> /etc/modules and conf.modules? If you use the 'kernel-package' Debian package, things will work exactly this way. You just run make menuconfig/config/xconfig/whatever, and then run: make-kpkg --revision whatever.1.0 kernel_image (you can replace 'whatever' with whatever ;) and voila! you'll have a new .deb package of your kernel and modules and other assorted goodies in /usr/src. :) End of quote: // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Crazy Ultrasound card
You probably only miss the right kind of connector on the Ultrasound or the right kind of cable, depends on how you think about it :). My SB 32 has two CD audio connectors. They both have left, right and two ground pins but the shape of the connector and the pin assignment is different. The CD-rom also has left, right and two ground pins. I think there are three things you can try. Ask for a working cable from the place you bought the Ultrasound from, ask your friends for any extra cables or try to fix the existing cable. Both my CD-rom and SoundBlaster manuals have the pin assignments printed on them. // Heikki Jason Killen wrote: > > I just bought a Ultrasound sound card, used. It's up and working fine > but the audio cable that came with my cdrom drive wont fit the pins on > the card. > > Did I miss something? -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Announce: Yet another Debian package finder
See http://sunsite.tut.fi/finder/ for my Debian GNU/Linux package finder. You can use the finder to do for example these things: o search for packages with a keyword o search for packages owning the file you're looking for o browse stable/unstable/non-US sections and see what's in them. The download site can be selected from the sites listed in README.mirrors. Downloading the non-US packages also works. Enjoy :-) // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Second Ethernet Card
Hi, Maybe I can help. Check http://cesdis.gsfc.nasa.gov/linux/drivers/3c509.html which talks about Linux and 3com 3c5*9 Ethercards. I just got a new version of the 3c509 driver for my card to get a fix for the 'Waiting for 3c509 to discard packet, status 2001.' nag some of you might have seen. The page has a link to the 3c509.c file which has this in the comments: FIXES: Alan Cox: Removed the 'Unexpected interrupt' bug. Michael Meskes: Upgraded to Donald Becker's version 1.07. Alan Cox: Increased the eeprom delay. Regardless of what the docs say some people definitely get problems with lower (but in card spec) delays v1.10 4/21/97 Fixed module code so that multiple cards may be detected, other cleanups. -djb The URL straight to the driver source is http://cesdis.gsfc.nasa.gov/linux/drivers/3c509.c // Heikki Drazen Lalovic wrote: > > THANK YOU FOR YOUR HELP. DO YOU KNOW THE PATCH NAME? > > THANKS, > DRAZEN > -- >From: Philippe Troin >To: Lalovic, Drazen >Subject: Re: Second Ethernet Card >Date: Thursday, June 26, 1997 3:45PM [cut] > > The 3c509 driver in 2.0.30 can only cope with one card. You need to > > patch it to get more than one. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with mtab?
This is what I have in my /etc/init.d/boot, version @(#)boot 2.10 26-Apr-1997 [EMAIL PROTECTED] [cut] # # Remount rootfs rw (default), but do not try to change mtab because it # is on a ro fs until the remount succeeded. Then clean up old mtabs and # finally write the new mtab. # mount -n -o remount,rw / rm -f /etc/mtab~ /etc/nologin : > /etc/mtab mount -o remount,rw / mount /proc [cut] So according to this, /etc/init.d/boot should take care of removing /etc/mtab~ during the boot. The /etc/init.d/boot file I have comes with sysvinit version 2.71-2 that was installed with Debian 1.3. // Heikki Sylvain Briole <[EMAIL PROTECTED]> wrote: > > Hello all, > > I use Debian 1.2/Linux 2.0.29. > I have three partitions on my hard disk : > > /dev/hda1 : linux > /dev/hda2 : linux swap > /dev/hda3 : ms-dos 6.22 > > And I have a problem with the mount command : > > nathalie# mount /hda3/ > can't lock lock file /etc/mtab~: timed out > > So, I delete /etc/mtab~ and it runs. > But, I have to delete this mtab~ at each boot. > I have three computer running Debian, and it is the only one with this > problem. > > Any idea? -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: FAT 32
I don't think there's a .deb for fat32. However, see http://bmrc.berkeley.edu/people/chaffee/fat32.html for info about Linux and fat32. A friend of mine patched his kernel and has had no complaints about it so far. If you decide to go for a custom kernel, check kernel-package in the misc section. The instructions how to build a kernel with it can be found from /usr/doc/kernel-package. When you have your fat32 aware kernel up and running, the filesystem type for mount is vfat. Try something like 'mount -tvfat /dev/fat32partition /mnt/fat32'. At least that's what my friend said :) compwiz wrote: > > I just upgraded my FAT 16 partition to FAT32, but now I can't mount it in > Linux, be the filesystem is unknown. Is there a package or module that > will let me mount my partionion? > Thanks. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Still trying to install debian through ftp.
Hello again, You probably have to say 'ifconfig eth0 down' before you try to remove the 3c509 driver module. I have the pnpdump utility in /sbin directory and you can get it from http://www.cs.tut.fi/~hessu/pnpdump if you trust my binaries :) NOTE: pnpdump resets the PnP cards. It hasn't been harmful for me but there's warning about it on the pnpdump's manual page. I was able to construct the same kind of arp table when I pulled the cable off my machine and then tried to ping a host via the eth0 device. That caused the HWaddress entry in the arp table to be filled zeroes. Since your setup looks otherwise ok and the arp HWaddress entry for your gateway is zero filled I belive the ethernet card isn't working right. If your 3c509 card has more than one interface (twisted pair, coaxial, AUI) you might want to check you are using the right one. You can check this with the 3c5x9 utility, binary and source available from http://www.cs.tut.fi/~hessu/3c5x9 and http://www.cs.tut.fi/~hessu/3c5x9setup.c respectively. Before you run 3c5x9 you need to do 'ifconfig eth0 down' first. 10baseT is the twisted pair (giant phoneplug) interface and 10base2 is the coaxial cable interface. If it can't find the card from the IO base address 0x300 you have to supply the correct IO base address on the command line like './3c5x9 -p 0x210'. Do you have any other cards installed besides the 3c509? // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Still trying to install debian through ftp.
If you still have the line starting 'lookup ...' in your /etc/resolv.conf try
removing it completely. I checked the man page and there was no mention about
lookup option.
My resolv.conf looks like this:
search ton.tut.fi cs.tut.fi atm.tut.fi cc.tut.fi
nameserver 193.166.80.16
nameserver 130.230.24.10
The search option simply means that when I type e.g. 'ping ftp' the resolver
[1] first tries to see if there's ftp.ton.tut.fi and if the host is not found
then tries ftp.cs.tut.fi and so forth. The search option is not needed and you
can leave it out if you don't mind typing the hostnames in their full length.
There's also no 'domain' option in my /etc/resolv.conf since it's not
necessary either. See 'man resolv.conf' for more info about the search and
domain configuration options.
You probably don't need reverse arp (RARP) since it's only needed by disksless
workstations and some other hardware which knows it's Ethernet (MAC) address
and wants to know the respective IP-address. That's one of the main uses for
reverse arp.
I also have a 3c509b at home and plug'n'play is causing small problems. To get
the 3c509b working I need to have the driver as a loadable module and prior
loading it I need to run 'pnpdump'. The reason for this is that I have a PnP
Sound Blaster 32 which somehow masks the 3c509 and runinning pnpdump before
'insmod 3c509' magically cures the problem.
'pnpdump' is an utility which resets all the PnP cards into the inactive
state. See the pnpdump man page for more details. Pnpdump is part of the
isapnptools package which is installed by default. Even the bootdisks have
pnpdump.
Before you try pnpdump do 'ifconfig eth0' first. This should show the
current status of your ethernet card. You might need to do 'rmmod 3c509'
before running pnpdump. My card looks like this:
eth0 Link encap:10Mbps Ethernet HWaddr 00:20:AF:C4:60:B2
inet addr:193.166.80.212 Bcast:193.166.80.255 Mask:255.255.255.0
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:5834274 errors:7 dropped:7 overruns:1
TX packets:259211 errors:0 dropped:0 overruns:0
Interrupt:10 Base address:0x300
Note that the IRQ and base I/O address are also shown in the output. You
probably don't have 'PROMISC' flag shown in your output but that's ok.
If this looks ok, check the output of 'route -n' command. Mine looks like
this:
Kernel IP routing table
Destination Gateway Genmask Flags Metric RefUse Iface
193.166.80.00.0.0.0 255.255.255.0 U 0 0 23 eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 01 lo
0.0.0.0 193.166.80.10.0.0.0 UG1 0 50 eth0
The last checkpoint this time is the arp table. Try running 'arp -a' or
'arp -an' if 'arp -a' hangs. The output of arp should show at least the
HWaddress of your default gw after you have tried to initiate some outbound
traffic. Once again, here's my arp table:
Address HWtype HWaddress Flags Mask Iface
193.166.80.1ether 08:00:02:04:6A:78 C * eth0
The network diagnostic application number one is ping. Try pinging other
hosts and see what happens. Also traceroute might output something useful.
If this doesn't help send me the output of 'ifconfig eth0', 'route -n',
'arp -an' and the results of running ping and traceroute against
ftp.debian.org (130.207.7.21).
1. Resolver: a chunk of code in the c-library responsible for the
hostname <-> ip-address conversion
PS. Just two more thoughts. When you reboot your machine hit the reset button
during the boot.
Also try changing 'route add default gw ${GATEWAY} metric 1' to 'route
add default gw ${GATEWAY} metric 1 dev eth0' in the /etc/init.d/network
script. This should work without 'dev eth0' but one never knows...
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: help with man pages
Isn't it 'set -a' not 'set -e'? Otherwise I think what Karl is saying should fix the bug. Karl M. Hegbloom <[EMAIL PROTECTED]> wrote > > >>>>> "frankk" == frankk <[EMAIL PROTECTED]> writes: > > frankk> I have installed Debian 1.3 and have a problem to see > frankk> man pages. If I try to see any man page (e.g man man), I > frankk> can only see the following : > > There is a bug in Bash-2.0 that causes this. Remove any calls to > `set -e' and `set +e'[1] in your "/etc/profile", and "~/.bash*" files, > and explicitly `export' each of the variables that would have been > exported implicitly by the `set -e'. After you log back in, `man' > should work again. > > > Footnotes: > [1] To find out what that did, do `help set' at a shell prompt. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: mirroring web pages
Besides wget you might also want to try WebCopy. The url is ftp://ftp.inf.utfsm.cl/pub/utfsm/perl/ and here's a small snippet from the documentation: WebCopy is a perl program that retrieves the URL specified in a unix-like command line. It can also retrieve recursively any file that a HTML file references, i.e. inlined images and/or anchors, if specified with an option. It can be used as a "mirror" program to retrieve a tree of documents from a remote site, and put them on-line immediately through the local server. The license agreement says this, I hope it's free enough: The author of this program is Victor Parada <[EMAIL PROTECTED]>. This program is "Freeware", not "Public Domain". This program must be distributed for free, and cannot be included in commercial packages without prior written permisson from the autor. This program cannot be distributed if modified in any way. This program can be used by anyone if the copyright and this notice remains intact in every file. If you modify this program, please e-mail patches to the the author. [cut, more to follow] // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: web documentation
I'll throw in one web server too. Today I found the thttpd server written by Jef Poskanzer. I haven't tried it, only unpacked it and looked around a bit. The url is http://www.acme.com/software/thttpd/ and here's the head of the README file from the source distribution: thttpd - tiny/turbo/throttling HTTP server version 1.90a of 15nov96 thttpd is a simple, small, portable, fast, and secure HTTP server. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: How to change in.fingerd -w banner?
It looks like the banner has been hard coded in the in.fingerd source file. In.fingerd is in the netstd package and if you view the source/net/netstd_2.13-1.diff.gz file and search for the string 'fingerd' in your pager you'll find where the banner is printed out. I guess the easiest way to have a different banner is to change the finger daemon like somebody already suggested on the list. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: ssh for Debian?
Have you checked the non-US section? From there you can find .debs for ssh and other crypto related packages. README.non-US in Debian mirrors lists all the sites that carry the section. George Bonser wrote: > > Where can I find information baout setting up a couple of debian boxes > with ssh? I want to do some sysadmin over the net on a remote system and > would like to make it a little difficult for someone to watch what I am > doing incase they want to swipe passwords, etc. > // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: man hangs with less and more
You probably have 'set -a' in /etc/profile or some of your bash startup files.
For some reason 'set -a' causes this problem. It's also reported as bug #8390.
The only cure I have is to remove 'set -a' from the startup files and
logout/login again.
Mark Boyns wrote:
>
> When I run:
>
>$ man ls
>
> it just sits there. ^C reveals:
>
> man: command exited with status 2: /bin/gzip -dc '/var/catman/cat1/ls.1.gz' |
> { export MAN_PN LESS; MAN_PN='ls(1)'; LESS="$LESS\$-Pm\:\$ix8mPm Manual page
> $MAN_PN ?ltline %lt?L/%L.:byte %bB?s/%s..?e (END):?pB %pB\\%.."; less; }
[cut]
// Heikki
--
Heikki Vatiainen * [EMAIL PROTECTED]
Tampere University of Technology * Tampere, Finland
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
