2.1.x version of ipmasq!
For those of you running the bleeding edge development kernels, I modified
the /etc/rc.boot/ipmasq script (from Debian's ipmasq package) so it can be
used with either ipfwadm or ipchains.
You'll need to add variables "EXTERNAL" and "INTERNAL" to your
configuration (/etc/ipmasq.conf). These variables should be set to the
corresponding interface, in the same manner as the IPs are configured.
Here is an example:
EXTERNAL=eth1 <- added
EXTERNAL_IP=198.109.162.43
EXTERNAL_NETMASK=255.255.255.224
INTERNAL=( eth0 eth2 )<- added
INTERNAL_IP=( 192.168.100.2 192.168.200.2 )
INTERNAL_NETMASK=( 255.255.255.0 255.255.255.0 )
I made up eth2 to show how multiple adaptor could be configured.
... it also doesn't require the /sbin/ipmasq program included in Debian's
package.
-Paul
BTW- I've sent a bug report with my updated version to the maintainer.
#!/bin/bash
#
# ipmasqSet up IP Masquerading for Debian systems
#
# v2.1 14-June-1998
# support for both ipfwadm and ipchains
IPFWADM=/sbin/ipfwadm
IPCHAINS=/sbin/ipchains
# Source configuration
. /etc/ipmasq.conf
# helper function
function xto {
if [ $1 -eq -1 ]; then
return;
else
xto $(($1 - 1));
echo $1
fi
}
# ipfwadm compatible kernels (2.0.x and 2.1.x(x <= 103))
function ipfwadm-rules {
echo -n "Initializing IP Masquerading..."
# flush all
# Incoming, flush and set default policy of deny.
$IPFWADM -I -f
$IPFWADM -I -p deny
# Outgoing, flush and set default policy of deny.
$IPFWADM -O -f
$IPFWADM -O -p deny
# Forwarding, flush and set default policy of deny.
$IPFWADM -F -f
$IPFWADM -F -p deny
# set rules
for i in `xto $(( ${#INTERNAL_IP[*]} - 1 ))`; do
$IPFWADM -I -a accept -V ${INTERNAL_IP[$i]} -S
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
$IPFWADM -I -a deny -V $EXTERNAL_IP -S
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]} -o
$IPFWADM -I -a accept -V $EXTERNAL_IP -D $EXTERNAL_IP/32
$IPFWADM -I -a accept -W lo
$IPFWADM -O -a accept -V ${INTERNAL_IP[$i]} -D
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
$IPFWADM -O -a deny -V $EXTERNAL_IP -D
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]} -o
$IPFWADM -O -a accept -V $EXTERNAL_IP -S $EXTERNAL_IP/$EXTERNAL_NETMASK
$IPFWADM -O -a accept -W lo
$IPFWADM -F -a masquerade -V $EXTERNAL_IP -S
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
done
}
# linux 2.1.x (x > 101) kernels
function ipchains-rules {
if [ "$EXTERNAL" == "" -o "$INTERNAL" == "" ]; then
echo '
Invalid configuration. Kernel 2.1.x (x > 101) no longer supports the use of
IP addresses in its rules. The interface name must now be used instead. To
fix this error message, please re-run ipmasqconfig.
You may also add the variables "EXTERNAL" and "INTERNAL" using the interface
name of variables "EXTERNAL_IP" and "INTERNAL_IP" in a similar manner.
'
exit 1
fi
echo -n "Initializing IP Masquerading..."
# flush all
# Incoming, flush and set default policy of deny"
$IPCHAINS -F input
$IPCHAINS -P input DENY
# Outgoing, flush and set default policy of deny"
$IPCHAINS -F output
$IPCHAINS -P output DENY
# Forwarding, flush and set default policy of deny"
$IPCHAINS -F forward
$IPCHAINS -P forward DENY
for i in `xto $(( ${#INTERNAL[*]} - 1 ))`; do
$IPCHAINS -A input -j ACCEPT -i ${INTERNAL[$i]} -s
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
$IPCHAINS -A input -j DENY -i $EXTERNAL -s
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]} -l
$IPCHAINS -A input -j ACCEPT -i $EXTERNAL -d $EXTERNAL_IP/32
$IPCHAINS -A input -j ACCEPT -i lo
$IPCHAINS -A output -j ACCEPT -i ${INTERNAL[$i]} -d
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
$IPCHAINS -A output -j DENY -i $EXTERNAL -d
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]} -l
$IPCHAINS -A output -j ACCEPT -i $EXTERNAL -s
$EXTERNAL_IP/$EXTERNAL_NETMASK
$IPCHAINS -A output -j ACCEPT -i lo
$IPCHAINS -A forward -j MASQ -i $EXTERNAL -s
${INTERNAL_IP[$i]}/${INTERNAL_NETMASK[$i]}
done
if [ -e /proc/sys/net/ipv4/ip_forward ]; then
echo "1" > /proc/sys/net/ipv4/ip_forward
fi
}
# linux 2.1.x (x > 101) kernels
if [ -e /proc/net/ip_fwchains ]; then
test -x $IPCHAINS || exit 1
ipchains-rules
else
test -x $IPFWADM || exit 1
ipfwadm-rules
fi
echo "done."
Re: Home Network TCP/IP with PPP ?
On Sun, 14 Jun 1998, Cormac McGuinness wrote: > Hi, > > Can anyone give me a concise description of how to set up a home-network > of a linux machine (with modem) and Windows 95 machine, so that the > Windows 95 machine can access the internet through the PPP connection > of the linux machine (start it and stop it if necessary). > > I have the two machines networked and have assigned 192.168.2.1 for > the linux machine and 192.268.2.2 for the windows machine... ^^^ Did you mean 192.168.2.2? > beyond this I am a little at sea as to what to do ? > Can someone help me ? or point me to what document I should be > reading > I have basically the same thing (plus an additional Linux box). Since you are using 192.168.2.x addresses, you will want to recompile your kernel with IP-masquerading so the Win95 box can access the internet. On the Windows 95 box, under Control Panel, Networks, enable Client for Microsoft Networks and TCP/IP protocol. Since you already have the two machines networked, I won't describe that. On the Win95 box, set the gateway address to 192.168.2.1. You may also want to install the Debian diald package, so when the Windows 95 computer tries to access the internet, the Linux computer will automatically dial up the connection. You will find much of the information you need in the IP-Masquerade mini-HOWTO and the NET-3-HOWTO (both in the doc-linux-text package). Bob Bob Nielsen Internet: [EMAIL PROTECTED] Tucson, AZ AMPRnet: [EMAIL PROTECTED] DM42nh http://www.primenet.com/~nielsen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
GECM.COM: Delivery Notification: Delivery has failed
Either debian.org or someone on this list has or relays to a system called 'gecm.com'. I've been receiving lots of messages saying that a delivery has failed, which is not true. These messages are usually over a day later than when I sent them, have been successfully received by the destination, and sometimes they've even replied! .. so if this is your server, please fix it! -Paul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt weirdness
On Sun, 14 Jun 1998, Igor Grobman wrote:
> > Timeout at /usr/lib/perl5/Net/FTP.pm line 334
> > Carp::croak('Timeout') called at /usr/lib/perl5/Net/FTP.pm line
> 1060
I'm sure Manoj can say more, but it sounds to me like this is a perl
FTP library glitch - doesn't handle timeouts well.
Jason
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On Sun, 14 Jun 1998 16:46:45 -0500 (CDT), [EMAIL PROTECTED] wrote: > >/lib/libc.so -> /lib/libc.so.5 >^^^ >MAY-DAY! MAY-DAY!!! > >/lib/libc.so.5 -> /lib/libc.so.5.4.36 >/lib/libc.so.5.4.36 >/lib/libm.so -> /lib/libm.so.5 >/lib/libm.so.5 -> /lib/libm.so.5.0.9 >/lib/libm.so.5.0.9 > > >I have now removed /lib/libc.so link and 'gcc -o hello hello.c' >worked > >Other compiles were still complaining about > >ld: warning: libc.so.5, needed by /lib/libncurses.so, may conflict with libc.s >o.6 > >I looked at /lib/libncurses.so links: > >/lib/libncurses.so -> /lib/libncurses.so.3 >/lib/libncurses.so.3.0 >/lib/libncurses.so.3.4 > >I removed /lib/libncurses.so and /lib/libncurses.so.3 and all seems to >be well now. You probably want to remove _all_ /lib/lib*.so links. You definitely want to remove the ones that point to libraries compiled with libc5. ldd will tell you which those are: $ ldd /lib/libm.so.5 libc.so.5 => /lib/libc.so.5 General principle: you can not mix libc5 and libc6 code. That goes for libraries as well -- any nontrivial shared library (such as libncurses) will use C library functions. >> The key thing here is that the libfoo.so links need to point to the >> libc6 libraries. Check that there is no /lib/libc.so link. In fact, >> there should probably be no xyz.so files in /lib at all, except >> /lib/ld.so. (xyz-2.0.7.so files are ok.) Check that /usr/lib/libm.so >> points at /usr/lib/libm.so.6. > >I have: > >/lib/libm.so -> /lib/libm.so.5 *** Is this trouble? Yes it is. Programs that use libm will get the same sort of conflict. >/lib/libreadline.so -> /lib/libreadline.so.2 >/lib/librl.so -> /lib/librl.so.2 These are also trouble; they're the libc5 versions of libreadline and librl. Delete the links. >With this all fixed I just did a kernel compile and it went without a >hitch.(I did not change the /lib/libm.so link above). > >This all started when I upgraded from bo to hamm. >Something must not have cleaned up its links when being upgraded. > >This machine has been upgraded in place since Debian 0.93 days but >this is the first problem I have had like this. What seems to have happened is that the hamm libraries put the .so links into /usr/lib, but the libraries from some pre-FSSTND debian put them in /lib. ld searches /lib before /usr/lib, so it uses the libc5 libraries. bo puts .so links in /usr/lib, so it was earlier than that. You never noticed before because you never had a C library major version change. (And hopefully it'll never happen again :) One thing that bothers me is that your kernel compile was affected. The kernel shouldn't use anything from libc. If the kernel boots it's fine, but there might be something strange going on there. Also I wonder what libm is doing in /lib. That's probably legitimate, but I'd think nothing needs it before /usr is mounted. zw -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Home Network TCP/IP with PPP ?
Hi, Can anyone give me a concise description of how to set up a home-network of a linux machine (with modem) and Windows 95 machine, so that the Windows 95 machine can access the internet through the PPP connection of the linux machine (start it and stop it if necessary). I have the two machines networked and have assigned 192.168.2.1 for the linux machine and 192.268.2.2 for the windows machine... beyond this I am a little at sea as to what to do ? Can someone help me ? or point me to what document I should be reading Previous to purchasing 2nd computer and network cards i had ppp setup on the linux machine with a minimal amount of configuration Thanks in advance a debian user -- --- Cormac McGuinness e-mail: [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] === Experimental Physics Department,Tel:+ 353 1 706 2205 University College Dublin, Fax:+ 353 1 283 7275 Belfield, Dublin 4, Ireland. Also: School of Physical Sciences,Tel:+ 353 1 704 5840 Dublin City University, Fax:+ 353 1 704 5384 Glasnevin, Dublin 9, Ireland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
libBrokenLocale - what is it?
On Fri, 12 Jun 1998 11:11:40 -0500, Ed Cogburn wrote: (snip) >The install of the Slackware package has overwritten the file permissions of >apparently several files and directories. and Steve Lamb wrote: > Stupid question, is your passwd file there? That is about the only time > I ever see numbers in ls is when the passwd file is fuqed. Thanks to you both for the helpful suggestions (yes my password file is still there and not - as far as I can tell - fuqued, and yes, Slackware had altered the directory permissions on /etc and /var). However: In my /lib directory I have: libBrokenLocale.so.1 -> libBrokenLocale-2.0.6 2776 libBrokenLocale-2.0.6.so I have tried re-installing libdb2.3.16.deb and doing ldconfig, without result, and am unsure how to repair the above error - presumably a linking error. I also ried renaming the above files and then re-installing libdb2, but dpkg then complains that it cannot find libBrokenLocale, which the fuquer apparently needs. Any suggestions? Many thanks IA. Adrian Monk Bedford -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
XFree86 problems
I have been having problems with X, the last time I installed Linux, I installed the XFree86 packages that as far as I can see are required to run X (the Standard Fonts, FVWM, the svga server, and what ever eles dselect told me I needed), but after runinng XFree86Config, when it goes to graphic mode, my screen goes crazy, it has many moving, blue lines, and it even messes up the other terminals. Should I fire up dselect and try to get a new (if ther is one) copy of XFree86, and if so what should I get? Or use what I have and find out it was some stupid little thing I for got/did not see? HW: Oak Tech SVGA card AST SVGA monitor SW Debian 1.3.1 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
WE HAVE A WINNER!!! SEE BELOW!!! On 14 Jun, Zack Weinberg wrote: > > /usr/lib/libm.so -> /usr/lib/libm.so.6 > /usr/lib/libm.so.6 -> /usr/lib/libm-2.0.7.so > /usr/lib/libm-2.0.7.so > /usr/lib/libm.a /usr/lib/libm.so -> /lib/libm.so.6 /lib/libm.so.6 -> /lib/libm-2.0.7.so /lib/libm-2.0.7.so /usr/lib/libm.a > > The C library is a little weird, because the real thing lives in /lib > and libc.so is a file. That should be set up like this: > > /usr/lib/libc.so > /lib/libc.so.6 -> /lib/libc-2.0.7.so > /lib/libc-2.0.7.so > /usr/lib/libc.a > /usr/lib/libc.so /lib/libc.so.6 -> /lib/libc-2.0.7.so /lib/libc-2.0.7.so /usr/lib/libc.a > Alongside this will be the files for libc5, which will be like this: > > /lib/libc.so.5 -> /usr/lib/libc.so.5.4.44 > /usr/lib/libm.so.5 -> /usr/lib/libm.so.5.4.44 > /usr/lib/libc.so.5.4.44 > /usr/lib/libm.so.5.4.44 > /usr/lib/libc5-compat/libc.so -> /lib/libc.so.5 > /usr/lib/libc5-compat/libm.so -> /usr/lib/libm.so.5 > /usr/lib/libc5-compat/libc.a > /usr/lib/libc5-compat/libm.a > /lib/libc.so -> /lib/libc.so.5 ^^^ MAY-DAY! MAY-DAY!!! /lib/libc.so.5 -> /lib/libc.so.5.4.36 /lib/libc.so.5.4.36 /lib/libm.so -> /lib/libm.so.5 /lib/libm.so.5 -> /lib/libm.so.5.0.9 /lib/libm.so.5.0.9 I have now removed /lib/libc.so link and 'gcc -o hello hello.c' worked Other compiles were still complaining about ld: warning: libc.so.5, needed by /lib/libncurses.so, may conflict with libc.so.6 I looked at /lib/libncurses.so links: /lib/libncurses.so -> /lib/libncurses.so.3 /lib/libncurses.so.3.0 /lib/libncurses.so.3.4 I removed /lib/libncurses.so and /lib/libncurses.so.3 and all seems to be well now. > > The key thing here is that the libfoo.so links need to point to the > libc6 libraries. Check that there is no /lib/libc.so link. In fact, > there should probably be no xyz.so files in /lib at all, except > /lib/ld.so. (xyz-2.0.7.so files are ok.) Check that /usr/lib/libm.so > points at /usr/lib/libm.so.6. I have: /lib/libm.so -> /lib/libm.so.5 *** Is this trouble? /lib/libreadline.so -> /lib/libreadline.so.2 /lib/librl.so -> /lib/librl.so.2 With this all fixed I just did a kernel compile and it went without a hitch.(I did not change the /lib/libm.so link above). This all started when I upgraded from bo to hamm. Something must not have cleaned up its links when being upgraded. This machine has been upgraded in place since Debian 0.93 days but this is the first problem I have had like this. THANKS!!! Brian -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt weirdness
Some time around Fri, 12 Jun 1998 17:56:04 EDT,
[EMAIL PROTECTED] wrote:
> I am doing 'apt-get update' and it dies saying:
>
> Get ftp://ftp.debian.org frozen/contrib Packages
> 31% [Packages 2920/21.4k 13%]Uncaught exception from user code:
>
> Timeout at /usr/lib/perl5/Net/FTP.pm line 334
> Carp::croak('Timeout') called at /usr/lib/perl5/Net/FTP.pm line 1060
> Net::FTP::I::read('Net::FTP::I=GLOB(0x8365cb0)', 'M-&M-m^JL^[M-V ^_;
> M-pN
> kM-%M-/JUM-]6M-H?M-^TM-YM-^Q1M-DM-{mM-q^HM-\&[EMAIL PROTECTED]@lM-A
> M-1^
> Zd^M1M-mM-)7M-iVM-^E?^M_MM-^OZg...', 1024) called at /usr/lib/perl5/Net/FTP.
> pm l
> ine 334
> Net::FTP::get('Net::FTP=GLOB(0x830d498)', '/debian/dists/frozen/cont
> rib/
> binary-i386/Packages.gz', 'ftp.debian.org_debian_dists_frozen_contrib_binary
> -i38
> 6_Packages') called at /usr/lib/apt/methods/ftp line 476
> main::handle_requests('Configuration', 'HASH(0x81c50d4)') called at
> /usr
> /lib/apt/methods/ftp line 545
> main::main('ftp://ftp.debian.org/debian/dists/frozen/contrib/binary-
> i386
> /Pac...', 'ftp.debian.org_debian_dists_frozen_contrib_binary-i386_Packages',
> 'ft
> p://ftp.debian.org/debian/dists/frozen/main/binary-i386/Packag...', 'ftp.deb
> ian.
> org_debian_dists_frozen_main_binary-i386_Packages', 'ftp://ftp.debian.org/de
> bian
> /dists/frozen/non-free/binary-i386/Pa...', 'ftp.debian.org_debian_dists_froz
> en_n
> on-free_binary-i386_Packages') called at /usr/lib/apt/methods/ftp line 549
> Fetched 29.0k in 2m55s (165b/s)
> ERROR ftp://ftp.debian.org/debian/dists/frozen/contrib/binary-i386/Packages.
> gz
> Bad return code from subprocess
>
> My apt is version .16, Any ideas anyone?
Just got the same thing while updating. It looks like the ftp session timed
out, but I don't think it should complain this loud about it ;-). Hence, I am
cc'ing this to apt developers.
--
Proudly running Debian Linux! Linux vs. Windows is a no-Win situation
Igor Grobman [EMAIL PROTECTED] [EMAIL PROTECTED]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
perl broken install (bo)
Hi all, installing bo on a rex machine left me with this problem: perl won't install, saying that "subprocess post install script returned error 123" or something like this. What's error 123 and how can I fix it? without perl, other packages are left uninstallable. Thank you. -- |||| ||| Marco Frattola Microsoft is not the answer ||`..'|| |||... Piacenza, ItalyMicrosoft is the question ||| ||| |||''[EMAIL PROTECTED] "No" is the answer ||| ||| ||| www.enjoy.it/users/~mk/index.html Live Linux, live free! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: swaping - raid
On 14 Jun 1998, Deniz Dogan wrote: : Hi, : : > "Nathan" == Nathan E Norman <[EMAIL PROTECTED]> writes: : : [...] : : Nathan> Yes. Using RAID0 on swap partitions is more or less a : Nathan> waste of time, since the kernel supports round robin : Nathan> swapping. You need to add a "pri=" to the options : Nathan> field in /etc/fstab. : : [...] : : Thanks for your answer. And one more question: What is the difference : between round-robin swaping and raid (raid0)? According to the documentation, there really is no difference. Memory will be swapped to the partitions equally (if I understand it correctly). The documentation is a bit sparse :) I also recall reading some software RAID docs that said the kernel round robin method was just as efficient as RAID0 on swap, but that you need kernel 1.3.16 or better for that. This is all off the top of my head, so I could be mistaken. -- Nathan Norman MidcoNet - 410 South Phillips Avenue - Sioux Falls, SD 57104 mailto://[EMAIL PROTECTED] http://www.midco.net finger [EMAIL PROTECTED] for PGP Key: (0xA33B86E9) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: swaping - raid
Hi, > "Nathan" == Nathan E Norman <[EMAIL PROTECTED]> writes: [...] Nathan> Yes. Using RAID0 on swap partitions is more or less a Nathan> waste of time, since the kernel supports round robin Nathan> swapping. You need to add a "pri=" to the options Nathan> field in /etc/fstab. [...] Thanks for your answer. And one more question: What is the difference between round-robin swaping and raid (raid0)? -- Deniz Dogan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Linux install problem
On Sun, 14 Jun 1998, Shanta McBain wrote: > I have a HD 350 meg Caviar 2340. I cannot change the partitions. Fdisk > says that there are two partition The dos primary and a extended > secondary. It reports that there are logical partitions on the > secondary. Cant delete the secondary partition as the logical partitions > exist. likely the linux primary and swap. It will not delete them as > delete partitions dose not see the partitions. > Is there a linux tool that will fix the problem? I think you should try cfdisk first. It's far easier to use, and it displays you the logical partitions within your extended. Another way is to download a low level formatter like DiskManager or EzDrive. As stated by the ATA specification, a hd has to do such a low level format without any problems. Bye Daniel -- Daniel Gross eMail: [EMAIL PROTECTED] Hollarstrasse 2 [EMAIL PROTECTED] D-85053 Ingolstadt, Germany -- An anthropologist at Tulane has just come back from a field trip to New Guinea with reports of a tribe so primitive that they have Tide but not new Tide with lemon-fresh Borax. -- David Letterman -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: swaping - raid
On 14 Jun 1998, Deniz Dogan wrote: : Hi, : : To use kernel`s raid0 support for swap partitions, I added : these two lines to /etc/fstab : : /dev/hdc2 noneswapsw 0 0 : /dev/hda1 noneswapsw 0 0 : : But boot messages says these partitions have different priorities : (-1 and -2). : : Any ideas? Thanks in advance. Yes. Using RAID0 on swap partitions is more or less a waste of time, since the kernel supports round robin swapping. You need to add a "pri=" to the options field in /etc/fstab. On one of my systems the swap partitions are described thusly: /dev/sdb7 noneswapsw,pri=1 /dev/sda7 noneswapsw,pri=1 As long as the priorities are the same, round robin swapping will occur. User assigned priorities have a "non-negative" value, according to `man 2 swapon'. It looks like the priority field is a signed integer, so you have over 32000 values to play with for user assigned swap areas :) Cheers, -- Nathan Norman MidcoNet - 410 South Phillips Avenue - Sioux Falls, SD 57104 mailto://[EMAIL PROTECTED] http://www.midco.net finger [EMAIL PROTECTED] for PGP Key: (0xA33B86E9) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On Sun, 14 Jun 1998 11:39:30 -0500 (CDT), [EMAIL PROTECTED] wrote:
>On 14 Jun, Zack Weinberg wrote:
>>
>> It looks to me as if the gcc drivers aren't doing anything wrong. But
>> this is a big fat clue:
>>
>>>ld: warning: libm.so.6, needed by /usr/lib/libstdc++.so, may conflict with l
>ib
>>>m.so.5
>>>ld: warning: libc.so.6, needed by /lib/libm.so.6, may conflict with libc.so.
>5
>>
>> The linker called by gcc/g++ 2.7 is attempting to link your program
>> with the crt1.o from libc6 but the dynamic library from libc5. That
>> implies you have strange environment variable settings or mangled
>> linker config files. egcc has a wrapper for ld which is smart enough
>> to straighten this out.
>>
>
>% cat /etc/ld.so.conf
ld.so.conf is unrelated.
>% ld --verbose
>GNU ld version 2.9.1 (with BFD 2.9.1)
> Supported emulations:
> elf_i386
> i386linux
>using internal linker script:
>==
>OUTPUT_FORMAT("elf32-i386", "elf32-i386",
> "elf32-i386")
>OUTPUT_ARCH(i386)
>ENTRY(_start)
>SEARCH_DIR(/lib); SEARCH_DIR(/usr/lib); SEARCH_DIR(/usr/local/lib); SEARCH_DIR
>(/
>usr/i486-linux/lib);
>
>Hmmm. I don't even have a /usr/i486-linux/lib directory!!
It's a hook for if you want to do cross compilation. Don't worry
about it.
I bet you have spurious library symlinks in /lib or /usr/lib. For a
normal library, like libm, the only files that should exist are
/usr/lib/libm.so -> /usr/lib/libm.so.6
/usr/lib/libm.so.6 -> /usr/lib/libm-2.0.7.so
/usr/lib/libm-2.0.7.so
/usr/lib/libm.a
The C library is a little weird, because the real thing lives in /lib
and libc.so is a file. That should be set up like this:
/usr/lib/libc.so
/lib/libc.so.6 -> /lib/libc-2.0.7.so
/lib/libc-2.0.7.so
/usr/lib/libc.a
Alongside this will be the files for libc5, which will be like this:
/lib/libc.so.5 -> /usr/lib/libc.so.5.4.44
/usr/lib/libm.so.5 -> /usr/lib/libm.so.5.4.44
/usr/lib/libc.so.5.4.44
/usr/lib/libm.so.5.4.44
/usr/lib/libc5-compat/libc.so -> /lib/libc.so.5
/usr/lib/libc5-compat/libm.so -> /usr/lib/libm.so.5
/usr/lib/libc5-compat/libc.a
/usr/lib/libc5-compat/libm.a
(The shared libraries -- libc.so.5.4.44 -- may also be in
libc5-compat; I don't have a debian 2.0 system here to check on.)
The key thing here is that the libfoo.so links need to point to the
libc6 libraries. Check that there is no /lib/libc.so link. In fact,
there should probably be no xyz.so files in /lib at all, except
/lib/ld.so. (xyz-2.0.7.so files are ok.) Check that /usr/lib/libm.so
points at /usr/lib/libm.so.6.
>% /lib/libc.so.6
>GNU C Library production release version 2.0.7, by Roland McGrath et al.
>Compiled by GNU CC version 2.7.2.3.
Hmf, I guess Uli backported that to the 2.0 release. Your
/usr/lib/libc.so is fine then.
zw
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On 14 Jun, Zack Weinberg wrote:
>
> It looks to me as if the gcc drivers aren't doing anything wrong. But
> this is a big fat clue:
>
>>ld: warning: libm.so.6, needed by /usr/lib/libstdc++.so, may conflict with lib
>>m.so.5
>>ld: warning: libc.so.6, needed by /lib/libm.so.6, may conflict with libc.so.5
>
> The linker called by gcc/g++ 2.7 is attempting to link your program
> with the crt1.o from libc6 but the dynamic library from libc5. That
> implies you have strange environment variable settings or mangled
> linker config files. egcc has a wrapper for ld which is smart enough
> to straighten this out.
>
% cat /etc/ld.so.conf
/usr/lib/tkstep
/usr/X11R6/lib/neXtaw
/usr/X11R6/lib/Xaw95
/usr/X11R6/lib/Xaw3d
/usr/X11R6/lib
/usr/lib/libc5-compat
/lib/libc5-compat
/usr/local/X11/lib
/usr/local/lib
> If you have environment variables named LD_somethingorother
> (LD_RUN_PATH, LD_LIBRARY_PATH, etc.) unset them and try again to
> compile with gcc 2.7. Also, do this:
No LD_* environment variables.
>
> ld --verbose | more
>
% ld --verbose
GNU ld version 2.9.1 (with BFD 2.9.1)
Supported emulations:
elf_i386
i386linux
using internal linker script:
==
OUTPUT_FORMAT("elf32-i386", "elf32-i386",
"elf32-i386")
OUTPUT_ARCH(i386)
ENTRY(_start)
SEARCH_DIR(/lib); SEARCH_DIR(/usr/lib); SEARCH_DIR(/usr/local/lib); SEARCH_DIR(/
usr/i486-linux/lib);
Hmmm. I don't even have a /usr/i486-linux/lib directory!!
/usr/local/lib doesn't have anything concerning libc in it.
>>% cat /usr/lib/libc.so
>>/* GNU ld script
>> Use the shared library, but some functions are only in
>> the static library, so try that secondarily. */
>>GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a )
>
> *blink* I thought that was only in development libc. What do you get
> if you execute /lib/libc.so.6 as a program?
>
% /lib/libc.so.6
GNU C Library production release version 2.0.7, by Roland McGrath et al.
Compiled by GNU CC version 2.7.2.3.
Copyright (C) 1992, 93, 94, 95, 96, 97, 98 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.
Compiled on a Linux 2.0.32 system on 1998/04/15.
Available extensions:
GNU libio by Per Bothner
BIND-4.9.6-T1A
UFC-crypt, patchlevel 1e by Michael Glad
linuxthreads-0.6 by Xavier Leroy
Report bugs using the `glibcbug' script to <[EMAIL PROTECTED]>.
Thanks,
Brian
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Linux install problem
Hi I have a HD 350 meg Caviar 2340. I cannot change the partitions. Fdisk says that there are two partition The dos primary and a extended secondary. It reports that there are logical partitions on the secondary. Cant delete the secondary partition as the logical partitions exist. likely the linux primary and swap. It will not delete them as delete partitions dose not see the partitions. Is there a linux tool that will fix the problem? Thanks Shanta -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Holy Water in my system? Daemons keep dying . . .
Something keeps splatting my poor daemons. lpd tends to die within a day or so, and pland dies within minutes, announcing (incorrectly) an application logout. rick -- These opinions will not be those of ISU until it pays my retainer. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: VIM
On: Sat, 13 Jun 1998 10:28:54 -0700 (PDT) Nick Moffitt writes: > > On 12 Jun 1998, Torsten Hilbrich wrote: >> It is a very useful feature called backing up files ;-) >> >> You can convince vim to not create these files by either editing >> /etc/vimrc (for all user) or ~/.vimrc (on a per-user basis) and add a >> "set nobackup" line or edit the existing "set backup" line. > > Is there any way to get vim to write numeric backups? The GNU > indent program pays attention to an environment variable, and will > make backup files like foo.~1~ and foo.~2~ if it is set right. According to the help page there is none. These are the only configuration options regarding backup I found there: |'backup'| |'bk'| keep backup file after overwriting a file |'backupdir'| |'bdir'| list of directories for the backup file |'backupext'| |'bex'| extension used for the backup file |'writebackup'||'wb'| make a backup before overwriting a file Torsten -- Whenever a system becomes completely defined, some damn fool discovers something which either abolishes the system or expands it beyond recognition. Fortune Cookie PGP Public key available -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
It looks to me as if the gcc drivers aren't doing anything wrong. But
this is a big fat clue:
>ld: warning: libm.so.6, needed by /usr/lib/libstdc++.so, may conflict with lib
>m.so.5
>ld: warning: libc.so.6, needed by /lib/libm.so.6, may conflict with libc.so.5
The linker called by gcc/g++ 2.7 is attempting to link your program
with the crt1.o from libc6 but the dynamic library from libc5. That
implies you have strange environment variable settings or mangled
linker config files. egcc has a wrapper for ld which is smart enough
to straighten this out.
If you have environment variables named LD_somethingorother
(LD_RUN_PATH, LD_LIBRARY_PATH, etc.) unset them and try again to
compile with gcc 2.7. Also, do this:
ld --verbose | more
This will produce much incomprehensible output; you care only about
the first few lines, which should read like this:
GNU ld version 2.8.2 (with BFD 2.8.1.0.25)
Supported emulations:
elf_i386
i386linux
using internal linker script:
==
OUTPUT_FORMAT("elf32-i386", "elf32-i386",
"elf32-i386")
OUTPUT_ARCH(i386)
ENTRY(_start)
SEARCH_DIR(/lib); SEARCH_DIR(/usr/lib);
SEARCH_DIR(/usr/local/lib); SEARCH_DIR(/usr/elf_i386/lib);
If it doesn't say "using internal linker script", or if any of the
SEARCH_DIR entries have the word `libc5' in them, you have a problem
with your linker installation.
>% cat /usr/lib/libc.so
>/* GNU ld script
> Use the shared library, but some functions are only in
> the static library, so try that secondarily. */
>GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a )
*blink* I thought that was only in development libc. What do you get
if you execute /lib/libc.so.6 as a program?
zw
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
fvwm95
I've just run dselect to try to download xview (I'm running hamm) and it offered to install ~15-20 packages. I said go ahead, and now my system, which had had a nifty W$95 style desktop, seems to be running a blank X system with one xterm. I can use it to spawn new xterms and run netscape (which I'm using to send this), but I miss being able to click an Icon to run netscape, and the clock, etc. ps shows that fvwm95 is running for me. What must I do to restore order. I'm a command-line kind of guy, I'm afraid I find the X environment a bit daunting. TIA for any tips you might have. -- - Ralph Winslow [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: can't locate module
Sounds like you are trying to load the UMSDOS filesystem as a kernel
module, but you didn't build and install the modules yet, or UMSDOS
wasn't selected as a fileystem type in the config stage.
Go through the kernel config ('make menuconfig' is a quick and easy text
menu), go down to "Filesystems", make sure there is an 'M' in front of
UMSDOS.
When that is done, exit and save your new configuraion. To insure
everything is rebuilt and set up properly, do the following:
make dep ; make clean ; make bzlilo ; \
make modules ; make modules_install
(the 'make modules_install' puts the modules under /lib/modules/
directory so insmod has a standard place to find them).
Sidenote: I am finding that 'make bzlilo' is becoming necessary
otherwise I get the 'System too big' error and it won't finish. You
only need to do this if 'make zlilo' quits with a "System too Big"
error.
Hope this helps,
troy
Marc van der Vossen wrote:
>
> Hi all.
>
> I've got a simple question for ya. I'v compiled a new kernel 2.0.34 When i
> boot with it all goes well, until it wants to load a module. At that point I
> get a message like:
> umsdos can't locate module umsdos
>
> How do I solve this ?
>
> Thanks in advance.
>
> Marc
> [EMAIL PROTECTED]
>
> --
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On 14 Jun, Zack Weinberg wrote:
> On Sat, 13 Jun 1998 22:51:30 -0500 (CDT), [EMAIL PROTECTED] wrote:
>>On 13 Jun, Zack Weinberg wrote:
>>> What do you get if you do this:
>>>
>>> gcc -v -o hello hello.c
>>
>>This is what I get:
>>
>>% gcc -v -o hello hello.c
>>gcc version 2.7.2.3
> [...]
>> ld -m elf_i386 -dynamic-linker /lib/ld-linux.so.2 -o hello
>> /usr/lib/crt1.o /usr/lib/crti.o
>> /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtbegin.o
>> -L/usr/lib/gcc-lib/i486-linux/2.7.2.3 /tmp/ccd01600 -lgcc -lc -lgcc
>> /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtend.o /usr/lib/crtn.o
>>/usr/lib/crt1.o(.text+0xe): undefined reference to `__libc_init_first'
>>/usr/lib/crt1.o(.text+0x18): undefined reference to `_environ'
>
> Hunh. It is pulling in the C library (-lc). Next thing is to see
> what egcc and/or g++ do differently; can you try the same thing with
> either of them? (I only need to see the link line, so you could do
>
> gcc -c hello.c
>
> egcc -v -o hello hello.o
> g++ -v -o hello hello.o
>
> to cut down on the noise a little.)
>
% egcc -v -o hello hello.o
Reading specs from /usr/lib/gcc-lib/i486-linux/egcs-2.90.29/specs
gcc version egcs-2.90.29 980515 (egcs-1.0.3 release)
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/ld -m elf_i386 -dynamic-linker
/lib/ld-linux.so.2 -o hello /usr/lib/crt1.o /usr/lib/crti.o
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/crtbegin.o
-L/usr/lib/gcc-lib/i486-linux/egcs-2.90.29 hello.o -lgcc -lc -lgcc
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/crtend.o /usr/lib/crtn.o
(note this the the egcs g++)
% g++ -v -o hello hello.o
Reading specs from /usr/lib/gcc-lib/i486-linux/egcs-2.90.29/specs
gcc version egcs-2.90.29 980515 (egcs-1.0.3 release)
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/ld -m elf_i386 -dynamic-linker
/lib/ld-linux.so.2 -o hello /usr/lib/crt1.o /usr/lib/crti.o
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/crtbegin.o
-L/usr/lib/gcc-lib/i486-linux/egcs-2.90.29 hello.o -lstdc++ -lm -lgcc -lc -lgcc
/usr/lib/gcc-lib/i486-linux/egcs-2.90.29/crtend.o /usr/lib/crtn.o
Note: Package conflicts wouldn't allow both g++272 and egcc to
co-exist on Debian. So I had to remove egcc, g++ in order to install
g++272.
(note this is the gcc g++)
% g++272 -v -o hello hello.o
gcc -v -o hello hello.o -lstdc++ -lm
Reading specs from /usr/lib/gcc-lib/i486-linux/2.7.2.3/specs
gcc version 2.7.2.3
ld -m elf_i386 -dynamic-linker /lib/ld-linux.so.2 -o hello /usr/lib/crt1.o
/usr/lib/crti.o /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtbegin.o
-L/usr/lib/gcc-lib/i486-linux/2.7.2.3 hello.o -lstdc++ -lm -lgcc -lc -lgcc
/usr/lib/gcc-lib/i486-linux/2.7.2.3/crtend.o /usr/lib/crtn.o
ld: warning: libm.so.6, needed by /usr/lib/libstdc++.so, may conflict with
libm.so.5
ld: warning: libc.so.6, needed by /lib/libm.so.6, may conflict with libc.so.5
> Also, another thing that might possibly produce this error is if
> /usr/lib/libc.so is mangled. Can you check that that's an ordinary
> file containing something like
>
> GROUP { libc.so.6 ld-linux.so.2 libc.a }
>
> please? It ought to affect both compilers if that's wrong, but it's
> worth checking.
% cat /usr/lib/libc.so
/* GNU ld script
Use the shared library, but some functions are only in
the static library, so try that secondarily. */
GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a )
Thanks,
Brian
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install/Uninstall dramas (novice user)
In article <[EMAIL PROTECTED]>, Hamish wrote: >I have been running Linux for some two years and Debian for more than one, >and certainly haven't required MC yet. Even standard would be inappropriate >I think -- standard is for things which are standard on a Unix system, >and should probably be as strict as possible to keep the size down. >Optional remains appropriate. Not that I am experienced enough to suggest a list, but perhaps Debian should have a "Recommended" standard setup for inexperienced users? -- Carl Fink [EMAIL PROTECTED] Q. Why do some people take astrology seriously? a. Because they have unusually small brains.--Dave Barry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Will the new ssh get into hamm?
On Sun, 14 Jun 1998, Liran Zvibel wrote: : A new version of ssh was released (today or yesterday). : Will it get into hamm? Due to cryptographic laws in the USA, that's not as easy at it looks like. Anyway, it is currently at nonus.debian.org in the Incoming directory. Hope that helps, -Remco -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Will the new ssh get into hamm?
Hi. A new version of ssh was released (today or yesterday). Will it get into hamm? (BTW: Since it is the exam season, I unsubscribes (I can't cope with the bandwidth wile studying to exams) so please post answers to me too.) Liran Zvibel. --- http://www.math.tau.ac.il/~liranz/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
netscape stuff
I have Navigator 4.05, the standalone version, and also Mozilla (from the .deb package). Navigator 4.05 I installed with the provided script, as there does not seem to be a .deb package for the Navigator alone, only Communicator. OK, what I want to know is: How do I get Netscape to open an xterm running pine with an appropriate address when I click on a mailto: item? There doesn't seem to be any appropriate helper application entry to edit. How come Java isn't working properly? No applet execpt for trivial ones seem to work. The status bar says they are running, but nothing happens. Any ideas? Andrew Tarr "We were so close to heaven --- Peter came out and gave us badges, proclaiming us `The Nicest of the Damned'" -- They Might Be Giants ___| (http://multinet.co.nz/personalhomepages/locusmeus/antechamber.html) ~~~| -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
swaping - raid
Hi, To use kernel`s raid0 support for swap partitions, I added these two lines to /etc/fstab /dev/hdc2 noneswapsw 0 0 /dev/hda1 noneswapsw 0 0 But boot messages says these partitions have different priorities (-1 and -2). Any ideas? Thanks in advance. >From mdutils docs: 2 Swapping over RAID The kernel has native support for distributing swap space over several disks. Just add all swap partitions to /etc/fstab and use 'swapon -a' to activate all of them. The kernel uses striping (RAID-0) for them. Here's a sample setup: /dev/sda3 noneswapsw /dev/sdb3 noneswapsw >From swap(2) man: Swap pages are allocated from areas in priority order, highest priority first. For areas with different priori- ties, a higher-priority area is exhausted before using a lower-priority area. If two or more areas have the same priority, and it is the highest priority available, pages are allocated on a round-robin basis between them. -- Deniz Dogan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HELP! Spammer using my system !
>> "mt" == matthew tebbens <[EMAIL PROTECTED]> writes: mt> I suspect a spammer is using my system to relay or queue spam ! mt> I'm using smail 3.2-3. The smail in hamm has anti relay measures. In the logfile you should see a Received FROM line for the [m0ykxmp-000NUHC] ID. If it is non-local, It is most linkely a spammer. mt> How do I stop this ASAP ? You could add the host the spammer uses to /etc/hosts.deny mt> How do I find the spammer ? Maybe you could set smail to queue only, so that messages will only be stored. Then you can make a copy of some of the messages, so you can inspect the headers. Ciao, Martin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
can't locate module
Hi all. I've got a simple question for ya. I'v compiled a new kernel 2.0.34 When i boot with it all goes well, until it wants to load a module. At that point I get a message like: umsdos can't locate module umsdos How do I solve this ? Thanks in advance. Marc [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ISP Connection + Misc.
>> "DD" == Dennis Dixon <[EMAIL PROTECTED]> writes: DD> I'm guessing that the first time I installed Debian I accidentally did DD> something to ppp.chatscript to make it wait for the ISP to request the DD> username, but this time the username needed to be in the ppp.options_out DD> script also since by the time the ISP requested chatscript was DD> already closed. DD> This is merely a wild guess. But since nowhere in the instructions does it DD> say to add your to ppp.options_out I thought it was worth mentioning. The two common ways to log into your isp are pap/chap authentication and terminal login. pon is preconfigured for terminal login. From the errormessages you got I believe your isp doesn't support terminal login but only pap. In this case one has to substitute the username und password lines in the chapscript with a connect line and add "user xyz" to the pppd options. It looks strange to me that it worked for you without removig the lines from the chatscript. I don't have the pppd package installed, so I can't tel you if this is described in the docs. Should be in /usr/doc/pppd/debian*. It is described in the faq-o-matic at http://www.debian.org/fom/20.html DD> 1) In the error message above, does "peer" refer to my ISP, not my Debian DD> machine? It appears that way to me, but I'm not sure. Me either, because for pppd, both sides are peers. You could add "debug" to the pppd options and look in /var/log/ppp.log for a line like TermReq. Then you will see, if it was sent or received. DD> 2) "pon" (and also 'exec pppd') only works when logged on as root. Is it DD> supposed to be that way? Shouldn't you be able to access the Internet when DD> logged in as a 'user'? If so, how do I change this? This is a feature. Add a user to the dialout group, and he will be able to use pppd. adduser theuser dialout DD> 3) In 'dselect' how do you safely remove packages without destroying all DD> your previously installed files. I think this is why I had to reinstall DD> Debian in the first place. I went to 'deselect' and then 'remove packages' DD> and it started deleting like crazy. What is the proper procedure? DD> (e.g. go DD> to the 'select' option and put a '+' near only those to delete and '-' DD> everything else?) You go to the select option and mark every package you want to install with + and every package you want to uninstall with - (remove, but leave the configfiles) or _ (also remove config files). Then you coose install / remove to perform the actual action. Maybe you pressed - on a "chapter" like "installed packages" ? Or you pressed - for a important package and agreed to remove all dependant packages on the conflict resolution screen. [ other questions answered by someone else ] Ciao, Martin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HELP! Spammer using my system !
On Sat, Jun 13, 1998 at 10:49:30PM -0700, Steve Lamb wrote: > On Sat, 13 Jun 1998 21:07:32 -0400 (EDT), matthew tebbens wrote: > > >How do I stop this ASAP ? > > killall smail > > Not quite the answer you're looking for, but it works. ;) Actually, it probably won't, because the default way to run smail is via inetd. Hamish -- Hamish Moffatt, [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Latest Debian packages at ftp://ftp.rising.com.au/pub/hamish. PGP#EFA6B9D5 CCs of replies from mailing lists are welcome. http://hamish.home.ml.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Install problem
Hi all Here is the problem I have been trying to install Debian from a set of floppies. Things go well until disk 5 is done and it try's to decompress the final file for install. It gives and error that it cant install the file and goes to the next step. I then reboot the system. It will go through a hole series of incremental segmentation faults to 33 ending with /etc/fstab: no such file or dir. fsck failed Please repair manually Control-d will exit this shell and continue system start up give root password for maintenance (or type control-d for normal startup): Control-d takes one back to the same point. Password rm: /root.bush_profile: Read-only file system mv: replace '/root/.bush_profile' , overriding mode 0755? y takes one asks to change password cannot lock the password file; try again later passwrd[47]: can't lock password file. locked into change password What now? I have tried fdisking the disk but it will not remove the partions. to start again. drive has dos and extended dos (Linux). What is worng here? Thanks for your time Shanta -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HELP! Spammer using my system !
On Sat, 13 Jun 1998 21:07:32 -0400 (EDT), matthew tebbens wrote: >How do I stop this ASAP ? killall smail Not quite the answer you're looking for, but it works. ;) -- Steve C. Lamb | Opinions expressed by me are not my http://www.calweb.com/~morpheus| employer's. They hired me for my ICQ: 5107343 | skills and labor, not my opinions! ---+- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HELP: w/ grep on a log file.
Thanks Michael for a good explanation. Johann. -- | Johann Spies Windsorlaan 19 | | [EMAIL PROTECTED]3201 Pietermaritzburg | | Tel/Faks Nr. +27 331-46-1310 Suid-Afrika (South Africa) | -- "I beseech you therefore, brethren, by the mercies of God, that ye present your bodies a living sacrifice, holy, acceptable unto God, which is your reasonable service." Romans 12:1 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ipmasq w/ ipchains?
Does anyone have a modified version of ipmasq, which uses ipchains instead of ipfwadm for 2.1.x kernels? ... or maybe some examples on converting from ipfwadm commands to ipchains commands ... I'm not seeing any obvious equivalent arguments. Maybe it's too late. Thanks -Paul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
problem using staroffice 4 (fwd)
The following was sent to webmaster. Hopefully someone here on debian-user will be able to help them. Jay Treacy - Forwarded message from Phillip Neumann - >From [EMAIL PROTECTED] Sat Jun 13 07:52:08 1998 Delivered-To: [EMAIL PROTECTED] Message-ID: <[EMAIL PROTECTED]> X-Originating-IP: [206.48.134.101] From: "Phillip Neumann" <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: Re: Hi... Content-Type: text/plain Date: Fri, 12 Jun 1998 23:12:16 PDT X-UIDL: cc05ea2bfe0cf1fb480bf353a91bba19 Hi again, Im Phillip Neumann. Im sorry for email you for novices question. Actually i cant subscribe to the mail list bescause i have some problems with my ISP. Well the other email i wrote your, said that i had problems with the instalation of one program. It was StarOffice 4. I was working with glibc and today i change to the none-glibc version of debian. I could install staroffice without problems. When i read some documents, there said that Staroffice cant work with glibc, but i had also programs that requiered glibc. What can i do?? I want to work with staroffice, and with my other programs too!!. Thanks... [EMAIL PROTECTED] __ Get Your Private, Free Email at http://www.hotmail.com - End of forwarded message from Phillip Neumann - -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: e2defrag-problem
First, I cannot verify this, but I am almost certain that you should UNMOUNT before running any defrag program! Secondly, there is no need for defrag on ext2fs systems unless you have a REALLY weird setup. John Wolfgang Gernot Bauer <[EMAIL PROTECTED]> writes: > Im having a problem with the e2defrag-program (debian2.0). I always get > the following message: > > > mother# e2defrag -Vr /dev/sda1 > > e2defrag 0.73 > > RCS version $Id: defrag.c,v 1.4 1997/08/17 14:23:57 linux Exp $ > > > > e2defrag: Error seeking to end of filesystem > > mother# > > The debian-system itself works well... :( > > Here are my mounted drives: > > mother# df > > Filesystem 1024-blocks Used Available Capacity Mounted on > > /dev/sda12478138 1824826 525196 78% / > > /dev/sda41019856 711792 308064 70% /msdos/c > > Anyone had luck with defrag? > > Gernot > -- > - > Gernot Bauer Salzburger Kredit- und Wechsel-Bank AG > eMail: [EMAIL PROTECTED] (Office) Makartplatz 3, 5020 Salzburg > Home: [EMAIL PROTECTED] Austria/Europe > Phone: ++43-662-8684-364 > The answer is yes, me. Fax: ++43-662-8684-23 > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > -- John Goerzen Linux, Unix consulting & programming [EMAIL PROTECTED] | Developer, Debian GNU/Linux (Free powerful OS upgrade) www.debian.org | + Visit the Air Capitol Linux Users Group on the web at http://www.aclug.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On Sat, 13 Jun 1998 22:51:30 -0500 (CDT), [EMAIL PROTECTED] wrote:
>On 13 Jun, Zack Weinberg wrote:
>> What do you get if you do this:
>>
>> gcc -v -o hello hello.c
>
>This is what I get:
>
>% gcc -v -o hello hello.c
>gcc version 2.7.2.3
[...]
> ld -m elf_i386 -dynamic-linker /lib/ld-linux.so.2 -o hello
> /usr/lib/crt1.o /usr/lib/crti.o
> /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtbegin.o
> -L/usr/lib/gcc-lib/i486-linux/2.7.2.3 /tmp/ccd01600 -lgcc -lc -lgcc
> /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtend.o /usr/lib/crtn.o
>/usr/lib/crt1.o(.text+0xe): undefined reference to `__libc_init_first'
>/usr/lib/crt1.o(.text+0x18): undefined reference to `_environ'
Hunh. It is pulling in the C library (-lc). Next thing is to see
what egcc and/or g++ do differently; can you try the same thing with
either of them? (I only need to see the link line, so you could do
gcc -c hello.c
egcc -v -o hello hello.o
g++ -v -o hello hello.o
to cut down on the noise a little.)
Also, another thing that might possibly produce this error is if
/usr/lib/libc.so is mangled. Can you check that that's an ordinary
file containing something like
GROUP { libc.so.6 ld-linux.so.2 libc.a }
please? It ought to affect both compilers if that's wrong, but it's
worth checking.
zw
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ISP Connection + Misc.
On Sat, 13 Jun 1998, Dennis Dixon wrote: > 4) Which command starts XFree86, 'xdm' or 'xstart'? When installing XFree86 > it asks if you want 'xdm' to start at boot time. Then warns that if the > monitor isn't configured yet this could cause problems. Will answering yes > actually bring up a graphical interface at boot time? The safe answer would > be 'no', but then what initialization file do I change later when I want it > to start at boot time? you are after startx. Yes, answering yes to the xdm question will bring up a graphical login screen at boot. It would be a good idea to get X configured correctly for you video card first though. BTW, Your original VT's are still there when it is started, just type CTRL-ALT-Fx to get to them. To get xdm to run on boot, change the line "no-start-xdm" to "start-xdm" in /etc/X11/config You can then directly call '/etc/init.d/xdm start' to get xdm to start without a reboot. > 5) Last, and this may seem really silly, but what is the correct way to log > off and turn off the machine. I've been just turning off the computer at > the 'login' prompt, but when I reboot I get messages such as '... not > cleanly unmounted' leading me to believe I'm not doing things correctly. BIG NO-NO!!! type CTRL-ALT-DEL. thats stops everything and unmounts the root partition correctly. other ways: (when logged in) shutdown -h now (halt) halt (halt) shudown -r now(reboot) reboot(reboot) Michael Beattie ([EMAIL PROTECTED]) --- If it can't be fixed with Vise-Grips & duct tape, it can't be fixed. --- Debian GNU/Linux Ooohh You are missing out! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install/Uninstall dramas (novice user)
On Sun, 14 Jun 1998, Hamish Moffatt wrote: > On Sun, Jun 14, 1998 at 11:55:37AM +1200, Michael Beattie wrote: > > > files/dirs in the other panel. MC should be required software, its that > > > good. > > > > DEFINITELY! > > I have been running Linux for some two years and Debian for more than one, > and certainly haven't required MC yet. Even standard would be inappropriate > I think -- standard is for things which are standard on a Unix system, > and should probably be as strict as possible to keep the size down. > Optional remains appropriate. Sorry about my tone... it was supposed to be a :) ... :) Anyway, you are right, optional is better, but it is such a good utility. I am proud to say that I dont know all the tricks of the trade, so something that does it for me is excellent.. In that respect, it would be good for new/novice users. Michael Beattie ([EMAIL PROTECTED]) --- WinErr: 007 System price error - Inadequate money spent on hardware --- Debian GNU/Linux Ooohh You are missing out! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
On 13 Jun, Zack Weinberg wrote: > > C vs. C++ aside, there's a real problem here. The symbols > __libc_init_first and _environ ought to be defined in libc.so; I'd > guess that gcc isn't telling the linker to pull in libc. > > What do you get if you do this: > > gcc -v -o hello hello.c > > [Note I'm not actually on the mailing list, please cc: me on replies.] > > zw Zack, This is what I get: % gcc -v -o hello hello.c gcc version 2.7.2.3 /usr/lib/gcc-lib/i486-linux/2.7.2.3/cpp -lang-c -v -undef -D__GNUC__=2 -D__GNUC_MINOR__=7 -D__ELF__ -Dunix -Di386 -Dlinux -D__ELF__ -D__unix__ -D__i386__ -D__linux__ -D__unix -D__i386 -D__linux -Asystem(unix) -Asystem(posix) -Acpu(i386) -Amachine(i386) hello.c /tmp/ccb01600 GNU CPP version 2.7.2.3 (i386 Linux/ELF) #include "..." search starts here: #include <...> search starts here: /usr/local/include /usr/i486-linux/include /usr/lib/gcc-lib/i486-linux/2.7.2.3/include /usr/include End of search list. /usr/lib/gcc-lib/i486-linux/2.7.2.3/cc1 /tmp/ccb01600 -quiet -dumpbase hello.c -version -o /tmp/ccc01600 GNU C version 2.7.2.3 (i386 Linux/ELF) compiled by GNU C version 2.7.2.3. as -V -Qy /tmp/ccc01600 -o /tmp/ccd01600 GNU assembler version 2.9.1 (i486-linux), using BFD version 2.9.1 ld -m elf_i386 -dynamic-linker /lib/ld-linux.so.2 -o hello /usr/lib/crt1.o /usr/lib/crti.o /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtbegin.o -L/usr/lib/gcc-lib/i486-linux/2.7.2.3 /tmp/ccd01600 -lgcc -lc -lgcc /usr/lib/gcc-lib/i486-linux/2.7.2.3/crtend.o /usr/lib/crtn.o /usr/lib/crt1.o(.text+0xe): undefined reference to `__libc_init_first' /usr/lib/crt1.o(.text+0x18): undefined reference to `_environ' Thanks for any input. Brian -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Complex passwords
Only way I found: [not extremely hard, but not basic]
Get the source - shadow_980403_orig_tar.gz
in the file shadow-980403/src/passwd.c
Change line 402 so that it looks like below(get rid of the "I am root"
test part of the strength test). Then run the configure script, and
rebuild the passwd suite, then use your new weak-proof passwd.
to rebuild, in the shadow-980403/src directory, type ./configure, then
'make', after that, type 'make install' (must be done as root)
Here is what the lines should look like after you fix the root kludge.
if ((!obscure(orig, pass, pw) || reuse(pass, pw))) {
printf (TRYAGAIN);
continue;
}
later,
troy
Nathan E Norman wrote:
>
> Hello,
>
> I'm writing a script to enable non-UNIX types at my work to be able to
> administrate customer accounts. I'm fairly pleased with what I've got,
> but I've got a question.
>
> The passwd command runs a complexity check on what is entered, and tells
> you if you're picking a password that's too easy. However, entering
> that password again makes passwd accept it. This isn't so good - I'd
> like to force these "administrators" to choose complex passwords. As it
> stands, one can hit the enter key three times and passwd will enter a
> null password!
>
> The manpage doesn't show me an obvious way to disable this behavior of
> the passwd command (accepting easy passwords). Is there a way?
>
> Thanks!
>
> --
> Nathan Norman
> MidcoNet - 410 South Phillips Avenue - Sioux Falls, SD 57104
> mailto://[EMAIL PROTECTED] http://www.midco.net
> finger [EMAIL PROTECTED] for PGP Key: (0xA33B86E9)
>
> --
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Complex passwords
On Sat, 13 Jun 1998, Nathan E Norman wrote: > that password again makes passwd accept it. This isn't so good - I'd > like to force these "administrators" to choose complex passwords. As it > stands, one can hit the enter key three times and passwd will enter a > null password! I don't know of any way to force passwd to do this. But, there are a number of password generators around. I use one written in expect. Maybe you can just make it a policy that the generator has to be used. We used to have fun with passwords... things like b0ne.hEad or Uf0o1 :-) -- Stuart Krivis[EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
accessing an Authenticated proxy server
This is probably off the topic, but here goes I am trying to access an authenticated proxy server for doing a POST operation. I have done this using perl and a normal proxy server, by sending the following to the proxy: POST http://machine.edu/cg-bin/test.pl HTTP/1.0 Host: machine.edu Pragma: no-cache Cache-Control: no-cache Referer: http://machine.edu/index.html Content-type: application/x-www-form-urlencoded Content-length: 16 val=var&val2=var What do I have to add in to authenicate myself?? Thanks for any help Chris -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install/Uninstall dramas (novice user)
On Sun, Jun 14, 1998 at 11:55:37AM +1200, Michael Beattie wrote: > > files/dirs in the other panel. MC should be required software, its that > > good. > > DEFINITELY! I have been running Linux for some two years and Debian for more than one, and certainly haven't required MC yet. Even standard would be inappropriate I think -- standard is for things which are standard on a Unix system, and should probably be as strict as possible to keep the size down. Optional remains appropriate. Hamish -- Hamish Moffatt, [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Latest Debian packages at ftp://ftp.rising.com.au/pub/hamish. PGP#EFA6B9D5 CCs of replies from mailing lists are welcome. http://hamish.home.ml.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gcc can't compile, egcc can!??
C vs. C++ aside, there's a real problem here. The symbols __libc_init_first and _environ ought to be defined in libc.so; I'd guess that gcc isn't telling the linker to pull in libc. What do you get if you do this: gcc -v -o hello hello.c [Note I'm not actually on the mailing list, please cc: me on replies.] zw -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
HELP! Spammer using my system !
I suspect a spammer is using my system to relay or queue spam ! I'm using smail 3.2-3. How do I check, know its being used as a relay ? How do I stop this ASAP ? How do I find the spammer ? Thanks, Matthew Some data from /var/log/smail/logfile might help - 06/13/1998 21:03:17: [m0ykxmp-000NUHC] destination supports esmtp PIPELINING 06/13/1998 21:03:17: [m0ykxmy-000NUIC] destination supports esmtp PIPELINING 06/13/1998 21:03:18: [m0ykxmp-000NUHC] Delivered VIA:smtp.infoasis.com TO:[EMAIL PROTECTED] ORIG-TO:<[EMAIL PROTECTED]> ROUTER:inet_hosts TRANSPORT:smtp 06/13/1998 21:03:18: [m0ykxmy-000NUIC] Delivered VIA:indy.discovery-intl.com TO:[EMAIL PROTECTED] ORIG-TO:<[EMAIL PROTECTED]> ROUTER:inet_hosts TRANSPORT:smtp 06/13/1998 21:03:26: [m0yl1Ce-000NXFC] Received FROM:[EMAIL PROTECTED] HOST:ALDERSSTUDIO.COM [206.175.102.17] PROTOCOL:smtp PROGRAM:in.smtpd ORIG-ID:<> SIZE:4199 06/13/1998 21:03:30: [m0ykxmp-000NUHC] destination supports esmtp 8BITMIME SIZE 06/13/1998 21:03:31: [m0ykxmp-000NUHC] Delivered VIA:mail.diamondpeak.com TO:[EMAIL PROTECTED] ORIG-TO:<[EMAIL PROTECTED]> ROUTER:inet_hosts TRANSPORT:smtp 06/13/1998 21:03:32: [m0ykxmp-000NUHC] destination supports esmtp, but is buggy (250-wwwebzone.iqtinc.com 250-HELP 250-EXPN 250-XREMOTEQUEUE 250-PIPELINING 250 SIZE) 06/13/1998 21:03:34: [m0ykxmy-000NUIC] Delivered VIA:smtp.discusdental.com TO:[EMAIL PROTECTED] ORIG-TO:<[EMAIL PROTECTED]> ROUTER:inet_hosts TRANSPORT:smtp 06/13/1998 21:03:35: [m0ykxmp-000NUHC] Delivered VIA:iqtinc.com TO:[EMAIL PROTECTED] ORIG-TO:<[EMAIL PROTECTED]> ROUTER:inet_hosts TRANSPORT:smtp -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Install/Uninstall dramas (novice user)
On Sat, 13 Jun 1998, Ed Cogburn wrote: > Terence Kearns wrote: [snip] > Are you using the Midnight Commander (MC) file manager? If so the easy > way > is to "step into" (using MC's VFS capability - with cursor line on tarball, > press enter) the tar file on one panel so you can see all the files/dirs of > the tarball, and use the tarball itself as a guide as you delete related > files/dirs in the other panel. MC should be required software, its that > good. DEFINITELY! Michael Beattie ([EMAIL PROTECTED]) --- Curiousity may kill the cat, but a 12 gauge is quicker! --- Debian GNU/Linux Ooohh You are missing out! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ISP Connection + Misc.
For reasons explained later in this post I needed to reinstall Debian from scratch. This time around I had trouble connecting to my ISP. Although I finally fixed the problem it appears to be either a (sort of) bug or something nonstandard with my ISP. Anyway, I thought it was worth mentioning. What I did: edited - /etc/ppp.chatscript inserted - /etc/ppp.options_out changed '/dev/modem' to '/dev/ttyS3' - /etc/ppp/pap-secrets inserted ' * ' above ' os * password' - /etc/hosts.allow added 'ALL : LOCAL' new file - /etc/resolv.conf'search myISP.org nameserver xxx.xxx.xx.xx nameserver xxx.xxx.xx.xx' The Problem: worked manually: exec pppd connect \ 'chat -v -f /etc/ppp.chatscript' -detach crtscts modem defaultroute \ user \ /dev/ttyS338400 with 'pon' wouldn't connect, various error messages: IPCP timeout sending Config-Requests LCP terminated at peer's request serial line is looped back sometimes locks up and need to reboot The Solution: edited /etc/ppp.options_out by inserting 'user ' before '/dev/ttyS3' I'm guessing that the first time I installed Debian I accidentally did something to ppp.chatscript to make it wait for the ISP to request the username, but this time the username needed to be in the ppp.options_out script also since by the time the ISP requested chatscript was already closed. This is merely a wild guess. But since nowhere in the instructions does it say to add your to ppp.options_out I thought it was worth mentioning. Being a new user of Debian I also have some miscellaneous questions, which might seem simplistic to others but would be helpful if answered. 1) In the error message above, does "peer" refer to my ISP, not my Debian machine? It appears that way to me, but I'm not sure. 2) "pon" (and also 'exec pppd') only works when logged on as root. Is it supposed to be that way? Shouldn't you be able to access the Internet when logged in as a 'user'? If so, how do I change this? 3) In 'dselect' how do you safely remove packages without destroying all your previously installed files. I think this is why I had to reinstall Debian in the first place. I went to 'deselect' and then 'remove packages' and it started deleting like crazy. What is the proper procedure? (e.g. go to the 'select' option and put a '+' near only those to delete and '-' everything else?) 4) Which command starts XFree86, 'xdm' or 'xstart'? When installing XFree86 it asks if you want 'xdm' to start at boot time. Then warns that if the monitor isn't configured yet this could cause problems. Will answering yes actually bring up a graphical interface at boot time? The safe answer would be 'no', but then what initialization file do I change later when I want it to start at boot time? 5) Last, and this may seem really silly, but what is the correct way to log off and turn off the machine. I've been just turning off the computer at the 'login' prompt, but when I reboot I get messages such as '... not cleanly unmounted' leading me to believe I'm not doing things correctly. Thanks for any help. Dennis Dixon P.O. Box 1896 Fort Bragg, CA 95437 (707) 964-2979 [EMAIL PROTECTED] http://www.dixonadvise.com -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
