> /usr/include/linux/param.h:4: asm/param.h: No such file or directory
Looks like you don't have the kernel headers installed. I'd go looking for a
package that says kernel-headers or something in the name and install it.
Also make sure that /usr/include/asm either doesn't exist or points at
[EMAIL PROTECTED] said:
> Anyway, pouring through the strace info did the trick. That's a new
> one for me - I haven't used that tool, but I see it is incredibly
> useful. The output is amazing - a bit too much, but very informative.
You can use the -e switch to filter strace output. Also logg
[EMAIL PROTECTED] said:
>From memory there is a Linux specific one:
I probably am not going to be the only person to post a pointer to the whole
thing, but maybe I'll be first :-)
http://lark.cc.ukans.edu/~pauljohn/linux/airline.txt
Jeff
I think we're being trolled here, and reasonably well, too.
To support this, we have have our troll making inflammatory claims, completely
ignoring any kind of logic, and tossing in extras like comparisons to Windows.
Also, cute female physicists tend not to go advertising that fact on the net.
[EMAIL PROTECTED] said:
> However, I can't upgrade glibc using rpm, because I need a higher
> version of rpm.
> I tried installing rpm 4.0.2 from source, but it immediately barfed
> and told me I needed a higher glibc.
Have you tried rpm --nodeps in one of these strategic places? The new rpm
[EMAIL PROTECTED] said:
> I 'make bzImage'd, replaced /vmlinuz (which appears to have been a
> symlink to something in /boot) with /usr/src/linux/arch/i386/boot/
> bzImage0 and ran lilo, then tried to reboot.
Someone else said it, but it bears repeating. If you have a booting kernel,
DO NOT TOU
[EMAIL PROTECTED] said:
> man: can't create index cache /var/qmail/man/index.bt: Permission
> denied
> I've checked permissions on /var, /var/qmail, and /var/qmail/man, and
> they all the same as another system where this error doesn't happen. I
> can also touch a file in /var/qmail/man.
Try ru
[EMAIL PROTECTED] said:
> Unfortunately, there isn't any directory notification on Unix.
It's not clear whether you're including Linux as Unix, but Linux (2.4) does
have directory notification. See Documentation/dnotify.txt in the kernel pool.
It has some limitations. Updating a multiply-link
[EMAIL PROTECTED] said:
> the directory (/usr/src/linux) is correct
/usr/src/linux is WRONG.
Don't build new kernels in it. The reason is that the headers in
/usr/src/linux/asm have to match the system (i.e. libc), because
/usr/include/asm is a link to /usr/src/linux/include/asm. Dumping a n
[EMAIL PROTECTED] said:
> A!
> Any day in which you learn something is a good day.
Any day you manage to teach someone something ain't so bad either :-)
> Thanks, Jeff, for making this a good day for me. :)
Jeff
___
[EMAIL PROTECTED] said:
> echo 1 > /proc/sys/net/ipv4/tcp_syncookies
> for which I am getting the following error message.
> /proc/sys/net/ipv4/tcp_syncookies: No such file or directory
What Angela said.
To explain that error above, /proc is special. It is not a normal filesystem
in which yo
[EMAIL PROTECTED] said:
> You were right on that the automount process' stat was D... once I
> removed that process from the startup and it was no longer running, my
> disks mounted fine.
So, it turned out to be exactly the same problem.
So much for my "mount can't possibly care about the netw
[EMAIL PROTECTED] said:
> (see the "Found the Culprit!" section) http://www.linuxcentral.com/
> linux/lg/issue43/tag/4.html
Oh, the automounter was getting in the way on that one. That's something I
don't know much about.
So, do you have an amd in state 'D'? If so, disabling it or fixing mig
[EMAIL PROTECTED] said:
> i've traced the command using strace, and found that it hangs right
> on the the mount() system call.
If the mount syscall is hanging, then DNS can't have anything to do with it (I
don't think even in the case of NFS, because the kernel doesn't know about
resolv.co
> free_one_pmd: bad directory entry 0020
> Is this a kernel error and how severe is it? Should I worry?
This looks like a NULL which has had a one-bit error in it. If so, you have
some flaky memory.
It could crash your system if it happens in the right place. The error above
happened in
Have you read Documentation/oops-tracing.txt in your kernel tree? If not,
that would be a good place to start.
> Do i need to run the ksymoops from the same computer which barfed?
It needs access to your System.map, which ought to be in /boot.
> from what i've been able to tell, the output f
[EMAIL PROTECTED] said:
> Not computer geek library is complete with Donald Knuth's seminal
> series on the Art of Programming.
Yup. That series isn't going out of date any time soon.
> Another one of my all time
> favorites is Barb Liskov's "Abstraction and Specification in Software
> Develo
[EMAIL PROTECTED] said:
> But is 2.4 going to be release with any of the bugs they know about
> now, say?
Yes, it will. As of test10, Linus was only taking critical bug fixed.
The CRITICAL bugs on Ted's list will be fixed. Most of the rest will be in
2.4.0 and be fixed later. Most of them a
> I'd say free of all known bugs - who would willingly, knowingly, release
> buggy software as anything other than an alpha or a beta release?
That would be stupid, not to mention impossible.
If you're going to try to do right by your users, you want to release
something as soon as the benefits
[EMAIL PROTECTED] said:
> Well I think Linux kernel is declared stable when there are no
> *showstoppers* - I don't know about no *bugs*.
I forgot to mention one thing. The current kernel bug list is maintained by
Ted T'so at http://linux24.sourceforge.net/. The ones that have to be fixed
bef
[EMAIL PROTECTED] said:
> Well I think Linux kernel is declared stable when there are no
> *showstoppers* - I don't know about no *bugs*.
There will be bugs. They will be squashed in succeeding 2.4 releases, as is
happening with 2.2 right now. The thing with a new stable series is that the
us
[EMAIL PROTECTED] said:
> It just hangs the box when trying to boot into the newly compiled
> kernel.
You're going to need to be a lot more informative if anyone's going to help.
What manages to appear on the console would be a good start. Also what makes
it into the log if that's not on the
[EMAIL PROTECTED] said:
> Unfortunately this is a different problem from mine.
Sorry about that. I think I understand the problem now.
I can't reproduce this on either 2.2.5 or 2.2.14 in the simplest case (add a
virtual interface, ssh out to another machine, look at where it thinks I came
fro
[EMAIL PROTECTED] said:
> so far i know parent and child process should have separate address
> space. but here the addresses are same in both the processes. r they
> sharing the same address space?
They are in different address spaces, which is why those addresses can be the
same without them i
[EMAIL PROTECTED] said:
> Then we upgraded the kernel to 2.2.14, and the behaviour changed: now
> it suddenly started to announce the name/IP on the last virtual
> interface on eth0 as its name, fex. when being used for irc or when
> NFS-mounting directories.
There was a complaint about this on t
[EMAIL PROTECTED] said:
> using group read and write is not an option.
Why not? The obvious thing is to chgrp file foo to group foo_group and anyone
who's allowed to modify foo is a member of foo_group.
Jeff
___
tec
[EMAIL PROTECTED] said:
> I'm wondering what disk utilities are available under Linux--or are
> they necessary? For instance what about a disk defragmentation
> utility like Norton Speed disk?
As far as I understand, they're not really necessary. The ext2 filesystem
isn't as prone to fragm
[EMAIL PROTECTED] said:
> Can someone point me to documentation on the available linux kernel
> params and how I go about querying and modifying them?
Documentation/kernel-parameters.txt in your favorite kernel pool. Looking at
them and changing them varies. The most popular ones have entries
[EMAIL PROTECTED] said:
> >> find . -type f -print | xargs grep /dev/null
>
> I don't get it. I doubt it hurts, but why the /dev/null?
To prevent grep from reading stdin (and just sitting there) if is
empty.
Jeff
___
[EMAIL PROTECTED] said:
> How do I find out what version of library is on my system?
On RH (and other rpm distros), you can do this:
~ 1009: rpm -q -f /lib/libc.so.6
glibc-2.1.1-6
You have to peel off the "1-6" to see that this is glibc 2.1. Something
similar should work or whatever librarie
[EMAIL PROTECTED] said:
> I thought the -p option preserved the permissions. How can I preserve
> my files original permissions scheme (in directory B)?
The permissions being preserved are those of the copied file. Any files that
get copied over don't matter.
Did you try not preserving permiss
Also check out mason (http://users.dhp.com/~whisper/mason/). It's not a GUI,
but it does have the feature that you can put it into learning mode, make
connections to the firewall, and have it generate the rules needed to block
that connection.
Jeff
[EMAIL PROTECTED] said:
> However, we have observed that instead, it just acts as if there is
> no memory available, and doesn't free up buffer/cache allocated
> memory when new programs are started. We're using RedHat 5.2, kernel
> 2.0.36.
So, do you start swapping when there is a ton of mem
[EMAIL PROTECTED] said:
> There are certain things that you will and should never use an rpm
> for.. for example.. apache.
Why not? If you've got a server that you take seriously and want control over
what modules (and what versions) go into it, you build from source. But for
your generic s
[EMAIL PROTECTED] said:
> hey, anyone have a minute to explain the differences/advantages
> disadvantages of source vs rpm? I've installed things both ways, but
> i really only know barely enough about them to be dangerous. whenever
> i "make" something and it doesn't work i just kinda sit ther
[EMAIL PROTECTED] said:
> I am automatically disconnected. Is there an easy way to either a)
> disable the auto disconnect feature or b)extend that to a more
> reasonable time constraint (20-30 minutes would be fine).
I use diald, which I got because it provides demand dialling, but it also
c
[EMAIL PROTECTED] said:
> No not those windows specific directories... all my DOC's etc... like
> I want windows specific stuff on my system, that would be similar to
> hitting my head against the wall and well.. I am not into self
> destruction
In that case, go ahead. That convoluted command th
[EMAIL PROTECTED] said:
> thanks for the info I will try it out, Actually what I am doing is
> Migrating everything to Linux, so all the Windows File thingy's are
> copied over to a Linux filesystem and well I kind of want to get rid
> of Windows all together.
What do you expect to do with your
[EMAIL PROTECTED] said:
> Is there a file I can create that will do this? I am not a programmer,
> so I have limited knowledge in creating files but I can sure try...
> Any help you could provide would be amazing.
I don't think I could pass up a chance to be amazing :-)
find /dos -name "* *" -
[EMAIL PROTECTED] said:
> I'm wondering how to send email messages that query the reader's app
> and display as text if that's what you're reading with (ie, pine or
> mutt), or display as html if you have an html-capable reader (ie,
> Netscape)?
You don't, AFAIK. If you really want to do that,
> Once in a while Linux (RH 6.2) will insist that a cdrom is busy and
> thus will not let me umount it and thus will not let me eject it.
Another handy thing is lsof. lsof | grep /mnt/cdrom will tell you what
processes have open files in /mnt/cdrom (I think - my system seems not to have
lsof
[EMAIL PROTECTED] said:
> I approach you now to ask, as I am running Gnome and Windowmaker,
> where can i find out which parts of the GUI i see are Gnome's fault
> and which are the fault of my window manager?
Kill your window manager and see what parts of your GUI go away.
Also, "Gnome" is not
[EMAIL PROTECTED] said:
> Unfortunately, modules/standard/mod_so.o doesn't exist. The file is
> mod_so.c. Do i need to compile this to mod_so.o?
This might be system-dependent. I know squat about Solaris, but this is the
sort of thing that might work on one Unix, but not another.
Also, does
[EMAIL PROTECTED] said:
> When I get to Disk Druid and try to partion the root directory, I get
> an error that says Boot Partiton is too big.
I think what it's talking about is the LILO 1024 cylinder limit. The kernel
has to be within 1024 cylinders of the beginning of its partition. That's w
[EMAIL PROTECTED] said:
> The project in question: a generic ratings and comments engine for a
> website. The things being rated could be anything, but site visitors
> should be able to vote on various survey questions for each
> (book|movie|restaurant), and add their comments. Visitors should als
[EMAIL PROTECTED] said:
> On Fri, 12 May 2000, alissa bader wrote:
> > Specifically, anything containing the
> > string "QAA." Tried doing a grep, but I got the
> > entire line again, not just the portion of it that
> > contained "QAA."
> what part of the line do you need to grab, other than QAA
> What's ya'll's more educated opinion?
I'd call it more of a wild guess, but when I see where I've got an
applet-widget.h and then see what rpm it belongs to, I get this:
% locate applet-widget.h
/usr/include/applet-widget.h
% rpm -q -f /usr/include/applet-widget.h
gnome-core-devel-1.0.4-34
S
> We kept getting random signal 11's with an old 486, but the memory
> check didn't find anything. The cpu fan wasn't working, though, so
> could that also cause sigsev? Just interested...
I think so. If the cpu is overheading, that could cause the same sorts of
things. Although, I've mostly
> I need some help making a small script file. What I need it do is
> copy the directories on one HD to another.
I usually do this with 'cd srcdir; tar cpf - | (cd destdir; tar xpf -)'
cp -a will apparently do the same thing.
Jeff
__
> A g++ program is encountering problems during execution. There are no
> error messages as it gets linked and compiled. The executable file
> does not run though. These are the error messages observed: 'Program
> received a signal SIGSEGV, segmentation fault' OR 'Segmentation fault
> (core dumpe
> What does signal 11 mean?
Signal 11 is SIGSEGV. It means that the process accessed (tried to, anyway)
memory that it had no access to, either because the address didn't exist in
its address space or because it didn't have the right permissions.
The question you probably wanted answered more
> I heard a rumor that about six months ago or so there were kernel
> patches for hard disks larger than 34Gig.
> Have these been incorporated into the kernel,
It has been. I believe that was the problem with cylinders being a short and
34 G disks overflowing it.
> and, if so, which version?
> What's the best way to monitor disk space usage? Ok, the most
> efficient way to monitor disk space usage?
du and df are the basic tools.
To find the disks-space hogs, I usually start where I already know the problem
is (like ~ :-), and do "du -s * | sort -rn" and recurse.
> Surely a script
> Apr 15 14:40:33 trinity fetchmail[924]: 5.1.0 querying maple.he.net
> (protocol APOP) at Sat, 15 Apr 2000 14:40:33 -0400 (EDT)
> Apr 15 14:45:33 trinity fetchmail[924]: timeout after 300 seconds waiting to
> connect to server maple.he.net.
> Apr 15 14:45:33 trinity fetchmail[924]: client/ser
> When I try to telnet to it on 110 from my shell account (this mail
> account), QPOP answers me. So it's somehow related to the firewall.
I hate to ask stupid questions, but can you make sure that your shell machine
and firewall agree on the IP address for maple.he.net?
> Could this have some
> Item1="0 \"Midnight Commander\" off "
> Item2="1 \"Mouse support\" off "
> Item3="2 \"Dos-Tools\" off "
> Items=$Item1$Item2$Item3
> whiptail --title "$Titel" --checklist "$Text" 15 40 10 "$Items"
The quotes around "$Items" is forcing the $Item1$Item2$Item3 to be a single
argument, whereas
> it is only: michelle1:/# myscript -parameter /path/filename
> A small example will be nice.
#!/bin/bash
if [ $# -eq 0 ]
then
echo You gave me no parameters
elif [ $# -eq 1 ]
then
echo You gave me one parameter, which is \"$1\"
else
echo You gave me $# parameters which are \"$*
[EMAIL PROTECTED] said:
> I'm a little scared of trying it - a file system is more importatnt
> than a window manager (or TV card driver), so I'm not too keen on
> using something unstable...
[EMAIL PROTECTED] said:
> But It is not stable. I use it only for squid cache now. I
> would not r
> My problem is that I'm looking for "eth0" under /dev but it's not
> there. I compared with work and it isn't under /dev/ either so I'm
> assuming it shouldn't be there. So how do I configure a nic?
Network interfaces don't get /dev entries.
> And can
> someone explain where linux is getting
> When I try to shutdown the machine to reboot into Windows when I need
> to, it no longer shuts down. It stops at a bash prompt. I always used
> shutdown now and had no problems, but it doesn't seem to work now.
> Any suggestions as to how to fix this?
I think the thing to do is figure out wh
> You can comment it out ir remap it to another program, like a script
> that logs who and when pressed it.
> If the console is in a place where other people can easily get access
> to it and you are worried about security there are many programs, for
> X and tty*, that lock the console out with
> Can anybody tell me which is the file where the routing decisions/
> algorithms are implemented in Linux and how to get it's source code.
You can get Linux sources from http://www.kernel.org.
The code you're interested in is in net/*.
Jeff
_
> Any more ideas?
Send in some complete error messages. Console logs. Anything that looks out
of the ordinary...
Jeff
> Strings? How do I do that? Then how do I get the stuff from the
> output?
strings is a very handy little utility. 'strings filename' goes through
filename printing out anything that looks like text.
It is a zero-order approximation to any number of blah2txt utilities :-)
In this case, it
[EMAIL PROTECTED] said:
> AHA! My DVD is bootable! And I have a CD-RW
> Will let you know if this works tonight... thanks Jeff!
Cool. Since it's a bit larger than a floppy, you can put a whole root
filesystem on it if you want. I'm not sure that's a great idea, though. /dev
likes to be
> In any case, I can't boot Linux without the boot disk, and when I do,
> "starting linux" it takes about 10 minutes for all the dots to go
> across the screen. Any hints on how to speed up my boot?
Boot off something other than a floppy. Zips can be made bootable, I think.
> Or on how
> to
> It got upset and when I boot up, I get a lovely screen with 01 01 01
> 01 ... It's a Dell Dimension L500r with two EIDE drives in there,
> Win98 on the first one and Linux occupying the second one. Can
> anyone give me any hints as to what the booting heartattack is over?
At the bottom of the
> The system seems to be finding it just fine, but
What do you mean by this? The system normally just sees the serial port, and
doesn't know what if anything is attached to it.
> the modem won't respond to commands (ATZ, ATDT, etc.).
Are you sure you've got the right port? If not, just ru
> Hm, maybe this should be a FAQ. Both indicate that youre accessig
> memory that you shouldn't.
A segfault does.
I did a quick check of the kernel (core kernel + arch/i386), and found two
places where processes get SIGBUS:
swapping out a process failing to free any memory (swap_out in
> Would anybody please let me know if the "noauto" should be "auto" or
> something else?
It should not be noauto (i.e. leave it out) if the media is always there, like
a Windows disk partition, and you want it mounted at boot time. If it's
something like a floppy, "noauto" would be reasonable
> But Jeff, I've seen situations where ping has been turned off for
> security reasons.
What security reasons were those? ping is a really basic tool without too
many security implications (besides DOS, which can be done with any protocol).
If ping has been disabled, then I'd start looking
> I can ssh to any server (there are dozens) from my puter but I can't
> telnet to my puter from the servers.. I'm thinking it has something
> to do with how security is set up but I don't know.
I think your site has network connectivity problems. Regardless of security
policy, all machines
> I'm not at that computer now and can't run ifconfig until tomorrow but
> I've tried pinging from one of our server and this is what I get:
> chipshot # ping 10.254.254.245 PING 10.254.254.245 (10.254.254.245):
> 56 data bytes
Your box is off the net somehow. Make sure the wires are hooked up
> I'm trying to get my desktop computer at work (which is connected with
> an ethernet card
Run ifconfig and make sure that it has the right ip addr:
eth0 Link encap:Ethernet HWaddr 00:B0:D0:20:06:FC
inet addr:198.99.130.20 Bcast:198.99.130.255 Mask:255.255.255.0
which
> I am using Red Hat 6.1 and was wondering if there's there a way to
> boot directly into gnome without having to log in (anywhere).
This is off the top of my head, so you'll probably have to play around before
it actually works. The basic idea is that you replace xdm in the boot process
with
> It is the print function for the kernel routines. I don't know the
> workings of the output of this fuction yet.
It's just like printf, without a lot of the bells and whistles (no
floating-point, numbered arguments, etc).
Jeff
[EMAIL PROTECTED]
> What exactly is printk()? I haven't found any information about that
> particular function.
And the reason that this internal kernel function is appearing in user-level
docs is that printk'd stuff shows up in the log.
Jeff
[EMAIL PROTECTED] h
> It is always the same inode-- When fscking, inode 818763 is always
> deleted with zero dtime. What does this show?
This could just be a temporary file that some daemon has opened and unlinked.
When the system crashes, the one reference to that file (the process) has
disappeared, so fsck se
> Is there some command or function that I can use that tells a hard
> drive to spin down, etc?
Try 'man (heh) hdparm'.
The -y switch looks like what you want. -S and -C also look interesting.
However, this is an ide command, but there is this at the end:
Although this utility is in
> Or, can I just not bother with initrd? Is it worth having?
You don't need it.
You should also probably change the kernel version number if it would
otherwise be the same as what you've already got. What I did was make
EXTRAVERSION=a, so that the new kernel version was 2.2.5a rather than th
[EMAIL PROTECTED] said:
> I got same weird message every time I tried to boot, so I wimped out
> and reinstalled. :-) Guess I need to buy a book eh?
Pay attention to the chapter called "Single-user mode". :-)
Jeff
[EMAIL PROTECTED] http://www.
> Well, i've already downloaded the latest versions of those libraries
> (source rpmd), unpacked, configured and installed them according to
> the manual, and those programs still prompt the same msg
You built them from source, not from a binary rpm, so rpm doesn't know about
them.
So, just
> The next question, of course, is: Can anyone recommend any rackmount
> servers from companies that support both Linux and NT?
dcginc.com. They sell rackmount servers, your choice of Linux or NT. A
popular local co-lo service swears by them. An extra attraction for us (YMMV
on this) was th
> > Also, mysql (mysql.com) and msql, both free. Freeness makes them good
> > for
> > learning what you want in a "real" database. And they might turn out
> mysql is not free according to the DFSG[1] (or opensource
> according to the OSD[2], if you take that side). See
True. I was being a b
> Hey girl,
I'd probably get flayed if I said something like that :-)
> What I use is postgres.
Also, mysql (mysql.com) and msql, both free. Freeness makes them good for
learning what you want in a "real" database. And they might turn out be be
"real" enough for you.
> I remember when I was first learning linux (redhat 2.something, I
> think), there were a couple of times when I screwed something badly
> enough that it was easier (and faster) to reinstall than to learn
> enough (on an unusable system) to make the system usable again. ;)
For those of us with s
> Because I screwed up my linux install, and had to start over yesterday
> :-)
BTW, what did you do? I'm having trouble thinking of anything short of
trashing a disk that would call for a complete reinstall.
Jeff
[EMAIL PROTECTED] http://www.l
> Because I screwed up my linux install, and had to start over yesterday
> :-)
Gotta hate it when that happens...
Jeff
[EMAIL PROTECTED] http://www.linuxchix.org
> I am beginning to believe that I sent to wrong place, say
> [EMAIL PROTECTED] or something.seems strange I did not get
> them back though.
Why not go poking through /var/logs/maillog to see what really happened?
Jeff
[EMAIL PROTECTED] http:
> xauth has been partially overwritten with bad data
xauth or your .Xauthority?
> .Xauthority for my personal account is listed when I do ls, but I
> can't get a text editor to read it.
It's (at least partly) binary data. Mine has some readable strings in it, but
it's definitely data.
> So t
> > grep hdb /var/log/messages|more
> > is better. It tells you what the disk told the OS when it booted.
> Bingo. Thanks, Jeff!
Actually, this is probably the right way to do it:
hdparm -g /dev/hda
/dev/hda:
geometry = 1048/16/63, sectors = 1056992, start = 0
Multiply sectors
> dired is ^X ^B.
Sorry to follow up on my own blatant disinformation, but the above is wrong.
dired is ^X d; ^X ^B is the buffer editor which is real similar.
I stand by the rest of it. (heh)
Jeff
[EMAIL PROTECTED] http://www.linuxchix.org
> > Also, I seem to remember there's a way to find out how large a disk I
> > have without having to open my box and look at it. Anyone?
> Try the command df
That only tells you about filesystems on the disk which isn't necessarily what
she wants.
echo p | fdisk /dev/hdb
is a little messy, b
> To delete such problem files, if you use emacs, you can use the
> directory editing feature in emacs is useful (^X ^F from memory).
dired is ^X ^B. ^H m will give you a little buffer which tells you what all
the commands are, but the most interesting ones are the movement commands (^N
^P - u
> I can't find it spelled out plainly anywhere. Anyway, I want to:
> 1) load the kernel onto a floppy so I can make sure linux will boot
> with the new kernal 2) backup the old kernal 3) install the new
> kernal
You shouldn't need the floppy. When you build your new kernel, just make sure
i
> Obviously you would select n for items that you don't need included in
> your kernel, but what is the difference in saying y to support for my
> scsi or m?
For the average user, it doesn't make any difference. You need to say 'y' to
anything that you need to boot the kernel (so if you're boot
> > My ppp log shows that all 7 bits set to zero not 8 bit clean, is there
> > any way to resolve this?
> It's often a problem with the ISP, I don't remember what the exact
> cause is. So it's not your problem, unless you've also changed
> something in the ppp configuration.
> It goes withou
> If so, how can I install it (or do I need to reinstall RedHat?).
> Install directions seem to want to use gcc to do the installation.
You didn't way whose install directions want gcc. If the system is up enough
to run rpm (and it should be if you can do 'which foo'), check to see if it's
in
> In the /etc/fstab, the entry the corresponds to the filesystem you
> want to have mounted at boot time should have a 1 in the fifth field.
> A 0 in this field indicates that it should not be mounted at boot
> time.
This is what man fstab has to say about field # 5:
The fifth field, (f
> I am sure I chose a wrong option someplace along the way because now
> my long file names are truncated.
Is this doing 'ls' on a windows filesystem?
If so, I don't see how it's related to the new kernel. That's (in my
experience) controlled by how you mount it. If it's mounted as an msdos
1 - 100 of 126 matches
Mail list logo
