Linux-Misc Digest #994, Volume #18 Fri, 12 Feb 99 07:13:14 EST
Contents:
Re: UPS restores power after system is halted; will it boot? (Robert Nichols)
wmppp + getmodem speed problem ([EMAIL PROTECTED])
Re: Linux on a Sparc (Tadpole) (Robert Barnes)
Quantum Incompatible? ([EMAIL PROTECTED])
linux can't mount root fs (Frank Berwanger)
how to rename GNUstep? (Alex Postnikov)
threads (Du Jinsong)
Re: oops "rm -r *" Missing part of tree, or at least a shrub (Jason Bell)
Re: The Importance of Stable URLs. (William Burrow)
System command /usr/bin/gunzip -c /usr/man/cat1/date.1.gz exited ("Joe (tWdy)
Philbrook")
Re: one thing that sux about Linux.... (William Wueppelmann)
Gateway problem! (Michele Beltrame)
Re: /lib/modules/preferred deleted on boot ???????? (Villy Kruse)
Re: Need help with hosed rpm database ("ne...")
Re: hacked login (gus)
Re: deleting kernel sources okay? (gus)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Robert Nichols)
Subject: Re: UPS restores power after system is halted; will it boot?
Date: Fri, 12 Feb 1999 01:04:29 GMT
In article <[EMAIL PROTECTED]>,
Anthony Christofides <[EMAIL PROTECTED]> wrote:
:More or less the subject says everything. Consider an unattended Linux
:server powered through a UPS. Power fails, the UPS signals the system,
:and the system starts the power fail procedure. This waits for a while,
:in case power is restored, then begins to shut down the system, and
:finally the system is halted. Suppose that the UPS has not cut the power
:yet, and the power is restored. The UPS signals that power is restored,
:but the halted system can't respond to the signal. So the system will
:remain halted.
The proper way to use a dumb UPS is to have the power fail procedure
reboot the system, not halt it, and have the system activate the control
signal that shuts off the UPS just before entering the BIOS for the
reboot sequence. If the AC power has already recovered, the system will
simply reboot.
--
Bob Nichols [EMAIL PROTECTED]
Finger [EMAIL PROTECTED] for PGP public key.
PGP public key 1024/9A9C7955
Key fingerprint = 2F E5 82 F8 5D 06 A2 59 20 65 44 68 87 EC A7 D7
------------------------------
From: [EMAIL PROTECTED]
Subject: wmppp + getmodem speed problem
Date: Thu, 11 Feb 1999 00:47:09 GMT
I have finally installed wmppp on a RedHat 5.2 machine almost correctly...
The only problem is getmodem speed doesn't work. I used the one out of
example-scripts/ and it was suid-root. But upon running it, the echo is
simply that the binary cannot be executed. Reading /var/log/messages, I find
this...
Feb 10 19:38:46 nivdome modprobe: can't locate module binfmt-17791
What is this?
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Robert Barnes <[EMAIL PROTECTED]>
Subject: Re: Linux on a Sparc (Tadpole)
Date: Thu, 11 Feb 1999 20:00:48 +0000
Paul Hovnanian wrote:
>
> I (we) have a stack of Tadpole "SparcBooks" which are notebook
> computers with Sparc-compatible processors, running SunOS 4.1.2.
>
> Is anyone familiar with these things? Would it be possible to
> run Linux for a sparc on them (they are binary compatible with
> an old Sparc 1+ we used to have)? The display hardware appears to
> be some sort of VGA-compatible LCD with a 640x480 VGA output. They
> have built-in ethernet, modem and parallel port.
It depends which model Sparcbook they are. If they are Sparcbook3
(sun4m architecture), then the answer is yes, but they can't use
the LCD display (I guess that you could use them as "servers").
Development for the Weitek P9100 video controller (which the
Sparcbook3 series uses) is under serious consideration.
-bob
------------------------------
From: [EMAIL PROTECTED]
Subject: Quantum Incompatible?
Date: Fri, 12 Feb 1999 08:18:32 GMT
I am attepting to install Redhat 5.2. I have a 10.2gb Quantum Fireball
harddrive that has 3 logical drives on it. The drives are 2gb a peice for a
total of 6gb and the rest of the harddrive is open. When i try to install
Redhat 5.2 workststion package, it tells me that there is not enough disk
space for this type of installation. I have windows 98 as a secondary OS too.
Does Linux not recognize my harddrive? Do i need to install additional
partitions? Please reply ASAP Thanks, Baffled
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Frank Berwanger <[EMAIL PROTECTED]>
Subject: linux can't mount root fs
Date: 12 Feb 1999 08:27:13 GMT
Hi!
I've got a boot problem.
During the boot process linux writes:
VFS: Cannot open root device 30:2b
Kernel panic: VFS: Unable to mount root fs on 30:2b
The interesting part is, that I can boot, when I use the root
disk, I#ve made under dos via setup. I click through the linuxrc-program
and enter /dev/hda2 as root filesystem to boot.
I use Suse linux 6.0 with kernel 2.0.36. Linux was installed on an existing
Linux partition, but I had to install /usr on an SCSI Drive (don't ask).
What's the problem??????
Please mail to:
[EMAIL PROTECTED]
--
Ciao,
FRANK:-)
*******************************************************
* Who's that riding in the sun? *
* Who's the man with the itchy gun? *
* Who's the man who kills for fun? *
* Psycho Dad, Psycho Dad, Dad. *
*******************************************************
------------------------------
From: [EMAIL PROTECTED] (Alex Postnikov)
Crossposted-To: comp.os.linux.x
Subject: how to rename GNUstep?
Date: 12 Feb 1999 10:24:11 GMT
Reply-To: [EMAIL PROTECTED]
Hello all,
I use WindowMaker and I am happy with it except one little thing. I like to
use small letters for file names in my home directory and prefer when
configuration files are hidden (i.e. start with a dot). The only exception
from this rule is the ~/GNUstep hierarchy? For AfterStep there is an option
'-f config_file'. But what should I do with WindowMaker? Is there a way to
rename default ~/GNUstep directory? If this feature is not implemented yet,
then I address this question to the developers of WindowMaker, AfterStep, etc.
Why not use a more reasanable name like ~/.GNUstep ?
Thanks,
Alex
------------------------------
From: Du Jinsong <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.development.system,comp.os.linux.help,comp.os.linux.networking
Subject: threads
Date: Fri, 12 Feb 1999 17:24:51 +0800
Hi, would you please tell me whether I can call pthread functions(e.g.
pthread_create()) in Linux kernel source .c files?
Thanks a lot!
------------------------------
From: Jason Bell <[EMAIL PROTECTED]>
Subject: Re: oops "rm -r *" Missing part of tree, or at least a shrub
Date: Fri, 12 Feb 1999 00:54:15 -0800
Frans Gumpu Slothouber wrote:
> Yah Right <[EMAIL PROTECTED]> wrote:
> : I seem to have accidentally (methodically) disposed of the "local"
> : directory under /usr on my rh5.2 box. I was reinstalling (ech) win95
> : and and storing data on the linux side and wiped unknown goodies along
> : with the ms data. I had "installed everything" from the rh5.2 disc,
> : what I need to know is how much did I shred? Any "easy" ideas for
> : reloading the tree below /usr/local/ ? Oh well at least I know that
> : installing Linux is much easier than win95 whether I go from scratch or
> : not!
>
> I think rpm might help here. rpm can tell you what files and packages
> are missing. It does sound like a lot of work though... you
> might be better of just reinstalling...
>
> Have fun,
> Frans.
>
>
> --
> ______________________________________________________________________
> /Frans Gumpu Slothouber <[EMAIL PROTECTED]>
> \___/\ GalaxyNG Game Master. http://gumpu.student.utwente.nl/~galaxyng
> / \/
> Implementation: The fruitless struggle by the talented and underpaid to
> fulfill promises made by the rich and ignorant.
Chalk this up as a lesson on "Why you should never operate as root, unless
you absolutely need to!"
;)
------------------------------
From: [EMAIL PROTECTED] (William Burrow)
Subject: Re: The Importance of Stable URLs.
Date: 11 Feb 1999 00:45:13 GMT
Reply-To: [EMAIL PROTECTED]
On 10 Feb 1999 09:41:32 -0800,
Fred Flatstone <[EMAIL PROTECTED]> wrote:
>Linux Web Masters: Please do your fellows and all WWW users a big
>favor and try your best to avoid changing the URLs of your pages.
...
>If you're an ISP hopper, use one of the free or cheap dynamic domain
>services or this particularly interesting free service. You don't get
>your own domain, just a sort of link between their directory tree and
>yours. They sound "persistant". Their org is associated with Libraries,
>etc.I haven't tried them, but I think I will. They're called PURL
>(Persistant URL) and are at:
>
>http://purl.org
This is an _interesting_ resource. The question that pops to mind is,
is there a top level PURL domain for Linux? Who will administrate it?
Perhaps it is time people popped over there and investigated this...
>Examples of often worse-than-worthless unnecessary changes:
>-- Home page changed from index.htm to index.html or home.html
>-- Changed spelling like speler.html to speller.html.
Unix has filesystem links. Use them. (I once misspelled a link in a list
message (d'oh!), I solved by making a link. Easy.)
>-- ISP hopping.
PURLs would be real handy.
>-- Changed words like /searcher/ for /locator/. (Actual example.)
>-- Changed hierarchies like /breadbaskets/ to /baskets/bread/.
Symlink or PURL, though sometimes no remapping is possible.
--
William Burrow -- New Brunswick, Canada o
Copyright 1999 William Burrow ~ /\
~ ()>()
------------------------------
From: "Joe (tWdy) Philbrook" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux.slackware
Subject: System command /usr/bin/gunzip -c /usr/man/cat1/date.1.gz exited
Date: Thu, 11 Feb 1999 21:34:39 +0000
System command /usr/bin/gunzip -c /usr/man/cat1/date.1.gz exited with status 13.
Mnnnnn...
...and all I wanted to do was a...
~$ man date|lessv:!su
$PASSWD
date 0211194599
>>sickbiosbatteryHOW-TO
But the man command didn't just happen tell me how to get to a live :!
command prompti emulation from within a less session which was invoked on
the stdout of the man command....
What did I do wrong???
So I couldn't test the form of:
~$ man date|lessv:!su^M$PASSWD^Mdate 0211194599^M
to make it worth my while to find the pathname of the HOW-TO I'd want to
define $sickbiosbatteryHOW-TO just prior to doing an:
/home/jtwdyp $ export $sickbiosbatteryHOW-TO
in an:
case $itWORKED in
yes*)itWORKED="YES";;
esac
case $itWORKED in
YES*)echo "lets make it real please <g>";
echo '~$ man date|lessv:!su^M$PASSWD^Mdate 0211194599^M'>>$sickbiosbatteryHOW-TO;
echo "why can't the man $COMMANDname output always contain";
echo "at least one line containing the full local pathname of the most";
echo "apropiate HOW-TO document AND the command line syntax of how to";
echo "explicitly call up the apropiate [ LaTeX? ] document with the";
echo ' "info" '" system??? And of course"' "$ man info
" '"should lead to";
echo "the local path to the HOW-TO on installing the <explitive deleted>";
echo "info system, especialy since some man documents explicitly inform";
echo "the poor innocent former win9x user who is probably looking at some";
echo "OLD book that came with an earlier version of linux that was packaged";
echo "as a"' "complete LINUX kit" '"That somebody paid $29.99 for just before";
echo "the slackware 3.5 was being promoted on alt.os.linux.slackware newsgroup";
echo "and the book thinks man [command name] is my friend...";;
*)$itWORKED="NOT !";;
esac
echo " <sigh>"
| ~^~ ~^~
| <?> <?> Joe (theWordy) Philbrook
| ^ J(tWdy)P
| \___/ <<[EMAIL PROTECTED]>>
------------------------------
From: [EMAIL PROTECTED] (William Wueppelmann)
Crossposted-To: linux.redhat.misc
Subject: Re: one thing that sux about Linux....
Date: Thu, 11 Feb 1999 01:45:53 GMT
In our last episode (Tue, 9 Feb 1999 16:40:38 -0500),
the artist formerly known as Southam said:
>You expect a CEO to learn e-mail on Linux? Right. His time costs about $200
>an hour - too valuable to spend learning a new OS, when Windows is ready and
>able to handle at least this simple chore ( for a while anyway, until it
>chews the FAT table on the disk and explodes like a roman candle) and he
>mostly knows how to use it. I wish everyone in here would drop the "holier
>than thou" attitude. You all seem to either have a superiority complex or
>want to divide the world into technology haves and have nots.
I suspect that learning how to use email on Linux is just as easy as learning
how to use email on Windows. I also suspect that most people who use email
don't really understand email all that well to begin with. Pine is Pine,
whether it's on Linux or Windows. Netscape Mail is Netscape Mail, regardless
of the platform it runs on.
More people use computers today than ever before, but the percentage of
computer users who are functionally computer-illiterate is probably higher
than ever before. With the level of knowledge that most people have of their
computers, and with the level of sophistication with which most people use
their machines, I suspect that most people can be migrated from one system to
another with only a couple of hours of training. And if two or three hours of
training under Linux is the price for almost zero downtime, that's probably a
worthwhile investment for most executives.
Remember, we're not talking about being able to install the system or new
hardware or to write shell scripts or anything like that. We're talking about
using email, a word processor and a spreadsheet on a system that's been set up
and preconfigured, just as is currently done using Windows.
Personally, I think most people would do best with something less complicated
than a Mac. A general-purpose computer with a powerful operating system just
isn't necessary for people who use a computer as an appliance and for only a
few limited tasks. A stable system and software that doesn't need to be
upgraded every year just to be able to keep doing the same things, howerver,
would be a tremendous advantage.
--
William
It is pitch black. You are likely to be spammed by a grue.
------------------------------
From: [EMAIL PROTECTED] (Michele Beltrame)
Subject: Gateway problem!
Date: 10 Feb 1999 23:29:19 GMT
Hi!
I've got a Linux machine which, beyond being a gateway for Internet
access for a couple of LANs, is also a gateway *between* these two
LANs, which are 192.168.2.0 and 192.168.12.0.
However, I've got a problem: although the hosts on network 12.0 are
able to see (=ping) the hosts on 2.0, there are problems in data
exchange between them (the two hosts are a bar code reader on net 12.0
and an AS/400 on network 2.0). Routing in the Linux machine (which is
2.10 and 12.10) is configured as follows:
=====
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
194.184.42.7 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0
192.168.2.0 0.0.0.0 255.255.255.0 U 0 0 0 tr0
192.168.1.0 192.168.2.221 255.255.255.0 UG 1 0 0 tr0
192.168.0.0 192.168.2.221 255.255.255.0 UG 1 0 0 tr0
127.0.0.0 0.0.0.0 255.255.255.0 U 0 0 0 lo
192.168.12.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
0.0.0.0 194.184.42.7 0.0.0.0 UG 0 0 0 ppp0
=====
and ipchains (I use kernels 2.2.x):
=====
Chain input (policy ACCEPT):
target prot opt source destination ports
DENY all ----l- layla.leader.florida.it anywhere n/a
DENY all ----l- layla.leader.florida.it anywhere n/a
DENY all ----l- 127.0.0.0/8 anywhere n/a
DENY all ----l- 127.0.0.0/8 anywhere n/a
Chain forward (policy DENY):
MASQ all ------ localnet/24 anywhere n/a
MASQ all ------ 192.168.2.0/24 anywhere n/a
MASQ all ------ 192.168.12.0/24 anywhere n/a
Chain output (policy ACCEPT):
=====
If I remove the chain for 12.0, also ping stops working. I *think* that
maybe linux masks (masquerades) the addresses of 12.0, as ipchain is
launched with "-j MASQ" option and this may cause some problem: maybe
192.168.12.100 (bar code reader) is masqueraded as 192.168.2.10 (the
Linux host). How can I do (supposing that the problem is this) to
configure ipchains so they only forward WITHOUT masquerading between
nets 2.0 and 12.0? I tried removing the "-j" parameters, but then nothing
works anymore (ping also stops working).
Suggestions?
Thank you in advance, Mickey.
--
Michele Beltrame
[EMAIL PROTECTED]
http://www.io.com/~mick/
------------------------------
From: [EMAIL PROTECTED] (Villy Kruse)
Subject: Re: /lib/modules/preferred deleted on boot ????????
Date: 12 Feb 1999 12:17:06 +0100
In article <[EMAIL PROTECTED]>,
Robert Lynch <[EMAIL PROTECTED]> wrote:
>If you are on a RedHat system, the problem is that the 2.0.34 modules
>directory does not contain a "dot" file:
>
>.rhkmvtag
>
>You can make this file by just doing:
>
>cat /proc/version > .rhkmvtag
>
>This will cure the problem.
>
Unless we are talking about RH5.1 first edition, which haven't yet
implemented the file .rhkmvtag
The old problem was that /etc/rc.d/rc.sysinit at some point did a
'depmod -a preferred' . The script does check for the existense of
the name 'preferred' but the broken version would do 'depmod -a preferred'
in both cases. Changing this to 'depmod -a' would fix the problem.
This is an extract of the current and fixed version of rc.sysinit.
echo -n "Finding module dependencies... "
if [ -e /lib/modules/preferred ]; then
depmod -a preferred
else
depmod -a
fi
echo "done"
Villy
------------------------------
Crossposted-To: comp.os.linux.setup,comp.os.linux.x
From: "ne..." <[EMAIL PROTECTED]>
Subject: Re: Need help with hosed rpm database
Date: Fri, 12 Feb 1999 11:19:47 GMT
On Feb 12, 1999 at 05:21, Dennis Putnam eloquently wrote:
>On Thu, 11 Feb 1999 15:40:02 -0500, Ben Russo
><[EMAIL PROTECTED]> wrote:
>
>>
>>get the packages that should be installed.
>>
>>do a
>>
>>rpm -Uvh --replacepkgs --replacefiles name-of-package*.rpm
>>
>>Then if you want to erase them you should be able to rpm -e them.
>>
>>-Ben.
>>
>One additional point with the XFree86 upgrade. As far as I can tell no
>one has put that into an rpm file. I downloaded all the tgz files and
>installed them using the instructions but nothing updated the rpm
>database for that package. When I do a verify many files come up with
>a checksum error and the database only has 3.3.2 in it. Anyone know
>where I can find an rpm file of XFree86 3.3.3.1 so I can install it to
>get the database correct?
They are on the RedHat errata page.
http://www.redhat.com/support/docs/rhl/rh52-errata-general.html
Else try http://rufus.w3.org/linux/RPM/
>Am I being overly concerned with the integrity of the rpm database in
>the first place?
Either that or you are not querying it properly.
--
Two wrongs don't make a right, but three lefts do.
------------------------------
From: gus <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.admin,comp.os.linux.networking
Subject: Re: hacked login
Date: Fri, 12 Feb 1999 10:33:50 +0000
Reply-To: [EMAIL PROTECTED]
Graffiti wrote:
>
> In article <[EMAIL PROTECTED]>, Rafael Marcus <[EMAIL PROTECTED]> wrote:
> >Someone probably hacked the login program in my Linux system and it
> >doesn't record te remote host address for a remote telnet login in the
> >"wtmp" file. I recompiled the login program but the problem persists.
> >Any idea what else I should check to solve the problem.
> > Thanks
> > Rafael.
>
> Never, ever, ever, *ever* continue to use a compromised system if you can help it.
> Re-install.
> No, I'm serious.
> You never know what *else* was compromised. libc? syslogd? cc? ftp?
> You might try to grab an update to replace binaries that you *think* are
> compromised, but that's completely useless if, say, ftp was compromised to grab
> a trojan'd binary. Unlikely, yes. Impossible? No.
> Re-install.
> And when you do, pick different/new passwords for *all* your accounts.
>
> -- DN
If you have an RPM managed system, RdHat, SuSE, etc, then you can use
flags to RPM to determine which, if any, files have been modified. I
forget the flags, but try it. IIRC it is -v or something, for verify. It
can do MOD5 checksums, etc.
All the best
gus
------------------------------
From: gus <[EMAIL PROTECTED]>
Subject: Re: deleting kernel sources okay?
Date: Fri, 12 Feb 1999 10:39:52 +0000
Reply-To: [EMAIL PROTECTED]
Gordon Vrdoljak wrote:
>
> Hello,
> I recently upgraded my kernel to 2.2.1. I was wondering after
> successfully installing and
> running the new kernel - can I now delete the source files under:
> /usr/src/linux?
> I don't have a ton of room on the hard disk. What would be the
> implications of removing
> the source too? Would anything break or would it give me problems
> later?
> Could I also delete the old source files in:
> /usr/src/linux2.0.36
> and
> /usr/src/redhat?
> Any comments appreciated. Please send them to:
> [EMAIL PROTECTED]
> Gordon.
Having looked through all the already posted replies, you may just want
to do "make clean" in all the /usr/src/linux-* directories which will
remover all files which are the result of compiles, without removing the
source itself. This is a compromise solution in that it will clear about
half of the used space. It is a very safe alternative.
Just a thought.
gus
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
