I've been mercilessly !#/bashing away at this linux box now for nearly two
years.

In that time I've destroyed the bootblock, corrupted the ext2 filesystem,
wiped and reinstalled the partition during an unfortunate foray into
ReiserFS, mistakenly downgraded components, thoughtlessly screwed with file
dependencies, messed up hundreds of symlinks, totally screwed file
permissions, altered the entire contents of the /etc directory to an extent
that can only be described as "mind buggering", config'ed installs with the
wrong prefix (and ended up having multiple orphaned copies of different
versions of the same packages), recompiled the kernel on a seemingly hourly
basis ... and after all that ...

It's still standing !!!

Ok, so it's moved from one disk to another then back, but it's still the
same installation of RH.

And it's not just "getting by" either.

It's got a highly adaptive self correction facility ... but unlike the
bollox that comes with Win2K this facility has two legs and a hairy chest
... me. 5 minutes reading a FAQ, HOW-TO or man page and a few of seconds of
vi'ing/sh'ing scripts and the system is back to normal. Oh, with the help of
some very expensive backup software of course ... namely - this one line
script: "time nice -n 19 find / -mount -depth -print | nice -n 19 cpio -ovaH
crc | nice -n 19 bzip2 | nice -n 19 pgp -fes u...@ftc.com u...@ftc.com >
/mnt/roswell/linux-backup-19.10.2001.cpio.bz2.pgp" and two others - one for
verifying and one for restoring. God knows I've maliciously tried to destroy
this beast, but its not having it, I give up. Its invincible.

I've learned more about ... well ... everything, in the past two years using
linux, than the rest of my adult life put together, the point is that you
can screw linux up, like you can with anything else, but unlike certain M$
products, you never need to reach the stage where "a new clean install is
necessary because there is actually no way to fix the problem" - everything
is "fixable" under linux.

I find it ironic that nobody anywhere - including M$ support and the whole
M$ ng community can explain to me why I get ...

'The configuration information of the performance library
"D:\WINNT\System32\w3ctrs.dll" for the "W3SVC" service does not match the
trusted performance library information stored in the registry. The
functions in this library will not be treated as trusted.'

... type errors (not just that library - but about 20 others - not all
related to IIS) scattered throughout my Win2k event viewer, nor how to fix
them. There's a single document on M$ support that basically explains one
circumstance under which this situation can arise, but gives no true
explanation and no permanent solution. The circumstance doesn't match mine
and the "workaround" .. well .. doesn't. Anyway with Apache on the linux
box, why the hell would I want IIS ( presumably standing for too (two)
shit ) infesting my system. IIS Sux ... it's official - we have photos.

[H]omer
A member of "Script Kiddies R Us" since 1792

***
cat /lost+found/sock | /dev/washingmachine
***

 [snip about invincible Redhat installation]

Funny you should bring it up; just today I was talking with a buddy of
mine and she told me the story of her box that is running Slackware 2.
It began life in '96 with a 1.2.13 kernel, and it runs a 2.4.x kernel
now complete with glibc 2.2 and everything in between.  She killed the
box once doing an upgrade of libc (and ever since, she's kept every
version of libc inbetween -- its probably the only Linux box in the
world capable of running *any* Linux binary dynamically linked).

No packages, no Xconfigurator, nada.  Amazing.

:)  Yep, she did that too.  Her system has >100gb of storage now, but
its running on an older K6 (lost her posh machine in the dot-bomb).

 [snip]

I can add another hundred, or so, like-wise error "messages" to that
one.  My favorite is a WINS error we got and the error message in the
event viewer went like this (I kid you not):

 "WINS encountered an error.  The error is in the data."

Nice.

 [snip]

--
It is financially more expensive to go to prison than to attend Harvard.
Craig Kelley  -- kellc...@isu.edu
http://www.isu.edu/~kellcrai finger i...@inconnu.isu.edu for PGP block

On 18 Oct 2001 23:12:56 -0600, Craig Kelley <i...@inconnu.isu.edu> wrote:
|
|  [snip about invincible Redhat installation]
|
|> It's still standing !!!
|
| Funny you should bring it up; just today I was talking with a buddy of
| mine and she told me the story of her box that is running Slackware 2.
| It began life in '96 with a 1.2.13 kernel, and it runs a 2.4.x kernel
| now complete with glibc 2.2 and everything in between.  She killed the
| box once doing an upgrade of libc (and ever since, she's kept every
| version of libc inbetween -- its probably the only Linux box in the
| world capable of running *any* Linux binary dynamically linked).
|
| No packages, no Xconfigurator, nada.  Amazing.
|
|> Ok, so it's moved from one disk to another then back, but it's still the
|> same installation of RH.
|
|:)  Yep, she did that too.  Her system has >100gb of storage now, but
| its running on an older K6 (lost her posh machine in the dot-bomb).
|
|  [snip]
|
|> 'The configuration information of the performance library
|> "D:\WINNT\System32\w3ctrs.dll" for the "W3SVC" service does not match the
|> trusted performance library information stored in the registry. The
|> functions in this library will not be treated as trusted.'
|
| I can add another hundred, or so, like-wise error "messages" to that
| one.  My favorite is a WINS error we got and the error message in the
| event viewer went like this (I kid you not):
|
|  "WINS encountered an error.  The error is in the data."
|
| Nice.

My fave "error" message is from win2k. A service hadn't started at boot
and I was investigating why. I attempted to start the service manually.
The error that this generated was something like:

"The service is not running because no attempts have been made to start
it". (paraphrased, from memory, as I am not near the box in question)

Remember, this error was *caused* by an attempt to start the service!

--
Reverend Paul Colquhoun, ULC.    http://andor.dropbear.id.au/~paulcol
     Asking for technical help in newsgroups?  Read this first:
        http://www.tuxedo.org/~esr/faqs/smart-questions.html

news:slrn9svfd9.ojr.postmaster@tux.andor.dropbear.id.au...

Yeah I remember some error message oddities from long ago, not as good as
those above, but:

"Setup Error - Mouse driver not found. Mouse not detected. Click OK to
continue"

and

"Error Deleting File - Cannot delete 016: There is not enough
free disk space. Delete one or more files to free disk space, and then try
again".

Then there was that classic "Office Configuration Box" spoof that has a
config box with options like "Hidden Settings - manufacturers' default
settings (not to be edited). Endurance options - Crash every xxx
hrs/mins/secs, crash after every xxx bytes of unsaved changes. Save
Options - Create incredibly large files, Allow me to carry on typing during
autorecovery saves. Autorecovery takes - bloody ages/forever/out the whole
system. Other options - Piss me off by applying the same stupid attributes
to every table picture and object I insert, Annoy me with that sodding paper
clip - constantly/when I least expect it."

Ah class, pure class.

Hmmm, thats put me in the mood for an Easter Egg hunt ...

[H]omer

In <slrn9svfd9.ojr.postmas...@tux.andor.dropbear.id.au> Paul Colquhoun posted:

From a Cadnetix CAD workstation circa 1986:
"Fatal error #xyz: We should never get here."

See, Microsoft *does* learn from history.
Now *what* they learn ...

--
| I'm old enough that I don't have to pretend to be grown up.|
+----------- D. C. Sessions <d...@lumbercartel.com> ----------+

Or From most PC's when switched on with no keyboard connected
'No Keyboard detected, Press F1 to continue'

In comp.os.linux.advocacy, Craig Kelley
<i...@inconnu.isu.edu>
 wrote
on 18 Oct 2001 23:12:56 -0600
<m1ofn4xlev....@inconnu.isu.edu>:

Who *writes* this stuff??  :-)

To be sure, this might be engineering codewords for "The error is in
the data *framing*" or "data+checksum are not consistent" or "CRC
failure" or "your network is total crap and the gremlins have decided
not to give you the time of day until fed a chocolate chip cookie;
thank you; please drive through". :-)

But I do wonder -- and probably so does the poor user who sees this.

[snip]

--
ew...@aimnet.com -- please put all clear concise error messages here: [_]
EAC code #191       99d:05h:09m actually running Linux.
                    Hi.  I'm a signature virus.

The Ghost In The Machine wrote:

But how is this any better than the cryptic error messages I often see
issued by FreeBSD and linux?

People who live in glass houses...

There is one fundamental differece - Linux and BSD users can grep the
source to find the error msg and determine from that what will lead to it.

When you see a cryptic error message on closed source stuff, you're
in trouble.

--
Mark Kent
                                               Take out the ham to mail me.

In comp.os.linux.advocacy, drsquare
<nowh...@nowhere.co.uk>
 wrote
on Sun, 21 Oct 2001 11:31:02 +0100
<sm65ttkhqvcn09jm0id5j6s9i89mq3s...@4ax.com>:

"Splunge!" comes to mind.  That's about as cryptic as they come.

To be sure, I found that in the kernel *first*.  It has something
to do with Adaptec, if I remember correctly -- probably when
the card enters an unexpected state and sends a response that
confuses the module.

--
ew...@aimnet.com -- insert random misquote here
EAC code #191       102d:20h:03m actually running Linux.
                    Be paranoid.  Everyone else is.