Copy Link
Add to Bookmark
Report
Net Vandal 3
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*VOL:1* NUMBER 3.000199, Jan. 2, 1995 ALL WRONGS DESERVED TORONTO
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VAND
ET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDA
T.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL
.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|
VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL||
ANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||
NDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||N
DAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NE
AL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET.VANDAL|||NET
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
An exercise in irritainment and mathematical mind-blanking
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NET.VANDAL vol 1, number 3.000199
|||||||||||||||||||||||||||||||||
Brought to you by: The Most Reverend Lucifer Messiah
CONTENTS
ontent
nten
..
* F E A T U R E *
Where Are We Going?
- The Most Reverend Father continues bashing iNTEL
* G O S S I P - R U M O U R - F E E D B A C K *
- Work, Stealing a Byte,
* C A B A L T R I C K S *
- The iNTEL UnrePentium Story
* S P O R T S *
- Windows 3.1 Redefining Mathematics
"We are Intel of Borg. Accuracy is irrelevant. Mathematics
is irrelevant. Your division will be estimated"
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* F E A T U R E *
Where Are We Going?
- The Most Reverend Father continues bashing iNTEL
In the last issue, I wrote an article entitled 'Where have we been?'. This
time around, I thought I would look a little into the future, and ponder
what it has to offer. At this point, we need not look too far ahead. The
answer is staring at us right in the face.
In light of iNTEL's most recent misfortune, looking at the future of home
based computing has become a very easy task indeed. In case you haven't
heard, (and if not, where have you been?), the IDIV function, responsible
for dividing floating point numbers, in the Pentium processor is bugged,
and known to give faulty results.
Nobody actually noticed that there was a problem, until some poor hapless
soul posted the following message on a small local BBS in Dildo,
Newfoundland:
---
Hello,
Could anyone help?
I upgraded my motherboard yesterday from a 486DX4 to a pentium, but now
I am not longer able to start WinWord. It terminates exectution with
the message "wrong Windows version: 3.095"
Any suggestions?
---
Okay, maybe that isn't the way it all started. No matter. Somebody screwed
up, and nobody is willing to take the blame.
What is this with iNTEL? My last article dealt with how long they have
taken to reach standards already reached some years back, and now we see
that when they try to invent their own standards, everything runs amuck!
When news of this bug got around, iNTEL offered (under extreme pressure, we
can be assured) to replace the chips. Arguing that this bug is so minor
that the average user wouldn't see its effects, that only one in 27 billion
or so divisions will produce incorrect results (an arguable estimate, BTW,
since it was worked out on the above said Pentium chip), iNTEL would replace
your Pentium chip _if_ you could prove that you truly needed it. Maybe the
iNTEL developers should only recieve their pay if they can prove that they
really need it. This has apparently changed, yet again, and iNTEL will
replace the chip no questions asked (using a $2000 credit card bond, and
a very difficult process to bear, but that is another subject unto its
own)
To add more dry wood to the fire, it appears that iNTEL has known about this
bug since the summer. In what was probably a move to save face, iNTEL chose
to hide the problem, and then when confronted by it, low-played its
severity. At the very least, their extreme reluctance to replace the
defective chips is a sign of the quality of service we can come to expect
from iNTEL in years to come.
Many people think the hubbub isn't worth raising sweat over. Lets make a
parallel to the story. Say Toyota releases a car, in which they find that
once in every 27 billion revolutions of the axel, the wheels just might fall
off. What would the public reaction be if Toyota decided to hide the
problem, and then upon its exposure, show similar reluctance in replacing
the defective parts? What if the axel was replaced only if you prove that
you will reach that 27 billionth revolution at some date in the near future?
Or that you actually _do_ need 4 wheels to drive on, in lieu of the fact
that motorcycles only require 2 of them. Chances are, Toyota would have
quite a problem on their hands. The buying public wouldn't stand for such
sloppiness and such obstinance. Toyota would be forced out of business.
But this would never happen. In the auto industry, recalls or replacements
have always been immediate. None of this hiding behind shadows, hoping that
the wheels will hold on for enough years that customers never find out what
the industry was keeping from them.
It is, of course, quite interesting to see exactly how severe this problem
is. For this bug has been so lowplayed that it will reportedly affect a
system only one out of so many billion potential passes. Taking this into
account, it should also be an equally complex matter to reproduce the
problem. The mathematical equations would have to be great.
Apparently not.
Pentium users, try dividing 5505001 by 294911. Your trusty dusty Pentium
will tell you that the answer to this simple division is 18.66600093. The
real answer? How does 18.66665197 sound? That wasn't so hard afterall.
And worse, several other similar equations, each very simple, have been
developed which give incorrect results. If it is this easy to trip off,
then why does iNTEL deny it so?
The real problem was not with the bug itself. It was with iNTEL's refusal
to respond to it in a professional manner, or even with some form of
business ethics. In this day and age, where technology is outracing our
capacity to keep it all together, are we losing touch of values that we once
held onto with all our might?
One parting shot. I am easily reminded of the most recent Star Trek movie,
Generations, where Data is completely frazzled by the same emotion chip that
had previously led to his own brother's demise. Do you think, could it be
that the chip was a Pentium-based processor?
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* G O S S I P - R U M O U R - F E E D B A C K *
Many apologies go to the numerous people who wrote in wondering where the
hell NET.VANDAL had disappeared to. I went through a busy period at work,
and had to put my time to other use. Things are back to normal now, and
that's only so far as normal gets in a mag like this, and so here we are
with a shiny new edition of NET.VANDAL.
Let's see, what happened while we waited for this issue...
Landon Dyer played the dating game, telling me that he has Byte magazine #2,
and had #1 before it was stolen. (I doubt it will be found on the "Back
Issues" page!). Landon says that the 32032-based computer in the
advertisement I ranted about in NET.VANDAL vol 1 issue 2 ran like a pig,
albeit, elegantly. I'm not sure how an elegant pig runs, but I'm sure it
would be interesting to watch.
The lack of segmented memory was part of that elegance, to which I firmly
agree.
Another reader, who wished to remain anonymous (how goofy), asked me to put
together an issue about cryptography. I've begun compiling information for
just such an issue. If anyone is interested in contributing, please send me
mail at lucifer@csis.pcscav.com. Good information on this topic is rather
hard to come by, so any submissions will be greatly appreciated.
That's about it. It's a new year now, and lord knows there will be all
kinds of new NET.VANDAL type material invading our homes.
(See, in refusing to take the glaring opportunity, I have avoided printing
out yet another blank, unemotional wish for a happy new year, or even a
belated Happy Ho Ho.)
Chao, kids.
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* C A B A L T R I C K S *
- The iNTEL UnrePentium Story
Keeping in the spirit of things, I compiled a few of the better Pentium
jokes, and other interesting responses, found on the net lately. Remember,
To Err is human. To really fuck it up, you need a Pentium. Enjoy!
Q) How many pentium designers does it take to change a lightbulb?
A) 1.9999999943424 but iNTEL says that's close enough.
Q) Why did they name the 586 'Pentium'?
A) Because when they added 100 to 486, they came out with 586.465798654.
Q) Why did the Pentium cross the road?
A1) It was trying to stay on the sidewalk.
A2) Because someone threw it.
Q) Anyone heard about the rename of the RU486 abortion pill?
A) They'll rename it RU-Pentium since it prevents cell division to occur
correctly. Still under development at this time, the pill is not
reliable, causing spontaneous abortions only once in 27 billion
attempts. The company, however, feels it is good enough to sell to
anti-abortion groups and third world nations who have less need
for accuracy.
Q) What's the difference between an intel salesman and a used car salesman?
A1) A used car salesman KNOWS when he's lying.
A2) It's easier to tell when a car salesman is lying.
I'd be interested in knowing how this affects the average nuclear family.
What was that? 2.200011 children, or somewhere thereabouts? Close enough.
Coco Pentium machine Winner
------------------- --------------- ------------
Accurate division Who knows? Coco
Doesn't run Windows Runs Windows Coco
Costs $10 Costs $2500 Coco
CPU can't ... CPU can be used as coffee mug warmer Pentium
The good news is that pi, now becoming an integer, will be easier
to work with. The bad news is that now we'll have to watch Beverly
Hills 90209.9878733469...
One group calculated that Pentium errors should occur every 27 years.
Funny, I got every 26.999999789 years.
iNTEL's new logo: United we stand, Divided we fail.
Some TV shows that have sprung up in response to the Pentium fiasco:
- My 3.0154 Sons
- Beverly Hills 90210.999534
- Babylon 5.000000016
- Deep Space 8.7
- 3.076's Company
- Death of the Net, .GIF at 10:58:02
- Thirty.000458-Something
- Hawaii 50.2
- Rocky 7.1
TOP TEN NEW INTEL SLOGANS FOR THE PENTIUM
--------------------------------------------------------------------------
9.9999973251 It's a FLAW, Dammit, not a Bug
8.9999163362 It's Close Enough, We Say So
7.9999414610 Nearly 300 Correct Opcodes
6.9999831538 You Don't Need to Know What's Inside
5.9999835137 Redefining the PC -- and Mathematics As Well
4.9999999021 We Fixed It, Really
3.9998245917 Division Considered Harmful
2.9991523619 Why Do You Think They Call It *Floating* Point?
1.9999103517 We're Looking for a Few Good Flaws
0.9999999998 The Errata Inside
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* S P O R T S *
- Windows 3.1 Redefining Mathematics
Sure, the Pentium may be bad, but there is another candidate for the
Mathematical Stupidity Prize: Windows Calculator.
It took a lot of work for Microsoft to make Windows something that people
would consider worth buying. I still remember all the hooplah regarding the
Windows "look", and how it infringed on Macintosh copyrights. It took until
version 3 before anyone would even look at it as a worthy computing
environment.
Now, after some years of people using this "environment" (I'm trying really
hard to avoid calling it an operating system), it has now become an office
staple. There probably isn't a modern office setting in North America that
doesn't have at least one Windows based PC.
But it, like the Pentium processor, is completely inane (or is that insane?)
when it comes to doing simple mathematics.
I came across this 2 years ago, when I used the Windows Calculator to
decrypt a text string in the 'Natas' computer virus (Written by NuKe). The
algorithm was simple. Each letter in the message was XOR'ed by 1. After
going through each character one at a time, and noticing that the result was
a completely incoherant jumble of letters and characters, I did the math by
hand. Oddly, done by hand produced results that could be understood.
Windows Calculator had correctly decrypted some of the letters, but not
others. The ratio of hits and misses was really sad. Needless to say, I
was quite glad my own system wasn't running MS-anything!
I called Microsoft, and they told me there was no such bug with Windows
Calculator, and that I must have been doing something wrong myself. I can
easily understand how a person who can do binary arithmetic by hand might not
have it in them to operate a calculator properly. Gee, thanks guys.
But that's not all. Binary arithmetic is one thing, but Windows Calculator
has been proven so useless that it can't even subtract properly!
Ironically, this bug also produces floating point errors, but much more
severe than with the Pentium IDIV problem.
We all know that 0.11 - 0.10 equals 0.01. We know this, because even
Windows Calculator will tell us that. Thank god, or at least Bill Gates,
for that.
But we also know that 1.11 - 1.10 also equals 0.01. We can even verify this
with Windows Calculator, can't we? Err, no, we can't. Windows Calculator
reports that it equals 0! Even weirder, if you multiply that 0 by 100, you
will get 1! This also works with 2.11, 3.11, etc.
Now let's add another decimal place to this game. Good old MicroSoft tells
us that 3.111 - 3.110 equals 0.00099999999999. I must have looked like
quite an idiot for all these years I thought it equalled 0.001.
But everything gets even stranger and stranger as you play around with the
decimal points. Adding yet another decimal point, and changing the pattern
a little, lets do 3.0001 - 3.00. Anybody with higher than a grade 4
mathematics will tell you that the answer is 0.0001. Not Windows
Calculator. The answer it spews back at you is 0.0001000000000002.
And to top it all off, the errors in Windows Calculator seem to be rather
inconsistant. Why can it subtract 0.10 from 0.11, but it can't subtract
1.10 from 1.11?
I find it very distressing that in a world of inferior PC's invading the
market, we allow ourselves to be sold such inferior software as well.
END ISSUE ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Addendum:
The reason so many Bible students allow themselves to become deceived is
that they study the word of God, and not the works of God.
Personal mail to:
lucifer@csis.pcscav.com
all wrongs deserved, granted "post-everywhere" status by the teenage buddha.
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
<<<<<<<<<<===NET.VANDAL===>>>>>>>>>>
will appear whenever I get to it
(good forgeries are welcomed)
How do I JOIN NET.VANDAL?
************************
Join the list at any time by sending a "SUBSCRIBE NET.VANDAL"
command in the body of a message to net.vandal-request@hack.pcscav.com
How do I LEAVE NET.VANDAL?
*************************
Leave the list at any time by sending an "UNSUBSCRIBE NET.VANDAL"
command in the body of a message to net.vandal-request@hack.pcscav.com
How do I SUBMIT INFO to NET.VANDAL?
**********************************
Send your articles addressed to net.vandal@hack.pcscav.com
<<<<<<<<<<===NET.VANDAL===>>>>>>>>>>
--
The Most Reverend Father Lucifer Messiah
"If you act like a dumbshit, Subscribe to NET.VANDAL
they'll treat you as an equal" Send "SUBSCRIBE" to:
- J.R. "Bob" Dobbs net.vandal-request@hack.pcscav.com
Patty scored us 40 pounds of dynamite on Sunday. If they give us an
openning then we'll blow them all to hell.
--------
For more information about this anonymous posting service,please send mail
to remailer@csua.berkeley.edu with Subject: remailer-info.
This message contains automatically generated keyword blocks
that have been designed to resemble a threat. These blocks
are not a statement of intent by the remailer operator or anyone else.
--------
To respond to the sender of this message, send mail to
remailer@soda.berkeley.edu, starting your message with
the following 7 lines:
::
Response-Key: the-clipper-key
====Encrypted-Sender-Begin====
MI@```%5^&2?(E<U9BPG1]W7L'=>KH_Q;E8W+E;IXU=*:[SXO!SD7]U">92TN
M3>54U<KKEMBX-07*OV(U$:9XCY5Z!S#0;CG=H,:]3C9%7IVL2Z:66&N
