Year2000.com Announcement List, Special Mailing,
September 2, 1998
by Jon Huntress
You are receiving this e-mail because you subscribed to
Year 2000 Announcement List.
This mailing is to provide you the first of our three part
special editorial coverage of the "Texas Year 2000 Workshop"
on the Year 2000 Computing Crisis that took place in
Houston, Texas on August 10, 1998.
Our next mailing will be our regular newsletter, which
should go out within the next few days.
Our special coverage in this issue is sponsored by Fidelity
Technology Solutions. They provided the resources necessary
to cover the conference, so please read their insert below
and check out their web sites when you have a chance.
If you want to leave the list, please do not "reply" to
this message. It won't work. Instead, use the form at:
year2000.com
to get off the list. If you don't have web access, please
see the note at the end of this message to unsubscribe using
e-mail.
------------The Texas Road Show: Embedded systems: Dave Hall
In order to raise year 2000 awareness and give local program
managers and lease holders information on dealing with the
expected problems, the state of Texas sent a team of program
managers and presenters to address the cities of Austin,
Houston, Harlingen, Dallas, Lubbock and El Paso. Shannon
Porterfield, the Statewide Year 2000 Director in the
Department of Information Resources ran the show. Because
the issues addressed affect organizations and municipalities
beyond the State of Texas, I am sharing some of the coverage
of this event with this list over three mailings. The next
report will be on testing and the final report will be on
municipal and county issues and readiness.
The main presenters at the Texas Y2K Road Show were Dave
Hall from the CARA Corporation and Don Estes from 2000
Technologies Corporation. Dave Hall talked about finding,
assessing and fixing non-IT (embedded) systems. I will
cover Dave Hall's presentation in this article and Don Estes
in separate reports because both men had some very
interesting and important things to say about the year 2000.
Dave Hall started by restating what so many have said: "How
could just two digits cause such a big problem?" He pointed
out that the problem lies at the very foundation of the
logic of the way the computers were designed. The fact that
computers were programmed with 2 digit date fields from the
sixties in order to save memory is mentioned in most of the
articles on the year 2000 problem, but most non-IT people
still don't understand how 2 little digits can be such a big
deal. At the risk of being redundant, I will offer one more
explanation for your friends or business associates who
still don't quite get it. Computers do math. They know how
to add, subtract, multiply and divide. I remember the very
first IBM computer I ever worked with in college, the first
thing to do was to load the stack of cards that would teach
the computer how to do those four math processes. The year
2000 mistake is that we forgot to tell the computers that in
certain circumstances, and for our own convenience, we want
00 to be larger than any number between 01 and 99. So when
the computer subtracts a bigger number from a smaller number
to find someone's age for example, it will get a minus
number, (wrong answer) or it can crash. One way to fix it
could be adding the code to tell the computer what to do
when it gets a minus number in an age calculation. This has
to be done every place where a time calculation is made. In
an age calculation, a minus answer is impossible, which
might cause the program to quit. But this would probably
not be the case with an interest calculation, where the
computer could easily calculate that you owe 99 years of
interest on your savings account. No rules of logic were
violated when it figured the interest. What happens depends
on the way the programmer wrote the code for the program
years ago.
Dave asked the audience if they had ever known a programmer
who knew the very best way of doing something and no
argument could persuade them otherwise. There was laughter
at this because everyone in the audience knew many
programmers who fit this description. Dave painted the
picture of thousands of individual programmers, absolutely
sure of the right way of doing something, programming all
the computers their way. He then pointed out that modern
computers, sometimes as small as a fingernail, have more
computing power than the mainframes of the sixties, but they
act just like those mainframes. An embedded system acts
just like any other computer and the programs for embedded
systems are just as individual and just as unique as the
programs for the mainframes, and there are many more of them
out there. He went on to say that the year 2000 embedded
systems problem is not a technical problem. "Show me a year
2000 impact and I'll fix it! Easily!" He said. "But I want
you to show me where they are! This is an inventory
problem." Then he gave the audience the numbers. Out of
40,000,000,000 chips, between 1% and 10%, or 400 million to
4 billion could have a date problem. Finding them is the
problem.
He covered the three kinds of ways that computers can break.
The first is when they quit, and this is actually good
because now you know you have a problem and a pretty good
idea of where to look to find it. The second way of
breaking is when they make large observable errors. This is
good too because the errors are obvious and will be seen and
fixed. The problem is with the third way computers can
break and that is small accumulating errors. Dave asked how
long it would take you to notice if your cruise control
added one mile per hour to the speed of your car every
minute. How about if the program at your bank deducted 1%
from your account every day? "How often," he asked the
audience, does someone go ahead and do something because the
computer told them to do it?" I was reminded of the recent
story of someone who got a tax bill from the IRS for more
than a billion dollars. They took the bill to the IRS and
showed the clerk the bill. The clerk checked it on their
computer and told the stunned couple that the bill was
correct! The clerk believed the computer because the
numbers in the computer were the same as the numbers on the
tax bill! How many people out there still believe the
computer because it is a computer? Want to bet that number
takes a nose dive in the next few years?
"How many of you think you would get the correct change at
McDonalds if the computers go down? Dave asked. It
reminded me of a time I went to a McDonalds in Santa Fe.
The bill was for something like $8.79 and I gave her a ten
and four pennies so she could make the change a dollar bill
and a quarter. (I hate pennies) The girl looked at the
money then at her cash register, then back to the money,
frowned, turned and yelled to the others behind her, "Does
anybody here know math?" |
