Klarinet Archive - Posting 000773.txt from 1998/08

From: "Tim Roberts" <timr@-----.com>
Subj: Re: [kl] Re: Y2K
Date: Fri, 28 Aug 1998 10:08:42 -0400

I started to send this privately, and I probably should have continued to do
so, but I finally decided this was an important point to clarify. I didn't
realize when I started that I was going on a crusade....

On 26 Aug 1998 12:00:41 EDT, <GrabnerWG@-----.com> wrote:
>In a message dated 98-08-21 17:13:44 EDT, timr writes:
>>
>> Yes, but there has been far too much hysterics generated about this topic.
>> Let's take a reality check for a minute.
>
>Tim, I'm sorry, but your "reality check" is based on your opinion only.
>
>Again.....I will state that the biggest problems will be in equipment that
>has embedded microchips. Some of these microchips are 286's....and
>I'm sorry, but they will not work, probably will not boot, after 12/31/98.

Unless you're talking about a chip other than the Intel 80286, this is just
exactly the kind of hysterics to which I was referring. The 286, like EVERY
x80 and x86 processor ever produced, has no inherent date capabilities at
all. As such, all the 286s in the world will be completely oblivious to the
change of century. Even 20 year old 8080s will continue to work right across
the century boundary without a hiccup.

I suspect what you MEANT to say is that there are older SYSTEMS (which might
be based on 286s) which have software that will not handle the date change
correctly, but that is VERY different from what you said. The chips will not
stop working. They don't care.

All the computers in the world are going to boot on 1/1/00 (assuming they
were bootable on 12/31/99!). Some of them have operating systems which will
read the date incorrectly during their boot process, but the computers are
NOT going to roll over and die. Older PCs might think the date is 1/1/1980,
but they'll work. Inconvenient, not catastrophic.

>Many organizations
>realize that they will not even be able to Identify all the equipment that
>they use, let alone replace, test, and put back into production, where
>obsolete microchips are still in use.

Obsolete microchips are NOT the problem. Microprocessors almost never have
time and date functions built-in. The problem is old SOFTWARE.

>Example.....a manufacturing company in Milwaukee where I have done some
>consulting work will have their business computer Y2K compliant by next
>spring. However they are just NOW realizing how many machines on the
>shop floor rely on old microchips THAT SIMPLY WONT WORK.

Of COURSE they will work! What part of them do you think is going to stop
working? To rephrase what I said in my last ranting, if you never tell a
piece of machinery what the date is, then IT DOESN'T CARE ABOUT THE CENTURY
CHANGE. It CAN'T care: it won't know! It will continue to run happily
along, forever. Very few pieces of factory automation eqiupment care about
the date, because they don't NEED to. This is just business common sense.
It costs money (and cycles) to add and maintain calendar capabilities in a
piece of equipment. If it's not needed, it's not added.

>Cars, toasters, telephones, toys, radios, alarm clocks, medical
>equipment......they all can have microchips!!!!!!!!!!!!! Think about it.

Yes, they all have microchips, but NONE OF THOSE THINGS CARE ABOUT THE DATE!
THAT'S my point. You don't tell your toaster what the date is. Your radio
doesn't know what year it is. Your doctor's EKG machine doesn't know what
year it is. And if they don't know the year, they CANNOT fail when the year
rolls from 99 to 00. That is not opinion, that is fact.

--
- Tim Roberts, timr@-----.com
Providenza & Boekelheide, Inc.

-------------------------------------------------------------------------