Re: April 15, 2000. Time to end income tax once and for all?

James Rogers (jamesr@best.com)
Sat, 11 Jan 1997 22:02:26 -0800


At 09:52 PM 1/11/02 -0500, Michael Lorrey wrote:
>Chris Hind wrote:
>>
>> Fast approaching is tax day and I'd like to remind everyone that when 2000
>> strikes, machines across the globe will go haywire creating wrong time/date
>> stamps. Wouldn't this be the optimum time to end it once and for all?
>>
>
>Your intrepid reporter from the future here with a commentary on what
>were widespread rumors prior to the millenium on the so called "reset
>effect". Many not very well informed individuals proclaimed that since
>MS-Dos was written with a clock that ended in 1999, that everyone's
>clocks would reset to 1900 upon the millenium (even though the real
>millenium wasnt to happen for another year, Jan 1, 2001), causing
>widespread computer chaos.
>
>However, as you can see in the date and time of this email message, that
>with the fixes in Win95, as well as other operating systems and BIOS
>upgrades that occured in 1995-96, that this prediction never occured.

You are very mistaken.

The systems people are worried about are not PCs. The *real* problem only
really manifests itself in old mainframe systems and older, large database
systems (like the IRS?). In some mainframe cases, the problems are caused
at a very low level and would essentially require replacement.

The biggest problem is software where dates were hardcoded with a two-digit
year. In some systems, this was the only way it could be done at the time.
The result is that *a lot* of program logic will become invalid when the
year 2000 rolls over. It has already been accepted that for some
companies/agencies and systems, there is physically not enough time left to
do the necessary conversions/modifications. The number of lines of affected
code is absolutely enormous. There aren't enough available programmers to
tackle the problem, and companies that were created specifically to
specialize in this area are making MILLIONS.

The year 2000 will bring a lot of big database and mainframe software
systems to their knees, due mostly to the unresponsiveness of organizations
to the immediate nature of the problem. Whether or not the IRS will be one
of the victims remains to be seen. In any case, the world will not come to
an end as we know it. To most people, it will most likely manifest itself
as occasional short-term inconveniences, inefficiencies, and inaccuracies
that will disappear after a year or two.

You will probably see the same type of problem occur a few decades into the
21st century when 32-bit epoch time rolls over, but probably not as serious.

-James Rogers
jamesr@best.com