Re: year 2000

Don J. S. Adams ( (no email) )
Sun, 27 Sep 1998 00:22:23 -0500

Hi Ken,

> Now.. how does this topic relate to Keelnet list in anyway
> shape or form? Agreed that it is a problem and BIG one
> at that....

Considering that its related to computing science I felt it was relevant.
Especially when you consider that Keelynet is made possible by
the internet (linked computer network) that could come to a halt
within about 500 days. If the worst case scenario does occur...then perhaps
a known timeline where we all need to work together to get some real answers
fast might be useful in knowing about. Right now the many bright minds
of Keely are connected, while they are, perhaps some great good can be produced.
In a year and a half they may no longer be connected should the systems go down.
In my mind this is VERY relevant but perhaps the possible demise of such communication
services isnt important enough or relevant enough to mention on Keely.

> So.. I guess this is fair game here.. my two cents..

any thing I say is always fair game <g>!

> 75% of the major programs (financial) where programed
> in Cobol where the code only allowed 8 bit year fields..or
> ascii.. two digits. This is the major problem... reading the
> compiled Cobal program and decifering it.. or decompling
> as they say. Then rewriting and compiling the code.. doing
> all this while the system is in use. Now.. how many people
> now currently know or use for that matter.. Cobal! Id you
> do know it.. and can decompile/compile.. and write in it...
> then you can have a job most anywhere for $100,000/yr!

true... however this problem also relates to many other lingo's
not to mention psuedo-lingos, i.e. auto-scripting in biz apps.
Another really nasty problem though is the chips, wherein coding is
hard wired or burnt in... that means they have to be manually fixed
and cannot be as easily 're-coded' as software code in writable storage
mediums like drives etc...

> The major problem will steam from dynamic searches and
> not static events.

think about the loop back self running diag and main. tests that
PLC's perform or have performed on them. You get a large clustering
of nodes on an industrial WAN that regulate power distribution...
well you only need a critical number of these to get 'stuck' in their
'overdue' servicing flag to have the whole grid jeopardized.

As for fincanial institutions...they're screwed. Citycorp, if I understand
correctly has about a half a billion lines of code that still have to be searched.
No way they will even come close by 2000. Aussie govt., doesnt know if they even
be able to have essential serives online for sure.