Re: (not entirely...) OT: OPINION... chicken entrails, runic stones, and crystal balls... WAS CoBOL moved to OO

From: Peter E.C. Dashwood (dashwood_at_enternet.co.nz)
Date: 01/02/04 

Date: Sat, 3 Jan 2004 11:34:09 +1300

"Donald Tees" <donald_tees@nospam.sympatico.ca> wrote in message
news:t1iJb.14155$Vl6.3179913@news20.bellglobal.com...
> Judson McClendon wrote:
> > "Howard Brazee" <howard@brazee.net> wrote:
> >
<snip>

> Why would anybody do
> something like year end tax statements one transaction at a time, or be
> interested in a financial positions on a minute by minute basis 20
> minutes before midnight on new-year's eve?
>

In 1904 the Editor of the London Times, when shown the newly invented
telephone, said: "Why would I buy that when I can send a copy boy?"

If transactions attain speeds that are unthinkable by today's standards (and
the indications are that they will...) then there is no reason not to have a
real time system.

Such a system models reality. As events occur in the Real world, they are
mirrored on the system. It makes sense.

The objections raised here have studiously avoided (or maybe not seen) the
post I made that says transactions can be scheduled whenever you want them
scheduled (in addition to mirroring random events as they happen).

There is NO problem with achieving synchronicity in a transaction based
real-time system. It can achieve everything currently achieved by Batch
Processing, and then some. The only problem is that (at the moment) the
technology is not quite powerful enough to ensure that ALL transactions
(both random and scheduled) are processed immediately.

My speculation is predicated on that changing.

Pete.
 main argument for having a real time system is that it models reality

> Donald
>
>
>