Nice Theory!
By Michael A. Marullo, Contributing Editor
Once upon a time, there was an operating system called “DOS.” Although DOS wasn`t all that elegant or robust, it was very easy to understand and use, readily available and perhaps best of all, inexpensive. As we all know, DOS made quite a name for itself, and within a relatively short time, it became the operating system of choice for a majority of personal computers being sold around the world.
Along with the exuberance and euphoria surrounding this phenomenal proliferation, however, came the notion that DOS could (and should) be adapted for virtually any application that could be imagined — including real-time applications. Never mind that DOS is “single-tasking” and that a real-time environment was probably far from Mr. Bill`s mind when he wrote the code for his multi-billion dollar legacy.
For years, legions of clever and innovative programmers attacked the frailties and limitations of DOS with a vengeance, confident in their ability to “software around” virtually any obstacle that might present itself and miraculously transform DOS into a real-time contender. But alas, their valiant attempts to defy its inherent limitations merely proved that DOS was, at best, a poor real-time platform. Not surprisingly, companies that had invested heavily in DOS were eager to embrace the long-awaited DOS successor: Microsoft Windows.
Indeed, Windows was a big improvement over DOS … no argument there. Besides, there was already this huge installed base of “performance-challenged” real-time systems in the field just waiting for an operating system that could update an alarm display in less time than it takes to drink a cup of coffee. But alas, Windows was not the panacea either! It still had limitations in demanding real-time applications and fell far short of expectations in the “upwardly mobile” department, harnessing users to the scalability boundaries of PC architecture. Not to worry though. Like Henry Ford, Mr. Bill had a better idea!
Enter, Microsoft Windows/NT … now there`s a panacea, right? Runs on a PC, runs on a workstation, runs on whatever floats your boat. Just a few little kinks to work out and NT (“New Technology”) would surely cure all of our problems — or so we thought. But, it soon became obvious that NT was in danger of flunking its first “real-time” test: Availability! So Mr. Bill came up with a gap filler product, originally dubbed “Chicago” (i.gif>., Windows 95), to quiet the rising tide of developers charging that what NT really meant was “Not There.”
Now, here it is 1996 and NT has indeed become a reality. It does lots of clever and wonderful things and even does some real-time stuff fairly well. However, there seems to be a confluence of people who think (wish?) that NT immediately displaces and destroys all of the other real-time operating system alternatives that are out there in one fell swoop.
As nice as NT is, it is useful to realize that although Mr. Bill`s market objectives may include SCADA and other real-time applications, rest assured that he is far more concerned about whether NT satisfies the K-Mart and Wal-Mart corporate computing networks than he is about whether you can click a relay within two seconds from a hundred miles away. Nonetheless, I`m sure that there are those who still believe that NT was created specifically for real-time applications and that it will solve all of the problems they have faced in the past, and then some. To those people, I would offer another possible definition for NT: Nice Theory!
Author bio:
Michael A. Marullo is managing director and CEO of cfar international, a global automation marketing and technology consulting firm. Questions or comments may be directed to P.O. Box 641177, Kenner, LA 70064-1177; phone: (504) 733-5504; fax: (504) 733-0754 or via e-mail to [email protected].
