There Will Always Be An Operating System
Yet another story about the End Of The Operating System As We Know It, based more on wishful thinking than cold reality. Hope springs eternal, more now with the Browser as the ultimate program, instead of another dig at Microsoft.
Neil McAllister’s blog has good stuff, timely and useful. But not everyone can bat a thousand, and his recent entry, Are Operating Systems Doomed?, attributed too much to browsers, and blaming the operating systems for problems outside their control.
As was entertainingly demonstrated in the movie Tron, the management of machine resources will always require a central arbiter. The alternative is for each application do this directly, like the bad old days of having to code for various printers and video drivers, et al.
The economic advantages of eliminating an OS have drawn a lot of smart people and piles of venture capital for some time now, with nothing to show for all the effort to date.
Browsers, as well as run time emulators such as Adobe’s AIR or Sun’s Java, are neither autonomous nor sovereign, but applications that operate under the aegis and control of an operating system, be it Apple, Bea, Linux, Pick, Solaris, Windows, or the myriad Unix flavors. Even .NET is a collection of wrappers for Win32 API calls, done with varying degrees of efficacy and quality.
Neil’s article does illuminate the problem of applications’ look and feel, and how they are under designers’ control, often at the expense of ease of use for ordinary mortals otherwise known as customers. These myriad skins are much like the Babel of CP/M, and the highly fragmented mobile applications and services markets.
The other issues mentioned in Neil’s article, like automated and incremental updates, tighter security, reducing training costs in corporate environments, etc., have been solved some time ago by successful software vendors. None of these are unique to standard desktop software; web-based systems have these problems as well, in addition to a few of their own making.
CP/M’s Tower Of Babel was the result of three main items. The first was incompatible disk formats – every manufacturer had its own.
The second was the use of both the Intel 8080 and Zilog’s Z80, with different instruction sets and vastly different architectures, resulting in incompatible software or dual code bases that were impossible to keep in sync.
The third was that CP/M, perhaps the first occurrence of open source, could be and was modified by the hardware manufacturers, and sometimes even the software publishers.
This Babel was a main contributor to Microsoft’s rise, reflected in a somewhat famous argument between Bill Gates and Gary Kildall of DigitalResearch. Gates felt there would only be one OS; Kildall thought there would be many.
All of this has been documented many times and is now part of the industry’s lore. Perhaps the best general overview of all of this was done in a PBS documentary, and a companion book, Triumph of the Nerds: A History of the Computer by Robert Cringely.
I’d always thought that Microsoft was first and foremost a management and marketing company, certainly not a technology leader. It seems to me that they’d do better by getting out of technology and managing their investment portfolio.
Microsoft’s’ moment has passed. But the widespread adoption of Microsoft’s operating systems, including DOS 6.2 which is still being bought in OEM quantities, has resulted in a uniform platform that has created a number of huge markets. Like DOS 6.2, Windows will not go away for a long time, not because they are better, but because they’re here. I’d like to think that someone will do the same thing for the fractured mobile market soon.
A Word From Our Sponser
Stop Toxic Email From The Bad Guys Before It Can Get Close To Your PC! Our 
ePAL for email security is a barrier between the outside world and your regular email program
Tags: Adobe, Application Development, Google, Java, Microsoft, open source, Platforms, Sun Microsystems, User experience, Web applications