"Those who cannot remember the past are destined to repeat it."  Santyana

During the rise of the mainframe industry from 1951 through about 1980, Sperry “Univac” (it went through various name changes reflecting organizational shifts over the decades) often had faster computers and more efficient operating systems than did IBM. 
 
Sperry, true to its scientific and engineering origins, had a much narrower focus on large scientific, airline and military customers.  Commercial customers were not a major target for Sperry.  This allowed Sperry to streamline its operating systems, rather than load them with a lot of features necessary to accommodate a wider customer base, but that were just overhead for any given segment of that base.
 
IBM, on the other hand, true to its primarily business origins, designed its operating systems to accommodate the widest possible swath of customers, from commercial to scientific.  Its operating systems were loaded with features, many of which were not used by any given segment of its customer base.  It was a one-size-fits-all strategy.
 
These two principle players in the history of the computer industry “duked  it out” to see which strategy would be left standing when history had its say.  It turned out that history had a different view of things.
 
While Sperry, especially in the beginning, had superior technology, it wasn’t technology that selected the winner.  This bout was fought in the arena of the maturing market driven economy following the Second World War, although that wasn’t necessarily understood in the 1950’s through the 1970’s by everyone.
 
IBM was definitely superior in the marketing field.  For every salesman Sperry would put on an important customer account, IBM would place ten.  IBM had a document publishing division second only to the U.S. government’s publishing capability in size.   IBM’s strategy was to always be first in marketing, whereas Sperry (Univac) counted on the customer to recognize technological prowess.
 
The users of Sperry systems had to be much more technically savvy.  Sperry did not subscribe to hand-holding, nor did it have to with its chosen customer bases.  It was joked around the campus that “Real men don’t need software development tools or documentation” like IBM provided.  Many of the large accounts modified Sperry’s software extensively, at the gut level.  Lufthansa (German Airlines) is a case study. 
 
With upwards of eleven huge Sperry 1100 series systems, Lufthansa had a lot of clout with the company.  They were instrumental in the development of the Sperry USAS (Univac Standard Airline System), a huge software system targeting airline operations.  When Lufthansa wanted special features added to this “standard” software and supported by Sperry, they oftentimes got it.  It was a “tail wags the dog” story.
 
But this dog had several tails: Cathay Pacific (Hong Kong airlines), Sabena (Belgian airlines), Air France, QANTAS (Australian airlines) … - you get the picture.  These were all significant accounts, and they were all wagging the dog.  USAS as it matured became more like IBM’s software because these giant customers ran their operations differently – sometimes significantly so.  This became a real overhead burden for Sperry, but it couldn’t afford to drop the load.
 
That was one problem for Sperry’s strategy.  Another was that, having less than ten percent of the market, no independent software to speak of was ever developed for the company’s systems.  The company made moves belatedly to broaden its customer appeal, but it was hesitant, halting, and sometimes missed the sense of the marketplace altogether. 
 
As an example, it tried at one point to promote development with a development center for third party developers in Minneapolis.  This didn’t survive for long.  Sperry shot itself in the foot by charging potential developers for the use of the center.  You can guess how many takers that strategy produced.
 
There was little interest in developing independent software for such a limited customer base.   When the unavailability of developmental systems (unlike the IBM world, where there were lots of systems that were not huge and locked up in a glass case in Frankfort) is added to the equation, the interest dropped to zilch.  Sperry recognized that belatedly, but its solution to make the development center a profit center was misguided.
 
So what, you say?  Well, they say history repeats itself.  In today’s world Microsoft’s Windows operating systems take a lot of flak.  There is a small but persistent choir singing the praises of Apple and UNIX.  But, perhaps they are a little like Sperry insofar as they are technologically excellent but play to a small segment of the market.
 
You can’t beat Apple for graphics and game playing, for example.  To be sure Apple has that market segment in its pocket.  But how big a market segment is that.  Is history repeating itself?  Is technology superiority, in and of itself, just not the deciding factor in today’s PC market?  It would seem from the evidence over the last two decades that “Those who cannot remember the past are destined to repeat it”, albeit with tremendous enthusiasm.
 
And UNIX.  Here is an example of Sperry’s problem with independent software development.  Ok, ok, I know that has been rectified to some extent over the last few years, but it is still a marketing deficit for UNIX.   Now UNIX does seem to be superior to Windows at a certain technical level.  It is for that reason an operating system of choice for servers, network work and those applications where the “end user” is savvy (see Lufthansa above).  And, for those applications, it is superior.  But it is a day late and a dollar short on the common man’s user interface (see Sperry’s development center above) when compared to Windows.  And, that is simply THE determining factor in the software wars!
 
Like it or not, Microsoft is today’s IBM on the software side of computer industry.  Both Apple and UNIX are improving in marketability all the time, but so is Windows.  And Microsoft has the “throw weight” to keep them “niched” for the foreseeable future.
 
I am not ecstatic about some of the features of Microsoft Windows (XP Professional), but for availability of software to accommodate my needs as a “general purpose end user”, I will take it over Apple.  And I need to get the job done, so playing “techy” with UNIX or LINUX is not in the cards. 
 
Now, another argument made by the aficionados of  Apple and UNIX is that one does not encounter the virus and other problems that one does in the world of Windows.  Yes, but why is that?  The claim to tighter code as the reason is a thin argument to me.  I think it might have more to do with the target base, or dearth thereof, for “independent developers” of mischief software.  I’m not convinced, although I am not sure, that design is the difference.   It could be a market driven, "bang for the buck", ROI (return on investment) consideration rather than the technical difficulty of breaching the Apple or UNIX design and coding.