% fortune -ae paul murphy

Windows vs. Unix: the resolution

Be it resolved that Microsoft's success is a marketing, not a technical, phenomenon.

There's a two step argument for this:

  1. first it's possible to argue that Microsoft has never yet introduced a significant technological innovation - that everything they've done with respect to technology has been either a copy or an extension, admitted or otherwise, of someone else's innovation or a technologically illogical response to a marketing need;

  2. and secondly it's possible to argue that Microsoft's financial success has been entirely a return to monopoly - at first IBM's monopoly on the hearts and minds of data processing managers, then in response to illegal monopolistic tactics by Microsoft, and most recently in response to a "Microsoft effect" distorting both the information available to decision makers and the valuations they attach to that information.

The point of step one is to show that Microsoft's dominance isn't due to its ability to generate and popularise new technology - and the broad strokes on this are pretty clear.

To see this at the most superficial level, just compare Microsoft's primary OS interface products to those from Apple - Vista is just the latest and possibly the most blatant example: every Microsoft interface "innovation" since Compaq added hard disk support to MS-DOS has duplicated something Apple already had.

MS-DOS itself was derived from CP/M via Patterson's quick and dirty DOS; the Windows 3/9 series ran as DOS applications on the SunView model, and the current NT series developed from VMS - so lots of creative coding, but no substantive technical innovation, in any of those.

The Office components, of course, are like IE: traceable to acquisitions - it's the key reason they've had so much trouble achieving file and command level integration.

Microsoft maintains an official list of corporate acquisitions back to 1994 (for earlier ones see: the "Nearly whole Microsoft Catalog" ) but doesn't match that list to derived products. Thus to know for sure whether they've originated anything you'd have to look closely at the background to every single one of Microsoft's 6000 or more software products (and their 500 or more hardware products), but from aircraft simulation to database management and ERP, its easy to see that all the big products came from third parties pushed out of the way by a Microsoft check - so my guess is that if you did the detailed work you'd find that over 99% of Microsoft's money comes from selling stuff derived from work somebody else did first.

The point of part two is to ascribe their success largely to monopoly rents (i.e. unearned returns) rather than legitimate sources like better design, efficient resource management, product integration, or financial accumen.

For "phase one" it's kind of obvious: back in 1984 people had real choices: Apple offered the MacXL - $5,495 at list complete with 1MB, a 16/32bit, an 8Mhz (5 in early releases) MC68000, a floppy drive, 10MB disk, an integrated screen, a GUI, and a suite of core applications; Tandy and a host of others offered Unix variants for less, and IBM offered the PC-AT at $5,500 at list for a 4.77Mhz, 16bit chip with 128K, a 10MB disk, and nothing more than BASIC - but Apple sold 50,000 units while IBM cranked out nearly four million.

A mystery - until you look at the one characteristic separating Mac buyers from PC buyers: the Mac people didn't have pre-existing systems organizations and the PC buyers did. Thus it wasn't Microsoft's monopoly back then, but their early success was due to a monopoly nevertheless.

There's no mystery for "Phase Two" either: in the Jackson round of anti-trust trials Microsoft ended up convicted of numerous criminal counts before the government's lawyers (led by the same David Boies who seems to be having some difficulty proving the obvious in the SCO case) proved unable to demonstrate that Microsoft, whose revenue rose from around $9 billion to $23 billion during the trial period, benefited from its criminal activities during that period.

"Phase three" is, however, a very different kind of animal. There are people (some at the Department of Justice among them) who think Microsoft has simply continued its monopoly practices, but my guess is that they've become considerably more cautious about flaunting the law since the Jackson trial.

So if the major driver for Microsoft's continuing near monopoly isn't coming from illegal practices, and it isn't coming from product quality either, then what's driving it?

My answer starts with the idea that Microsoft has generally focused on making money for the guy selling Microsoft because that guy's commitment is far more important to Microsoft than end-user satisfaction. Thus the ideal Microsoft product is one that's easy to sell but requires lots of relatively simple maintenance -basically something that costs as much as the customer will pay, but doesn't work too well.

Follow that up by recognising that the first two phases left Microsoft with a functioning monopoly, and the picture you see is one in which sellers are active, committed, and enthusiastic about selling while buyers are turned off, apathetic, and largely mis-informed about their options because everywhere they turn, someone is lying hard to sell them the same stuff.

"The thing about Windows, that you don't seem to understand," one of my friends said to me the other day, "is that a lot of people make a lot of money off Microsoft."

I don't doubt that for an instant - and that's the bottom line on protecting an established monopoly against obviously better, obviously cheaper, alternatives.

Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.