If Windows Had Never Happened ...

By Peter Coffee  |  Posted 2005-11-13

If Windows Had Never Happened ...

In Fritz Leibers 1958 short story, "Try and Change the Past," a man gains unauthorized access to a time machine.

He tries to rearrange events to prevent his own death from a bullet between the eyes.

He gives up when he sees himself die of a micrometeorite impact in precisely the same place. He decides that events have an inertia that people cant hope to overcome.

Trying to imagine a world without Microsoft Corp.s Windows invites the same suspicion of inevitability. Anything so transforming and so pervasive was perhaps simply meant to be—but perhaps not.

Windows certainly wasnt greeted as inevitable when it was proposed in the autumn of 1983. It seemed an outrageously ambitious goal to construct a universal software layer on top of the IBM PC platform, infamous for its low-level quirks.

Click here to read more of Peter Coffees views on Microsofts Windows empire and its effect on innovation.

That was a time when a PC planning department could order a copy of Microsofts breakthrough Flight Simulator by calling it a "graphics compatibility diagnostic suite"—an insiders joke, but not a lie because thats the role it played by exploiting low-level hardware behaviors.

Likewise, Digital Equipment Corp. and Texas Instruments Inc. pitched their DOS-compatible (but not IBM-compatible) PCs by touting their superior machine-specific versions of Lotus Development Corp.s 1-2-3 integrated spreadsheet, enabling use of more memory—up to 896KB on DECs Rainbow—by using less fragmented memory maps.

It was a competitive environment where hardware makers sought a hardware edge.

Nor was Microsoft yet the feared competitor it would become. An editorial cartoon in the earliest days of this publication (then called PC Week) depicted a boxing glove labeled "IBM" punching a caricature of Bill Gates, with an observer of the battle saying that Gates needed to learn who really set the industrys standards. As it turned out, there were plenty of learning opportunities to go around.

Next Page: The power of printers and mice.

The Power of Printers

and Mice">

The win in Windows

If Microsoft hadnt done Windows, the most immediate need would have been for someone to solve the printer problem. By the time that Windows emerged, a DOS application might come with more floppy disks of printer drivers than of application code.

Application-specific printer drivers werent entirely bad for users. Even after Windows-based word processors had been on the market for years, DOS-based WordPerfect was still far faster in a mail-merge benchmark during a PC Week Labs Shoot-Out.

WordPerfect was slow to fade from print-intensive environments, no doubt in part because it put out hard-copy documents better than early versions of Windows printing services.

Most application vendors would likely have tired, though, of staying abreast of printer vendors product introductions. They would have sought a universal protocol for document description.

Hewlett-Packard Co.s LaserJet dominance could have made it a standard setter, but other printer makers would not have readily crowned it as first among equals. Adobes PostScript might have taken on that role to an even greater extent than it has—recall that Display PostScript, co-developed by Adobe with NeXT Computer Inc., came along a few years later.

eWEEK Labs analysts recall their experiences testing earlier versions of Windows. Read more here.

Its ironic, then, to find Microsoft compelled by customers to support Adobes PDF in the forthcoming Office "12." Perhaps, in the realm of document description and printing, Windows will turn out to have been a mere (if lengthy) detour.

Another important contribution of Windows has been in device support—specifically, the combination of USB with plug and play. This was no easy accomplishment: Witness the crash heard round the world when Bill Gates Chicago Comdex demo of Windows 98 went blue-screen upon connecting a USB scanner.

Things are better today. Consider the simple choice of a comfortable and task-appropriate pointing device at a time when most laptops already come with built-in touch-pads.

Third-party mouse or trackball devices, even cordless ones with radio transceivers that plug into USB ports, are promptly and quietly recognized by Windows XP—without even the minor inconvenience of loading drivers from a CD. This is no small thing.

Its not clear, though, that the industry needed Windows to achieve either technical capability or marketplace ubiquity of broad peripheral device support.

Every Macintosh, beginning in 1984 when Windows 1.0 was still a year away, has offered some kind of device bus. Only a Mac devotee could recite the overlapping timelines of AppleTalk, SCSI, ADB, USB and FireWire availability, but each of those technologies enabled a broad array of third-party device choices—and many were not specific to the Mac.

Next Page: Microsofts mastery of innovative development tools.

Microsofts Mastery of Innovative

Development Tools">

Developing a relationship

Finally, theres the matter of applications.

Its impossible to overstate the importance of Microsofts mastery and promotion of innovative development tools, plus ardent and effective courting of application developers, in achieving and maintaining the dominance of Windows.

The key question, however, is what the application development space would look like if Windows had never emerged.

When Microsoft started talking about Windows, Apple was a few years away from the release of its groundbreaking HyperCard—but it wasnt until 1990 that Windows 3.0 was actually useful, with another year before Visual Basic 1.0 made Windows development feasible for users.

By 1991, NeXT Computer had introduced an object-oriented development platform; Apples HyperCard 2.0 was already a year old and proving a fertile environment for innovative applications.

Will the launch of SQL Server 2005, Visual Studio 2005 and BizTalk Server 2005 be one of Microsofts last big launches? Click here to read more.

Visual Basics model was a GUI with behaviors built behind it, while HyperCards was an extensible data structure with powerful but approachable GUI tools.

The HyperCard model might have been a better choice for the baby-duck imprinting of a generation of programmers. HyperCard "stacks" could easily have become an approachable model for distributed platforms and concurrent processing engines, and there were several competing Windows and cross-platform development tools that resembled HyperCard by the time Visual Basic arrived.

Visual Basic had the edge, though, in exploiting the Windows APIs—and that was the advantage that mattered.

If there had been no Windows, wed still be printing, wed still be plugging and playing, and wed still be developing applications. Windows was brilliantly positioned, however, so as to replace the problem of how to do things on PCs with the problem of how to do things on Windows.

Thats a problem that Microsoft always solved better than anyone else.

Peter Coffee can be reached at peter_coffee@ziffdavis.com.

For reader response to this editorial, click here.

Check out eWEEK.coms for Microsoft and Windows news, views and analysis.

Rocket Fuel