What Next?

More Delays for Microsoft Windows

BYTE Magazine

February 1985

In early October 1984, Microsoft Corporation announced that it was postponing the introduction of its long-awaited Windows software-integration package until June 1985. Leo Nikora, Windows product-marketing manager at Microsoft, said that the company was undertaking "a major redesign," in part because Windows' code currently takes up too much space and also because several functions are not running fast enough.

As recently as this spring, Microsoft was hoping to achieve a minimum recommended system size of 192K bytes. The most current technical information available on Windows states that Windows together with the operating system occupies 156K bytes of memory; thus the currently recommended 256K bytes leaves only about 100K bytes for applications software—not much by today's standards.

Nikora said that almost all of Windows is now written in the C language and that Microsoft plans to rewrite as much as half of the program in 8088 assembly language. Apparently Microsoft is happy with the window-management functions of the program but feels that text management is inadequate. Nikora said that Microsoft expects a twofold increase in text performance after the code is rewritten, although he feels that the performance of the product is already satisfactory on the IBM PC AT.

Microsoft is clearly worried that its decision to delay Windows will lead to a negative attitude in the marketplace. "We have to be careful that Microsoft doesn't get the reputation of giving up in the face of TopView," claims Nikora, referring to IBM's entry in the window-management fray.

He also maintains that Microsoft's decision to delay the product introduction hasn't led to mass desertions on the part of companies developing applications software for Windows. On the contrary, he said that there was a general feeling of relief that they were being given more time to get their applications ready for market.

Microsoft is also looking for a way to differentiate Windows from TopView, and the company appears to have found one because the current version of TopView is designed for a character-based display. This will, at least temporarily, be a selling point for Windows, which functions only in a bit-mapped environment.

Will Windows face the same fate that befell Visi On? Nikora says that he is certain that it won't—his evidence is the fact that a number of the manufacturers of IBM PC-compatible computers appear to have a sizable stake in the success of Windows. Still, Microsoft is starting over again after investing more than a year in attempting to develop a user interface for the IBM PC.

Copyright 1985