Programming Windows: Pre-History (Premium)

In the 1970s, the software wizards at Xerox Parc pioneered the future of computing by implementing a graphical user interface, or GUI, that allowed users to manipulate on-screen objects—windows, buttons, and other widgets—using a mechanical device called a mouse instead of just the keyboard. In an age of text-only displays, Xerox had accomplished something that was truly magical.

But the firm failed by not successfully commercializing its inventions.

It did so by showing these inventions in 1979 to Apple, which promptly took Xerox’s ideas, expanded on them, and developed its own GUIs. First for the Lisa, which was overly expensive, aimed at the business market, and a failure. And then later for the Macintosh, which likewise failed in the market until desktop publishing and the first laser printer rescued it from oblivion in the late 1980s.

Apple announced and then released the first Macintosh in 1984, but the product had been in development for at least two years before that. (An earlier, entry-level computer that was also named Macintosh notwithstanding.) Its innovations, especially in the personal computing space, span a long list. Instead of a text-based display, the Macintosh used a bitmapped graphics display. Individual applications appeared in floating, resizable, and overlapping windows, instead of fully occupying the screen. The interface was optimized for manipulation via mouse, not the keyboard.

As a major vendor of developer tools for microcomputers and a blossoming software application provider, Microsoft played a very visible role in the Macintosh launch. The firm, at the time, supported every personal computing platform available with its software development tools, especially BASIC, and was quickly expanding the reach of its productivity software as well. So, it quickly ported both Multiplan and Word to the Macintosh, and then later released Excel first on Macintosh.

“Macintosh allows us to write software that is significantly easier to use,” Microsoft CEO Bill Gates said in 1984, referring to that system’s GUI. “What it delivers for its cost is really great. If this machine can’t make it, I don’t know what can.”

Gates, as excited by anyone about the future on display with this platform, predicted that half of Microsoft’s 1984 revenues—as much as $50 million—would come from the Macintosh.

That didn’t happen. But less well-known than Microsoft’s early support of the Macintosh, perhaps, was that the company also worked secretly with Apple to help it define the Macintosh user interface. It defined such widgets as radio buttons and dialog boxes, and it suggested the double-wide outline around default buttons. This gave it both an outsized influence on the new system and important experience with GUIs, which were still quite new at the time.

This partnership made sense: Gates really did believe that the Macintosh would be successful and would drive future revenues. And because it had elected to support this unknown new platform, its experience building high-quality apps gave the firm insights into how the interface should improve.

But Gates was also a cagey businessman and knew from his earlier dealings with IBM and the first PC that owning the software platform would be even more lucrative. So, Microsoft announced in 1983 that it would release an operating environment for MS-DOS that would bring a GUI and mouse support to its own platform as well. It promised Apple that it would not infringe on the Macintosh user interface, wink, wink. Apple, to its credit, was correctly and immediately suspicious.

Regardless, the success of Window was not guaranteed, and it would enter a market crowded with competitors. There was the Apple Lisa and the then-secret Macintosh, of course. And many software vendors, including most notably VisiCorp, had released or at least announced various DOS-based graphical shells of their own in a bid to capitalize on the GUI trend and make PCs more efficient. Indeed, the very existence of this competition on PCs is what triggered Microsoft’s preemptive announcement for a product that, at the time, didn’t even exist. Microsoft was worried that it would lose its foothold on the PC and it had a lot of catching up to do.

Worse, Windows would be among the least sophisticated of those shells because of Microsoft’s secret deal with Apple. It would not visually resemble the sophisticated Macintosh user interface, and it would lack some of that system’s more advanced and useful features, like support for overlapping windows. Instead, on-screen windows could only be tiled onscreen. At least at first. The vision for Windows 1.0, which was originally named Interface Manager, was somewhat humble.

But then two years went by without a release. Two. Years.

During the intervening years, so much time elapsed that Windows was, according to Gates co-author Stephen Manes, responsible for the now-common industry term vaporware. But the Macintosh had launched and, after some initial excitement, had bombed because of its constrained resources, high cost, incompatibility (with Apple II, III, and Lisa, as well as with competing platforms), and lack of a high-quality application library.

So, Windows seemed to have a chance. That it was made by the creators of MS-DOS seemed to ensure that it would achieve at least some level of success, and would perhaps even define the market for GUIs on PCs going forward.

But when it was finally released on November 20, 1985, Windows 1.0 impressed no one.

More soon.