The Secret Sabotage Behind Your Worst DOS-to-Windows Memories | Comuniq
This site requires Javascript to work properly. Please enable Javascript in your browser.
-->
/Technology
Gadgets, apps, inventions and everything that involves the world of technology. Share your links here and see what the guys have to say in the comments.
Members: 17 Join
Moderated by: mozzapp
The Secret Sabotage Behind Your Worst DOS-to-Windows Memories
Manon_code<br>1783666631<br>[Technology]<br>0 comments
If you started out on a PC back in the era of the black screen with green lettering, you know exactly what I'm talking about. Switching from DOS to Windows wasn't a smooth upgrade, it was almost a collective trauma, the kind everyone who lived through it still brings up with a slightly haunted look. But why exactly did it hurt so much? The answer has less to do with silly nostalgia and more with badly resolved technical decisions, broken promises, and, get this, at least one corporate scandal that only came to light decades later.
MS-DOS started life as a project literally called, with zero irony, "Quick and Dirty Operating System." It was bought in a rush by Bill Gates from a barely known little company, Seattle Computer Products, and hastily licensed to IBM for their first PC. It was already old the day it was born, honestly. Historian Jimmy Maher describes how the system was stuck with a 640 KB memory ceiling, had commands that were far from intuitive, and no real multitasking, which forced every software company to reinvent the wheel on its own, even for something as basic as printing a document ([The Digital Antiquarian, filfre.net](https://www.filfre.net/2018/06/doing-windows-part-1-ms-dos-and-its-discontents/)). So when Windows showed up promising windows, a mouse, and nice little icons, it wasn't competing against a mature system. It was trying to patch over a foundation that was already considered archaic on the day the first IBM PC went on sale, back in October 1981.
And the first version of Windows was a lot thinner than people tend to remember. In November of 1983 there was, no exaggeration, exactly one real application you could actually run on it, a little paint program that looked a lot like MacPaint. Everything else was skeletal: a calculator, a very basic text editor. Windows 1.0, which didn't even ship until 1985, was still using block based graphics, nothing like the smooth bitmap graphics Apple already had on the Lisa and the Macintosh. The "future" Microsoft was selling was, in practice, several years behind what the competition was already showing off ([The Rise of Microsoft Windows Part 1](https://retrotechreads.substack.com/p/the-rise-of-microsoft-windows-part), [Microsoft Windows version history, Wikipedia](https://en.wikipedia.org/wiki/Microsoft_Windows_version_history)).
There's a funny detail in all of this. For a while Microsoft actually believed Windows should be mouse only, no keyboard at all. They backed off because they realized nobody was going to drop DOS's text commands overnight, and, supposedly, because the US military complained it would be tricky to operate a mouse while "driving a tank." That one course correction alone cost somewhere between three and six months of delay.
But the part almost nobody knows about, and honestly the part I find most telling in this whole story, is the AARD code affair. Behind the scenes of the Windows 3.1 beta, in 1991, Microsoft was facing a real competitor in the operating system market: DR-DOS, from Digital Research, which already had built in disk compression and multitasking before MS-DOS 5.0 even shipped. And Microsoft's response wasn't just to compete on the product. Engineers at the company built in a piece of code, internally nicknamed "AARD" after the programmer's own initials, Aaron R. Reynolds, that checked whether Windows was running on top of DR-DOS and, if it was, triggered a fake, alarming error message. The code was XOR encrypted and self modifying, a technique that today we'd associate far more with malware than with commercial software ([AARD code, Wikipedia](https://en.wikipedia.org/wiki/AARD_code)).
Internal emails that surfaced years later, during the United States antitrust case against Microsoft, showed executives openly discussing how to trip up DR-DOS. One of them wrote that the plan was to detect the rival and simply refuse to load, with an error message along the lines of "invalid device driver interface" ([The Register, 1999](https://www.theregister.com/1999/11/05/how_ms_played_the_incompatibility/)). Microsoft did disable that check in the final retail version, but left the dormant code sitting right there, reactivatable by flipping a single byte. Sit with that for a second.
Caldera, the company that later bought the rights to DR-DOS, sued Microsoft over it. The case wasn't settled until 2000, and the agreement stayed sealed and secret until 2009, when it finally came out that Microsoft had paid 280 million dollars to make...