From 1984 to 1987, I worked at Microsoft. For about half that time, I was a member of the Windows 1.0 development team. I joined Microsoft straight out of college. Although I’d had several summer jobs doing programming since high school, this was my first permanent job in the software industry. Here's my version of the Windows 1.0 story.
Working at Microsoft in those days was somewhat like living in a college dorm. Because so many of us had relocated to the Seattle area straight out of college, most of our friends and acquaintances were our coworkers. This, combined with the pressure to get this already late product out the door resulted in us all working lots of nights and weekends. Frisbee, playing guitar, and throwing bouncy balls were all popular hallway activities. With few responsibilities outside work, work was fun, although it could definitely wear you down over time.
Even the hiring process was fun: Send resume. Get letter back. Send some sample source code. Get phone call inviting me out to Seattle for a visit. Fly out to Sea-Tac and go through a full day of Microsoft's famous interview process. Get job offer at the end of the day. Try not to show too much excitement on the spot. Spend another day checking out Seattle. Fly back to school to finish senior year. Get mailgram from Bill Gates (i.e. the Microsoft HR department) congratulating me on my job offer. Get another letter from HR with a copy of Time with Bill Gates on the cover.
(I didn't happen to be a beneficiary of one of the best Microsoft interview perks, a trip to dinner in the San Juan islands with Microsoft programmer and private pilot Peter Stevens. Later on, as an employee, I was lucky enough to go on a couple of these with other interviewees.)
At that time I joined, Microsoft headquarters was a building on Northup Way in Bellevue. All software development was done in this building, although growth soon required relocation of the other groups to some buildings on the other side of Highway 520. One of the best features of the Northup Way building was its proximity (across the parking lot and down a few steps) to Burgermaster (“Dine in your car”!), source of excellent burgers and shakes. The Microsoft phone system was programmed with a speed-dial to the Burgermaster number.
It was an exciting time to be at Microsoft and on the Windows team. Of course by the time I got there in June 1984, Windows had already been under development for some time. It was announced in November 1983 at Comdex in Las Vegas, complete with a working demo. The demo was impressive enough to convince most people that Microsoft was close to delivering the actual product, but in reality it was just a prototype and the real product was far from done (although some of the code from the demo may have actually survived to make it into Windows 1.0).
I worked on GDI – the “Graphics Device Interface,” which provided a device-independent interface for rendering shapes and fonts on screens, printers, and so on. My manager was Marlin Eller at first, and later Rao Remala (after Marlin moved to Japan). The other major functional areas were User, headed by Neil Konzen, which implemented Windows’ user interface, and Kernel, led by Steve Wood, which did the memory management and other underlying system services such as message infrastructure, and the ability to run MS-DOS applications.
The Windows development team was led by Scott McGregor, previously from Xerox PARC. One of his most noted contributions to Windows history was the tiled window interface (as opposed to the overlapping window style used in every version of Windows since 2.0). Before Windows ultimately shipped, Scott left Microsoft, and Steve Ballmer took over management of the Windows development team until the 1.0 product shipped.
We spent the summer and fall of 1984 trying to get Windows to the point of being functionally complete. During this time, we were told two things by Steve Ballmer: the final product had to fit on a single 360K floppy disk, and the product would ship amidst great fanfare at Comdex that November. Unfortunately, we all pretty quickly realized that neither of those would become true.
Later in 1984, Radio Shack introduced the Tandy 2000 PC. This was a non-IBM-compatible PC that ran MS-DOS. It had a nice color display, Tandy developed it to be the ideal machine on which to run Windows 1.0. Unfortunately for Tandy CEO John Roach, Windows didn’t ship for another year, by which time the Tandy 2000 was no longer a leading edge PC. In the meantime, Tandy was nice enough to run an ad featuring Bill Gates and a Tandy 2000 running a pre-release version of Windows.
Windows 1.01 finally shipped in November 1985. Steve Ballmer's final ship-date promise to the press was "before the snow falls," and we just made it. We were rewarded with a trip to COMDEX – the first business trip ever for many of us, where we got to explore Las Vegas and soak up the attention surrounding the long-awaited release of Windows. One of the highlights was the Windows roast where everyone made jokes about how long Windows had been vaporware. John Roach from Tandy was there, enjoying the opportunity to rib Microsoft for how he had been burned by all the delays.
After Windows 1.0 shipped, the Windows project was adrift for a while. Although it was assumed that we’d be building Windows v2, there was no schedule, and no committed list of features to implement. IBM and Microsoft had a joint development agreement, but that had not resulted in anything other than a requirement to start putting a big IBM header template into any source file that you happened to edit.
In early 1986, we moved from the buildings on either side of Northup Way and Route 520 to our new campus in Redmond. Not long after, I moved across “Lake BillG” to the applications group, where I joined Chris Peters’ team to work on Excel for Windows.
More information:
Windows 1.01 screenshots
Press kit from November 1983 Windows launch
Windows 1.0 and the Applications of Tomorrow by Charles Petzold
Windows 1.0 from Wikipedia
Windows Desktop Products History from Microsoft (there are a couple errors in the early history)
Lots of fascinating Windows and Microsoft history can also be found via the History archives on Raymond Chen's and Larry Osterman's respective weblogs.
