In-Depth

Happy Anniversary Windows: The Windows Operating System Turns 25

After a quarter-century, Windows has gone from a long-awaited experiment to the dominant software product in the world. But the operating system's past isn't necessarily a prologue for what could be an uncertain future.

• By Jeffrey Schwartz, Lee Pender
• 10/02/2010

Marsha Collier recalls purchasing a copy of Windows 1.0 when it was first released 25 years ago. But, like many users of MS-DOS, Collier had second thoughts. Upon looking at its features, Collier wondered what she would do with it -- and she returned it right away.

"I seriously read the packaging and thought to myself, 'Why would I ever want to run more than one program at a time?'" recalls Los Angeles-based Collier, who is no tech neophyte. Today she hosts a tech radio show and is author of various books, including "Ebay for Dummies" (For Dummies, 2004). Like many at the time, she had no idea how big Microsoft Windows would become.

Indeed, it's now arguably the most dominant and influential single product -- of any kind -- of the last quarter-century. As few as 20 years ago, Windows wasn't a sure thing to succeed. But succeed it did, and beyond the imagination of most observers -- and maybe even beyond the imaginations of the creators of the OS itself.

Windows runs more than nine out of 10 of the world's computers. Huge corporations run on it; sole-proprietorship businesses get off the ground with it. It runs stock markets, media empires and governments. Go to India, China, Russia, Australia, Texas or Brazil, and almost every computer in all of those places will run on Windows. No other single product has the global reach of the Microsoft client OS. The sun never sets on the Windows empire.

Microsoft, on the back of its OS, has built one of the most recognizable brand names in the world. Bill Gates is a hero to millions and, despite his amazing charity work, still a villain to others. Windows is a target for critics, hackers, bloggers, journalists and competitors, but everybody knows which product rules the roost in its category.

Love it or hate it, there's likely no other 25-year-old as influential and pervasive as Windows. But is the mighty OS at the peak of its powers? Has it peaked already? With the nearly 10-year-old Windows XP still dominating the market worldwide, where will Windows head as the 21st century stretches into its second decade? Nobody is sure, but one thing is certain: The world sees computing through Windows ... for now.

Two Years of Waiting
Windows 1.0 got off to an auspicious start on Thursday, Nov. 10, 1983, at the Plaza Hotel in New York City. That's when Microsoft founder and CEO Gates took the wraps off the first version of the OS. Gates' original name for Windows was Interface Manager, but Rowland Hanson, a Microsoft marketing guru at the time, is credited with convincing Gates to go with Windows.

Invitations to the launch were sent to the press in a box with a squeegee. The header read: "For a clear view of what's new in microcomputer software, please join Microsoft and 18 microcomputer manufactures for a press conference..."

But like many versions of Windows that would follow it, the first release didn't ship until two years after that fateful press conference, leading many to refer to it as "vaporware." Finally, Microsoft released Windows 1.0 in November of 1985 at Comdex. William Zachmann, at the time a prominent IDC analyst and now an independent consultant with Canopus Research Inc., was at that Comdex launch.

"It was by no means clear who was going to win during the period of the early '80s when the rush to GUI was on," Zachmann says. "The success of the PC in 1981-82 was a character mode interface, and graphics were handled only by applications -- and rather tentatively."

That competitive rush included Apple Lisa (the precursor to the Macintosh), Xerox Star, VisiCorp VisiOn, IBM Top View, Compaq Computer Presentation Manager and Digital Research Graphics Environment Manager, Zachmann recalls. "It wasn't clear that Microsoft was going to be successful with its approach because there were a lot of competing alternatives," he says. "Windows was viewed as the most likely to succeed, [but] there was no guarantee of it at the time."

Besides adding a primitive graphics layer to Microsoft MS-DOS, Windows 1.0 wasn't regarded as a feature-rich environment, according to Zachmann. But in Redmond's defense, he says the weakness of Windows was more a function of the limits of hardware than of the Microsoft software.

"PC hardware, up until the early '90s, really didn't have the capability of adequately sustaining a GUI environment; it was very limited compared to what it does today," Zachmann says. "Remember, in 1990, 512MB was a lot of memory."

Microsoft released Windows 2.0 in November 1987. With improved graphics and support for dynamic data exchange, this release took advantage of 286 processors; a version called Windows 2.03 added extended memory and 386-processor support.

In May 1990, Microsoft released Windows 3.0, which most observers consider to be the first major release of the OS. Windows 3.0 offered full 386-processor support, and Microsoft introduced the File Manager and Print Manager. It was the first release to get widespread support from software developers and hardware providers.

A key fork in the road for Windows came after that release. Originally, Microsoft and IBM Corp. had agreed to co-develop the next version of DOS called OS/2. OS/2 had a character-mode and the Windows-like OS/2 Presentation Manager. But the relationship between the two companies was dicey at best. Zachmann says IBM's goal was to make OS/2 proprietary and link it to the company's Micro Channel Architecture.

He points to three reasons Microsoft objected to Big Blue's ideas: "IBM really had this underlying agenda of wanting to keep it proprietary," he says. "Second, IBM never accepted the notion of Microsoft as an equal partner; it wanted to treat Microsoft as a supplier that it could tell what to do. The third issue was that IBM was incredibly bureaucratic in its software development and, frankly, not very good at it."

With the two companies far apart, each decided to go its own way: Microsoft promoted Windows clients and built Windows NT for servers, and IBM stuck with OS/2. At the time, Zachmann believed that IBM had the upper hand and would prevail with OS/2.

"I was identified as one of the few idiots that actually thought OS/2 was still viable," he says. "I've made some pretty good calls at times; that was definitely not one of them."

Windows for the enterprise started to take off with the release of Windows 3.1 and its successor release, Windows for Workgroups 3.11, which offered peer-to-peer networking, support for Novell NetWare and remote access service (RAS). Windows would solidify its dominance with the release of Windows 95, which added 32-bit TCP/IP support and introduced Plug and Play.

Eyes on the Future
When it comes to Windows 1.0, Microsoft prefers not to look back. When Redmond magazine reached out to Microsoft to talk about the company's first edition of Windows,

Microsoft declined to make available anyone who was with the company during the development of Windows 1.0.

"We're focusing our anniversary efforts on the Windows 7 first birthday, so unfortunately we won't be able to provide a briefing from someone from the Windows 1.0 days," a spokesperson for Microsoft said in declining our request.

Indeed, the release of Windows 7 has been well-chronicled in the pages of this magazine over the past two years, and there's no question it dispelled concerns that Windows has hit the end of the road. In addition to fixing many of the bugs in Windows Vista, Windows 7 adds performance improvements, slightly faster booting and support for Windows PowerShell, as well as some UI enhancements.

Since its launch a year ago, a stunning 150 million Windows 7 licenses have been sold, Microsoft says. This brings up the question: When will Microsoft release the next version of Windows? Look for the first service pack of Windows 7 to ship imminently. As for the next full-blown release of Windows, Microsoft is keeping mum.

"To tell you the truth, I'm not sure [Microsoft] knows at this point," says IDC analyst Al Gillen. "Windows 7 is so important to [the company's] current roadmap that the last thing it can afford to do is talk about the next version. What it needs to do is drag its feet and not rush the next product out. All that does is add confusion. Windows 7 is going to be widely adopted and as the momentum increases for Windows 7, it's going to be very hard for Microsoft to change and redirect that momentum to Windows 8 or whatever it winds up being called."

Rather than a full-blown release, Gillen suggests Microsoft might be better off adding an upgrade to Windows 7, something akin to a 7.1 release. "Microsoft should release a version with incremental functional improvement that doesn't disrupt any applications or break any of the installs," Gillen says.

That said, Gillen points out that his advice is what he thinks Microsoft ought to do, not what it likely will do. Observers widely believe that the next version of Windows will be Windows 8 and that it will ship in 2012. Blogger Steve Chapman published an extensive report consisting of leaked Microsoft documents that show details of Windows 8. Originally published on the Italian Web site Windowsette, Chapman wrote in his own post: "I'm quite confident these are the real deal. I just feel bad for the poor sap who either leaked these or inadvertently shared these with the world."

Among features under consideration for Windows 8, according to the leaked slides, are support for facial recognition as a means of providing user access, assumed connectivity to the Internet and cloud resources, and the ability to power up instantly.

Windows 8 will also target three key form factors: slates, notebooks and all-in-one PCs. Furthermore, it will support customization of applications, devices and content, according to the leaked documents. To what extent these documents will represent what Microsoft ultimately delivers, of course, remains to be seen.

Slates
While Windows 8 will bring better support for slates, many had hoped OEMs would be able to port Windows 7 to slate form factors. While Hewlett-Packard Co. back in January promised a slate-based PC by the end of this year, it backed off on that promise after agreeing to acquire Palm Inc. for $1.2 billion, and is planning a device based on the Palm webOS platform. Now HP is indicating that a slate-based PC running Windows 7 may come after all, perhaps by the next quarter. However, the HP Windows-based slate is likely to be significantly more costly than its webOS or the Apple iPad device, and the HP slate will likely be geared toward enterprise business users and not consumers.

Meanwhile, Dell Inc., Acer Inc. and others appear to be focusing their slate efforts on the Google Android platform. But if slate OEMs are passing on Windows 7, does that mean that Microsoft will miss out on this market? IDC's Gillen doesn't believe so.

"Certainly Microsoft has failed to be a leader and innovator in this space; there's no question about that," Gillen says. "Apple may be on a run-rate to sell 5 million iPads a year, but that's still a fraction of the PCs sold, and for now it's additive, not replacement," he says.

Enter 64-Bit Windows
Windows 1.0 was a 16-bit OS. It wasn't until Windows 3.11 was introduced in 1994 that it supported 32-bit processing. Seven years later, Windows XP had a 64-bit edition. But Windows 7 promises to bring 64-bit processing to the mainstream, according to Gillen.

"We see it as a fundamental part of what's being deployed with Windows 7," he says. Why? Currently, most desktops and laptops come with at least 4GB of RAM, and often support much larger amounts. But the 32-bit versions of Windows can only use a maximum of 3GB of RAM, thereby limiting the power a user can get out of today's hardware.

"64-bit allows you to access that memory and has almost no downside impact on the average business application," Gillen says. It does have an impact on peripherals, but that will not likely be a showstopper for those who want better-performing systems down the road.

He explains further: "It future-proofs you better if you move to 64-bit."

The Incredible Shrinking Windows
While 64-bit Windows might be the immediate future of the OS, its long-term future is, to introduce a metaphor, cloudy. Microsoft is straddling two worlds with its reliance on Windows as a major revenue producer and its commitment to be "all in" for the cloud. Cloud computing emphasizes lean over fat, cheap over expensive and -- most telling of all -- the browser over the traditional OS.

Indeed, the future of the OS itself as a concept is, to work the cloud metaphor again, up in the air. But that doesn't mean that Windows, or the OS in general, will go the way of the dinosaur anytime soon.

"The first thing to acknowledge is that, while OSes are going to change significantly in the near future, they're still going to exist," says Michael Cherry, an analyst at Directions on Microsoft and a former employee at Microsoft for 11 years, primarily in the 1990s. "Something has to sit between the applications people want to write and the hardware. Even on the iPhone, you have hardware that has a certain set of attributes, and the average developer that wants to develop an application doesn't want to deal with that."

The structure of Windows, though, might change. Thus far -- and especially since its re-tooling in 1990 -- Windows has been a catch-all. Given its universal nature, it has to run just about any application and include just about any capability that any user might need.

"Windows is fundamentally a general-purpose OS," Cherry says. "It runs on a wide variety of platforms, and it runs a wide variety of applications. It provides such a wide set of services that it's big."

Windows might need to shrink, at least for IT customers, suggests Don Jones, a consultant, Redmond magazine columnist and noted technical author. It's fine for consumers to have the whole raft of Windows functionality, but Jones says that businesses don't need it.

Microsoft should "focus on making [Windows] an OS and not a bundle of applications," Jones says. He says that Microsoft should do much more to differentiate between consumer versions of the OS and enterprise versions. And he has an idea for how the enterprise version should look.

"We already know what that looks like," Jones says. "It was DOS. There should be a control panel and an empty start menu when you get the thing, and that's it. Every moving part is just another thing you have to worry about being patched. Something like 30 percent or 40 percent of patches have nothing to do with a business computer. That's insane. Microsoft has to let that go."

Jones breaks the idea down even further, saying that with Windows, Microsoft "should focus on different audiences, not on feature sets." For instance, he says, he might be tempted to buy a copy of Windows if he knew that it included BitLocker, but most of the other features in the OS wouldn't matter to him. Jones, as it happens, is a Mac user.

"I really think [it] has got a fantastic OS," Jones says of Microsoft. "It's got a great core there. [But] the world doesn't need another raft of features laid on top of the features we've already got. The Windows OS is tremendously noisy. [Microsoft] needs to back down. It tries to make one OS for every possible audience, and that's its problem."

Jones' business partner and fellow Redmond magazine columnist Greg Shields offers similar suggestions, but with a different focus. Shields says the key to the future of Windows is ease of deployment on multiple platforms -- another factor that could lead to Microsoft shrinking the OS, as it has tried to do to an extent with Windows Phone 7 for mobile devices.

"Where's the OS going?" Shields asks, and he answers: "It's going into your phone; it's going into your PDA; it's going into your laptop; it's going into your [Virtual Desktop Infrastructure]. As the world starts becoming more app-centric, the concern is: What happens if apps start getting hosted on other platforms?"

Shields continues: "That's where Windows could potentially be marginalized. Where Microsoft stands to keep itself relevant is in making that OS as deployable as possible. The company's relevance stays as long as it becomes supremely easy to deploy that OS to any delivery mechanism possible."

But Directions on Microsoft's Cherry isn't so sure that Windows will shrink. It was necessity, he notes, that made it large to begin with.

"You don't go to work every day and say, 'Let's build a big, fat, bloated, dumb OS,'" says Cherry, who helped design features for Windows 2000. "It happens day by day. You start off with a really great design. People comment on your design and they say, 'This is great, but it would be really great if it did ...' And you can fill in the blank with your favorite capability there. Any request to make a change in the OS appears reasonable. You add it in, and that creates a set of dependencies."

Cherry does, however, echo Jones' notion of different versions of Windows -- as well as competing OSes -- being targeted at different audiences. "In the future, we may see far more specialized OSes that target a specific device or a specific set of functionality than the Microsoft general OS does," he says.

Here and Now
As for the threat of the browser replacing the OS as cloud computing grows, experts remain skeptical for the time being. "I think that the OS as it stands today is the OS we're going to see into the medium-term future," Shields says.

Cherry offers a more blunt view: "There's nobody out there running a centralized service that I trust with my data," he says of cloud providers. "I may use some Web apps. I may park some files in the Web so that I can access them from other machines. But in terms of saying, 'I'm not going to have local storage,' I don't trust any of these people. I don't want them to have all of my data. I'm not prepared to go to a totally lightweight OS and not have data on it."

So, Windows remains the standard for now, and it probably will for a while to come. But what the next 25 years will bring is still anybody's guess. And Microsoft can't afford to make missteps with its flagship product, which is in many ways the linchpin of its entire operation.

"I don't think there's anybody at Microsoft who has a solid, long-term vision for what Windows is supposed to be," Jones says. "There's no vision for where the truck is driving. The danger, if Microsoft doesn't do the right thing with Windows, is it stands to start losing other pieces of the enterprise. Windows kind of locks you into a Microsoft solution."