Original URL: http://www.theregister.co.uk/2008/06/30/farewell_gates/
Bill Gates has gone, what's his legacy?
From hippy to hanger-on
This week marks another first in the 33-year history of Microsoft - life without Billg. The company and the man who co-founded it and rose to become the world's richest geek have parted ways. Bill Gates is no longer chief software architect and will be checking in only as company chairman.
Gates is hailed as the visionary who changed our lives by delivering on the vision of a PC in every home. Certainly, Gates and Microsoft came along at the right time. Ken Olson, former chairman and president of computing pioneer Digital Equipment Corp (DEC), is famed for saying in 1977 he saw no reason why anyone would want a computer in their home. It was also IBM's lack of interest in building software for PCs that gave Microsoft its first break.
Had it been left to companies like DEC and IBM, computing today would likely be a different, analogue, green-screen world.
The challenge of Gates, though, is to put him - and Microsoft - into context. The dictionary defines a visionary as someone “given to fanciful speculations and enthusiasms with little regard for what is actually possible” or “a person with unusual powers of foresight”.
Gates turned computing into a mass market by his focus on “experience” on a low-priced Intel box. He had the “vision” to see what Olson couldn’t when it came to PCs: that given the accessible and affordable tools ordinary people - not just those individuals staffing corporations or the engineering and scientific communities served by IBM and DEC - could do great things with computers. His success, though, turned him into another Olson, a man whose beliefs became contained and defined by the market his company had carved out.
Behold: your future
OK, so we all know Gates was fascinated with the idea of the computer from an early age. Before there was Microsoft, there was Trafo-data, which he started with Microsoft co-founder Paul Allen at high school to monitor traffic patterns across a roadway. The system used an Intel 8008 chip priced $360. The love affair with Intel, software and the power of what could be achieved was born.
The 11 hippies and loons that went on to start Microsoft got their big break when mighty IBM asked them to write the operating system, Microsoft Disk Operating System (MS-DOS), for its PC in 1981.
Here’s where one company’s lack of vision showed. That company? IBM. The giant left Microsoft free to license MS-DOS to others.
That happened as businesses began buying more and more PCs during the 1980s to digitize manual processes, and computing became affordable thanks to Intel-based PCs running Microsoft’s software. The 1980s saw Microsoft go from MS-DOS to the first version of Windows, an extension to DOS and a major step forward in terms of the way people interfaced with the PC.
It was the 1990s, though, that were the golden age for Microsoft in terms of consolidating its place on the desktop and in going beyond the PC and on to the server. Among the high points: Microsoft launched the first widely popular versions of Windows for the PC, version 3.0 in 1990 and Windows 95 in 1995. Both helped make the PC something for ordinary end users, with the latter something for consumers, not just business.
The decade also saw the launch of Microsoft’s first server operating system, as well as Windows NT, database, SQL Server, email system, Exchange and Outlook. There was also Office in 1989, when Microsoft bundled its separate word-processing and spreadsheet applications.
Here’s where the idea of Gates the visionary needs its first qualification. From office productivity suites, server operation systems and databases to email and collaboration, Microsoft was never first. But Gates’ famously obsessive, verging on rude, attention to detail meant that Microsoft delivered products that were generally better put together from the end-users’ perspective. They were easier to use. Oh, and by running on Intel-based hardware, affordable.
Security was another matter - this was before networks provided the kind of connectivity that meant bad people could abuse systems or data on Windows machines. Most PCs and servers-based software ran in isolated, secure islands. The focus therefore was on empowering the person sitting in front of the PC running Microsoft's software.
Completeness of product was also a joke in IT circles. For all Gates’ claimed nit-picking attention to detail and intolerance for incompetence, the standing joke was you waited for version 3.0 of the Microsoft product. Anything before that was considered unstable or sub-par on a features basis compared to the competition.
Banking on Windows NT
How could you take something like Windows NT, launched in 1993, seriously when it was rooted on the desktop and when the emphasis was on empowering the end user, not preventing bad people from doing the wrong thing, or on providing massive amounts of scale?
These early days were stories of failure for Microsoft’s newer software. The UK’s NatWesk Bank running more than 2,000 machines on a Windows NT branch network suffered a massive outage in the late 1990s. You wouldn’t get that if you’d been running Novell’s NetWare. Windows Exchange Server was notorious for needing a quick reboot every few days. Not so IBM's Lotus.
History will record, though, that for all the early stumbles and sniffy dismissals of the IT pros, Microsoft closed the gap. Office, Windows NT, SQL Server and Exchange Server matured and came from behind to take a number-one or two slot against IBM, Lotus, Oracle, Novell in addition to a range of dead and forgotten Unix vendors.
For all Microsoft's strengths, and Gates’ supposed laser-sharp vision, the story of Microsoft’s success is as much down to the strategic weaknesses and blind spots of the of the competition. Only too late did they take Microsoft and its under-featured software seriously.
Novell wrestled with the snooty concept of the super network operating system while Microsoft closed the feature gap on NetWare. Unix vendors believed in the innate superiority of Unix over Windows and preferred to squabble for the crown of defacto Unix in a shrinking kingdom. Spreadsheet manufacturers dickered with the idea of putting their products on the then-new Windows 3.0. In mobile, Palm, like others, failed to learn from Mahatma Ghandi: first it ignored then it laughed at Windows CE, then it chose to fight - too late in the day only to see its lead in hand-held computing eroded.
Where Microsoft really shone during this time was in building an army of individuals outside Redmond developing and selling applications for Windows, thereby sustaining it and increasing its usefulness.
Developers and channel partners saw the potential Windows and all those applications offered: a mass market using computer software that was easy to use and more integrated than software from rivals. And all that on low-cost, Intel-based PC hardware. Developers were signed up to the Microsoft Developer Network (MSDN) and fed resources so they kept targeting Windows. Channel and retail partners got financial incentives, business and technical support.
One man, one Microsoft, one vision
The year 2000 saw a major strategic gamble on Microsoft’s part. With so many platforms in place, Microsoft decided it was time to unify everything around a common framework, tied into one development environment. Also, it was time to answer the growing challenge of Java, which aimed to be write-once-deploy-anywhere software.
Arguably, Gates’ last major achievement therefore came eight years ago. He oversaw .NET, designed to let multiple languages run equally on Windows using a virtual machine environment and to also unify different Windows' products development environments. Also, Gates introduced C#, the object-oriented language that offered a quicker and cleaner way of building code, and that is today Microsoft’s fastest growing language.
These followed a decade during which Microsoft under Gates had ridden the PC boom and reached a position of unparalleled power and influence without anybody really noticing or taking it too seriously. No longer the outsider, such was the ubiquity of Windows and Office and the company's influence in tech that - for example - no standard was an industry standard unless Microsoft was on board with it. It was became clear that Microsoft was no longer the outsider scrapping with giants like IBM to get ahead. The opposite was, in fact, true: Microsoft was becoming an IBM in terms of its size and power.
As such, Microsoft began serving its own interests and perpetuating its power and position. With this Gates, too, began losing his status as a computer revolutionary.
Microsoft was now milking the market demand for personal productivity that Windows had delivered by tying OEMs into restrictive licensing deals when it came to using Windows on their machines.
Microsoft was also starting to play politics with industry consortiums that were clearly intended to consolidate Microsoft and Windows and to wrong foot rivals, regardless of the implications for the end users. Microsoft drove formation of the Web Services Interoperability (WS-I) consortium with IBM to establish its early work on SOAP and the cumbersome WS- specs to deliberately exclude Sun Microsystems, the industry's steward of Java. It was no coincidence that Microsoft had only just unveiled .NET, its own web-services friendly stack and Microsoft's answer to Java, which had been developed six years earlier and was offering an alternative development platform to that offered by Microsoft.
IBM, long chafing at Sun's control of Java and eager to assert an industry leadership role on Java, quickly signed up to the WS-I to become Java's biggest spokesperson in the group. That was until Sun - finally - got voted in.
Not a benevolent monopolist
Microsoft also moved to crush browser startup Netscape through licensing and pricing pressure on OEMs. The idea was PCs should run Microsoft's Internet Explorer browser not Netscape.
This landed Microsoft in hot water with the US Department of Justice and led to a long-running court case. The result? Microsoft was judged a monopolist. This is not an illegal activity in US law, it's just what you choose to do with your monopoly power that can be a problem. Microsoft was judged to have abused its power. Not the actions of a true industry visionary.
The trial lifted the lid on a number of other unsavory practices. I personally met with the inventor of a paint application who explained how the Microsoft suits had came to see him one day wearing their angry eyes: they offered to buy his application for Office at a price of their choosing, or they could build their own and shut him down. That was the deal. He chose the money.
The trial also revealed Gates as an ungracious individual: stubborn, defensive and unwilling to admit Microsoft had done wrong.
During a testimony given by video instead of in person, like any normal person subject to the laws of the land might do, Gates offered obfuscatory answers and said "I don't recall" so many times even the presiding judge had to chuckle.
It was a blow to brand Bill. His kudos as the modest geek who was above corporate politics, who played only in the impartial world of technology, who believed in building great software to empower the end user was gone. Bill was revealed as a corporate bastard screwing the little guy while his company became the Evil Empire.
Then there were the technology mispredictions from the Gates visionary. Microsoft's success has come from refining ideas others pioneered. This was a trait that served it particularly well in the world of the 1990s, where Microsoft defined its position as the quick innovator against slower PC and server software companies. When it came to going beyond the PC and server Gates and Microsoft have stumbled.
The TabletPC, a Gates' cause celebre, has failed to take off. The concept of Windows running everything in your life, not just the PC, is struggling against Java and Linux. After 10 years and $10bn investment IPTV has come to nothing, while the Windows-powered HD-DVD alliance with Toshiba collapsed in the face of the Java-powered Blu-ray alternative. The Windows-powered SPOT watches have failed to generate any consumer excitement.
Gates works out what's next after the PC
Microsoft and Gates failed to anticipate or respond to the internet. MSN was launched in 1995, along with Internet Explorer, as other companies began moving to the web. MSN was an internet portal, the flavor of the time, and was soon filed away under the somnolent leadership of Mr Internet and former SGI chief executive Rick Belluzzo. The future was still a PC and server future as far as Microsoft was concerned.
Interestingly, Google popped up with something called internet search in 1996, one year after MSN launched, and was led by two up-and-coming kids from Stanford University, not some systems exec just jobbing it. Today Microsoft is paying the price for its lack of vision and poor execution. MSN is a number-three player with a third-rate search service and chief executive Steve Ballmer is trying spend billions of dollars to buy Yahoo! to catch up to Google.
The 2000s also saw Microsoft begin to pay the price for Gates’ focus on user “experiences” and lack of vision when it came to network-based security in the software he oversaw during the 1990s. The Windows' security architecture was exposed in a series of catastrophic viruses in the early 2000s. In 2000 there was the ILOVEYOU worm while Code Red ravaged inboxes a year later.
It took a year for Gates' to publicly address the problem and promise action. He issued his Trustworthy Computing memo to employees, which launched an internal initiative to make Microsoft focus on security. The danger, though, was already out there with years of legacy code unsuited to deployment in the networked environments of the internet age. The year following Gates' memo, the SQL slammer worm caused widespread denial of service attacks in SQL Server, hitting more than 70,000 victims in just 10 minutes.
Better to double down on what you know best. Focus on the user experience on the desktop. That led to Windows Vista. Gates unveiled Windows Vista in 2003, as Microsoft's "biggest release of the decade" - the biggest since Windows 95 we were told. It was an ambitious work, featuring a brand new XML interface and architecture, a web services stack and a search system that integrated with the server, called WinFS.
Reworking the desktop was too big a job, though. Faced with slipping deadlines and re-writing the code base, Microsoft re-set the vision and ripped out certain features, making them available as add-ons to Windows XP. WinFS disappeared all together. In a video interview on Microsoft's Channel 9, Gates admitted: "We tried to do too much we learned from that."
Break glass for emergency vision
During the twilight years of his technical leadership at Microsoft, Gates must surely surrender his status as a person able to chart the future course of the industry and computing.
Two big challenges face Microsoft, and there is little evidence to suggest he saw either coming or acted in accordance with his visionary status. Those challenges? Open source and - yet again - the internet. This time, it’s phase two of the internet, dare we say Web 2.0.
His reaction to both, though, suggests he is not reading the tea leaves of trend and technology to chart a direction to the future as he did in the 1970s. He is cutting to fit something that suits Microsoft’s corporate objectives. As a company founder and former chief executive, that's not just his prerogative but is to be expected. This does not make him a genuine visionary, though. It makes him a man with an agenda.
That makes him another person like DEC’s Olson or Oracle’s Larry Ellison, serving their company’s interests. Not a Tim Berners-Lee, James Gosling or Tim Bray who contributed something we could all use - the web, Java or XML - and whose talk of what’s possible is not constrained by the agendas of their employers, if they have them.
The biggest issue facing Microsoft is how to deal with open source. Microsoft must square the ideas of what to release under open source licensing and its lingering issues over intellectual property claims with what it considers are its responsibilities to shareholders.
Gates' words on open source have been few and far between. When he has spoken, though, he does not sound convincing. His definition of open source is unrecognizable to everyone else. According to Gates: "There's free software and then there’s open source” and under open source “nobody can ever improve the software".
Far from being a visionary, Gates has been dragged into an uneasy compromise on this subject. In 2001 he called GPL "Pac-Man-like" for eating authors' intellectual property. According to Gates, Microsoft invented open source. The reality is that open source has forced Microsoft to lower its prices so that those in the developing nations now flocking to open source can afford to pay for Windows.
This contorted definition does more than leave Gates looking out of touch. It has left Microsoft with an uneasy internal truce on the subject of open source. A range of lower ranking engineers and execs in the server and tools division are working slowly to convince and cajole senior management of the virtues of open source. They must fight against the vested interests of the closed-source Office and Windows businesses, who want to keep things shut and charge for licensing of software.
It's a fragile and painfully public debate that's left a house divided and could have been avoided if Gates, with his supposed vision and clout, had identified the trends in open source and staked out a credible stance for Microsoft early on. Microsoft is hiring open source engineers on one hand while letting Ballmer rattle the saber of intellectual property rights on the other. Community developers outside watch the spectacle, not taking Microsoft seriously when it talks of working with open source. Lower ranking execs now admit Microsoft is 10-years behind on open source.
Gates was also out of touch on the latest chapter of the internet, software as a service. In this respect, Gates has more in common with executives at Oracle and SAP who are responding and trying to protect their core licensing businesses against the threat posed by online services. He has little in common with the true believers, such as Salesforce.com’s chief executive Marc Benioff who are challenging the established order and the facts as we know them.
When it came to software delivered as a service, Microsoft - like Oracle and SAP - initially ignored it. Finally, Gates - with his chief software architect heir Ray Ozzie - concocted a hybrid alternative. Something that combines service in the cloud with processing on the PC, outside the browser - markets where Microsoft is strong. That vision? Software plus service, a mantra only Microsoft is chanting.
A generation lost to Mac
Open source and online services are real problems for Microsoft. In the 1990s Microsoft built a massive army of developers and partners around exciting new products that revolutionized computing. Windows did away with niche operating systems and Office set a standard on personal productivity, creating mass markets that individuals could buy in to and build for.
Years later, Microsoft is not generating that level of excitement anymore. Today’s generation of developers is excited about Google, Amazon and Facebook. Not just as online destinations, but also in terms of the raw code resources and support they are getting. That’s where Microsoft won in the 1990s - giving away for free or a low cost the tools and resources that helped developers build applications for Windows and Office. Now Google, Amazon and Facebook are doing the same.
John Lam, hired by Microsoft to lead its work on a version of Ruby for .NET, put it succinctly when he told a recent Silicon Valley scripting conference that Microsoft has lost a generation of developers.
They have not just gone online. Thanks to debacles like Windows Vista they have also gone over to the Mac as a development platform under an invigorated Apple. The foot dragging and self-serving politics that delivered a vendor-friendly but inflexible WS- architecture has seen developers adopt alternatives like Representational State Transfer (REST).
Save the eulogies
History will record Gates as a visionary. Gates was a bits and bytes man, a productized thinker whose zealous focus on code combined with his belief in empowering people through PCs based on commodity components gave us things like a good productivity suite and low-cost database at a time when the competition either sucked or was too expensive for the ordinary consumer user or business. He saw the potential for the PC when computing dinosaurs like DEC and IBM were out of step or thought software too small to care about.
Gates' vision, though, became constrained and defined by Microsoft’s early success on the server and client. When it came to "non-traditional" platforms and anything other than Windows, Gates has been flummoxed. His ability to see clearly or sense a genuine shift in the winds of computing, akin to the shift that happened with the PC in the 1970s, has been compromised.
Gates ended as a man fitting his ideas to the realities of the business he’d created - a $50bn business. This you expect from a businessman. Unlike Berners-Lee, Gosling, or Bray who also changed our world with their ideas, Gates' vision became one of preserving and extending the one company and the one technology stack that he made and that defined him: Microsoft and Windows. ®