Bill Gates’ announcement that he will step down as Microsoft’s chief software architect by 2008 triggered some fond memories among tech journos. And I feel compelled to join the choir in gazing back through the swirling mists of time.
The first time I saw Gates was in Singapore in 1993, when he popped in to launch the amazing, fully 32-bit Windows NT 3.1. Gates is an excellent programmer, has impressive business acumen and his legal mind has proved sharp enough to outfox sustained government and legal challenges. However, he is an appalling presenter. I was struck by how unconvincing and almost incoherent he was, as he nervously struggled to convey his vision in a quavering, Kermit the Frog-like voice.
Although 32-bit processors had been around for a while when Windows NT was launched, it hit the market at a time when 4MB and 8MB systems were common. Windows NT needed 16MB if it was to run half-decently, however.
Not too many people were willing to fork out the large amount of money required to upgrade their system memory to this level then, especially when the hybrid 16/32-bit Windows 3.11, and later Windows 95, did most of the things they wanted and had lower resource demands. Windows NT was relegated to high-end niche roles and enjoyed more success in the server space than as a high-volume workstation operating system.
It took patience, perseverance and several Windows permutations before Windows NT matured into the now ubiquitous Windows XP, in 2001. Even so, the development time-frame looks fast compared with the upgrade time for XP, to Windows Vista, which is already four years behind schedule.
However, Microsoft has deftly managed expectations, as well as the delays, and kept its partners in the fold and competitors out of the market, despite only having a five-year-old operating system with patches on offer. The OEM industry is firmly locked in and, unless it actively funds a Microsoft competitor, has no way out of Redmond’s technology cul-de-sac.
Times have changed, however, and operating systems have become less relevant, thanks to internetworking. A slim-line bootloader for a networked interface with hardware-accelerated 3D and audio capabilities for gaming is all most users want, not the Windows Vista behemoth.
The forward-looking platform for Microsoft would be a “mini internet”: a media-centric Longhorn server, with networked terminals to display the content stored on it and interact with it. Intel has been pointing the way towards such a platform with its virtualisation and hardware-partitioning technology.
However, Microsoft has decided it will copy others rather than innovate, something that’s all too apparent with Vista. Customers upgrading to Vista, in 2007, will finally enjoy the kind of security that open-source users have enjoyed for years. But the OS is desperately short of fresh ideas and is more about fixing XP.
Those who think Ray Ozzie will be able to radically change Microsoft’s direction and revitalise the software giant are probably being too optimistic. Ozzie’s influence is showing in some great Microsoft products like Windows Live Mail and the whole Live.com concept. Microsoft and Ozzie could take MSN and turn it into the world’s largest application and content-serving platform, with social networking on top.
Unfortunately for Ozzie, however, he will be saddled with Gates’ legacy, an irrelevant operating system that nobody really wants but which cannot be jettisoned any time soon. Windows Vista will swallow marketing and development budgets that Ozzie should be spending on more visionary platforms. The challenge for Ozzie will be to survive Vista, while rapidly developing networked product and service offerings. This challenge was too much for the brilliant Gates, so the question is: will Ozzie will be up to it?