By most measures, it might seem that Microsofts crucial customers are the PC builders who license Windows for bundling on new machines. Those hardware original-equipment makers, though, are not the companys most critical constituency—and Microsoft knows it.
Yes, PC OEMs depend on Microsoft to support new hardware features and to generate customer demand for more hardware resources—and Microsoft depends on those OEMs for preload licensing revenues. But Microsoft depends even more on software developers to channel customer demand toward Microsofts platform, so that OEMs will continue to consider it a no-brainer to stay with it instead of looking at lower-cost alternatives.
Its developers, therefore, who create the software mandate that drives the hardware industrys choice—but that developer constituency has now had its nose pressed to the glass of the Longhorn candy store for a long, long time, while the sign on the door promising imminent Grand Opening keeps getting replaced with announcements of continual postponement. Little by little, those developers are starting to look up and down the street to see where they might find more immediate gratification.
What those developers are seeing is an expanding array of choices that are defined by Internet standards, not by Microsoft road maps. Theyre seeing business environments like Amazon.com and its virtual storefronts, content environments like Google and iTunes, and on-demand computing resource environments like Suns $1-per-CPU-per-hour grid.
Theyre seeing a world whose axis no longer passes through the PC: one that increasingly revolves around media industries, such as music and photography, whose offerings are quickly becoming the greatest stimulus for buying a new computer or mastering a new application.
Developer excitement has been palpable at events like Microsofts Professional Developer Conference in October 2003, where the company demonstrated the exceptional power and conciseness of its Indigo communication framework and its declarative programming technologies. The post-browser world of Internet applications was arguably defined at those Los Angeles sessions, where major Internet players like Amazon.com showed their prototypes of applications that combined rich client-side visual interaction with pervasive behind-the-scenes Web services.
Significantly, Amazon.coms back end runs on Linux, but the companys developers still understood the likely future importance of the Windows client. That understanding might now be evolving, though, in more eclectic directions as Internet access growth becomes dominated by mobile-phone and automotive devices rather than traditional PC sales.
Microsofts public pitching of its long-awaited Longhorn update has focused on end-user power and convenience, with emphasis on features such as automatic maintenance of virtual file system views. What Microsoft knows better than any other software provider on the planet, though, is that end users dont actually make the important decisions that lead to platform dominance.
Dominance decisions are ultimately made by application developers, who determine which platforms will be the beneficiaries of their own crucial investments in developer training, tool acquisition and platform-specific feature exploitation. Those choices create the applications, some of them mass-market but many more of them custom-built or vertical-market tools, that drive platform choice and sell associated hardware.
Microsoft has demonstrated for many years a thorough mastery in maintaining developer mind share, and it has posted an impressive menu of dramatic forthcoming enhancements to developer productivity. Developers have responded to the promise that theyll be able to write less code but gain more leverage from database disciplines and pervasive Internet connectivity.
At some point, though, the door to the candy store had better be unlocked, and there had better be goods in the cabinets—or developers will sate their hunger elsewhere.
Peter Coffee can be reached at peter_coffee@ziffdavis.com.