What Lies Ahead
It takes a certain sort of idealist to find any cause for complaint in the history of Microsoft's relationship with developers. The overwhelming market share of Windows has made it attractive for developers to exploit that full-featured platform in great depth-more so than would have been feasible if there had been several such platforms, any of them representing too large a market to be ignored. The low cost and high productivity of Windows development tools have yielded a cornucopia of useful applications. It's clear, however, that what's worked for Microsoft in the past will be less effective in the future.The vision of application development that Microsoft promoted before the release of the Vista update to Windows relied on a three-part promise of superior communications, superior interactive graphics and a superior model of client-side storage. Delivering that combination, known at one time as the "three pillars of Longhorn," would have encouraged developers to think in terms of a rich Windows-specific application that exploited the client-side platform to define and deliver the user experience-while calling out to the Internet cloud for data and background services. That was a powerful vision, but it was not delivered in anything like the time frame promised, nor were all of the pieces put together to complete the picture that had been so attractively drawn. The focus of developers has therefore shifted into the cloud itself, sometimes combined with client-side cross-platform logic using mechanisms such as AJAX-and decoupling application access from client-side platform choice. The world in which Microsoft will compete for developer mind share going forward is therefore fundamentally different from the one in which Gates' superb developer courtship tactics have succeeded in the past. The fully connected, always-on multiuser world of the cloud is one in which security, scalability and robustness will have to be rock-solid-and Microsoft's sheer size and the complexity of the solutions it hopes to offer to the market will represent major challenges of their own. Peter Coffee, former technology editor of eWEEK and one of the founding members of PC Week Labs, wrote reviews and columns on development tools and practices in PC Tech Journal, PC Week and eWEEK from 1988 through 2007. His coverage ranged from Ada and APL to Smalltalk and XLisp, with milestones including the first published review of Visual Basic 1.0 and intensive reviews of many Java development tools, including Microsoft's Visual J++. Coffee is currently director of platform research at Salesforce.com.
When applications run on a platform in the cloud-such as any of the platforms being offered by Amazon.com, Google, Salesforce.com or literally dozens of others-then the user stops seeing an advantage in running the most popular client-side operating environment. Any standards-conformant browser, running on any client operating system, becomes as good an application delivery tool as any other.