At a time when many organizations are moving from Windows XP or 2000 to Windows 7, the last thing a lot of people want to think about is Windows 8. At the same time, it is incredibly clear that the future of client computing is undergoing a major shift thanks to the rapid growth of iPad and other tablet devices. Windows 8 is not only the next desktop operating system from Microsoft, but it is also Microsoft’s answer to this substantial shift toward low-power touch-based client devices.
This change impacts not only the operating system, but the underlying application development model. The new Windows Runtime (WinRT) development platform represents an evolution for .NET developers, and a significant shift for non-.NET developers.
Realistically, enterprises should expect two things. First, the enterprise desktop/laptop platform will most likely continue its shift to Windows 7, and may remain on Windows 7 for many years. Second, the enterprise is already forced to deal with iPad and other tablet devices, and they’ll need to deal with (or embrace) Windows 8 tablets in the same way.
To the first point, I think it unlikely that most organizations will roll out Windows 8 on a broad scale to desktops or laptops. Most organizations are just now moving from XP/2000 to Windows 7, only because those ancient operating systems will soon be entirely unsupported. It is not realistic to think that organizations will immediately move from Windows 7 to Windows 8. It is more realistic to think that they’ll be on Windows 7 for 5-10 years, and will then move to “Windows 11” or something along that line.
As a result, organizations will be building and maintaining applications using Microsoft .NET (WPF, Silverlight, Windows Forms) for many years to come. There is no WinRT for Windows 7, so that new development platform will be off limits for mainstream enterprise application development targeting the desktop/laptop space in the foreseeable future.
To the second point, the reality that end users in organizations will acquire and use tablets on their own, if not supplied by the organization, is already happening. The flood gates are open, and organizations are now left to deal with the results. A chaotic landscape composed of iPads, random incompatible Android devices, and soon Windows 8 devices.
This is where things get interesting. Windows 8 on Intel devices can run the same .NET applications as a desktop or laptop. They can also run WinRT applications, which (when built using .NET) are extremely similar to Silverlight applications. Windows 8 on ARM devices will only run WinRT applications.
iPad and Android devices require completely different application development using tools unlike .NET. No code or functionality sharing between existing .NET desktop/laptop applications and these platforms is possible. Truly embracing these platforms means building up duplicate development staff for Objective C and Java, or switching entirely away from traditional smart client development to a pure HTML5 model. Sadly, HTML5 isn’t compatible across all these devices either, so even that isn’t an obvious solution.
In reality, it might be less expensive for organizations to buy employees Windows 8 tablets than to pay developers to re-implement applications across multiple platforms, and to then support those multiple implementations over time. In fact, I suspect it will almost always be cheaper to spring for a few Windows 8 tablets than to pay for duplicate software development and maintenance forever.
To achieve the broadest reach, Windows 8 apps should target WinRT. That allows the apps to run on Intel and ARM devices. As I mentioned earlier, when using .NET to build WinRT applications, the development model is very similar to Silverlight. This means that existing WPF and Silverlight developers will have a relatively easy time shifting to WinRT, and substantial amounts of Silverlight application code will often just work on WinRT.
Frameworks such as CSLA .NET provide even more cross-platform compatibility. For example, the business logic code for applications written using CSLA 4 that target Silverlight will just recompile for WinRT, usually with no changes required at all. The vast majority of business logic from WPF or Windows Forms applications written using CSLA 4 will also just recompile for WinRT.
In summary, Windows 8 represents a major factor in enterprise application development strategy. In the short term, it might offer a lower-cost way to get users onto tablets without the high cost of duplicate software development, or dealing with the cross-platform HTML5 issues. In the long term, WinRT appears to be Microsoft’s new strategic development platform, so organizations need to be considering how to move to this platform over a period of years.