When an author compares an event to Y2K, the reader is prudent to attend with some degree of skepticism. The Y2K problem was large and affected multiple platforms across all industries. The threat of mobile/cloud computing (if it can even be considered a threat) must be large and wide-spread to stand against Y2K.
I will say up front that the mobile/cloud platform is not a threat. If anything, it is an expansion of technical options for systems, a liberalization of solution sets.
Nor does the mobile/cloud platform have a specific implementation date. With Y2K, we had a very hard deadline for changes. (Deadlines varied across systems, with some earlier than others. For example, bank systems that calculated thirty-year mortgages were corrected in 1970.)
But the change from traditional web architectures to mobile/cloud is significant, and the transition from desktop applications to mobile/cloud is greater. The change from desktop to mobile/cloud requires nothing less than a complete re-build of the application: new UI, new data storage, new system architecture.
And it is these desktop applications (which invariably run under Microsoft Windows) that have an impending crisis. These desktop applications run on "classic" Windows, the Windows of Win32 and MFC and even .NET. These desktop applications have user interfaces that require keyboards and mice. These desktop applications assume constant and fast access to network resources.
One may wonder how these desktop applications, while they may be considered "old-fashioned" and "not of the current tech", can be a problem. After all, as long as we have Windows, we can run them, right?
Well, not quite. As long as we have Windows with Win32 and MFC and .NET (and ODBC and COM and ADO) then we can run them. But there is nothing that says Microsoft will continue to include these packages in Windows. In fact, the new WinRT offering does not include them.
Windows 8, on a desktop PC, runs in two modes: Windows 8 mode and "classic" mode. The former runs apps built for the mobile/loud platform. The latter is much like the old DOS compatibility box, included in Windows to allow us to run old, command-line programs. The "classic" Windows mode is present in Windows 8 as a measure to allow us (the customers and users of Windows) to transition our applications to the new UI.
Microsoft will continue to release new versions of Windows. I am reasonably sure that Microsoft is working on "Windows 9" even with the roll-out of Windows 8 under way. New versions of Windows will come out with new features.
At some point, the "classic Windows compatibility box" will go away. Microsoft may remove it in stages, perhaps making it a plug-in that can be added to the base Windows package. Or perhaps it will be available in only the premium versions of Windows. It is possible that, like the DOS command prompt that yet remains in Windows, the "classic Windows compatibility box" will remain in Windows -- but I doubt it. Microsoft likes the new revenue model of mobile/cloud.
And this is how I see mobile/cloud as a Y2K-like challenge. When the "classic Windows compatibility box" goes away, all of the old-style applications must go away too. You will have to either migrate to the new Windows 8 UI (and the architecture that such a change entails) or you will have to go without.
Web applications are less threatened by mobile/cloud. They run in browsers; the threat to them will be the loss of the browser. That is another topic.
If I were running a company (large or small) I would plan to move to the new world of mobile/cloud. I would start by inventorying all of my current desktop applications and forming plans to move them to mobile/cloud. That process is also another topic.
Comparing mobile/cloud to Y2K is perhaps a bit alarmist. Yet action must be taken, either now or later. My advice: start planning now.
Showing posts with label software obsolescence. Show all posts
Showing posts with label software obsolescence. Show all posts
Sunday, August 19, 2012
Sunday, March 18, 2012
Windows 8 means a faster treadmill
The release of Windows 8 marks a change in Microsoft's approach to backwards compatibility. Microsoft has shifted its position from "compatible at just about everything" to "things change a lot and your old things may not work".
Windows 8 and its Metro interface re-define the programming of applications. On x86 processors, legacy applications can run in "Windows 7 mode". On ARM processors, legacy applications... cannot run. And while Windows 8 offers Windows 7 mode, Microsoft has made no promise of such a feature in future releases.
With the shift to WinRT and Metro, Microsoft has started a countdown clock for the lives of all Windows applications in existence.
In the past, Microsoft maintained compatibility for just about every application. Even vintage DOS applications would run under Windows XP (and probably still run under Windows 7). That compatibility came at no small expense, not only in development and testing costs, but at opportunity costs. (New development was constrained by the design decisions of previous releases.)
Users, developers, and support teams are on a treadmill, with new technologies and releases arriving faster than before. The good old days of decade-long technology planning have been replaced with a range of two or three years.
People can get upset about the faster pace of the treadmill, but they have nowhere to go.
Apple has "revved" its platform a number of times, changing the processor, the operating system, the user interface, and the device form factor. The folks working on Linux are working on similar changes.
If Microsoft believes that it can be more profitable in a new market, or if it believes that the current market is not profitable, then I believe that they will move to the new market. It's customer's problems with lack of backward compatibility are not Microsoft's problem.
Interestingly, corporations long ago lobbied for shorter depreciation schedules for computing equipment. They successfully got the depreciation for equipment reduced to ... three years. Now Apple and Microsoft seem to be agreeing, indicating that equipment really is obsolete after three years. (Except that they include software in the definition of "equipment".)
I'm not sure that this faster pace is a good thing. I'm also not sure that I like it. But I do know this: it's happening. The question is not how to stop it, or how to avoid it, but how to cope with it. How do we live in a world when technology changes (dramatically) every three or maybe two years?
Windows 8 and its Metro interface re-define the programming of applications. On x86 processors, legacy applications can run in "Windows 7 mode". On ARM processors, legacy applications... cannot run. And while Windows 8 offers Windows 7 mode, Microsoft has made no promise of such a feature in future releases.
With the shift to WinRT and Metro, Microsoft has started a countdown clock for the lives of all Windows applications in existence.
In the past, Microsoft maintained compatibility for just about every application. Even vintage DOS applications would run under Windows XP (and probably still run under Windows 7). That compatibility came at no small expense, not only in development and testing costs, but at opportunity costs. (New development was constrained by the design decisions of previous releases.)
Users, developers, and support teams are on a treadmill, with new technologies and releases arriving faster than before. The good old days of decade-long technology planning have been replaced with a range of two or three years.
People can get upset about the faster pace of the treadmill, but they have nowhere to go.
Apple has "revved" its platform a number of times, changing the processor, the operating system, the user interface, and the device form factor. The folks working on Linux are working on similar changes.
If Microsoft believes that it can be more profitable in a new market, or if it believes that the current market is not profitable, then I believe that they will move to the new market. It's customer's problems with lack of backward compatibility are not Microsoft's problem.
Interestingly, corporations long ago lobbied for shorter depreciation schedules for computing equipment. They successfully got the depreciation for equipment reduced to ... three years. Now Apple and Microsoft seem to be agreeing, indicating that equipment really is obsolete after three years. (Except that they include software in the definition of "equipment".)
I'm not sure that this faster pace is a good thing. I'm also not sure that I like it. But I do know this: it's happening. The question is not how to stop it, or how to avoid it, but how to cope with it. How do we live in a world when technology changes (dramatically) every three or maybe two years?
Subscribe to:
Posts (Atom)