What is the point of duplicating an operating system? More precisely, what is the point of producing an OS solely for the purpose of supporting another OS' exposed programming API? The hurdles facing the Chinese operating system markets are two fold. First, there is the support to the Chinese national characters, which consist of BIG5 (for traditional Chinese, applicable to Taiwan and Hong Kong) and GB (for simplified Chinese, applicable to mainland China). Windows 98 has very bad support of either encoding methods since it has no internal support for Unicode characters and therefore must rely on double byte character encoding. What that means is that the win98 subset of win32 API has no intrinsic support of the Chinese language which simultaneously supports other Unicode languages. Or more precisely, the support for both traditional and simplified Chinese requires some really sophisticated character recognition algorithms. That speaks nothing of the other languages which a generalized OS such as Linux and Windows NT/2000/XP should conceptually support. Even if the complex problem of language support is ironed out, it says nothing to the sophisticated use of fonts in the Chinese language including word processing (yes, word processing requires very sophisticated manipulation of fonts), typesetting, and inter-OS communications requisite in a modern working environment.
Instead of concentrating on creating another OS, China should focus its academic attention on augmenting the Chinese support of existing Linux variants. Which currently sucks imo compared to existing standards on WinXP and MacOS.
I think the fight between Microsoft and the Linux community is as much an ideological one as well as a practical one. From a computer programmer's perspective, the change is an evolutionary one where the profession gradually transforms from one which has more affinity to those of the novelist, and the playwright to ones like the doctor and the lawyer.
In other words, under the Linux paradigm, programmers today will earn their livelihood not based on their past achievements in the form of intellectual property, but on their ability to continually provide up-to-date service in the improvement of certain corporate or organizational functions. The movement is a natural one since programming talent is no longer so rare in today's world. While the works of T.S. Eliot and Richard Wright are irreplaceable treasures of the history of mankind, the work of a computer programmer can be readily replaced with a fairly accurate estimate on cost by the work of another. Extrapolate this analogy to a larger extent, it simply means that the product of one software firm can be replaced by the product of another software firm. In this scenario, charging royalty for an entire software package instead of particular innovations becomes a very stupid proposition. Intelligent customers would simply switch over to cheaper alternatives given that they receive the same functionality.
With this in mind, the current move by Microsoft to switch over to a subscription based system is not surprising. The model makes more sense in the modern development environment for the following reasons:
1. Periodic release of software packages provides the wrong incentives for software firms. In order to attract customers, the software package must provide a huge laundry list of "features", software developers come up with the most gratuitous and often counter-productive functions. Those so called features often do not cater to the needs of the customers, but rather hinder their productivity. The result is often a feature bloated software which only gets fatter rather than better as time goes by. The compounded result of this effect can be observed in many Microsoft products, such as MS-Office.
2. Cyclical development of shrink-wrap software is terribly stressing on developers. Traditional software firms such as Microsoft and Oracle are known to have "burned out" several generations of young, bright programmers. As the profession of programmers become a more formally established one, more people would treat it as a lifetime profession rather than a youthful stunt. It is imperative for software firms to recognize the need to make the life of a programmer not only a challenging one, but also a sustainable one. The retention of expertise of experienced programmers would mostly likely pay off in the long run not only in the personal lives of those programmers, but also on the bottom lines of the software firms.
3. A subscription service based software paradigm would generate immediate user feedback, both on the current feature-set, and on the future needs of customers. Traditional software firms spend millions of dollars on usability tests to determine a pre-set list of features in their products, and those lists often do not reflect the need of real customers. The process is both exponentially more expensive, and much less efficient than the Linux community model where a large vocal Linux advocates are actively involved in the maintainence of the current feature-set as well as the determination of future development directions. In the long run, the Linux community would be able to produce products that fit more closely with customer need. Microsoft must switch to a service model in order the compete.
4. The subscription model ultimately would give end-users more choices. After all, would one prefer to pay $300 every two years to purchase MS-Office or $15 per month? The financial cost is essentially the same (Given the inflation rate). The difference is that if the customer decides that the software is not worth his/her money, the person could simply choose to cancel the service.
Given the above considerations, I think Microsoft's decision to switch to a subscription based service model is a sensible one. It would be the only way for Microsoft to compete with Linux in the long run.
Slashdot Mirror
What is the point of duplicating an operating system? More precisely, what is the point of producing an OS solely for the purpose of supporting another OS' exposed programming API? The hurdles facing the Chinese operating system markets are two fold. First, there is the support to the Chinese national characters, which consist of BIG5 (for traditional Chinese, applicable to Taiwan and Hong Kong) and GB (for simplified Chinese, applicable to mainland China). Windows 98 has very bad support of either encoding methods since it has no internal support for Unicode characters and therefore must rely on double byte character encoding. What that means is that the win98 subset of win32 API has no intrinsic support of the Chinese language which simultaneously supports other Unicode languages. Or more precisely, the support for both traditional and simplified Chinese requires some really sophisticated character recognition algorithms. That speaks nothing of the other languages which a generalized OS such as Linux and Windows NT/2000/XP should conceptually support. Even if the complex problem of language support is ironed out, it says nothing to the sophisticated use of fonts in the Chinese language including word processing (yes, word processing requires very sophisticated manipulation of fonts), typesetting, and inter-OS communications requisite in a modern working environment. Instead of concentrating on creating another OS, China should focus its academic attention on augmenting the Chinese support of existing Linux variants. Which currently sucks imo compared to existing standards on WinXP and MacOS.
I think the fight between Microsoft and the Linux community is as much an ideological one as well as a practical one. From a computer programmer's perspective, the change is an evolutionary one where the profession gradually transforms from one which has more affinity to those of the novelist, and the playwright to ones like the doctor and the lawyer. In other words, under the Linux paradigm, programmers today will earn their livelihood not based on their past achievements in the form of intellectual property, but on their ability to continually provide up-to-date service in the improvement of certain corporate or organizational functions. The movement is a natural one since programming talent is no longer so rare in today's world. While the works of T.S. Eliot and Richard Wright are irreplaceable treasures of the history of mankind, the work of a computer programmer can be readily replaced with a fairly accurate estimate on cost by the work of another. Extrapolate this analogy to a larger extent, it simply means that the product of one software firm can be replaced by the product of another software firm. In this scenario, charging royalty for an entire software package instead of particular innovations becomes a very stupid proposition. Intelligent customers would simply switch over to cheaper alternatives given that they receive the same functionality. With this in mind, the current move by Microsoft to switch over to a subscription based system is not surprising. The model makes more sense in the modern development environment for the following reasons: 1. Periodic release of software packages provides the wrong incentives for software firms. In order to attract customers, the software package must provide a huge laundry list of "features", software developers come up with the most gratuitous and often counter-productive functions. Those so called features often do not cater to the needs of the customers, but rather hinder their productivity. The result is often a feature bloated software which only gets fatter rather than better as time goes by. The compounded result of this effect can be observed in many Microsoft products, such as MS-Office. 2. Cyclical development of shrink-wrap software is terribly stressing on developers. Traditional software firms such as Microsoft and Oracle are known to have "burned out" several generations of young, bright programmers. As the profession of programmers become a more formally established one, more people would treat it as a lifetime profession rather than a youthful stunt. It is imperative for software firms to recognize the need to make the life of a programmer not only a challenging one, but also a sustainable one. The retention of expertise of experienced programmers would mostly likely pay off in the long run not only in the personal lives of those programmers, but also on the bottom lines of the software firms. 3. A subscription service based software paradigm would generate immediate user feedback, both on the current feature-set, and on the future needs of customers. Traditional software firms spend millions of dollars on usability tests to determine a pre-set list of features in their products, and those lists often do not reflect the need of real customers. The process is both exponentially more expensive, and much less efficient than the Linux community model where a large vocal Linux advocates are actively involved in the maintainence of the current feature-set as well as the determination of future development directions. In the long run, the Linux community would be able to produce products that fit more closely with customer need. Microsoft must switch to a service model in order the compete. 4. The subscription model ultimately would give end-users more choices. After all, would one prefer to pay $300 every two years to purchase MS-Office or $15 per month? The financial cost is essentially the same (Given the inflation rate). The difference is that if the customer decides that the software is not worth his/her money, the person could simply choose to cancel the service. Given the above considerations, I think Microsoft's decision to switch to a subscription based service model is a sensible one. It would be the only way for Microsoft to compete with Linux in the long run.