Once upon a time there was a corporation which sought to have its product on every hard drive on every personal computer to the exclusion of all other similar products. M$ almost succeeded. They reached over 90% share of web-connected PCs but lately their share has fallen and the rate of fall is increasing with time.
With the current attach rate being less than 80% and the rise of ARMed PCs, tablets and netbooks, and smartphones which will do almost everything a PC will (more if you value portability and energy saving), this trend will continue and accelerate. M$ plans to release a version of their software for ARM but after years of rapid growth of GNU/Linux in that space they are likely never to dominate it. M$’s huge market share was accomplished by exclusive dealing. In the increasingly diverse space that is personal computing, M$ will become just another player on the field. There are niches where M$ is nearly complete but even those are under attack: business and gaming. This likely means M$ will not be eliminated from IT but that M$’s share will be much lower than it is now.
What killed M$’s dream was not the superiority of other operating systems or platforms but the burden of supporting M$ in the manner to which it had become accustomed. The world needs small, cheap computers but does not need the malware, downtime, and deadening performance of bloated software. There is a market for small, expensive computers and Apple has provided for that. GNU/Linux plays on all platforms: super-computers, mainframes, desktops, notebooks, netbooks and smartphones. It is hard to compete against a product that is so useful. ARM also has price, portability and heat advantages over x86 so this wave of change was inevitable. In spite of all M$’s machinations, it became impossible to suppress innovation.
Eventually the share that M$ has will be of the order of 1/3 of personal computing. That is because, given Apple, M$ and Linux as choices equally available in the market, consumers who are not expert in the technology will make random choices. If further operating systems develop, the target could be even smaller. One choice is a monopoly. Two could be two monopolies in different niches. Three or more is competition. I like it. One important side effect of this competition is that developers of software will need to produce software capable of running on any platform if they don’t want to lose share. That means M$ will never be able to regain its monopoly. Another side effect is that retailers will see drop in share if they do not include the competition on their shelves so 2011 is the last year you are likely to find that other OS dominating retail shelves.
Business is seriously locked-in to M$’s way of doing things, seeing lack of choice as “efficiency” because they don’t have to accommodate diversity. That has been changing as Apple’s products are seen as superior in performance, design or reliability, and it will change further as employees drag in to work the equipment they bought on retail shelves. There will soon be no “M$ shops”. The share of businesses using exclusively M$’s products will drop quickly but there will still be a large share of business who see IT as not their job and be willing to pay M$ whatever it takes to keep the software coming.
M$ will no doubt raise prices to keep these truly locked-in businesses working for M$ but there is a limit to how far M$ can go. With the move to cloud computing, businesses will have obvious alternatives. Google, for instance, can afford to provide service merely for the accompanying advertising revenue. Only businesses that care little about the bottom line compared to having a single source of supply will cling to M$. That could fall to a very small share of PCs eventually.
Schools, governments, banks, and many other lines of business are already in the process of migrating from M$’s limited line of products to GNU/Linux simply because of cost. Monopoly is too expensive a model for IT.