For 25 years M$ has been selling the vision of one licence for that other OS on each PC in the world. They managed, nearly, to achieve that in the beginning because IBM blessed them as the OS shipped on “IBM-compatible” PCs. Later, M$ used other exclusive deals to exclude other operating systems from the market. The result was, at its height, the monopoly was shipping on 95% of PCs. That share has slid to 72% recently although the installed base is somewhat larger than 72%.
That is all changing with the move to notebooks and smart thingies. Notebooks are over 50% of shipping units and smart thingies shipments are expected to exceed x86-like PCs in 2011. This has changed how people use their person computers and how IT is done. More people are using stuff from outside the monopoly at home and bringing it to work. I saw it last year, out in the bush… People were bringing all kinds of web-enabled gadgets to school even though there were rules against doing so for students. Many of the gadgets were from Apple but now Android/Linux and other Linux variations are appearing.
Lately M$ (“8″), Google (Android and Chrome OS), HP (WebOS) and Baidu (the Chinese giant of search) are all planning to change how personal computing is done. They are implementing a lightweight OS for portable devices that will run smartly on any personal computer.
This amounts to a total rewrite for M$. For the others, it’s merely the GUI and apps on top of GNU/Linux an existing OS already ported and running well on ARM and x86 and a host of other platforms. These new ways of doing IT are available today, except for M$ which is probably not going to be effective until 2013, at which time they will have to sell the revolution to their customers unused to thinking “thin”. M$ actively sabotaged the thin client when it was touted in the 1990s. They will have an uphill battle to undo all that baggage. Their competition will merely be extending what consumers know and love in smart-thingies to all personal computers, an easier task.
Applications? As of 2011, there are hundreds of thousands of applications for smart thingies and developers see small mobile computers as an untouched marketplace, the land of opportunity. By 2012, many must-have applications will be supported on ARM or have a web-enabled equivalent. There will always be some apps that require huge screens and local resources but they will become a small minority.
This is a revolution in IT. It started slowly in the 1980s with thin clients and mainframes, accelerated in the 1990s with thin clients and x86, PPC and Sparc servers, and is now a rocket with smart thingies and x86 servers. We are even seeing some ARMed servers. Even x86 clients are more often used to access web-based applications than local applications. I produce content and I have my own server but I have only 4 local apps running but dozens of remote apps are running for me.
Market share? If it were not for their large installed base locked in to M$’s way of doing things, M$ would have a 1/N kind of share where N is greater than 3. We see XP and “7″ currently competing with each other for a dwindling share. Imagine “7″ competing with “8″ and all these attractive options. It will be obvious to all except those who belive they must have something from M$ that if the computing is done elsewhere they do not need M$ locally. I see a rapid descent to 50% or so in the next few years with no way to defend the monopoly against the hordes of alternatives. After that, it will be a few years to get to 1/N. Essentially the monopoly is on its last legs and will be gone by the time “8″ emerges.
Besides the numbers, a huge disadvantage for M$ is that they have to do a rewrite. The bloat of adding layers will not work on ARM. That will take them a very long time during which time the viability of the alternatives will grow exponentially.
For an excellent analysis of the momentum of this revolution, apart from numbers, read “The Freight Train That Is Android”
Baidu is the largest search operation in China.
Shipment of ARMed mobile Internet devices (MIDs) is expected to exceed shipments of PCs in 2011.