Several points are mentioned by the Gartner analysts:
- Windows has become too big and complex, therefore innovation happens rarely and is mostly just incremental.
- The PC market's growth has slowed (less people need to buy new machines). The markets where it does grow (emerging markets) are more price sensitive, often relying on cheaper hardware. Linux has emerged as a strong competition there, for example on cheap sub-notebooks such as the Asus Eee PC or the OLPC XO. Microsoft had to extend the life of XP for that reason, since these machines (and cheaper, older machines in general) usually cannot run Vista.
- The increased move to web-based applications makes the operating system itself less relevant.
Today, 70 to 80 percent of corporate applications require Windows to run, but the Gartner analysts expect a tipping point in 2011, when the majority of these applications will be OS-agnostic, such as Web applications.There are a couple of points I would like to make, though:
- I'm not entirely subscribing to the idea of the slow speed of innovation and development in Windows being such a huge problem. When was the last time we saw some truly spectacular new feature (I mean: "Whoa, that is amazing!" kind of stuff) from any of the main operating systems? I include here not only Windows, but also Linux, OS X, FreeBSD, etc. They are all more or less incremental in their features. Some OSs have smaller release cycles with smaller sets of these incremental features. Others OSs have longer cycles and then dump a larger number of those increments out at that point. In the end ... same difference. Sometimes there are some nifty new applications, of course, but those are not really a feature of the OS. Considering how many people are now demanding that the life of Windows XP be extended, I really don't think that 'lack of innovation' is such a big concern, especially for corporate customers who often prefer stability over leading edge innovation anyway.
- It is true that there are many users (home as well as corporate) that are entirely OS agnostic in ther usage patterns. They browse the Internet via a browser. They read their mail via a browser. They do their communication (IM, Skype, etc.) via apps that are available on a large number of platforms. These users could switch to any of the mainstream OSs at any time without loosing a beat. I have seen this happen with the Asus Eee PC. It runs Linux, but those who use it barely notice or care.
- But the effects of this OS agnosticism apply equally to all operating systems, not just those of Microsoft. True, currently you have to pay for your Windows license, and you can get Linux for free. So, once you have freed your data from the OS and its specific applications, the switch today tends to happen from Windows to Linux, because Linux today is cheaper and runs on cheaper hardware. But once something better and smaller than Linux comes around then these same users can again switch without problem.
- This then is also Microsoft's big chance to stay relevant in the OS market: Make a smaller, lighter, better OS than Linux. People - and most importantly the device manufacturers - will gladly use it then, especially if it is comparable in total cost. As an open source advocate, I wouldn't say people should use it if it is not open source, but no matter what I think, people would use it, as long as it is smaller and better than Linux. Vista is not the ticket here, and extending XPs life span is a stop gap measure. This is also recommended by the Gartner analysts: Make specific versions of Windows for specific use cases. I think Microsoft has very much recognised this and I'm sure they are working on a solution.
In the past, Microsoft's business strategy centred on good applications, which essentially only run on Microsoft's OS and which run on their OS better than any competing application. Usually, where a customer demands the Microsoft application suite, the Microsoft OSs are a must have as well. And often vice versa. Combined with lock-in via proprietary formats, this made for a very good business for Microsoft. So, to continue its business model of combined OS and application relevance, Microsoft has to be an important player in that 'new OS' - the Internet - as well. It has largely missed the boat and allowed the emergence of Google as a dominating player in that field. In light of this it is understandable that Microsoft now wants to buy its way into that market by trying to acquire Yahoo. It is essential for Microsoft's business model that they retain dominance on the platform level.
In the end, I'm convinced they will follow both strategies: Windows XP will continue to fill the gap until a lighter, smaller OS from Microsoft can take up the fight in the 'device OS' market, especially in emerging markets or sub-notebooks. At the same time, they will continue to aggressively pursue a strategy of (re)gaining some relevance in the 'Internet OS'. If they don't succeed in this, Microsoft will be in trouble. But they are a resourceful company with lots of smart people. I'm sure that come 2011 we won't be talking about a 'Windows collapse'. Instead, we will have seen the emergence of a new OS - or even a whole set of specialised OSs - from Microsoft, plus some significant inroads into the Internet OS. Windows - and Microsoft - will not have collapsed, but they will have evolved to better face the new realities.
Other related posts:
PC World: Move your business to Linux, not Vista
And you thought your computer would do what YOU wanted...
Firefox and Google at Microsoft
Comment by freitasm, on 13-Apr-2008 11:25
And an interesting comment from inside...
Comment by uccoffee, on 16-Apr-2008 12:22
MS 's already got a small OS, which is CE!
They have been working on it for years.
And it's finally a bit stablier since WM5
If all app can be run from the web, then MS should be working very hard on its mobile web broswer, not just their OS ....
Comment by bkil, on 14-Jul-2008 23:24
If this is all true then M$ will go online and live happily ever after.
By the way, what is the 'Linux' that is being referred to here so often? Linux is just the drivers and kernel (so is ucLinux), so it would be more precise to compare a distribution to W (and it's third party installs). I have an old-hat link about innovation for you:
There already EXIST many small and light Linux distributions, see the "webpage" link for a list of _some_. Take for example the 50MB DSL that you can do your casual browsing*, e-mail and some other simple office needs with, next to being an almost fully capable Unix-like system.
*:(Perhaps you'd like to install the extension for Java and SWFlash additionally)
I've used that system for extended periods (upto 2007) on a Pentium (1) with 64MB RAM, and most apps run like a charm, some even with 32MB. (Note that I also had a standard Debian Etch install on it that was mediocre, but still useable)
I'm on Debian GNU/Linux now, but I'd be willing to migrate to Debian GNU/Hurd[Coyotos] in the future when it matures.