Mar 25 2004
Who will Longhorn appeal to?
Right now it seems to me that Longhorn is being targeted at three communities:
- Home users, particularly those looking for a great multi-media experience
- Knowledge workers, especially those at the top end, who aggregate, integrate and assemble lots of information from many different sources
- Mobile workers, for whom thin client computing solutions don’t work and to whom the blend or personal and corporate features will appeal.
It’s got lots of other features that will appeal to the mass of task and structured task workers in corporate environments, but true thin client approaches will probably appeal more strongly for these users IT managers, particularly with the current wave of smart client rich UI toolkits that run on top of a JVM.
So how might this pan out in reality:
- Microsoft might get 20% market share from portable users
- A maximum of 20% market share from high end knowledge workers, who are not mobile
- Maybe 20% that it picks up just so they can use the same environment as the rest of the people in the office
So maybe that leaves 40% of users who will either switch to thin clients, unless Microsoft can convince businesses to stick with them because of the benefits of standards, or consistency with their home environment. A tough sell as lots of them will probably want to ‘prove a point’ to Microsoft.
If it’s going to achieve capturing this wavering 40% what do Microsoft really need to do:
- Make sure they DO capture the home market in a big way with Longhorn
- Make sure they convince their current huge number of ISVs to create Longhorn applications
- Make sure that Longhorn and Office 12 really do deliver compelling benefits right out of the box to the simple stuff that real people do every day and not just the stuff that appeals to the techno geeks.
What about the competition:
- Well for me Linux is still trying to get where Microsoft is today in terms of doing the basic things VERY simply and easily and its going to take a couple more years to get there, and even then its unlikely to get the home market penetration it needs
- Linux does not have the application portfolio to allow it to compete head on in current migrations, and has an incredible battle on its hands if its to get that application portfolio in place in Longhorn timescales
- If the above are true then Linux is really competing for the Task and Structured task workers and the Developers, pretty much the same market that Star Office is going after today. But this is the same market that the thin client device is going after.
My view:
- I think Windows, in the form of Longhorn will continue as the main stream general purpose client OS
- I think Linux on the desktop will fail to complete with thin clients
- I think that Linux will succeed in the special purpose client OS space, much the same as it is in the special purpose server space today
- I think Microsoft has a HUGE battle on its hands to convince big enterprises that it can be trusted OR that its value proposition is so compelling that they can afford to take the risk.
I admit that I like some of the stuff that Microsoft is doing, but they have a long way to go before they convince me on the last point. Unfortunately change is required, and if Microsoft was not doing Longhorn or Longhorn did no succeed who would have the ability to drive the change through.
As a final thought, look at the web, and the very small number of things that it’s possible to assume as universally available reliable standards. DNS, TCP/IP, HTTP, HTML 3.2? maybe a bit more at the network level, but not much over a period of decades. Then look at what Microsoft are trying to establish as a pervasive set of services available on every client! However maybe that’s why nothing should be done at the client and it should all be server side, but that’s the subject of a different debate. As a lead in to the inevitable questions, I think that what Linux, or perhaps more accurately GNU has achieved in terms of a pervasive set of services available on every client is incredible so far!
