Mixed Signals
Any sufficiently advanced information is indistinguishable from noise
Monday 22 October 2007, 1:45 PM
Microsoft's death knell for Windows Mobile
Over on news.com, there's a story about a rare insight into Microsoft's ongoing Windows development. As part of the process of re-engineering the kernel code, the basic heart of the operating system, the Windows team has produced a very small (for Windows) kernel that obeys the classic OS design rules. Take the essential services, keep them lean and simple, and put them in their own box.
This is significant, not because it's an astonishing innovation -- the idea of kernels can be traced back to the 1960s, and has been at the heart of just about every mainstream OS -- but because the Windows team has managed to sanitise Windows enough to let it join the club. And at just 25MB, that MinWin kernel is a respectable effort.
But this will only increase the pressure on an internal MS faultline that's been there for a while - why bother with Windows Mobile? I know that this is a hot topic inside the company, with strident support for and against the idea of merging this last hold-out with the mainstream Windows development effort.
With MinWin, one of the big arguments about embedding Windows proper in hand-held devices goes away. It''s no longer too fat to fit. In an age where a gigabyte of Flash costs about the same as a packet of washing powder, 25 MB is plenty small enough.
The other argument, that hand-held devices are too architecturally different to the PC model, still holds water. But it's leaking badly: Intel's roadmap for portable x86 chips leads the PC model on a path where it becomes a real competitor for the ARM-based standard that currently rules. That battle has yet to be joined, let alone won, but I find the arguments compelling: once you have a proper, long-lived handheld device that can load and run all the x86 software which drives the client side of the Web, you don't want to mess with anything else.
And of course, there's Linux - which already runs on hand-held devices without much fuss. That'll be in a much stronger position if and when the x86 architecture becomes a potent part of portable life. Much the same can be said of OS X, which has already started on its own journey into the pocket. Does Microsoft really want to have to support two mutually competitive operating system choices against that class of unified alternative?
I don't think there's more than one answer to this question.
Windows Mobile. Dead OS running.
PreviousComments on this post
PS.
The timing of this post may seem perverse. Microsoft is, as of today, embarking on a major WIndows Mobile campaign, with Ballmer set to make a big announcement later.
I'll return when the smoke from that dies down , but perhaps I should make myself a bit clearer.
I don't expect the Windows Mobile name to go away. Quite the opposite. But it won't be the mutant non-Windows Windows Mobile we know today: it'll be real Windows in a mobile format, with a shared codebase.
