, wkempf wrote

*snip*

It was very competitive early on. Then the iPhone showed people what a touch interface could do for them. WinMo couldn't compete against that. So Balmer made the correct decision there, though one could easily argue he did it late and probably had no choice.

I think the problem all along has been Microsoft wanting to transitioning from a company that licenses software for a one time charge to one that gets a monthly income stream from the user's of  software. Device OS software is not hard to write. It just takes a lot of programmers, trial and error. If .NET turns out to not be performant enough have other approaches in the works. But the message from the top of the company has to be that MSFT is going to have an OS for every device out there. Wherever possible apps for one device will work or can be ported to another.