MS years ago was getting worried Adobe was inching into the cross platform space (remember Netscape Navigator?), putting a final nail into its Windows franchise. Therefore it launched its own plugin initially called WPFe, which runs a tiny CLR (about 5Mb) with a subset of WPF, targetting WinXP and up plus MacOS. It is based around scenegraphs rather than Repaint(), and promises to make application design more designer friendly.

All this changed when MS realized that Windows has already lost. The web, where we live today, is already cross platform.

Meanwhile, the now-named Silverlight isn't attracting the right developers/desingers - Adobe tools still retains people who love UI design, while Windows developers are primarily still trapped in the WIMP mindset.

Secondly, Apple showed MS that cross platform wasn't necessarily a threat. The threat today is a device that can browse the web and perhaps run some native apps. The cross platform adoption never really took off. Flash based games continue to dominate. Mac aficionados never liked Silverlight. Android tablets don't run Silverlight, and why would MS want to help a competitor anyway?

Meanwhile, the original stable of enterprise customers started using Silverlight in their enterprise apps, citing simpler deployment. In Silverlight 5, MS has finally admitted more Windows specific features like PInvoke, COM automation - all this is enterprise related.

So today, Silverlight never achieved its original cross platform promise, it never achieved consumer adoption, it didn't attract (enough) developers who made beautiful apps, Apple is decidedly against other runtimes running on their newly established beachheads in the iPads. Sales growth of non-PC devices outpaced PC sales. If you include devices (smartphones, slates), MS has already lost its dominance. Plus with ARM devices continuing to get more powerful - there is no future left for a tiny, portable CLR.