What would Windows be without Win32?

Would Windows still be Windows if Microsoft yanked out the Win32 programming interface and User GDI graphics-device-interface layers and replaced them with .Net managed code?

That's not a purely hypothetical question; there are a couple of skunkworks projects indside Microsoft investigating that very idea. And if the results of their incubations pan out, the first fruits of their labors could become part of Windows, maybe as soon as Windows 8.

I wrote a bit about these incubation projects -- codenamed "RedHawk" and "MinSafe" -- back in July. Since that time, I've gotten a bit more information on these projects and have summarized my findings in a new article I wrote for Redmond Developer News. (And just to be clear, Microsoft officials are still unwilling to talk to me about any of these incubation projects; I am hearing about them from various, unnamed but knowledgeable sources.)

A quick refresher: RedHawk and MinSafe (two different code names for essentially the same thing) are projects from Microsoft's Developer Division and Windows unit, respectively. The pair are paving the way for Midori, the distributed operating system project under development by Microsoft Senior VP of Technical Strategy Eric Rudder & Co. The word is that the Midori folks are looking to RedHawk as the best way for Midori to get a commercial back-end compiler and minimal runtime.

According to my sources, MinSafe is being led by Microsoft Technical Fellow Mark Russinovich. RedHawk's champion is Technical Fellow Patrick Dussud. Both projects are working with Jon DeVaan, head of Microsoft's Core Operating Systems Division.

I heard from one of my sources recently that it's not coincidental that MinSafe and MinWin have similar names. MinWin -- Microsoft's project to create a slimmed-down Windows core -- is related to MinSafe. If I were a betting woman, I'd guess that MinSafe is one piece of MinWin, and most likely, the piece most likely to be commercialized first.

Microsoft's goal with RedHawk, MinSafe and MinWin seems to be to find a way to make Windows less bloated and more manageable, while creating as little disruption as possible for its developers and customers. Based on some of the new information on the RedHawk and MinSafe projects, do you think Microsoft can pull this off?