Lindows.com is a venture to provide an OS that runs both Windows and Linux desktop applications for under $100.

More at http://www.lindows.com/ and http://www.linuxworld.com/ic_718346_6994_1-4001.html

----

I won't be buying any stock in Lindows.  Many Windows users will continue to run only Windows.  Some will avoid it due to perceptions about Linux that may be outdated or wrong, but many others will continue to run Windows because it makes the most sense for their business.  So be it.  It's fair to say that people who avoid Linux will just as likely avoid Lindows, especially since Lindows is proprietary.

Other Windows users will try out the Linux desktop.  Some, like me, will find that Linux provides perfectly good alternatives to Windows for most typical desktop tasks (surfing the web, reading your email, writing letters to your grandmother, etc.), and they will run Linux as their primary desktop, perhaps occasionally booting back to Windows.  Other users will look into software like VmWare.  Unlike Lindows, VmWare does not attempt to compete with the other OS'es.  Instead, VmWare truly enables the best of both worlds, as long as you're willing to throw a little hardware into the mix.  (I have heard, though, that VmWare is quite fast for an emulator, and I've heard good things about the product in general.)

''This is true.  VmWare only virtualizes certain "protected" CPU operations which the OS performs from time to time.  The rest of the code runs at full speed on the real CPU.  Most of the cycles an application spends are running "normal" code, only a small percentage is spent doing these protected operations so it doesn't matter if they are >30 times slower.''

Meanwhile, as Lindows either spirals intro bankruptcy or slowly evolves from being vaporware to being an established OS like Windows and Linux, Linux desktop applications will continue to improve.

Two other factors will undermine desktop-centric OS'es in general.  Increased bandwidth will make server applications more critical.  (I shouldn't oversimplify, though--there will also be more demands on desktop OSes as higher bandwidth enables more video applications, for example.)  Also, dropping hardware costs will make the embedded devices market more important.

Another curious thing may happen.  Enterprises do not like to run multiple operating systems when they can avoid it.  Suppose that server-based and embedded applications become more ubiquitous, and that one OS gains a foothold in both of those markets.  Then, enterprises may feel an incentive to run the same OS on desktops as well, insofar as the benefits of standardization outweigh the substantive differences from competing desktop alternatives.  As application vendors begin to recognize this trend among enterprises, they will begin to target the "standard" operating system more, which may narrow the perceived gap between desktop usability among OSes.  

-- SteveHowell

[later...]

I should also clarify that while I think Lindows has a risky strategy, I would not guarantee failure.  Also, if they succeed at all, they may succeed quite spectacularly.

-- SteveHowell