There is no Santa Claus. There is no Easter Bunny. Tinkerbell will not come back to life if you clap hard enough. And your OS, that you love so much, is dead. It is time to acknowledge and move on.
Without tons of software, and I mean NATIVE software, not just crude ports, no platform will prosper. And to get those, you need the hearts of developers.
Rudi got it right that developers are the most important thing. OS/2 development was nigh-impossible for independant developers (as I found out the hard way), and the main complaint I heard about Be was that it was difficult to design for. The #1 force driving Microsoft's OS-substitutes is MSDN and Visual *. It's dead easy to develop for them, so everyone does. Linux and FreeBSD already have the tools and information really hardcore developers want, and are finally getting decent IDEs for the less-hardcore (after 30, your brain turns to tapioca and you need a point-and-drool programming interface to survive, or you'll have to go into management. Yeah, I didn't believe it would happen to me, either).
I'm currently using VisualAge for Java in Linux instead of my usual bash+make+vim+jikes+jdk, and speed and memory issues aside, it's incredibly powerful, and IBM's actually committed to supporting Linux developers. The Java metaplatform has largely succeeded, IMO, not because of the hype, nor even because it's a good technology (it is, but we just got lucky, and had a lot of money thrown at the problem to make better JITs), but because it's incredibly well-documented. There are ten thousand Java programming books, from beginner to obscure expert deep magic stuff. When I was programming for OS/2, there were a half-dozen books on it, and then the $$$ pricey developer connection from IBM. And that was it. Look at the number of Mac books - not many, but more than that, and there are a few different dev env's for it.
His repeated refrain of "pray for ..." is just the death rattle. When you hear that kind of thing, run for the hills and learn a new OS.
-- Mark Hughes