Well, you then have to ask yourself if we are talking about the Java as a virtual machine, or Java as a language. I don't really like either one to be honest, but that is besides the point.
Java is pretty much designed to be a language, not a universal virtual machine. Otherwise, you'd have to start comparing TAO to things like PERL. After all, PERL is a bytecode interpretter just like Java, and its ported to more architectures.
If TAO could live up to the hype, I'd consider it a bit more advanced than Java. While you could technically create a C compiler than creates Java code and integrate a java JIT into the OS, having OS integration (including things like device drivers that java was never intended for and specifically is designed to NOT be able to do) designed in and ready to compile your favorite C apps is kinda nice.
Anyway, this is all way off topic. All I said was that I felt that Amiga had dropped the ball by not pursuing the technology within the original timeframe ... off by a few years and still counting! It looked like they had positioned themselves to getting real 3rd party development again. I think by selling the development kits, they should have comitted to the project instead of shelving it for some indefinate length of time. By catering to the existing hobbyist crowd and shelving the tao project and shafting everyone that purchased developer kits for an OS that may never get released they may have seriously hurt any chance of getting commercial support.
The underdog OS crowd can't afford to treat third party developers the way MicroSoft does.
