Did anyone read this open letter to us?

[SamuraiCrow]

Also does anyone know which company was making tentative moves to purchase Amiga Inc. ??

DiscreetFX

[SamuraiCrow]

I agree that open source is probably the only possibility of moving forward.

I, for one, disagree with this point.  If it becomes open source then it will have to also become non-profit since it won't have a way to enforce who pays and who doesn't.  AROS is short on developers and it is open source and x86-based as well as running on some PowerPC machines and others.  Bounties are nice but some programmers need a steady paycheck to be able to stay involved with a project for an extended period of time.

I think the only hope for reunifying the factions of the Amiga community is to offer a binary object model that will work on Classic Amiga and new Amiga formats also.  Since AmigaAnywhere 2 will never be ported to the classics or to AROS it is NOT a possibility.  GCC, though open source and ported to the classic Amigas, is not well documented and difficult to work with.  Java's bytecode stinks and is so tailored to one particular programming language that it is no longer a possibility even though it might have been at one time.  That leaves the Low-Level Virtual Machine.

The reason why I think LLVM is useful is that it already supports most non-68000 32-bit processors already and is open source with funding coming in from Apple and Adobe.  It supports JIT, interpreted, and static compilation of code through the same library.  Also, even if it doesn't unify the Amiga community, it will offer a JIT compiler for UAE and others that is written in C++ rather than PowerPC or x86 assembly so it can be ported from anywhere to anywhere else.  Lastly, Sidewinder and I am already targeting the LLVM for Mattathias BASIC, our sequel to AmosPro and other versions of BASIC to be supported later by use of extensions to the Mattathias language.

[SamuraiCrow]

Ok.  Well, who's going to make profits selling AROS?  It only fares well when compared to other 3rd party operating systems.  When compared to Linux it doesn't do so well, nor does any other Amiga-like operating system.  All Amiga's got going for them now is small memory footprint.

AROS can run hosted on other operating systems or under virtualization tools like VMWare so maybe it might catch on.  But it's going to take more than the bounty system to make it happen.  It's still got to compete with the BeOS-like operating systems like Haiku as well.

The point I'm trying to make is that there are lots of operating systems around but not so many decent emulators that will generate native code for their host operating system.  AmigaAnywhere wasn't a bad idea, it just was poorly instrumented.

The solution I proposed is a for-profit open-source venture as well.  And AFAIK it's making people money by the bucketload.  LLVM was used to emulate shaders in older Macintoshes and the Mesa graphics library under other platforms now the Mac is catching on big time with Linux gaining users all of the time!

Don't get me wrong, I think a highly optimized operating system is important, I just don't think it's enough of a novelty yet to catch on.