Just to note for anyone who finds this forum, and I know its been said before:
Windows won't run on the Pandora because Windows is at fault, not because the Pandora is at fault.
It's Micro$oft's fault you can't run Windows on a Pandora not ours/the Pandora's.
If Micro$oft wants to keep market share in future, it needs to support ARM. ARM does not need to support Windows (though an x86-compatible mode would be pretty sweet though just contradictory since the ARM is a RISC processor, including x86 (CISC) support as just something for compatibility is a huge waste of silicon)
I do hope Microsoft gets left behind in the ruins of x86 because it would be nice to make a clean break into a world running free and open source software, which would be a much happier place. And well, it would serve M$ right for STILL using their 16-bit MS-DOS code in the 64-bit seventh generation of Windows. I mean, what? Nobody say compatibility or I will shoot them in the foot. Compatibility is nonexistant.
For a bit of info on the x86 architecture, it was built up from an early 16bit Intel processor, the 8086, released in 1978. The 286(80286) was an upgrade of the 8086 which was much the same processor but included features like an integrated Memory Mangement Unit and protected mode as well as a larger address space (i.e. more memory can be used with it). Then came the 386 which had a 32-bit instruction set. Then the 486 with integrated cache and Floating Point Unit (aka math unit). then the Pentium which was superscalar(could execute more than 1 instruction a clock cycle) had a faster FPU and the MMX instruction set. And so on. All based on an original 1978 core.
However the ARM architecture, it seems more crap cleaning is done between releases (I can't be bothered proving this). And as I mention below it was 32bit to begin with so it never had to rely on ancient code to bootload.
The Acorn RISC Machine (ARM1) was RISC from beginning because they realised that with CISC architectures that old code would always run slow and never really improve, as well as compiler design needing to be a lot more complex to produce faster code, so they reduced the number of instructions in the set. This processor was first made on 26 April 1985. It was 32bit but all program code had to be within (the first?) 64MB of memory due to the program counter being only 26bits. ARM chips seem to only have been used in Acorn's machines before becoming Advanced RISC Machines. Which is disappointing for me personally as I wouldn't mind having a few older ARMs for my DIY stuff
.
The downside of ARM is that older code is pretty incompatible on newer ARMs.
But you shouldn't listen to me as i'm probably biased as I love ARM - and i've written an x86 OS(more like a shell 'on a stick') so I have reasons to hate x86
.