Wither Natami?

[freqmax]

@bloodline:
UAE is an software emulation. And software just won't get it cycle exact to realtime. And for some applications software emulation won't cut it.

@mikej:
As for 68000 softcore, someone mentioned a successfull run at 27 MHz. Maybe we can go higher with optimization.

I agree with mikej that expandability is desireable. However I think the real point is that of ral-clan "We desperately need fresh, healthy new Amiga clone hardware, soon.". And this re-implementation should occour before hardware to replicate is all non functional.

The second goal should be to eliminate any dependency of companies or persons that goes out of business etc.. A tarball of HDL sources won't "die" like physical hardware can.

[freqmax]

I think the point being made here is that Minimig/NatAmi/CloneA etc.. isn't doesn't match the behaviour of the original correctly to 100%.

But that said, authors of all those project try to do their best in that respect. So if anyone knows an efficient method that also can be carried out then please tell us.
The original chipmasks are most likely *gone*. But one could possibly xray the existing chips. And then analyze the matrix with software. Another possibility is to make some kind of analyser software that pokes the hardware. Another one is clocks in sync where differencies are logged etc..

Considering that the Amiga development is so far behind the only point personally is to be able to run old legacy software.

Another perspective is to make use of the "Most out of the available hardware" philosophy (at a specific price level). Then it would be multi-core, microkernel, etc.. Without ties to the original hw/sw except for the philosophy and ideas.