William Donzelli wrote:
Perhaps I just
have less faith than you in the technology being
available in
20 years to probe inside modern systems to figure out how they work
and keep
them running :-)
Do you think that tools will not evolve in 20 years?
Of course they will. But I suspect that there comes a point when you just
can't practically figure out the function of some unknown chip for every
possible state and input condition in order to emulate it - at least not in
reasonable timescales and at reasonable expense.
Think of trying now to reverse-engineer something like a 68000 CPU given just
a faulty example of the chip and no supporting documentation whatsoever.
I don't think that problem will get any harder or easier over the years; the
tools readily available will track the advances in hardware fairly closely.
But the point is that reverse-engineering a "complex" undocumented chip (from
any era) is a difficult task at best, and I think it always will be. We're
just fortunate at the moment in that *generally speaking* for "our" vintage
equipment there's often documentation available, and any given chip found a
home in lots of different systems, so there's a better chance of finding a
straight replacement.
cheers
Jules