24 Dec
2016
24 Dec
'16
10:39 p.m.
On Sat, Dec 24, 2016 at 8:10 AM, Peter Corlett <abuse at cabal.org.uk> wrote:
On Fri, Dec 23, 2016 at 04:14:28AM -0700, Eric Smith
wrote:
The details aren't identical, but on the 6502 it's abusing the fetch of
both an opcode and an immediate operand.
If there was nothing in it, then it's a mystery why the Z80 was selected.
The same trick works perfectly well with a 6502,
and in fact was
invented by
Don Lancaster using a 6502 years before the ZX80
was designed. That
doesn't
really explain the choice of the Z80.
Interesting. I was of the understanding that the Z80 video hack didn't
work on
the 6502 due to the latter being slightly pipelined and so the instruction
fetch cycle couldn't be abused in the same way. Perhaps Sinclair couldn't
get
it to work on some dodgy 6502s that fell off the back of a lorry.
However, Sinclair wasn't exactly one for making
rational design decisions
based
on technical merit or industry best practice.
As much as I like the 6502, the Z80 is better in some ways, particularly if
you want to do a lot of 16-bit add and subtract.