In article , Jules Richardson wrote:
On Sat, 10 Mar 2012 13:33:46 +0000, John Rumm wrote:
;-) Rather like talking with IMB mid range / mainframe types who seem
convinced that Assembler (note the capital A!) can only be known to
hallowed IBM programmers, and can't seem to get their heads round the
fact that every processor has a low level assembly language...

Every? Depends on the definition of assembly language, I suppose - is it
the binary opcodes which form CPU instructions, or is it the human-
readable mnemonics which correspond to those opcodes?

I usually think of assembly language as the latter (that being what
people program at the low level using), but I suppose it's entirely
possible to have a CPU where there are no official published mnemonics
and instead things are documented in a more long-winded "binary opcode xx
performs operation yy" form. You'd hand-craft[1] binary directly, or
perhaps cross-compile using a completely different system.

It's possible in theory, but I find it hard to believe that such a
system would last for long before someone produced an assembler.
(Or four or five incompatible ones - how many of
http://en.wikibooks.org/wiki/X86_Ass...x86_Assemblers
would you describe as "official"?)

(I did wonder about FPGAs, but while an HDL isn't a conventional
assembly language, an FPGA isn't a conventional processor either.)