-=> On 25 Jan 96  22:46:00 David Noon said to John Poltorak <=-

Hi David,

 DN> I was also brought up on IBM mainframe assembler -- over 20 years ago!
 DN> However, in the 1970's I also encountered the worst range of machines
 DN> ever built, the DEC PDP-11, and learned its assembler, Macro-11.

I really won't have this blasphemy here. The PDP-11 was one of the best
machines ever built, and without it we's not have had CPM, DOS or UNIX.
Macro-11 was easier to use than Intel assemblers. (I concede that the
memory management was the definitive 'kluge', but unless you were
writing disk drivers, you'd probably not need to know that.)

 DN> Since the designers at Intel all worshipped DEC, the Intel has the
 DN> same dumb little-endian architecture, very similar instruction
 DN> repertoire, same word = 2 bytes, and its assembler language is almost
 DN> a clone of Macro-11.

Little endian is logically correct, as it expands correctly when you add
more digits. It is big endian that was a horrendous mistake brought on
by the europeans adopting the arabic numbering system, and despite the
fact that arabs write right to left, while europeans write the other way,
the europenas failed to notice they needed to switch the digit ordering.

 DN> So, I guess the PDP-11 was useful after all. It gave me some
 DN> preparation for assembler under PC-DOS, Windows and OS/2.

It also brought us the hardware stack and popularised the concepts of
bus structure and open architecture.

I think John meant ICL mainframe. The ICL machines had architecture that
was an assembler programmer's dream. (Apart from the complete impossibility
of handling ASCII codes with it due to 6-bits per char.) It was also a great
target machine for code generation.

I could probably write an emulator for it without access to the docs if
anybody was preapared to pay me.

Andrew


... Monday is the root of all evil!
--- Blue Wave/Max v2.12 OS/2 [NR]