Article 163012 of comp.os.vms:
Linwood Ferguson <Ferguson@mag.aramark.com> writes:

>Gregory Travis wrote:

>> The 6600 was a beautiful machine.  I can't imagine there ever being anything
>> remotely like it again.

>And you're really showing you're your age if you can place the phrase:

>"This book is dedicated to A6 and A7 without whom none of this could
>have been saved".  (+/- my failing non-parity memory).

Then you can only be talking about "Assembly Language Programming for
the Control Data 6000 Series and the Cyber 70 Series" by Ralph Grishman
(1974).  A fantastic book and the book from which I learned COMPASS back in
the 1970s.

And I'm "only" 33!

greg

p.s.  The actual quote is "This book is dedicated to A6 & A7,
without which none of the results in this book could have been saved"

p.p.s.  For the great unwashed, the 6000 series had three distinct
registers:

The first set was B0-B7.  These were 18-bit general-purpose registers
usually used for temporary address storage, boolean values, conditional
registers, or as loop counters.

The second was X0-X7 which were 60-bit general-purpose registers which
were generally used for shift/mask and arithmetic operations

The third was A0-A7 which were purely address registers.  However, loading
an 18-bit value into A1-A5 caused the associated "X" register (X1-X5) to
be LOADED with the contents of the memory location which corresponded
to the address loaded into the "A" register.  Moving a value beyond your
program's "field length" into a A register resulted in what UNIX guys
would call a segfault.

Loading a value into A6 or A7 caused the value of the associated X register
to be STORED at the address specified in A6 or A7.

X0 and A0 were not connected to each other in either of these ways.  Loading
a value into A0 caused nothing to happen with X0.  A0 was used as a general
purpose variable to pass addresses to subroutines (such as the address in
which to store return values)

Ahh, for the days.
--

greg		greg@indiana.edu	http://gtravis.ucs.indiana.edu/