Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!decvax!harpo!seismo!hao!hplabs!sri-unix!Smith@cmu-cs-c
From: Smith@cmu-cs-c@sri-unix.UUCP
Newsgroups: net.micro
Subject: Ah, yes, the Pdp-7.
Message-ID: <1933@sri-arpa.UUCP>
Date: Wed, 8-Jun-83 10:06:00 EDT
Article-I.D.: sri-arpa.1933
Posted: Wed Jun  8 10:06:00 1983
Date-Received: Mon, 13-Jun-83 11:09:46 EDT
Lines: 31

From:  David Smith 


	It is slander to say the PDP-7 (and 1,4,9,15) looked like the PDP-8.
	I admit that one-address instructions aren't very elegant, and a
	single AC is a bother, but at least these machines didn't have
	those god-awful 128 word "pages"...

I think you should use the word "slander" a little more advisedly.

When I said the Pdp-7 looked a lot like an 18-bit Pdp-8, I assumed that
the perceptive reader would not think I was implying that DEC wasted the
extra six bits in the instruction word.  Indeed, they raised the "page"
size from 128 words to 8192 (while dispensing with the zero/current
page bit).  They also expanded the opcode from 3 to 4 bits.  That afforded
them the luxury of instructions for "load accumulator," "execute out-of-line
instruction," and "one's complement add" (i.e., with carry out of the high
bit being added into the low bit).  They still had no subtract instruction.
The subroutine jump instruction was the same: store the return address in
the first word of the subroutine, start executing at the second.  You had
to go through the same rigamarole to pick up subroutine parameters.  The 
"operate group" instructions (skip if zero accumulator, etc.) looked the
same.  The I/O instructions were very similar.  The extended arithmetic
option added more instructions than the Pdp-8's, but they were of a similar
flavor.  Either machine's option added a multiplier-quotient register.
With either one, you had to store the multiplier or divisor into the
instruction stream.

Now, I suggest you apologize for your use of the word slander.

		David Smith