please see imbedded comments below.
Dick
-----Original Message-----
From: allisonp_at_world.std.com <allisonp_at_world.std.com>
To: Discussion re-collecting of classic computers
<classiccmp_at_u.washington.edu>
Date: Friday, April 09, 1999 7:18 AM
Subject: Re: stepping machanism of Apple Disk ][ drive (was Re: Heatkit 51/4
floppies)
>On Fri, 9 Apr 1999, Richard Erlacher wrote:
>
>> Careful, now! He would have played hell trying to interleave memory
>> accesses between an 8080 and the video refresh process, since its various
>> cycle types were so different. It would have been worse YET with a Z-80!
>
>It was done though. The 6845 you generally talk to it's local ram through
>it not around it. The H19 terminal did the latter.
It was, indeed, but not when the Apple was designed, since the 6845 and 6545
didn't yet exist.
>> The 6502 also allowed him to proceed with his own DOS and his OWN version
of
>> BASIC, without which he mightn't have gotten the strangle-hold on the
>> personal-computers-in-business market. It's pretty hard to criticize his
>> choices, however little I liked the result from the standpoint of seeing
it
>> as a tool, but his (and his partner's) decisions were definitely
vindicated
>> in the marketplace.
>
>It proved software was more important than hardware. The best cpu with no
>software was still nothing. The z80 was maybe the best at the moment but
>the 6502 was as versitile and plenty fast enough to make up for it's
>limitations.
There were many people who disagreed with that performance comparison. The
two processors (I used them both extensively) were different enough that
comparison and contrast was not easy. The Z-80 had many registers and a
rich instruction set of which much was awkward and difficult to use. The
6502 had fewer registers and fewer instructions but they were amplified by
addressing modes not readily available to the Z-80 user, particularly if his
code was for CP/M which sort-of required you stick to the 8080-compatible
instruction subset. The fastest 6502 available in 1980 was a 4 MHz
processor which stroked memory for 125 ns almost every cycle in its typical
application. The Z-80 had just that year become available in a "B" version
(6 MHz) which stroked memory for slightly less than 3 clock ticks on a
memory cycle and barely 1 (166 ns) on an unmodified M1 cycle. Having said
that, and given (1) that the two processors executed code at about the same
rate for the same memory cycle length (hard to prove or disprove) and (2)
memory cost was more of an issue than system performance, hence the memory
cost was used to set the rate at which the CPU operated, I'd say there's
really no basis for comparing the actual performance of the two processors.
>Likewise the trs 80 would prove lowercase and some more speed were very
>desireable (based on the two most common mods!).
The TRS-80 could have been put out with (1) an 80x24 display rather than the
16x64, it could have solidly supported double-density FD's (in the model 3)
and it could have operated at about 4MHz rather than the 2.-something it
used, and it could have switched in and out the ROM so it could run CP/M but
for the greed of Tandy Corp. It would have cost them an additional $5 and
change to put those features in their model 3, but they thought "well, we're
going to sell a million of these . . . " and decided they preferred having
the dollars. Had they gone the other way, they probably would have had the
"personal computer" market all to themselves. They had a huge distribution
network, a huge sales force, a huge service network, unlike any other
microcomputer manufacturer of the time. Within a year, Apple owned the
personal computer market, particularly with respect to businesses, even
though Radio Shack had better packaging by the time the Model 3 came out.
That was also about the time it became common to see the 80x24 displays, the
8"disk drives, the nearly 4MHz Z-80 running CP/M, the extra 16K memory, etc.
on the Apple II. If Tandy had gone with the better design, which was on the
table, there probably would be no IBM PC today.
>Allison
>
>
Received on Fri Apr 09 1999 - 10:21:36 BST