History of Heath...when computers came on...

[Jack Crenshaw]

Guy G. Giacopuzzi DDS wrote:

> I was at the Los Angeles store when the first H8 arrived.  That would be
> late 1976.  The interest in the machine was incredible, and we had quite
> a few on a "waiting list"  who wanted to be contacted when it arrived.
> This was, as Pensen has said, a "geek" machine, in that there was no
> monitor, just some LED's and a keypad.  As far as I  knew, that was the
> main way of communicating with the microprocessor....and the only way,
> till the peripherals started arriving...tape drives, paper punch
> machines, and the like...

The H8 was definitely a hobbyist's machine, not a personal computer in the sense of the
current PC.  You're wrong about the monitor, though.  The H8 had a serial port (or two, I
forget how many) and could communicate to a Teletype or CRT terminal.

Hardware-wise, the H8 was, I believe, very solid.  It had good reliability, and plenty
enough RAM and stuff.  That front panel you saw, with the LED's and keypad, was a big
plus.  It had a built-in hardware debugger that ran in real time, that could display the
contents of any register, while the CPU was running or stopped.  That is to say, you
could select, say, register HL (or PC, or SP), and you could also stop the processor from
the front panel.  You could then single-step through the code, and watch the data
change.  It was like any hex debugger, but it was implemented in hardware, so couldn't be
defeated or broken by software bugs.

That front panel made debugging a snap.  After I left Heathkit in '79, I went to work for
Honeywell, and found that their $30,000 mini had exactly the same arrangement.

The big problem with the H8 was the software.  CP/M, which rapidly became the industry
standard, had the OS in high RAM, leaving low RAM for the interrupt vectors and the user
program.  Heath used exactly the opposite approach, putting the OS in low RAM and the
user software in high.  That meant that essentially no software written for CP/M,
including the OS itself, could be easily adapted to the H8.

Either way, high or low, the BIOS and other OS bits ate up part of the 64K memory space.
But since the ROM was in low RAM instead of high, it differed from the usual convention.
That's why all the software, such as CP/M and Microsoft BASIC, had to be adapted for the
H8.

Kaypro did it right.  They used a phantom ROM for the BIOS, which could be flipped in &
out by writing to an output bit.  This means that, while system calls were slow (overhead
of flipping the bit), the entire 64K of RAM was available to the OS and user programs,
none req'd by the ROMs.  Given enough time and enough elbow room (read:  Freedom from
obstinate marketeers), I'm sure we would have gotten around to fixing it so the H8 made
all 64K available, also.  Never got the chance.

Jack

Listserver Subscription:listserv at listserv.tempe.gov - "subscribe heath 'name' 'call'"
Listserver Submissions: heath at listserv.tempe.gov
Listserver Unsubscribe: listserv at listserv.tempe.gov - -"signoff heath"