A Multibus board just sold for over $2000

[Eric Smith]

On Tue, Nov 8, 2016 at 9:21 PM, dwight <dkelvey at hotmail.com> wrote:

> For a 432 board, I'm not all that surprised.
> It only needs the 43203 board to be a system.
>

The iSBC 432/100 doesn't actually need a 43203, nor can it work with one.

The iSBC 432/100 was an evaluation board that only ran OPL-432 ("Object
Programming Language"), a dialect of Smalltalk.  It was not suitable for
running "normal" 432 software, which was written in Ada and used the
iMAX-432 operating system.

WIth OPL-432, the 8085 in the MDS performed all I/O and supervisory
functions without a 43203 Interface Processor.

OPL-432 is *only* able to run with the Release 1 432 components, while the
"normal" 432 software only was supported on later releases.  Release 1 had
both architectural and implementation deficiencies, which were mostly
corrected by Release 2.1, but OPL-432 was never revised for Release 2 and
later.

The "real" 432 system was the 432/600 series, with the 432/670 being the
standard configuration. The 432/6xx boards had the same form factor as
Multibus, but except for the 432/602 Interface Processor board, did not
plug into Multibus. The 432/600 series had its own 32-bit backplane bus.

Eric S. would have loved to get that board.
>

I have one complete 432/100 board, though I don't yet know with certainty
whether it works. Unfortunately no one has turned up a copy of the 432/100
software. I'm working on my own software for it. The 432 has an object
oriented architecture, and requires quite a few complex system objects to
be set up in memory before it can execute a single instruction. I've been
working for a few years on a 432 Release 1 image builder, which will take
an XML description of a 432 software system and build a binary image for
execution on the actual hardware. The work in progress is on github, but
there's a lot yet to be done on it before it will be useful.

I would like to get a second 432/100 board to have a better chance of
having one working, but also because I think the only real chance of
building an accurate 432 simulator will be to decap a set of 42301 and
43202 GDP components and dump the PLA contents.  If I had two sets, after I
had verified that one worked correctly, I'd arrange for the second, working
or not, to be used for decap and photomicrograph.

I expected that this board might go for over $3000. In better times, I
would have been willing to pay that.


> I suspect 432 stuff is even rarer than Intel bouble memory.
>

Yes, very much so.