3 Jul
1999
3 Jul
'99
5:11 p.m.
There's good reason to cast about for a general purpose bus on the order of
the middle-period S-100. The PC's of tomorrow will have no expansion slots
at all and will rely on USB, SCSI, and the various parallel port protocols
to do "practical" I/O. That will be very limiting. I don't know what
folks
will do in cases where they have measurements, telememtry, process control
tasks, or whatever to do. The PC has never been particularly well suited
for such tasks, since there were such meager offerings in the way of general
purpose I/O.
The common microcontroller setups are pretty costly, often much more so than
a Pentium based PC with thousands of times the computing power.
By the time it was standardized, the S-100 was pretty well settled.
Unfortunately, the amateur computer enthusiasts presented a bigger market
than the measurement and control people, of whom I was one, and so it was
pretty important to be able to build one's own interface circuits. For
that, the S-100 was not as friendly as it could have been. Because of the
legacy of 8080 signals and signal timing, even though the system usually had
a sensible processor which could have worked very well, there tended to be
glitches as caused by the fact that it took maybe three signals and a
decoder to sense a local I/O cycle, yet the bus provided six or seven, and
various board makers didn't use the same ones, nor did they use them in the
same way. That's a mistake that should be avoided in the future. By
contrast, the Multibus-I had signals somewhat similar to those on the ISA,
and they were simple, easy to understand, and so on. Of course Intel led
the charge on MB-I.
See comments below, plz
Dick
-----Original Message-----
From: Allison J Parent <allisonp(a)world.std.com
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu Date: Saturday, July
03, 1999 10:14 PM
Subject: Re: OT: A call to arms (sort of)
In reality, building an I/O mux onto a current
generation PC parallel port
makes as much sense as anything. With EPP you can get up to 2MB of transfer
bandwidth, in bursts, of course. That's not bad . . . AND you have a
"real"
computer with "real" tools that's very fast and "real" cheap.
>Treat the card as a functional black box.
Herc, CGA and VGA video is well
>enough known and the addresses are not secret. It's not a requirement to
>knwo the tiny design details of the 8042 keyboard controller to get it to
>give keycodes. Most of the floppies are the base 765 circuits pushed into
>a chip, same for serial and IDE is not a secret. Based on what I've seen
>of some of those cards the less I know the better!
>Allison
<Why do you assume that ISA -> Intel processor?
It may be something
<totally different, something that doesn't have efficient block transfer
<instructions.
Like a z280?
You must really love that chip, Allison, but yes, even that, if you wish.
It's YOUR computer, after all, so it should be the way YOU like it.
<I see a _lot_ wrong with the ISA signal definitions. For one thing the
<IRQs are edge triggered, active high, when any sane designer would make
<them level triggered active low (had IBM done this it would have cost
<them an extra couple of TTL chips on the PC motherboard. It would also
<have allowed the sharing of interrupts). For another thing there's no
<proper bus request/grant (multiple masters are almost essential IMHO).
So the interrupts are upsidedown and stoopid, it's useful as is none the
less. The yabut is for small systems it's fine.
<As I understand it, the aim is to make a PC (meaning something that runs
<a useful open OS like linux or *BSD) and which has 'modern' features like
<a good video card. Not to make the equivalent on an S100 system
Consider possibility number 3, something that is hybrid, having the
features
of S-100 like system but modern I/O and a different
bus.
I personally would favor the 96-pin connector (per DIN 41612) as used in
VME, but only one, for a basic card and make it on the nominally 4.5 x 6"
form factor of the single slot EUROCARDS (e.g. VME). That connector is more
reliable than card-edge connectors and it's used enough that it's relatively
cheap. It's compatible with a 0.100" matrix so a card and a backplane could
( if you were REALLY desperate and impoverished ) be made by hand on a good
wire-wrap card. Those are a few pretty compelling reasons. Of course
unless you actually adopted the VME standard, you'd still be on your own and
unable to buy a serial card or such.
<> known, and one doesn't need a video board right off the top. The
WD1003-
<> board is well uderstood and the EIDE interface
emulates that pretty well
<
<Sure. Now where do you propose getting schematics for this I/O card, and
<where are you going to get a data sheet on the ASIC that almost certainly
<appears on it. This is supposed to be _open_ hardware. This implies full
<schematics, not undocumented PCBs.
This is where I'd recommend caution if you use q-bus. The schematics must
not only be available and complete, but they must be correct as well. PALs
must be fully characterized, something I've never seen in a DEC product,
and, in fact, I'd say you have MUCH less "open" information about q-bus
than
about ISA. The problem with ISA is that the information was usually "out
there" well in advance and then, when the product was shipped, wasn't
available any more, because there were too many competitors and the doc cost
an extra few pennies.