In development. Inspired by the Spare Time Gizmos STG1861, but not
based on that design. Rev. 0, not yet ready for production:
https://www.flickr.com/photos/22368471 at N04/sets/72157667299828132
It is 2.0 inch by 0.7 inch, with a 24-pin round-pin DIP header to plug
into a normal 24-pin DIP socket (vs. more the common square pins that
won't work with normal IC sockets).
The surface-mount components were assembled onto the board by a
commercial service, which does not do through-hole, so I had to solder
the DIP header by hand. I had to make the pads for the DIP header
very small to squeeze the TQFP CPLD between the rows, so it turns out
to be unsuitable for hand assembly by novices. Since I am not willing
to do the hand assembly for other people, I'm not sure whether this
board would actually be worth selling; I might have too many customers
that aren't able to assemble it successfully.
The CPLD programming is done by a "Tag Connect", which uses pogo pins
to contact the ten gold pads seen on the top of the board. There are
holes near those pads for the Tag Connect's steel alignment pins;
while there is enough clearance on the top of the board, I failed to
consider that the frame of the DIP header on the bottom of the board
would prevent two of the alignment pins from extending far enough. I
had to cut out part of the DIP header frame.
The CPLD code has been written but has not yet been debugged.
From: John Willis
Sent: Thursday, April 21, 2016 8:02 AM
> That's another thing I remember and miss from those days... your average
> ISP would provide NNTP and UNIX shell accounts, as well as a few megs of
> space to put up a personal web site in ~/public_html.
I still read Usenet newsgroups via GNUS under Emacs on my shell account on
Panix, an ISP located in Manhattan, and have a small web site hosted there
as well:
http://www.panix.com/~alderson/index.html
Some things are too important to relegate to a web browser.
Rich
Rich Alderson
Vintage Computing Sr. Systems Engineer
Living Computer Museum
2245 1st Avenue S
Seattle, WA 98134
mailto:RichA at LivingComputerMuseum.orghttp://www.LivingComputerMuseum.org/
I did not use the H800, but I cut my computing teeth on smaller Harris models in college (where my work study job was in the computer center, and I was also a computer science major) and then part-time employment afterwards with the Army Corps of Engineers, which was big on Harris computers at the time. This was in the late 1970s to mid-1980s. I used first a model H150 and then the H550 after they upgraded. I even worked for a contractor part-time who had a Harris H120 (I think that was the model number) in his basement for engineering computing. I don't remember what models the Army COE used at the time - H500s of one variant or another I believe.
I thought these Harris computers were all a great system, the bees knees as far as I was concerned, far better than any full-blown PDP-11 system at the time (and no doubt cost more as well at the time). There are documents on these systems on Bitsavers. Everything was blue in color, and the console was a CRT that ran in block transmission mode, grabbing either one full line or the entire page off the screen at a time, feeding into the DMCP board. The H150 we had in college had two 80 Mbyte CDC drives, and later we added a 300 Mbyte CDC drive when we upgraded to the H550 model. The tape drive was a non-vacuum 1600 bpi drive, and I spent many hours backing up the system onto tape, and then swapping drive packs and downloading everything again. I vaguely recall that we did that drive swapping once a week in the wee hours of the morning.
The Vulcan Operation System (VOS), which later was called VMS, I thought was a cool system, but then I didn't have anything to compare it against. We had Fortran, Basic, Cobal, RPGII, and assembler, plus a version of Runoff so students could write papers that got printed on a Diablo. I spent many hours on that system after hours (I had a key to the college computer center), and even started writing my own tape operating system for fun. We had terminals strung all over campus feeding into the system, connected by long runs of serial cables in the heating tunnels; I spent many hours in those heating tunnels as well, as we had to fix things every time lightning from storms would take out the RS232 chips at either the terminals or on the DMCP board in the computer. I got to know the area Harris field engineer pretty well - he was a chain-smoker that constantly had a cigarette in his mouth, even while working on the computer. I watched him do many a system upgrade to boards, which were all discrete TTL chips and parts that were wire-wrapped at that time.
I'd love to know if any Harris computers still exist today. The ones I knew were all scrapped out years ago. I know if I tried to use one today, I'd get really frustrated with the OS, being as used to Linux/Unix as I am today. Harris did come out with a Unix OS for their computers in the mid-1980s, but I never used it.
Fond memories.
Kevin Anderson
>
> Date: Wed, 20 Apr 2016 20:08:56 -0400
> From: Toby Thain <toby at telegraphics.com.au>
> Subject: Re: High performance coprocessor boards of the 80s and 90s -
> was Re: SGI ONYX
>
> On 2016-04-20 8:02 PM, Michael Thompson wrote:
> >
> > I have a quad-860 VME board for Sun systems in my collection.
> >
>
> Do you have the development environment for it?
>
> --Toby
>
Yes, but it is on a Sun 4/260 that us buried in my collection.
--
Michael Thompson
Hey all --
I resolved the weird failure I was seeing on my 11/750 with the Cache/TB
diagnostic and since it was fairly random I thought I'd share it to save
people from the future (hi, people from the future!) from going through
the same machinations I did.
Issue: ECKAL diagnostic loads, prints banner and halts after about a
second with:
00003488 06
No other diagnostic is provided, and since there don't appear to be any
listings or real documents covering the test, it's not particularly helpful.
What I tried (prior to tonight):
- Checked voltages.
- Double-checked backplane for bent/shorted pins.
- Cleaned and reseated every socketed chip (especially the gate
arrays). On *all* boards.
- Swapped in a spare L0003 (after cleaning, as above).
- Swapped in all the other spares I have (one at a time, again, after
cleaning).
- Cleaned backplane with contact cleaner.
- Removed 2nd UNIBUS card.
- Tried a *third* L0003 card (labeled "GOOD" as of 1996 :)).
No change in behavior whatsoever. Very odd. Very frustrating.
So tonight I thought, hey, why not disconnect the UNIBUS just in case
something odd is going on there. Pulled the Unibus jumper connecting
the two backplanes, replaced with terminator.
ECKAL diagnostic now runs and passes.
So: This particular fault (at least in this case) is due to some oddity
on the UNIBUS. I suspect a problem with NPG grants, but I'm going to
have to go over this with a fine-toothed comb, it could be a bad
controller in there doing something mean.
Hope this helps someone at some future date...
- Josh
Back when I spent a couple of years at UNLV in the late 80s, I had a class in which I was forced to use an account on a Harris H800 computer, if my memory serves me correctly. Being a BSD snob, I felt that was a terrible imposition, much like being forced to calculate compound interest on a stone-age abacus made from partially petrified dinosaur turds. *Without gloves.*
Now, of course, I'm a lot more easy-going, and downright curious about things that might not have been my first choice for a computing environment. Even VMS!
So, does anybody here know anything about that family of computers? I seem to recall getting a tour of the computer room once, and the two front panels of the machine were swung open to reveal two thick, mattress-like beds of twisted pair wires. That seemed nauseatingly primitive to me at the time, but now the memory seems fascinating.
I also seem to remember an operator's console with two round CRTs on it, but I might have fabricated that memory from whole cloth.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
I wonder if there are any Commodore people out here who could tell me what
practical differences would result from using a Gotek-type
flash-memory-based floppy emulator in place of the C1581's mechanism, vs.
using Jim Brain's uIEC-SD or similar.
I don't know if the thing would even work in a 1581 case, or if Commodore
DOS or JiffyDOS would work with it; but if so, I wonder if the DOS would
work slightly more like the real thing, because it would be actual
C=/JiffyDOS running on an actual 6502, instead of something new running on
a microcontroller. I understand that you wouldn't get any of the
directory-changing commands et al. from the SD2IEC firmware.
--
Eric Christopherson
> From: Josh Dersch
> It's actually a SCSI device the size of a refrigerator.
Given all the largish machines you have, you must have either i) a warehouse,
or ii) a very large basement and a tolerant SO! :-)
Noel
I have spent over 100 hours looking for viable alternatives to selling
computer stuff. Vintage does not have its own place. Bonanza.com is gaining
popularity, but it is not really IT or electronics oriented. Pricewatch.com
and similar sites only deal in reasonably current equipment. VCF does have a
marketplace section, but (and please don't take offense) it is so
rudimentary, and the listings are so outdated, that is does not seem to be
of much use. How do you find what is still available? True auction or
ecommerce software would eliminate stale WTB and WTS listings. There are
many free places to send XML product lists to be added to their comparison
shopping engines, and all of those are very well indexed by Google. If you
make it easy for people to see that it is an open place to buy and sell, 1
page registration, etc., then it will probably grow quite well. If you
really want it to grow, you can get free press releases with many of the
computer related online magazines and communities. If you don't try it, you
will never know J
Thoughts?
Cindy Croxton
Electronics Plus