A little story:
Seeing the tinkertoy computer reminds me of the mechanical "Computer" I got
for Christmas when I was about 7 years old (1961..ish). It was a long long
time ago and my memory could be a little distorted but, as I recall:
It was made primarily of plastic and was about a foot long, 6 inches high,
and 6 inches deep. This "toy" was built in horizontal layers or "planes"
where each plane was one bit in the computer. The planes could slide either
to the right or left to represent the logic states. If the plane was in the
leftmost position, it represented a "0". If the plane was in the rightmost
position, it represented a "1". I believe there were about 4 planes total.
There was a "flag" (literally a flag) attached to the left end of each
plane that provided the output. As the machine was cycled, the flags would
raise and lower with the change in states.
Along the rear of the computer there were a series of vertical wires that
pushed the planes either to the right or to the left. These wires were the
logic "gates" that drove the computer. As I recall there were about 8 of
the wires (gates) in total.
There were small projections from the rear of each plane that interacted
with the wires (gates). To program the computer you placed small sections
of a plastic soda straw, about 1" long, on some of the projections. The
straws provided the input for each of the gates. If a straw was on one of
the projections, it would interact with the wire (logical 1). If there was
no straw on the projection, the wire would not touch the projection
(logical 0).
To operate the computer, the user would move a selected plane from one
state to another. Any straws attached to the projections on that plane
would push against the wires (gates) which in turn would push against the
other planes in the machine. Some of the planes would change states, some
would not.
For each cycle of the machine, the user had to program in the next
instruction. This meant moving around the straws on the back of the
machine. Once the next instruction was loaded, the plane was slid back the
other way. I never thought about it before but, this thing actually got two
instructions per clock cycle.
Max clock speed was about .01 CPS :-)
The manual had detailed instructions for playing TIC-TAC-TOE and it
actually worked. Unfortunately, it did not provide enough other examples
so, I could never figure out how to make my own programs :-(
I'm still fascinated by mechanical computers!
Steve Robertson - QA Team Leader <steverob(a)hotoffice.com>
> > So, where is the gag ? Just don't forget, the Z1 was entirely
> > mechanical, to be powered by a crank (if necercary - of course they
> > prefered th electric motor :).
>
> Gag? Mechanical computers are serious stuff, Hans. Like Danny Hillis'
> tinker-toy tic-tic-toe machine:
> http://www.yowza.com/classiccmp/toy/tinkertoy.jpg
>
> ObOT: BTW, does anybody know approximately when tit-tat-toe got renamed
> tic-tac-toe? Or is a regional thing?
Don't our "Queen's English" friends (uk and au) call it "naughts and
crosses"?
David Hansen wrote:
> Ok, so it sounds like the general consensus on this list doesn't
> perceive any possible problems from systems that perform automatic
> disconnection/refusal of services due to systems that are based on
> time/date ranges? Here are some examples of situations that are believed
> to be possible...
No, it is not that we do not perceive any possible problems. I at least
believe that there will be problems. But I don't believe there will be
major ones. And I certainly don't believe that they will happen all
suddenly on 1/1/00.
> utilities (power/water/telephone):
> Say a disconnect period is 2 consecutive months of non-payment. A
> customer pays his/her bill regularly but if the date isn't interpreted
> properly the next time the system does a check for 'disconnect
> candidates' it will not be able to find a single payment from a single
> customer within a 2 consecutive months time period prior to 01/01/00
> (1900).
Possible. And may affect telephones. But water, gas and electricity
suppliers won't have the staff to go out and cut everyone off, since this
has to be done physically at the premises. Besides, such a huge number of
disconnections will be noticed and the cause identified.
> banks:
> Keeping in mind that banks routinely deactivate and -absorb- every
> account that is over x years idle (commonly 2 yrs. but every one I have
> come across has some variant of this), see how the above method applies
> to this situation.
Are you sure this is legal? A couple of years ago the National Savings
Bank (as it was in the UK) were saying that they had accounts that had been
inactive for over eighty (yes, 80) years with a few shillings in but they
couldn't close them because they couldn't trace the owners or their heirs.
(BTW Nat Savings never paid interest on balances less than a pound, so a
few shillings 80 years ago is still a few * 5p today)
AFAIK there is nothing in any contract I take out with a bank that says
that if I deposit money and leave it there for over 2 years they are
entitled to it. I come back 10 or 20 years later expecting to retrieve my
savings. Yes, I did this with my German bank account in 1995 having not
touched it since (I think) 1985. It was still there and I could still draw
money out when on holiday in Germany.
Yes, empty accounts may well vanish, and rightly so!
> vendors of perishables:
> Shipments of perishable items (food/medicine) are refused by automated
> systems that read the dates on the items as expired.
This has _already happened_ and will be addressed long before 2000. IIRC a
chain of supermarkets in the UK was asked by a tinned (canned) food
supplier why they were suddenly ordering three times the usual amount of
tomatoes. It turned out that the warehouse was accepting tinned tomatoes
each day; overnight the stock check found that they were out of date ("best
before end feb 00" or something) and ditched them, and the following day
more were ordered.
It is not something that suddenly happens in 2000.
> payroll:
> Employee doesn't have any hours during the 'new' pay period so no
> paycheck is issued. Also paychecks are issued with wrong dates and such.
> Clients aren't billed if there isn't anything in the billing period.
> etc...
May happen. But the first manifestations of the bug will have turned up
already - temporary contracts that expire in 2 or 3 years for example. Our
accounting, job control etc. system uses 2 digit dates. Last year (or the
year before - I forget) a lot of work was done on our system and a 2 digit
date is now assumed to lie between 1950 and 2049.
> security access:
> I'll use my company as an example. I have full access to the office
> building between the hours of 7am - 7pm M-F by way of a keycard. If the
> wrong year is being calculated then M-F can easily be Sat-Wed., etc...
As Tony and others have pointed out, this will probably occur. But it is
relatively easy to override manually until sorted out.
> credit cards:
> Accounts are deactived or non-existent.
Credit card companies - I forget who told this one from personal experience
- hit the bug a year ago with cards that _expired_ in 00. This made them
do all the necessary work then. While I concede that it is possible that
there may be a few bugs left in (spurious late payment penalties for
example), I doubt that accounts will disappear (and who will care? I'm not
going to fuss if a lot of money I owe is forgotten about ;-) ). I know
this is another "assumption" - that having hit the one bug they did all the
work to cure the others - but I think the evidence points that way.
Damn! This discussion is way off topic and I wasn't going to join in.
Oh well, I've typed it now.
Philip.
> Spiritusumdrucker! What a fantastic word! I think that would be a
> mimeograph?
Oh, a spirit duplicator! I've been wondering for ages what a Mimeograph
was!
In the UK such machines were generally known by brand names. For small
runs - a couple of hundred copies - one used a Banda machine (2-part sheets
on which typing or writing transferred a waxy, usually purple ink to the
back of the top sheet. The duplicating machine put solvent on this so it
would transfer the ink to the paper.
For longer runs one used a Gestettner (I don't know how to spell it,
either) machine. Cut stencils by typing with the ribbon disabled in your
typewriter, or drawing with a special stylus.
I heard of electronic stencil cutters that would transfer a plain paper
image to a Gestettner stencil, but I never saw one.
Philip.
Can modern IBM PC clones easily handle ST506 hard drives?
IE, is IDE a superset of this? Or does one have to find
an ST506 controller board?
Thanks,
Miles
All,
>By the time I was in high school dittos were a distant memory. To put it
>in (classiccmp) perspective, by that time inkjets were still on the
>horizon and dot matrix printers were still the <rage>.
As a matter of fact, mimeograph/ditto/Spiritusumdrucker machines
provided me with one of my favorite examples of "antique" technology's
applicability. My wife was teaching at a high school in the early 1990's
that had a way over-subscribed photocopier and an idle mimeograph machine.
She made handouts on our Mac Plus using WriteNow and SuperPaint (both
fantastic programs). She then printed them onto mimeograph masters using
our ImageWriter 1, an *impact* (dot-matrix) printer, at home the night
before class. She'd take the master to school in the morning, hook it up
to the mimeograph machine, and crank out 250 copies as fast as she could
turn the crank - then stroll to class past the line of frantic teachers
awaiting their turn on the photocopier, having saved the school appx. $10
in copying costs.
She could also save and modify files and print new masters at her
convenience. No need to carefully retype or coddle a master for a whole
year for next year's class, or try to stretch a dying master for 30 more
prints. It was a great system.
BTW, she later convinced me to *give away* the Imagewriter in favor
of a Stylewriter (inkjet) we recieved as a gift, just prior to moving to
Texas. Not one of my better decisions. Ah well.
- Mark
Courtesy ard(a)p850ug1.demon.co.uk (Tony Duell):
>> I've ever seen, and nonexistent development software. Words of advice:
>> If you want to learn graphics, go track down an SGI Indigo - the purple
> Look, (and this may sound strange....) if all you want to do is compute,
> go and get a PC or a Mac, right. Part of the fun of classic computers is
> dealing with 'warts and all' - the fact that there are kludges
> everywhere, but somehow we get round them.
Yup. With a great big hammer and a nearby dumpster. If I want warts,
fun, and interesting, I'll spend my time on the MCS-4 or even one of my
old DG Novas. They had their strengths and their weaknesses, but they
were both well-built and historically significant. A badly-designed
machine was a waste of time the first time around - you insult your own
short lifespan by wasting yet more. If you don't agree with me, I'll
be happy to send you a box of assorted unreliable S-100 DRAM boards that
will keep you busy well into your retirement.
As for fun vs. usable, who the hell says you should have to choose? Why
not have both? That's why we're after our first Cray, why we maintain the
NetBSD Build Lab (a single location in which we're trying to assemble one
example of each machine supported). Normally I couldn't give half a shit
for Amigas, but it's fun to rackmount one and help get unix running on it.
Maybe that's the primary philosophical divergence here: I consider few
obsolete machines to be inherently interesting, when there are so many
others that are merely _thought_ to be obsolete but are still utterly
useful given a current, robust OS.
> Having once tried to get spare parts out of SGI, never again. I don't
> want a totally unmaintainable machine, thank you very much. It was all
> custom silicon, and it was all unavailable....
Well, this is just silly. How much more luck have you had getting parts
for that Compucolor? SGIs are just like anything else - if you've got
enough bits you can maintain them yourself. And I'll tell you one thing
for sure: Gathering enough spares to do the support is easy, seeing as
how there aren't (yet) any collector scum driving the prices of Indigos
to 20 times their original price.
Jonathan
Speaking of the Horizon, does anyone know anything about a New
Jersey-based company called Dendron Amusements? They apparently produced a
number of games for the Horizon based on Avalon Hill's strategy board
games. This must have been around 1979.
I am interested in any kind of information about this company or its
games.
/Fredrik
Arfon Gryffydd <arfonrg(a)texas.net> wanted to know:
> I have a chance to grab some Silicon graphics Irix 31xx machines... They
> SEEM to need a boot tape. Anyone know anything about these boxes? What's
> the OS?
There were a number of mostly-accurate replies to this yesterday. Seeing
as how I wrote the FAQ, I'd like to present the mostly-definitive reply.
First, this is a 68020-based, Multibus machine. Nothing at all in common
with the later MIPS systems.
According to William Donzelli <william(a)ans.net>:
> You are going to want and OLD version of IRIX (around version 3), as these
> boxes were fairly quickly kicked off the support wagon (SGI seems to be
> kind of bad about that). These old things are not incredibly usefull Unix
> boxes, as X was never (to my knowledge) ported to them. The graphics are
> quite good even today, as it probably has a board stuffed with Geometry
> Engines.
IRIX 3.6 was the last OS distributed. There's some incomplete 3.7 stuff
kicking around, but probably not of interest to most. To my knowledge,
no open-source unix has ever been ported to one. X is not supported,
though TCP/IP is, so you can get it on the network, but it's going to take
you a bit of work to make it useful, like compiling bind, as they predated
DNS. Your server will offer you a delightful choice of drives: ST506,
ESDI, or SMD. No SCSI.
Note that when SGI put out the first 4D (MIPS R2K/R3K) machines, they called
the OS IRIX 3.something as well, just to create confusion.
> Well, there are 3 of them for about $30.00 each... Huge cabinets and huge
> monitors...
IMnot-so-HO, that's too much money. Stuff of that vintage should be GIVEN
to you for nothing. Pay 'em $90 only if they'll deliver.
> Are they worth grabbing? Can I use the monitors for anything? Can I
> connect them to my ARCnet/Ethernet/Parallel/Serial port network and be
> useful?
Yes, they all have ethernet. Want a unix box? If so, they're worth grabbing.
Monitors are fixed-sync RGB, and probably not useful on anything else you
have or might get.
As it happens, there's another (non-list) chap for whom I've been getting
my 3130 running again (a bad video board, then a blown /etc/passwd...) in
order to generate a boot/install tape. I'd be much more enthusiastic about
doing it if there were a few other people on the list who need them as well,
so I can do a bunch at one time.
Jonathan
In a message dated 1/5/99 12:40:49 AM Pacific Standard Time,
Its an interesting thought, and I know from personal experience that
a set of Kenwood Twins operating at 100Watts about 4 ft from a TRS80
Model 1 can cause it to crash. I would think its possible for a big solar
flare to at least affect an old TRS80. unlikely, though.
kelly
dastar(a)ncal.verio.com writes:
> > >>>
> > Every 11 years the sun sends storms of cosmic rays which disrupt all
sorts
> > of electrical devices. The last was in 1989, and they are due again in
<snip>
> > >>>
>
> As Bill Gates would say, that's the stupidest thing I've ever heard of. If
> cosmic rays were to blame for the EIs suckiness then every computer back
> then would have had similar problems. No, you can't blame Radio Shack's
> sorry-assed engineering on cosmic rays, but nice try.
>
> Sellam Alternate e-mail: dastar@siconic.
> com
>
>Jerome Fine replies:
>
>I realize this may be a very dumb question, BUT if I have an ST506
>Qbus controller (either DEC RQDX1,2,3 OR 3rd party compatible
>like Dilog or Emulex), would it be possible to change the 34/20 pin
>cables (NOTE: On the Emulex DM01, there is already ONLY a
>SINGLE 50 pin cable which is split 34/8/8) and connect those
>cables in some manner to an IDE drive?
Thats a dumb question.
The situation is that an IDE drive has it's controller on board the drive
and it is proprietary. The purpose of the IDE "controller" card is only to
provide a standard interface.
Hans
