In what way?<br><br>I've wondered about it too. I still remember the DigiComp
over many other toys.

How did Digicomp influence you?<br><br>I know
what it did for me...Digicomp and HAL-9000 were my
first introductions to computers. Digicomp taught me to
understand binary notation in 5th grade; gave me a distinct
leg up over classmates in problem conceptualization;
and allowed me to feel to both teacher and students.
<br><br>None of THEM grokked "binary." OF course, the students
were none too bright - that same show and tell day one
classmate alerted us to the fact that he had read Haley's
Comet was due to crash into the Earth in 1968. Duhhh -
I knew it wouldn't hit us until 1986. But I was
surprised and delighted to discover that I knew about
something that even teacher didn't know.<br><br>--Larry

In my case, there were many parallel influences
that shaped my decision to become a programmer of
which the DigiComp-1 and Think-A-Dot days were one
influence.<br><br>I liked sci-fi movies like Forbidden Planet, Day
the Earth Stood Still, The Time Machine, and
Fantastic Voyage. My favorite scientist was Albert
Einstein. I played with slide rulesand Magic Brain adders
in grade school and programmable desktop calculators
in High School (Olivetti).<br><br>Star Trek was
defining point, I wanted to understand logic the way Spock
did. I even began reading a lot of my father's logic
books from college which stressed syllogistic logic. I
read a book on boolean algebra and was able to do all
the problems. My friends and I played wff'n'proof
games until the principal banned them because he
thought we were gambling (playing with dice).<br><br>My
uncle began to teach me about mainframes (core memory,
gates...), showed me Gunner-IV on the GE Timesharing network
via teletype and around that time I got a DigiComp-1.
I also began designing primitive switch based
computers using multi-pole switches and relays. The relays
were too expensive so I tried the old wire wrapped
around a nail with tin can contacts all connected "Get
this" connected directly to house current. Boy were my
parents mad. I also was able to coverup my shocking
experiment (blowing a few fuses) but couldn't figure out how
to change the time on one bed room clock (my
parent's bedroom).<br><br>The DigiComp was alot simpler to
work with. I think the mystery of how it really worked
stayed with me and has only now been solved. I didn't
have all the cool Advanced Programming books that
would have kept me even busier.<br><br>The end result
was I joined the first Boy Scout Computer Explorer
Post (Post 635 in Schenectady NY<br>and run by GE)
learned Fortran from Mario Farina<br>one of the first
computer programming language book authors (if not the
first). Went to a four year school majoring in Physics
(no computers at all I was tired of keypunch
machines) and reading a lot about General Relativity and
the nature of the universe (grandiose
dreams).<br><br>One college episode, I remember was watching a Prof
explain his circuit wiring problem to another student.
They were both stumped until I said I could solve it,
and proceeded to write a truth table, convert and
reduce the boolean equation and then draw the circuit
that he needed and he was flabbergasted. Apparently,
he didn't know about boolean algebra and what it
could do.<br><br>Eventually I got hired right out of
college by my GE Post advisor to do scientific and
business programming in Fortran, COBOL, and GMAP (GE Macro
Assembler)<br><br>Now its 30 years later where OO and Java rule and I
do data mining programs for IBM. DigiComp has come
back again from the past when I thought about ways to
teach my kids who are jaded with computer game
graphics, about computers. Building a DigiComp-1 or
DigiComp-2 would be a good woodworking project.<br><br>I've
showed them how a slide rule works and how it can be as
fast or faster than a calculator sacrificing only
precision and without any need for batteries. <br><br>Now
I'd like to show them how a DigiComp works before its
too late (well I can always wait for the grandkids).
Of course kids really like the games that you can
load on a TI graphing calculator but that's another
story.<br><br>Next,<br>Jim

I sure I'm not you. I was in the 1st year of
Explorer Scouts. H. Sadler was the Post Advisor. J.
Pirrone was his assistant and helped me run my compiles
(I got 4 a night whereas others got 1 or 2 a night).
M. Farina taught some of the courses and T. McGovern
was my mentor since I knew more programming at the
time. <br><br>The 2nd year through 5th years I was at
college and didn't come to any meetings. In April 1974, I
got hired by H. Sadler as the first member of Post to
get hired by GE.<br><br>I acted as a Post Advisor and
instructor for several years thereafter. My specialty was
teaching the programming projects. One such project was
for a Firematics database of fire equipment of sch'dy
county. <br><br>Another was for the Dating game that they
eventually shutdown because non-profit orgs can't make
money. I think that involved some kids from Mont
Pleasant and Linton and Niskayuna. Boy were they self
motivated couldn't decide if it was for the money or the
girls they hoped to match up with. <br><br>My HS was
Mohonasen and I went to Union College majoring in
Physics.<br><br>I also did the Fortran card deck route and never
got my program to run. Years later I finally figured
out what was wrong, I forgot to initialize the char
arrays to spaces. The program was essentially a char
based plotting program.<br><br>As an advisor I setup
the password sign-on scheme, maintained the games,
and encouraged a lot of odd projects. <br><br>One
famous fopah? was the implementation of the birthday
algorithm in COBOL, FORTRAN, TEX, and BASIC for an Explorer
Open House. Well one mom tested them with her
daughters birthday and discovered that they all said she
was born on Sunday except for the BASIC program which
said Monday. When she asked me about it (I wrote the
pgms) I said maybe your daughter was born at Midnight,
wrong answer. She was the wife of the operations
manager and I never heard the end of it. The BASIC
program was written correctly, but the BASIC rounding
function had an error when rounding certain decimals. So
it became a case of poor testing on my
part.<br><br>Oh well enough of the stories from yesteryear.

I posted a message in response to TomL's
messages<br>but now I can't see TomL's message in the message
list.<br><br>I feel like I'm in a parallel universe or else
message responses aren't being linked together
properly.<br><br>Anyway, consider this a test post.<br><br>-- jim

toml's m,essage was message #170 so now I want to see what number this one
gets.<br><br>-- jim

It looks like TomL message is lost among the
stray<br>bits of the YAHOO Web server's database
engine.<br><br>Ground up and cast out for some unknown
reason.<br><br>Maybe when it was submitted to Yahoo they were doing
some sort of backup or were under a
DOS<br>attack.<br><br>So if you're listening TomL please repost your
message so I know I'm not a Slider.<br><br>-- Jim

Use the "Messages" link under "Our Pages" to see lost messages

We have 3 rows by 6 cols:<br><br> for each cell
you can have T, F, nothing = 3<br> and a control, or
nothing = 2<br><br> 2*3 = 6 combos per cell<br><br> and
18 flipflop cells 6**18 = ?
101,559,956,668,416<br><br>Then figure in the AND gate attachment (the Y
thingee)
tieing two cols together. I'm still thinkin': Could we
have more than one, could they overlap cols if so, ...

We have 3 rows by 6 cols:<br><br> for each cell
you can have T, F, nothing = 3<br> and a control, or
nothing = 2<br><br> 2*3 = 6 combos per cell<br><br> and
18 flipflop cells 6**18 = ?
101,559,956,668,416<br><br>Then figure in the AND gate attachment (the Y
thingee)
tieing two cols together. I'm still thinkin': Could we
have more than one, could they overlap cols if so, ...

Forgot to factor in the cols are
interchangeable<br>so you must divide by 6!<br><br>We could drop out
duplicate cols because why program the same A.B.C
expression with the same<br>control in another column
(redundant logic).<br><br>So we have 6 FlipFlop
columns<br>for each column 3 choices for A, B, C logic<br>and
for the control column we have 8 combos<br>1 FF col =
3*3*3*8 = 216<br>total # of combos = (216 // 6) = 216! /
(6! * (216-6)!) = 216*215*214*213*212*211 /
6!<br>which is = 131,513,824,548 combos<br>not including the
AND connector.

Here we go again! I've written a spiffy neato
keen DC-1 emulator in MS Access 97 (better than the
other two IMHO) and posted a message here to offer it
to the members of this group but my message was lost
again. Is it just me? Why doesn't Yahoo like me? 8-(

When I was checking out other YAHOO clubs I saw a
message about a hacker attack that changed some club
content. Maybe the lost messages are a part of this or
reaction to the attack.<br><br>Its that or YAHOO doesn't
like people with ZIP code ids.

Hi Folks,<br><br>I got a wonderful letter from
Bill Duerig which answers many of the questions that I
posted earlier.<br><br>I will try to post portions of it
as I get the time (teaser).<br><br>To begin with Irv
Lieberman was the guy with the <br>orginal idea to
construct a mechanical digital computer. He worked with
Bill and Dave Hogan at the same company. Bill and Dave
helped him to improve the idea and build a working model
for the annual March NYC toy fair. The original model
used six circular planes with clock rods on the
perimeter and was about 1ft in diameter. <br><br>Sears
became interested but wanted a simpler model. So they
redesigned into the form we know today. Irv was bothered by
the simplification because it reduced the problem set
significantly<br>but he came up with some pretty cool problems in the
booklet.<br><br>::: My guess is that the original was a 6 bit machine
with 12 clock rods and a clock lever that rotated back
and forth <30 degrees (1/12 of a circle).<br><br>I
wonder if they still have the plans for the original or
if they applied for a patent on it?<br><br>The
redesigned DigiComp was built within hours using stiff paper
material and shown to the Sears buyer again who said he
liked it. They then found a plastics engineer who
helped create the production DigiComp-1 and the company
had its first product.<br><br>They eventually sold
250,000 DigiComp-1 machines.<br><br>to be
continued...<br><br>-- Jim

I have sent feedback to YAHOO about the "missing
messages" problem. We'll see what resposne we get. In the
meantime, it appears to me that if you click on MESSAGES
from the menu at left, you should see ALL messages.
Only in the "overview/Home" page do you get the messed
up message thread.<br><br>--Larry

What happened to this club? Doesn't anyone else
<br>post anything?<br><br>Are there any thoughts on the
number possible DigiComp-I programs?<br><br>I'm sending
notes out to Irv Lieberman and Michael Ransom. From
info I got from Bill Duerig <br>I searched Calif near
LA and found a possible <br>address for Irv
Lieberman.<br><br>Michael Ransom a prolific DigiComp programmer in DigiComp
Romps lived in Tulsa. So a Yahoo search <br>found a
possible address. Who knows?<br><br>An intersting exercise
would be to see how many unique programs there are in
DigiComp Romps by <br>studying the programming grids It
occurred to<br>me that it succinctly describes the program
logic<br>kind of like a logic signature.<br><br>later,<br>--
jim

As I stated in prior messages there are a least 60 programs for DC1.  50 were in
this game book which I still have not found.

I know there are around 60 programs. What I was
wondering if some are permutation wise the
same.<br><br>When you code a column, you code a small
matrix:<br><br> T F C<br> :===========<br>A: x | x | x<br>B: x |
x | x<br>C: x | x | x<br><br>There are six columns
on the DG-1. Won't the DG-1<br>work the same when C1
is swapped with any other column? <br><br>So maybe
there a couple of logic problems<br>that on the surface
appear different are really the same set of equations
programmed in a different column order.<br><br>Consequently,
we may have a few less unique
<br>programs.<br><br>Lastly, are you referring to the DG Romps
manual?<br><br>-- Jim

Anyone have a Think-A-Dot manual available? I'd
like to get a copy of it. I understand there was a
piece about octal math and Think-A-Dot. <br><br>I kind
of remember it from my original Think-A-Dot
something about dropping the marbles in a particular order
and reading an 8-bit number from the blue & yellow
dot array.<br><br>-- jim

Hey, Maverick!<br><br>I got your letter today. I
am the same Michael Ransom who was published in
"DigiComp Romps" way back when. Talk about a blast from the
past!<br><br>My interest in DigiComp as a 13 year old did indeed
presage my career choice. I work at Sabre (computer
reservation system) here in Tulsa as a Realtime Coverage
programmer. Most of the code we look at is in IBM S/390
assembler, but newer programs are being written in
C.<br><br>I still have the old DigiComp (minus a couple of
springs) and the manuals in a shoebox. I owned DigiComp
II, but no longer have it. It was made of vacuformed
plastic and tended to get stuck a lot.<br><br>I once
owned Think-A-Dot and Dr. Nim (although I still have a
DigiComp program that simulates Dr. Nim).<br><br>John
Penn, who was also published in "Romps", was a friend
of mine in Jr. high. I talked with him in the early
90s and he was working in computers at that time,
too.<br><br>I have just begun to read the message board, but
thought I would get a message out here.<br><br>I have a
large web site for fun called "Tulsa TV Memories". It
was featured on a local TV station here...the
RealVideo clip is on the
site!<br><br><a href=http://tulsatv.tripod.com
target=new>http://tulsatv.tripod.com</a><br><br>I will now continue my
reading!<br><br>Thanks,<br>Mike Ransom

Welcome aboard. The clubs a little slow but
people do check in from time to time. My letters<br>have
also reached inventors Bill Duerig (DigiComp-1) and
John Godfrey (DC-2 and Dr NIM).<br><br>I'm still
trying to reach Irv Lieberman (man with the original
idea for DC-1) and CH MacDonald (wrote 10 pgms for
DC-1). I did see your friend John Penn listed in the
Romps manual and have an Oklahoma address that I'll
try. YAHOO People Search really works if you use it
carefully.<br><br>Cheers,<br>Jim

I just sent Kevin a new program for DC 1 to simulate the Dr. Nim toy on his DC 1
Javascript simulation!

Not directly on topic but here's a web
reference<br>to a German Zuse mechanical computer
simulation.<br><br>They have several applets that show a similar<br>style
of mechanical logic using sliding planes and
rods.<br><br>-- Jim

Oops forgot the web reference:<br><br><a
href=http://www.zib.de/zuse/Inhaltsverzeichnis/Simulationen/sim.html
target=new>http://www.zib.de/zuse/Inhaltsverzeichnis/Simulationen/sim.html</a>

Hi again,<br><br>Check out
<a href=http://www.epemag.com/zuse/
target=new>http://www.epemag.com/zuse/</a><br><br>This is an article about
Konrad Zuse who built
the<br>first mechanical programmable binary digital computers
in 1941 to relieve engineers of the tedium of
calculations.<br><br>-- jim

John Godfrey sent an great letter on Think-A-Dot
that you may find interesting:<br><br>I can find no
substantive mention, nor discussion, of the use of octal to
economically reference binary numbers in the manual. There is
only one quote which even remotely alludes to the
subject, and this in very indirect terms, even by the most
generous interpretation. It starts on page 5 at the bottom
and states:<br><br>"THINK-A-DOT can also count in a
manner similar to large computers. Remember the colors
of the top-right, middle-right, and
bottom-middle<br>dots. Now drop the marble into the right-hand hole
eight times. The eighth time all three dots will return
to their original colors! This shows that
THINK-A-DOT can count and remember that you dropped the
marble through the same hole eight times in a
row."<br><br>Without saying so explicitly, this quote references
THINK-A-DOT's inclusion of four functioning 3 bit accumulators,
along four major<br>diagonals, two downward to the
right and two downward to the left, each of which may
be interpreted as counting from zero to seven, (or
one to eight), when eight succesive marbles are
dropped in its top flip-flop.<br><br>Also, the manual
makes no reference to the following interesting points
about THINK-A-DOT, posibbly because they are regarded
as too involved or technical. In the main I think
the manual is very well done and addresses the
interests of most of its customers.<br><br> 1. If each of
THINK-A-DOT's" flip-flops are assigned a "0" if tilted to the
right or a "1" if tilted to the "left", there are a
total of 256 different "tilt" patterns that can be
identified and enumerated. (The number of "color" patterns
conceivably can be less, however, as more than one "tilt"
pattern could result in the same "color"
pattern.<br><br>This needs further examination, but preliminary
results suggest this in not the case.)<br><br> 2. These
256 patterns can, in turn, be divided into 128 Type
"A" patterns and 128 Type "B" patterns.<br><br> 3. If
an initial pattern is a Type "A" pattern, then only
Type "A" target patterns can be obtained by a
succession of "legal" moves. The same is true for Type "B"
patterns. In other words, no Type "B" pattern can be
obtained, by legal means, from a Type "A" pattern and vice
versa.<br><br> 4. Legal moves consist of dropping any number or
marbles sucessively in any of the top three holes, in any
order, and assuming the machine makes no errors.<br><br>
5. Type "A" patterns are defined as all those
patterns that have an even number of "1" bits in their top
row of "3" bits, plus their bottom row of "3" bits.
Type "B" patterns are defined to have an odd number of
"1" bits, in these two rows.<br><br>----- end of Part
1 -----<br><br>Part 2 coming up (YAHOO doesn't like
long posts)

----- start of part 2 -----<br><br> 6. As every
marble entering a top row bit must also enter one and
only one bottom row bit, the parity of this six bit
number comprised of the top and bottom two rows is
unchanged by any succesion of "legal" moves. (i.e. As two
bits of the six bit number are changed, an even six
bit number stays even, and an odd number stays
odd.)<br><br> 7. If either the initial or target pattern is
randomly chosen, there is a 50-50 chance the target
pattern can't be obtained from the initial pattern in any
number of legal plays. The manual's suggestion that the
initial pattern be obtained by "...either tilting it to
one side or the other or [shaking it to get a random
starting pattern].", is wrong in the latter case because
it accepts the even chance that no solution can
be<br>obtained. (The [] in the above expression are mine, not the
manual's.)<br><br> 9. If THINK-A-DOT is initialized by either
tilting it to the right or tilting it to the left, a Type
"A" pattern will be initially obtained. In the first
case the six bits will be all "0's" and the parity
will be "even" and in the second it will be all "1's"
and the parity will be "even". (Zero "1's" is an even
number of "1's", and six "1's" is likewise.)<br><br>(A
Type "B" pattern can be obtained initially by tilting
the machine to the right or left, then with an eraser
on the end of a pencil, changing one of the dots in
the top or bottom rows.) <br><br> 10. In the
solutions shown to specific problems the manual seems to
imply that the order of the marble dropped into the top
holes is important. It it is not! For example, the
manual gives the solution to obtaining the target
pattern "I" as "MLLLLM", where this means, first drop a
marble in the "Middle" hole above, then drop "4"
successive marbles in the "Left" hole above, then drop a
final marble again in the "Middle" hole above. The
solution could have been given as "MMLLLL", or "LLLLMM",
or even "LMLLML", as all will produce the same
result. In view of this fact, perhaps the convention
should be followed of always giving the solution as the
letters "L", "M", or "R" from left to right, having made
the point that order is irrelevant.<br> <br> 11. Any
Target pattern can be systematically obtained from an
Initial pattern of the same Type by performing the
following steps:<br><br>First, observe the top left dot. If
it is not the target color drop one marble into the
top left hole and make it so. Next, observe the
diagonal going downward from the top middle dot to the
left. Drop marbles in the middle hole until this
diagonal agrees with the target. <br><br>The top LEFT dot
will be unchanged by these moves. Next, observe the
diagonal going downward from the top right dot. Drop as
many marbles as required in the top right hole to make
this diagonal agree with the target pattern. These
marbles will not alter the top LEFT dot, nor the diagonal
previously obtained. The pattern will now either totally
agree with the target pattern, or only the bottom right
dot will be of the wrong. If it is wrong it is
because AN ATTEMPT HAS BEEN MADE to obtain a pattern of
the opposite Type, and this can't be done in any
number of moves, or the machine has made an error in the
course of play. Check the parity of the initial pattern
and the target pattern. If they disagree the problem
is impossible. The maximum number of marbles
required to solve the problem should be 1 + 7 + 7 =
15.<br><br>I hope you find this information about THINK-A-DOT
interesting and informative.<br><br>Best wishes, John

I've finally placed my Think-A-Dot emulator out
on the Web for anyone to play with. Its much quieter
than the original, so you'll have to imagine each
mouse click as a marble drop and think about whether
your hearing is fading away (applies to anyone over 40
only).<br><br><a href=http://www.prismnet.com/~mcardle/thinkAdot.html
target=new>http://www.prismnet.com/~mcardle/thinkAdot.html</a><br><br>-- Jim

Hey TomL,<br><br>I never saw a repost of your
experiences with DigiComp and Explorer Post 635. Were you
heavily <br>involed with it? Do you remember the Dating
Program that some kids were working on?<br><br>-- Jim

An easy way to check the parity of a THINK-A-DOT
pattern is to add the number of Blue, (or Yellow), DOT's
in the Top and Bottom rows (but not the Middle). The
pattern is "Even", or "Odd", if this number is Even or
Odd. Even patterns can only be obtained from Even, and
Odd from Odd, "legally", by dropping balls in the top
3 holes. Tilting Left, (or Right), always gives an
Even pattern to start. Odd patterns are hard to get,
(they can't be got legally), yet the posted photo, (1
of 2) on the WEB site called "THINK-A-DOT and
manuals", shows an illegal Odd pattern.<br> "You can't get
there from here", by starting from a Left or Right
Tilt. Isn't that Odd! - JG