My 'beginner level,
class E and lower' chess computers
Novag Chess
Champion MK I

Year:
1978
Programmer: David
Goodrich
CPU:
F8 3850 @1.78Mhz
ROM:
2Kb
Elo level:
723 (933
FIDE)
CMhz: 0.45
Rperf: 45%
I
really longed completing my collection with the famous Chess Champion
MK I, but I choosed to wait for a nice opportunity: a mint condition
device, with a very well preserved packaging, including the mains
adapter and offered at a very low price (€16 on French eBay, I
remained the only bidder). Because this chess computer is objectively
not worth much! It is famous, surely not because of the chess skills,
rather for the very primitive aspects of the device! First one in the
Novag line of chess computers (also sold using the Videomaster brand in
UK), using a clone copy of the CompuChess software born in 1977 (legal
proceedings occured, without consequences as the CompuChess was not
properly copyrighted), it does not even check the player's moves with
regards to the chess rules... So, despite the manual stating the
computer would rely on the player fairness, the fact is a simple input
mistake can easily make the game void. So you really need to focus, in
order to complete a whole game... Moreover, the user's manual puts much
emphasis on the need to track each move on a paper game log, to be able
to perform checks and so rectify potential mistakes. Another obvious
restriction, it is announced that it only plays as black. As a
matter of fact, it is possible to let it play as white, first of all
switching queen & king positions in the computer's memory
(thanksfully, an ep function standing for Enter Position is available),
then playing a dummy first move leaving the whole position unchanged
(leveraging the lack of rules control). The player can then simply
place his black pieces in a usual position in front of himself, and use
the algebraic notation without any rotation, meaning his pieces are on
first and second row, while the A column stands on his left side. The
computer considers his pieces to be black, but it does not matter... A
bit tricky, but it works! Some parts of the user's manual are very
funny to whoever is aware of the 'floor' level of play:
- 'It may look as though the CC MK I reacts illogically' (indeed!)
- 'in endgames that are very clearly to your disadvantage...' (can this
happen?) '...you must be prepared to resign... otherwise the end-game
will be very drawn-out...' a way to aknowledge the program is unable to
force a checkmate in a number of situations!
The
packaging is as well sort of a compilation of humour, mainly the
quotations supposed to be Anatoly Karpov's comments (on my box; but I
found a picture of another box ascribing the very same words to Boris
Spassky):
- '... an opponent
... playing surprising move variants'
- 'Makes chess immediately enjoyable'
- 'An incorruptible player, relentless with the slightest mistake of
his opponent'
This chess computer is a real anthological
master piece!
Novag Chess
Champion Delta-1

Year:
1980
Programmer: David
Lindsay (pirated)
CPU:
F8 3850 @2Mhz
ROM:
4Kb
Elo level:
936 (1094
FIDE)
CMhz: 0.5
Rperf: 58%
The
Delta-1 is a slightly mysterious oddity within the first chess
calculators from Novag. Designed in 1979, it has been released later on
the market, end of year 1980, after the Chess Champion MK II and MK III
( = Super System III presented in next category), both released in
1979. Authorship is still to be confirmed, but I actually spotted 40%
of Boris Diplomat's ROM code being present in the Delta-1. As a
reference, 43% of Boris Diplomat's code is present in its brother Boris
(shown hereafter), consanguinity levels are equivalent. According to
Michael, Hiarcs forum member, the analysis
of RAM state
while playing (enabled running the emulated code with smartly located
breakpoints) shows similarity with the Boris, especially the same board
representation in memory. Besides, both Boris and the Delta-1 share a
rare peculiarity: the last move played by the computer is not updated
on the internal board, until the next player's move is entered. I think
odds are high for a D. Lindsay
program (either licensed or pirated !). Update: a fragment of Boris (v2
& Master) welcome message (BORIS AWAITS YOUR MOVE) has
been spotted at offset 0878
of the Delta-1 ROM. This data is nowhere useful in this device
(limited to displaying coordinates) and provides an evidence
the
program is a hacking
of Boris' ROM. To my opinion, being scalded by
the trial regarding the MK I (the decision by the court of appeal
occured in September 1980), the firm held the release on the market for
this device, too close to the MK I from a hardware standpoint
(copyright protection was better regarding hardware than software in
those times), and featuring a program whose origin was not announced
(the reason why being now obvious).
In the meantime, as for it, the Super System III might be released
(designed using another CPU, 6502, and a known programmer, M. Johnson);
but Novag needed to feed the lower range market. This has been done
selling the MK II also using a 6502 (it is no use multiplying
technologies) and whose programmer has been announced as well (P.
Jennings, of Microchess program fame). At late 80, feeling reassured by
the court decision in favor of itself, Novag gets back to using the
Fairchild F8, less performing but cheaper than the 6502, fitting well
its market target; and released its delayed Delta-1 among
others
(Chess
Champion Pocket Chess and Chess Partner 2000).
Actually
deserving
the title of successor to the MK I, what is this Delta-1
worth? I
got it as a birthday gift, it was offered on French eBay, €20
'buy
now', with original mint condition package, and the necessary mains
adapter, original Novag one as well. It is much more user-friendly than
the MK I, the interface is still basic but has been hugely enhanced:
play with white or black, castle and en-passant moves without tricky
keys inputs, legal moves checked... The display continuously shows the
current best move while the computer is thinking, but holding a key
down reveals the remaining time down to moving. The level is set using
a timer (as it is done with the Boris, so ...). Lacking any
openings book and any endgame dedicated knowledge, the chess skills
remain quite low. The Delta-1 loves pushing its pawns, even at the
expense of developing other pieces! Hereafter, ROM sample showing the
1979 (c) ; a date one can also find on the packaging and the user's
manual:
Mephisto Solar King

Year:
1996
Programmer: Craig
Barnes
CPU:
KS56C220 @4Mhz
ROM:
2Ko
Elo level:
971 (1120 FIDE)
CMhz: 0.39
Rperf: 61%
Square size: 0.98"
A
'solar' chess computer, powered by photovoltaic cells? I was really
curious to test that. With regards to other features, including a tiny
2K program from Craig Barnes running on a 4 bits micro-controller paced
at 4Mhz, nothing much promising... But this Solar King was offered for
€15 on french "le bon coin", in nice condition with original
packaging; so, for science sake, I yielded to temptation. So, does
playing using solar energy work? Well, technically yes. Outdoor and
under the shade of an awning, you can play without batteries. First
cold start required to press the reset button (ACL), that's all. On
another hand, the playing skills revealed being... simply awful. The
user manual states the 'Eco' levels (solar mode) require more time than
the 'Turbo' ones (requiring a set of three 1.5V batteries);
and I
go into more detail (after using a stopwatch): they are slower by 30%;
but above all, the manual warns about a weaker play. An euphemism...
The manual even states, as an example, the program can overlook a mate
in one while thinking in Eco mode, while such a fail would barely
happen in Turbo mode, only with the few first levels. I can confirm: I
experienced a mate in one not found in Eco-H8 level (the maximum one,
using 2 minutes and 15 seconds search time), while in Turbo mode, the
mate is found as soon as using the A2 level (5.5 seconds). My preferred
pace (and reference one) for playing my chess computers being 15
seconds per move, in solar mode the program is nowhere near to give
checkmate, except a lucky move. It is vulnerable to mates in one it
should be able to avoid, it takes opponent's pieces without basically
checking for immediate counterstrike - such as sacrifying rook or queen
for a defended pawn. Aren't you fully convinced of the playing skills?
I had it play two games in solar mode against my weakest chess
computer, the Novag MK I... The Solar King has been dominated, losing
piece after piece, to end checkmated by the MK I, whichever black or
white it played! A weaker opponent than the MK I, this will be hard to
find out... Oh yes, Alpha
Chess 4,the
weakest PC program I have, my estimation being around 560 Elo points...
OK, this time the Solar King dominates both games, getting in a huge
material advantage, as far as two queens, one rook, one bishop prowling
around the isolated opponent's king... and it gives stalemate in both
games. I evaluate its strength in solar mode around 600 Elo points. At
least, Saitek's marketing department didn't cheat while announcing 128
levels: 64 Eco levels plus 64 Turbo levels, as they do not overlap
(Turbo mode A1 seems to be slightly stronger than Eco mode H8). The
'Turbo' attribute is as well deserved... as difference in speed is
glaringly obvious. Using Eco mode, the chessboard is significantly slow
to response any command, either pressure on control keys or pressure on
squares, letting it be disagreeable to use. In Turbo mode, all is fine,
despite the squares lacking a bit of sensitiveness under finger
pressure, everything is immediately responsive, and confirmed by a
faint beep which the Eco mode disables to save energy. To illustrate
it, I compared the response time when I checkmate the program (yes
indeed, I can!): Eco mode, 12.5 seconds to assess the game is over.
Turbo mode, uneasy to measure manually with my 1/10 stopwatch, my
results focused around 4 tenth of a second. And considering the program
spends 12.5 seconds just to ascertain it is checkmated, you may guess
it is nowhere close to checkmate you, while playing at 15 seconds per
move in solar mode... What can I say, in addition to that, regarding
the solar mode? Oh yes, the energy saving... Luckily, the service
manual of the Solar Star (same device, but mobile form factor) is
available here.
Typical current consumption is 1.2 milliamperes in Turbo mode, while it
is 90 microamperes in Eco mode, thus a 13.33 ratio. Despite the
photovoltaic cells being up to the job in Eco mode, inserting batteries
is recommended, to compensate any potential drop in available light,
this could waste an ongoing game (the computer powering down without
saved memory, for lack of energy). In Turbo mode, the cells (assuming
they receive enough light) slightly reduce battery use (they are
expected to provide above 3.3V under 600 lux). The service manual
provides another much valuable information: the troubleshooting chart
states the technician should check the 4MHz clock only in Turbo mode.
On the other hand, the secondary oscillator, paced at 96kHz, must be
checked whatever the mode. So, the previously observed slowdown may be
deduced: in Eco mode, the active clocking signal is 41.66 times slower
than in Turbo mode. This confirms the delay to assess checkmate: 12.5s
in Eco mode divided by this ratio results in exactly 0.3 seconds in
Turbo mode, which value lays within the accuracy range of my own
measure. A pleasant fact to point out, levels are scaled in a perfect
linear manner, whichever mode Eco or Turbo, each step forward adding
around two seconds to the previous thinking time. No position setup is
available, but you may take back your last played move. Lastly, be
aware the program has no openings book, so computes as soon as its very
first move. What is my final conclusion about the photovoltaic cells
powering? It will come in very useful if you are Robinson on your
despair island, without any opportunity to find batteries, and assuming
you would teach chess to a kiddy Friday, who despite being a beginner
will be happy to humiliate the computer. Except this situation, the
lack of reactivity from the chessboard in solar mode, and the loss of
400 Elo points as an order of magnitude, aren't much bearable:
batteries and Turbo mode are indispensable. Only remaining to its
credit, the demonstrative aspect of the working with no batteries (but
the wow! effect will not last), and the ecological participation of
light energy to extend batteries life.
TEC
Schachcomputer

Year:
1981
Programmer: Peter
Jennings
CPU:
6504 @1Mhz
ROM: 5Kb
Elo level: 1009 (1148
FIDE)
CMhz:
1
Rperf:
60%
Square size: 0.82"
Peter
Jennings has been a pioneer of chess programming for
micro-computers, with his famous MicroChess program, developped as
early as 1976; it will encounter a large commercial success. It was
initially made available for the Kim-1 (6502@1Mhz)
as a listing in hexadecimal machine code, to be manually punched by the
user, and later on provided on audio tape; the program offered three
playing levels (3, 10 or 100 seconds average per move); the faster
levels just disabling part of the code. It then did not manage
castling, "en passant" capture, nor pawn promotion; in order to fit the
available memory (circa 1.1Kb)! On another hand, the manual did provide
details about the program logic, presenting it as a basis, the hobbyist
could extend; should he own a computer with more available RAM than the
KIM-1 had. These restrictions were to be removed with the 1.5 version
for the TRS-80 (still limited to only three playing levels) then the
2.0 version for the PET Commodore and the Apple II (with eight playing
levels). This latest 2.0 version will as well be integrated in the
following chess computers: Commodore Chessmate (1978), Novag Chess
Champion MK II (1979), and TEC Schachcomputer (1981). I played
a
lot with MicroChess 1.5 on my TRS-80, I fairly couldn't
not add it to my collection. Well, which one should I choose, then? The
Commodore Chessmate is definitely too ugly; I already own many Novag
devices; the TEC was my obvious best choice. I much like its look,
reminiscent of the PET
Commodore.
By the way, powered by a 6502@1Mhz, once running the MicroChess 2.0
program, the PET offered actually the same game as the TEC did, except
for the graphical display of course (I checked the identical moves with
a
PET emulator!). The three clones (Chessmate, MK II and TEC) have a tiny
openings book available, and play electronic tones with each hit key or
displayed character. Fortunately, the sound level of the TEC is not too
high, as these sounds cannot be disabled with any of the three clones.
In addition to the appreciated PET-like design, the TEC Schachcomputer
has the advantage of providing a small integrated chessboard (the
squares are roughly 0.8 inches large; a passive board of course, as the
moves need to be input using from-square and to-square coordinates) and
it includes its own power adapter, letting it run directly from the
mains.
The keys are pleasant to press, with a very clear catch, and the large
LEDs display layout provides perfect legibility: some far more recent
chess computers will achieve definitely worse. The playing strength is
the one of MicroChess: really weak, it is mostly a sparring partner one
uses to train himself how to win at chess... Nevertheless, this program
is a must have! I got it as a Christmas gift, it was offered on German
eBay for €35.
Applied
Concept Boris Diplomat

Year:
1979
Programmer: David
Lindsay
CPU:
F8 3850 @1Mhz
ROM:
2.5Kb
Elo level:
1024 (1160
FIDE)
CMhz:
0.25
Rperf:
65%
Square size: 0.39"
When
I received the shipped Boris Diplomat, bought for €30
(complete,
but without the original user's manual and packaging), I doubted the
parcel would contain anything. With no inserted batteries, it is
definitely light weighted! It has been the very first travel electronic
chessboard (not a pocket computer, though). It can run a few hours on
batteries (but a connector for a mains adapter is as well available),
and it features a small chessboard (a passive one, the player inputs
his moves using the keyboard). The tiny chessmen hold their
position with pegs; aside the chessboard two rows offer additional
slots to keep the sixteen first captured pieces. Beyond this number,
one can temporarily keep the next ones using the clear casing, turned
upside down. The bezel-shapped display is smartly designed, easily
readable and relatively well protected. The red LED character
set
is indeed small, but really luminous, and so legible enough. As its
elder brother Boris, the Diplomat can display a full rank of the
chessboard squares, but using a simplified character set, less easily
identified. Fortunately, the keys labelled 1 to 6
(potentially used to edit the chessboard position) are as well tagged
with the pieces symbols, and act as an aide-memoire. This less
advanced display also led to remove the sarcastic comments the Boris
used to write; however the level setting keeps on using a
timer,
and the playing style and playing level of both devices are much
similar. As with the Boris, one can miss the absence of any book of
openings, resulting in a slow game as soon as the very beginning, a
lack of variation unless changing the thinking time (despite some
randomness applied to moves achieving equal scores), and a game often
developped outside best practice. The real asset the Diplomat provides,
with regards to the Boris, is the easy way you can get it ready for a
quick game!
Applied
Concept Boris

Year:
1978
Programmer: David
Lindsay
CPU:
F8 3850 @1Mhz
ROM:
2.5Kb
Elo level:
1036 (1168
FIDE)
CMhz:
0.25
Rperf:
66%
Square size: 1"
A
historical milestone! I could not fail to get the immodest challenger
of the first Chess Challengers (Boris is King!). It is so vintage with
its unusual form factor featuring a walnut box, and its red LEDs
display, reminding the ones used by programmable Texas Instruments
calculators at this time (TI-57, 58, 59). It is able to display the
chessboard row after row, making the position verification process, and
update as well, outstandingly comfortable at that time, a lot more than
the tedious square by square process featured in a CC7. Symbolic
display of pieces is an asset as well, compared to numeral coding used
by Chess Challengers at the same time. Playing level is set with a
timer, to the nearest second, which guarantees a complete use of
granted thinking time, with no overstep. On the other hand, thinking is
stopped with no regards to the analysis completion. The Fairchild F8
powering it is definitely not a fast CPU, and playing level is quite
low. Boris enlives
its games with occasional text speech, more or less funny, streaming
across the display during its thinking time. The messages
seem to be
chosen randomly, they are not consistent with the game position. I got
this computer as a Christmas gift, it was offered for €100 on
French website 'le bon coin'.
Novag Chess
Champion Chess Partner 2000

Year:
1980
Programmer: Mike Johnson
CPU: F8 3850 @2.8Mhz
ROM: 4Kb
Elo level: 1078 (1200 FIDE)
CMhz: 0.7
Rperf: 65%
Square size: 0.98"
This
electronic chessboard deserves attention as it is the first
Novag/Scisys (both brands sold it, just after being splitted apart) to
offer an integrated sensor board (push sensitive squares).
Nevertheless, it does not break itself of the calculator form factor
used with previous models, persisting in handling input-output with
algebraic coordinates. So, no LEDs to display its moves; the player
continues reading the code to be able to execute the computer's moves;
and pushing the corresponding squares only provides an error control.
Conversely, pushing the squares to enter a player's move only inputs
the coordinates, then, once satisfied with the display, the player
still needs to push 'Enter' to commit his move. The squares are, so,
mostly experienced as a keyboard extension, compared to a Sensory 8 for
example (launched same year on the market, just a few months before).
At that time, a computer was still far from being common, and
you
could be proud of using a protocol having all the characteristic
features of computers. At the same time, adding '2000' to goods or
companies names was higly in vogue, and provided a taste of
ultra-modern prospects. The CP2000 is much typical from this
period and, so, desirable despite its flaws: the sensor board not
sensitive enough (in addition to the rather clumsy protocol), the
unsuccessful design (massive shaping, board crossbars, color), and the
playing weakness. The 4K program is nevertheless slightly stronger than
the 2K
one used in the travel device from the same author
(Pocket Chess / Chess Traveller, launched same year), benefitting from
a few stored opening moves and a few code enhancements for the endgame.
The weak computing power of the F8 (despite being hotted-up to 2.8Mhz) however
limits the result. This
device is not rare, it is easily found at moderate price,
but often the blank keys show ugly yellow spots - so I awaited
a
nicely preserved one to enter it into my collection - I bought it
€32 on German eBay (without original packaging). The Novag
brand
is labelled on the mains adaptor, while Scisys is printed on user's
manual last page.
Saitek Calculator Chess

Year:
1992
Programmer: Craig
Barnes
CPU:
S3C7 @4Mhz (?)
ROM:
2Kb
Elo level:
1096 (1214
FIDE)
CMhz:
0.39
Rperf
: 68%
Square size: 0.33"
An
atypic device in my collection, I usually avoid chess computers
requiring batteries. But its tiny chessboard is so charming, and I got
it for €15! I wanted a Craig Barnes authored device, he most
often
seconded Julio Kaplan as a programmer. The 2K program is definitely
limited, especially regarding its unusual setting for playing levels,
using letter and numeral. The letter sets the primary search depth, the
numeral has an effect on the endgame search depth, when less material
speeds up the search. Proceeding by trial and error I assessed the 'B'
levels to be too slow and too variable to fit my 15 s/move standard,
even using 'B1'. As a consequence I use the 'A8' level which plays
rather fast while using more time during the endgame. A matter of
balance... The microcontroller model is not announced, I guess it to be
probably close to, or similar to a 4bits
Samsung S3C7 fitting well a pocket calculator, providing
advanced features to manage the LCD display.
Craig Barnes authored some other chess computers equipped with
this microcontroller series (Mephisto Solar King and Solar Star)
showing similar specifications (RAM, ROM, clock frequency); and
Mephisto produced as well the Calculator Chess renamed Schachcomputer
64. Thus a S3C7 series microcontroller or similar is plausible, and is
consistent with the playing strength. The gadget-like mini-chessboard
can be completely pulled out, making available a small drawer at its
back, useful to store the pieces captured during the game, to avoid
losing
some (a high risk!):
Scisys
Chess Intercontinental Traveler

Year:
1982
Programmer: Mike
Johnson
CPU: F8 3870 @3Mhz
ROM: 2Kb
Elo level: 1114 (1227 FIDE)
CMhz: 0.75
Rperf: 67%
Square size: 0.47"
This tiny
2Kb program from Mike
Johnson has been widely sold starting from 1980, labelled as SciSys
Chess Traveler (as well branded Acetronic
Chess Traveller or Prinztronic Chess
Traveller
in the UK), Novag Chess Champion Pocket Chess, and Tandy
Computerized Chess Game. Its cheap cost and travel form factor were
tempting assets for a starter in computer chess world. I thus wished to
add one to my collection, in order to represent this line of products!
But as a matter of taste, I didn't like much the design of the
computer, too colorful and disadvantaged by the protuberance of the big
red LEDs display. I so waited for an opportunity to buy this more
recent version, featuring a more discreet LCD display, which
is
more smoothly integrated in the design; a significantly higher clock
speed, and a slightly enhanced software. On another hand,
this Intercontinental Traveler revealed more difficult to find
out, ads I spotted were all from the UK, where I finally bought it
(including original packaging, mint condition) for €6.37 on
eBay
(I ended as the only bidder). Unlike the undermentioned Fidelity Mini Sensory,
the holes in the chessboard only provide holding of the pluggable
chessmen; moves are entered using the basic algebraic coordinates
system. The software has no openings book, but nevertheless proceeds
with quite correct development. Once middle game is reached, it lacks
inspiration and is tactically weak, showing quite poor reactions facing
threats - these drawbacks of course getting worse during endgame. One
cannot expect wonders from a 2K program!
Fidelity
Mini Sensory Chess Challenger + CAC

Year:
1981
Programmer: Ron
Nelson
CPU:
8049 @6Mhz
ROM:
2+4Kb
Elo level:
1123 (1234
FIDE)
CMhz:
0.46
Rperf: 69%
Square size: 0.48"
This
small travel chessboard entered my collection for three reasons: its
form factor different from others in the Fidelity family, its processor
distinctive from Z80 based chess computers designed by Ron Nelson, and
last but not least a very attractive price, €35 on French "le
bon
coin".
And for this price, a CAC module was included (Challenger Advanced
Chess) and... a Sensory 8 as well (see next category). All this stuff
including
original
packages, user manuals and the Sensory 8 mains adaptor, which can also
be used to power the Mini Sensory.
Featuring only 2Kb ROM providing four playing levels, the basic Mini
Sensory upgrades to the six levels of a Chess Challenger Sensory 6
thanks to the CAC module. This 4K module also provides a small openings
book (same as the Sensory 6's one), the ability to take two half moves
back or to display a hint move, and pondering on opponent's think time.
As far as I know it is the first device sold by Fidelity to be able to.
It is not a permanent brain feature, as the program does not dig
further than the level that has been previously set; but it
occasionnally speeds the pace pleasantly up. Once the CAC module
inserted, this program is clearly very close to the Sensory 6's one,
thus... weak.
VEB
Schachcomputer SC 2

Year:
1981
Programmer: Ron
Nelson
CPU: U880 @2.46Mhz
ROM: 9Kb
Elo level: 1136 (1243 FIDE)
CMhz: 0.83
Rperf: 68%
Square size: 0.87"
Erich
Honecker's chess computer! He owned one, now kept in Chemnitz
games museum.
It is a rustic copy of Fidelity Electronic's Chess Challenger 10C,
featuring a slightly fitted version of Ron Nelson's program, running on
a processor which is itself an East-German copy of Z80. The massive
design made prominent by the thick bakelite casing, and the in-box
mains power adaptor, result in a 2.2kg device weight. The keyboard is
large as well, but provides an excellent feeling, the best ever amongst
all my chess computers (keys trigger reed switches). A singularity to
point out, the 'Check' (Schach) and 'Mate' (Matt) LEDs: Check lit
notifies the SC 2 is threatening your king, while Mate announces the
loss of the SC 2 ('I Lose' with the CC10). To announce its matting
victory, the SC 2 display blinks. Another peculiarity, assuming you use
some time before entering your move, the SC 2 calls you to order with
a beep (same behavior as the one of the CC Voice, who
periodically
announces 'enter your move' or 'your move'). According to Rüdiger
Worbs, co-author of next RFT chess computers (Chess-Master
and Chess-Master Diamond), the SC 2 costed 3 months of an usual GDR
wage and roughly 1000 units were built. I bought mine €25 on
French 'le bon coin' website, in working order, without original box
nor manual.
Fidelity
Chess Challenger 10

Year:
1978
Programmer: Ron
Nelson
CPU:
Z80 @4Mhz
ROM:
4Kb
Elo level:
1158 (1260
FIDE)
CMhz:
1.35
Rperf:
68%
Square size: 0.87"
It
is truly pure nostalgia that made me decide, quite late (after 10 years
collecting!), to purchase a Chess Challenger 10 (10 stands for ten
levels of play). Sure, I did not lack variants of Ron Nelson's program,
already present in my collection with the Chess Challengers
6, 7, 8, Mini
Sensory, and not to forget the East German SC2 copy. But it is the
chess computer that initially did "inoculate" me, more than 40
years ago, with this passion. So, I finally got aware I missed it...
Its appearance is much more luxurious than the one of all the other
Chess Challengers, featuring a brushed-metal looking board set with
solid walnut, with its nice little boxwood chessmen that look close to
the Champion Sensory ones (displayed within next category; the chess
pieces were most probably provided by Chavet, at least in France where
the Chess Challengers were distributed by Rexton). It is a very nice
article, providing a great pleasure to play with! However, the skills
level is weak; I translate a report from the French
Europe-Echecs magazine, October 1979 edition (you might want
to read more on Blitzchess'
amazing website): 'The
other points that I had mentioned concerning CC 10 and the others seem
to be confirmed by the tests of all the users: tendency to rush on the
material, lack of sense of danger, irregular performance, extreme
weakness in the endgame...'. Nevertheless, let's point out
it is
the first chess computer to reach the market with a (small) openings
book. The first sold devices got immediate complains from their users,
as the program nearly never castled, but Fidelity Electronics
did
remediate this weakness in a very short delay; it is the usual practice
to name those fixed computers "B" version, while the very first are
named "A" version. Mine is a "10B", thus apt to castling, and that is
fine! (it is revealed after few played games, and can be confirmed
using a test, thanks to a slight difference within the openings book -
as no visible information is available on the computer itself). I
bought it complete, in its original plastic attaché case,
with
the exception of the user manual (a printed copy of a scanned PDF);
the price was €95 on eBay (Mondial Relay
€4.99 shipment
included, so the purchase price for the computer was €90.01).
Of
course it is a bit expensive, but cheap ones (you may find some ranging
around €30 to 50) all displayed redhibitory defects to my
opinion
(non working, missing pawns, poor cosmetic aspect,...). Well, I was
keen on not being disppointed!

Will you play with
me? ...this is what this likeable Knight seems to tell you.
Conic
Computer
Chess

Year:
1981
Programmer: Stephen
Chen
CPU:
6504 @1Mhz
ROM:
4Kb
Elo level:
1173 (1272
FIDE)
CMhz: 1
Rperf: 70%
Square size: 1.2"
I purchased
this chess computer (€90, complete with its mains
adapter, user manual and original packaging) mainly with regards to its
eccentric playing system, based on pieces that can be plugged in, and
one diod per square. While plug-in pieces are common for small travel
chess computers, the Conic definitely does not belong to this category:
featuring more than 14" long, more than 11" wide, and close to 2.2"
thick, and only mains to power it, indeed this is a tabletop computer, and
rather an imposing one. The male metal jacks under the chessmen are not
so for stability, rather for detection thanks to electrical connection
in the sockets featured in the center of each square. It is so a rather
crude alternative ("tracer
board") to systems
named "sensory
board" offered by some other brands, such as Fidelity
Electronics. The operation is not so convincing, due to several
reasons: handling is rather unnatural and slow (you need to use a bit
of strength to pull the pieces out, then you need to
accurately aim the target socket); the software sometimes misreads the
signals (legal move not acknowledged so, triggering the flickering of
the relative LEDs, requiring to move back the piece to its initial
position before successfully entering the same move again); requirement
to confirm your move by pressing the "Enter" key (this reminds the Novag/Scisys
Chess Partner 2000 operation mode). The move signals
misreading can happen if the player is plugging in or pulling
out the chessmen too slowly,
and is unlikely to happen using straightforward and fast handling. On
another hand, this system looks to be very sturdy and long-lasting,
alike the whole device! Plastic is nice, which is no surprise knowing
that Conic
has been fonded thru diversification of an initial business of plastic
casings molding for electronic devices. Only three chess computers were
produced by this company, over a short period of time (in addition to
the present Computer Chess: the 1979 Master I, and the 1982 Korchnoi).
Without sure evidences, the author of these three programs appears to
be Stephen Chen, named as so with the experimental version "Conic X"
participation to the 1981 WMCCC (World Microcomputer Chess Championship).
The Conic Korchnoi is supposed to be a close derivative to this Conic
X. The assumed author is as well listed in Bill Hall's page dedicated to chess programming pionneers.
Worth pointing out: the Conic Computer Chess (model 7012)
introduced here was also distributed under the other brands and
names Conic Tracer
7012, Conic Electronic Chess 7012, Conic Sensory
H-20, Hanimex Computer Chess HGM 1200 (Germany), and Westrak
Computer Chess CC1 (UK).

Picture showing the side drawers and the 0.6" metal jacks under the chessmen.
Novag Octo

Year:
1986
Programmer: Julio Kaplan / Dave
Kittinger (?)
CPU:
80C49 @15Mhz
ROM:
2Kb
Elo level:
1183 (1279
FIDE)
CMhz:
1.16
Rperf:
70%
Square size: 0.98"
A
real low-end one... This electronic chessboard is designed for lowest
cost, with minimal LEDs and Bristol board playing surface, very light
plastic pieces, low-cost microcontroller and tiny 2K program.
At
least, a Kittinger one(*)... But the
poor choice for setting the playing
level prevents from best use of granted thinking time. A funny thing,
the 80C49 has been widely used as a PC
keyboard controller. Actually,
it has been designed to support managing many input/output signals, so
it was a good choice for monitoring a keyboard, and for a
push-sensitive chessboard as well. Playing chess against the computing
power of a PC keyboard is a must! I bought it for its processor I did
not own until then, and because it was offered for €10 in my
neighbourhood (including the manual and the mains adaptor!). So I
adopted the Octo boy...
(*) Update:
a member of Hiarcs.net
and Schachcomputer.info fora, Michael,
intensively researched about the origin of this program, and did spot
some common
code shared with Julio Kaplan's Mattel
Computer Chess. Indeed Novag might have delegated to Heuristic
Software, Julio
Kaplan's company, the supplying of short programs for its low-end
range of devices. Maybe Dave Kittinger did supervise the integration,
or maybe he was credited just for marketing purpose; while saving time
for more profitable efforts aimed at Novag's competitive chess
computers. The 2k program used in Novag Micro II and relative devices
(so Octo; and as well Presto and Micro III) appears to be a slimmed
down version of the 4k Mattel
Computer Chess; thus a Kaplan.