Here is a nice little fairy tale, especially written for Thorsten

The year is 19xx. We have a brilliant engineer who knows manufacturing and who has sufficient programming knowledge to be able to program the turning “on and off” of switches and responses to key press impulses on other gadgets. One day he meets with a man who has a business and money and they discuss business opportunities in the new electronics industry.

At the same time one of them if not both are aware of all the hype around a famous man and his bet about chess programs never being able to beat him. They are also aware of the large public interest for computer chess.
And the engineer says: “I bet I could build a table chess computer!” and the financer says: “Well if you could, then we would both become extremely rich and famous! And I would finance this immediately if you can do this!”

So the engineer goes home and starts working on this project. And he works hard designing on paper how the table computer should look. LCD display, signals and responses to key presses and small chess board. The engineer is happy, he has a computer designed! Yippee! But, now he has to develop a chess program and he begins working on this all alone in his spare time because he still has a 9 to 5 job to go to in order to make money for his family.

After extensive research, the engineer sees that his idea for a table chess computer is not exactly unique and he sees that there are many other companies also already working hard on this.

This knowledge adds to his pressure of having to find a way to develop a chess program as quickly as possible all alone and in his spare time. And it is a big problem because he knows that it has taken whole teams of really intelligent and brilliant students and professors in Universities more than 10 years to develop their programs and the code for this is not accessible to the public and he has only 6 months to do all this from scratch. So he thinks and he researches and worries about this every day.

One day, he reads an article about a hand-made computer and a program that you can manually key into this computer! At first he is a little disappointed, since technically after all he might not be the first to invent one. But he soon forgets his disappointment because after all this computer is not the same! It is not a table chess computer!

So he takes this program, studies it day and night after work and weekends, pulls it apart and completely changes it and behold it works!

But just like the original that he copied, it is not perfect, but it now looks very different and plays better, and no one will be able to tell. However, it still cannot promote pieces, it cannot checkmate, the key inputs although very much changed and better than the original, still does not follow standard chess notations.

He needs more time but he has no more time, the financier is impatient because he is worried about the competition and he wants to be first and not second.

So the product is launched before it is ready, upgrades are promised to disappointed customers, but who cares the world has a chess computer!

Viele gruesse,

Nick

Ron Nelson had a chess program for Z80A. this is in many older dedicated chess computers we know.

in the end it was in sensory voice

Fidelity Voice Sensory Chess Challenger

then came spracklen.
they used the same body and put in the 6502 motherboard and sold the machine as

Fidelity Champion Sensory Chess Challenger

So what do you expect ron nelson to do ? throw his chess program into
the trash bin ?

of course not.
he made backups and waited until he could refurbish it.

and the day came.

a new company was founded. with the rest of the old company.
now if you have less money, would you BUY another program or take the program you have in your desk for free because you wrote it yourself ??

__________________
Die ganze Welt des Computerschachs

Come on, how do you think he managed to get a 1000 ELO Program up to 1900 with his old Challenger program? If he did he would have used them on the weaker computers instead of using David Levy's.

There is maybe a difference between 80 Elo between Ron nelsons and spracklens program if you consider that the hardware was twice as fast due to the 6502 beeing faster then the z80 cpu.

It was not 1000 but 13xy Elo and the h8 is many times faster then the z80.

__________________
Die ganze Welt des Computerschachs

Can anybody calculate the speed difference between the Z80/4MHz and H8/12MHz processor in ELO?

Active list:

Fidelty Sensory Voice z80 4MHz = 1179
Excalibur Igor H8 12MHz = 1834


@Nick We all know that you have tested the Horvath progs a lot. Is there any of the style that fits the Excalibur Igor moves? The "standard" style of the f.e. Regency plays totally different than the Igor.

__________________
http://chess-computer.blogspot.com/

This is not easy as there is no direkt comparision of instructions. The H8 can do more 16-bit, can do more bit manipulation, has MUL and DIV and is faster overall. One try:

The H8 ist faster per instruction, most need 2 states compared with 4+ on the Z80. A factor of 5 seems reasonable, this would take 10 states on a Z80 for simple instructions and 20 for more complicated ones.

A factor of 3 comes with clock speed so a H8 is 15 times faster in this calculation. If we take one doubling gives 50 to 100 Elo, this is about 200 to 400 Elo.

The Fidelity Voice Sensory Chess Challenger is given with 1394 Elo, adding the 200-400 for the faster chip this gives about 1600 to 1800 Elo compared to 1845 Elo for the Excalibur Igor. This seems reasonable.

The programs are different in size and can use varying memory 16+4 (ROM+RAM) K versus 32+1 K. Less RAM means less space for move lists so the older program cannot just be ported. I don't think a program from Nelson could be written for another processor and just a fourth of RAM with more ease than using another program or starting from scratch.

At The Time Ron nelsons programs circulated, iterative deepening was not found. IMO spracklen first put it in but even they did not use it on all levels.

In nelsons program the levels 1/3/5/7/9 are different then the other levels.

__________________
Die ganze Welt des Computerschachs

You cannot take the 1394 ELO. Because otherwise you will have to increase every computer in the Info Active list accordingly. You have to deduct 200 ELO from that. Active ELO is 1179 is really about the best you can do. Trust me I played U1400 tournaments with over 1000 games so I know these machines very well. So your logic would really not get it higher than about 1500 elo max.

ps... and they wouldn't achieve even that. They lack in many games the ability to find a plan and settle instead with draws by 3x repetition (even with a huge piece majority) and they have total inability to check mate from positions that even 1500 elo computers would find. So even 1500 ELO in true practice with speed doubling is too high.

I just took data from the wiki and I don't think a comparision of the two programs is possible with my calculation.

Even my factor of 15 between Z80/4MHz and H8/12 MHz is just a rough one.

In unserer Diskussion ob die Programme von Excalibur Ron Nelson (Igor/Grandmaster) und/oder Guyla Horvath (in allen Geräten die sich mit HELLO melden beim einschalten) beinhalten kommt immer wieder der unterschiedliche Spielstil als Argument.
Wie unkonventionell, merkwuerdig und unpositionell der Igor/Grandmaster von Excalibur (im Gegensatz zum positionellen passiven Spielstil vom Horvath Programm) spielen kann man m.E. sehr schön an folgender Partie aus meinem Königsgambit-Turnier sehen. SO habe ich den schwarzen noch nie spielen sehen.



Das schwarze Spiel ist bizarr. Obwohl nicht rochiert zu haben treibt er an BEIDEN Fluegeln die Bauern vorwärts, als ob er gar keine Positionskriterien hat die ihm sagen das der König geschuetzt werden muesste. Andererseits stet er dennoch nicht schlecht. Scheinbar ist das Spiel eher taktisch fundiert als positionell. Auch in anderen Partien gegen Morsch Programme zeigt sich das. Das Programm im Igor/Grandmaster ist durchaus in der Lage schnellen und taktisch starken Morsch Programmen Paroli zu bieten.
Und der Spielstil des Igor/Grandmaster ähnelt stark dem der ersten Fidelity Ron Nelson Programme. Wenn auch auf einem anderen Spielstärke-Niveau.

__________________
Die ganze Welt des Computerschachs