Author: Christophe Theron
Date: 17:29:11 09/13/02
Go up one level in this thread
On September 13, 2002 at 13:57:19, Lex Loep wrote:
>On September 13, 2002 at 13:26:58, Christophe Theron wrote:
>
>>On September 13, 2002 at 11:41:27, William H Rogers wrote:
>>
>>>The smallest chess program ever written was by Peter Jennings. It was originally
>>>written for a small handheld and later ported to the TRS-80, Apple and the Pet.
>>>The original code was only 1k. On the trs80 it was 4k and that was with 3k of
>>>graphics. Although not perfect it was the samllest.
>>>Bill
>>
>>
>>
>>In 1982 I bought a TRS-80 Level I with 4Kb of RAM.
>>
>>For this computer I bought a program named MicroChess 1.5. It was very weak but
>>was able to run in the 4Kb of RAM (actually less than 4Kb was really available
>>to the program).
>>
>>The chess engine was probably much smaller than that, because the program had
>>full graphics and interface. Maybe 2Kb for the chess engine?
>>
>>However I do not remember if the program was able to play with all the rules of
>>chess. I'm not sure about en passant capture, threefold repetition and 50 moves
>>rule.
>>
>>
>
>I still got the assembly listing of SARGON by Dan and Kathe Spracklen from 1978
>The program occupies 8kb of ram which includes 2k of data area, 2k of
>graphics and 4k of move logic. In that time it was one of the best
>programs around. I learned a lot from it.
>
>Lex
>>
>> Christophe
I have got this book from Ralph E. Carter who has been kind enough to send it to
me as a gift a few months ago (in May).
I dreamed to have this book since 20 years (I did not even know anymore if the
book existed for real or not).
Now that I have read it, I'm not sure it would have been a good thing for me to
have been able to read it early in my chess programming career.
Most of the methods used in Sargon 1 are not used anymore. I think Sargon 2
already dropped a lot of them in favor of modern approaches.
For example, Sargon 1 uses a static exchange evaluator instead of a quiescence
search. I almost certain Sargon 2 switched to a real QSearch, and that was an
anormous gain.
The move ordering method used in Sargon 1 is incredibly dense. They generate and
evaluate completely all the moves at any level, sort the list, and go on
searching with this move order. That's very inefficient because a lot of useless
work is done. Here again, I believe that Sargon 2 switched to a much better
method.
However the historical value of the book makes it a "must have". I glad to have
it now.
Thanks again to Ralph E. Carter!
Christophe
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.