Author: Dann Corbit
Date: 03:08:58 08/28/04
Go up one level in this thread
On August 28, 2004 at 04:09:22, Reinhard Scharnagl wrote: >On August 28, 2004 at 03:52:40, Dann Corbit wrote: > >>On August 28, 2004 at 03:23:56, Reinhard Scharnagl wrote: >> >>>On August 27, 2004 at 22:53:04, Stuart Cracraft wrote: >>> >>>Hello Stúart, >>> >>>personally I try to avoid inspecting foreign chess program sources, because >>>that only would corrupt one's ideas and phantasy. I claim that analysing chess >>>program code from other people mostly would be counter productive and would >>>give one no real benefit at all. >>> >>>There are exceptions, where special problems of a language like C++ would be >>>discussed, but that mostly is independent from any chess know how. Of course >>>there had been benefits and learning effects from such discussion threads. >>> >>>Creating a chess program by "patchworking" it from different sources hardly >>>could be recognized as a success. Thus doing so can easyly lead to situations >>>where such "recycled" code would appear as an "own" chess engine, frustrating >>>and unfairly pushing back the elaborates of serious people. >>> >>>I cannot see, how the publishing of chess program sources especially as a full >>>packet could produce any benefit. Instead it could be more helpful to publish >>>articles or books on that theme but not including any compilable sources. > >>Sometimes looking at someone's implementation of an algorithm is much better >>than reading about one. Especially when the algorithm in use is better than the >>book version and it also contains all the work arounds for the "gotchas" > >This might depend on one's goals. It is not my goal to avoid weak own approaches >which could be improved from time to time. If one has the goal to complete a >good chess program gathering all foreign ideas one could get, things may change. > >As I have tried to explain, efforts to understand special algorithms should be >ok for me (if not based on compilable source code). > >>I agree that a patchwork program is not going to do well. > >Having seen foreign code it is very hard to make secure avoiding this. > >>People do not help other programmers much. Look at all the open source chess >>programs on SourceForge. Every single one of them has been rusting for quite a >>while. > >That is a big indicator for me that publishing source code is not that helpful >to the idea of fair chess programming (in numerous cases), but may targeting in >maximizing one's personal image (where no one has heard from that person >before). This is an interesting statement. I wonder why chess programming should be different than other avenues of programming. For instance, I do not think we would say that the publishing of GCC's source code is bad or the Linux kernel. What is it that makes chess programs different?
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.