Author: Robert Hyatt
Date: 20:32:39 04/30/02
Go up one level in this thread
On April 30, 2002 at 15:20:05, Miguel A. Ballicora wrote: >On April 30, 2002 at 14:42:50, Jeroen Noomen wrote: > >>During the last 6 - 12 months I have noticed a trend that there are very clear >>indications that non-commercial programmers of chess programs make use of (parts >>of) commercially available books. As I was in Maastricht during the WMCCC in >>2001, lots of speculations and complaints were heard about this topic. From >>several sources I know that also the Rebel books have been ?hacked? by people (I >>will not mention their names, I am sure they know themselves whom I am talking >>about). As the Rebel books are my lifework I want to put a very clear statement >>in this forum. >> >>The Rebel books have been developped by me during a period of almost 14 years. >>Not only is it handmade, it also contains lots of stuff you will not find in >>other books. In the past it has happenend a few times that chess programs were >>caught as not being made by the programmer himself. Examples are Quick Step back >>in 1989, Greif, the several Crafty clones. As we can see from todays rules by >>the ICCA, such programs are not tolerated anymore. And in my view this is fully >>correct. >> >>What about the openingbook of a chess program? Of course programs using a very >>good book have an advantage over programs using a less good book. Here is a >>point where especially the commercially available programs have an advantage. I >>need only to name Alex Kure, Sandro Necchi and myself who have produced high >>quality books. One thing is sure: These books are wanted, because when using >>them, other programs will get the strength of these books as well. >> >>Of course the copying of books is morally not to be tolerated. You can even call >>it piracy. By simply copying (parts of) one of the best books available other >>programs suddenly get the same elo boost, without a single second of work to be >>done. Also this is very suspect in tournaments: Such programs suddenly have a >>big advantage against others, who make their own book and do NOT use this >>illegal copying. A trend that is very worrying in my view and also break the >>rules of sportmanships. I think that the ICCA should make strict rules for >>openingbooks as well. There is a clear rule that it is not allowed to use f.e. a >>crafty source code. In this respect IMO it should also be forbidden to use >>(parts of) books of authors that have not given permission to use them. >> >>As far as the Rebel books are concerned: It is very easy for me to verify >>whether somebody has stolen parts of my book. If programmers think they can use >>my book and hide this from me: Sorry, that will not work. In the past I have >>seen this happen and I have taken proper steps. >> >>I want to make it clear that if I find out (at a tournament or elsewhere) that >>somebody is using my book, that I will undertake steps. As far as I am >>concerned, such programs are banned from all the important events, or something >>like that. I will also suggest such a rule to the ICCA. >> >>Jeroen Noomen >>Bookauthor for Rebel and Chess Tiger > >I believe that the "bookmaker" is part of the "programmer's team" and it should >be recognized as such. So, I fully agree with you but I would take an extra >step. Since I believe that the book author is part of the team, and no >programmer can be included in more than one team in a given competition, I do >not think that it should appropiate to have books from the same author in more >than one engine. For instance, the name of a team in a competition should be >"Rebel X.X, author: Schroeder book: Noomen". >This automatically avoids any other engine to use a Noomen's book in that >tournament. > >Bruce Moreland thinks that this should be also extrapolated to egtb's. >Controversial, but he has a very valid point. The more I think about it, the >more I agree with him. So, one team could be >"Crafty X.X author: Hyatt, book: Hyatt, egtb: Nalimov" > >Note that this would mean that Tiger won't be able to use a Noomen book >if Rebel is present, and Rebel or Tiger won't be able to use the Nalimov's >tablebases if Crafty is present (assuming that Nalimov participates as part >of the Crafty team as I used in the example). > >Regards, >Miguel I think the book issue is clearly as you state. EGTBs are different. They contain the _same_ information whether I write the code to generate them or whether someone else writes it... Just a list of values, indexed by a position Godel number, to say mate or draw... Since books contain _different_ information by definition, they should be protected. But a table of numbers that will be _identical_ in content should not be protected. IE egtb's... Otherwise my table of random numbers for the Zobrist hashing should be protected and that would be ridiculous...
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.