Here is the Google Translator version into English:
German to English translation
SPIEGEL ONLINE
SPIEGEL ONLINE
03. March 2011, 16:03 clock
Rybka chess software
Programmers assume Intelligenzklau
Frank Patalong
Also, the computer chess scene has her copy and paste scandal. Pilloried Vasik is Rajlich, developers of Rybka, the strongest chess program in four years the world. Rybka is this, then the accusation, based on plagiarized code. Copies will be on the scene quite often.
The computer chess program Rybka, which was for four years as the strongest program in the world and since 2007 four times in a row, the computer chess world championship should be decided, based on an engine that is at least in part on copied code. The at least maintained, the programmer just this code, Fabien Letouzey signed, have a letter from the thirteen other internationally renowned experts and computer chess programmers. The letter is addressed as a complaint to the ICGA International Computer Games Association, which hosts the Computer Chess World Championship.
Diehitzige debate about the supposedly fairly widespread copying style now is like a religious war, in which professionals make to each other in some shifting alliances allegations or say but confidence. The bizarre wrangling throws a spotlight on one scene in which to decide today even minimal differences in quality sometimes for a long time ranking positions and the real innovation progress are more difficult to achieve.
Letouzey claims in his letter of complaint that the programmers have taken over since the Vasik Rajlich Rybka 1.0 beta (2005) he developed parts of the chess engine Fruit. At this time Fruit was 2.1 as one of the strongest engines of the world, prior versions of Rybka, however, were not noticed positive. With the version 1.0 beta program that changed abruptly: Rybka showed tremendous skill level that could be developed by the developers in the next two years - since 2007 Rybka dominates the competitive environment almost at will.
This, however, claims Letouzey, was revealed by the improper transfer of source code from the time under the open source GNU Public License (GPL) Fruit-Engine have been possible - a breach of the terms of the license. Since the publication of the allegations it cooks in the forums of the chess community.
Software can not simply look under the hood
Who now believes, however, those issues would be decided by comparing the source code to but slightly wrong. First, is trending in the scene well: In variants of "mutual learning" about the reverse engineering - trying nachzuprogrammieren functions of a program - extend to the takeover code. In addition, all programmers use public domain code snippets, which provides for similarities here and there. Absolutely you need for the comparison but look at the source code, and that is not in the commercial programs given.
Even if: a comparison of source code is extremely complicated, says Matthias Wüllenweber from the Hamburg company Chessbase chess computer, which acts in addition to his own program Fritz as a distributor of competing programs such as Shredder and Rybka. Principle is based on athletic competition brushed chess programmers scene calibrated on confidentiality: a scientific exchange on how to programmatically achieve something, so think "instead of just on the back burner."
Wüllenweber: "The launch of Fruit was unusual, because this two groundbreaking ideas were known seems to me that all good chess programmers tried an adaptation of these ideas when Vas has Rajlich done the same, then he is probably the only one who managed to.. because its program from the beginning played so much stronger than Fruit. "
But it is also copied? to clarify this question is not so easy. The technical problem explains Wüllenweber, lies in the fact that a program whose source code is not available one, is very difficult to understand. The source code ("Source") is itself not even the program but must first be converted by the compiler to individual machine instructions that the processor can execute step by step.
But you can not close back on the implementation of a command to the source code - an equal-appearing effect can be achieved in different ways. Similarities in the performance of a program are therefore less significant than similarities in source code. Striking parallels are there, however, difficult to explain by chance - they are more evidence of "copy & paste".
A shot that backfired
Such was evidently the Rajlich himself when he made against his program until the process was initiated: Rajlich when comparing his program had discovered the competition program Strelka striking similarities and alleged that this was copied from Rybka. Not at all, to defend themselves then the developers: they would rather copied the open source Fruit 2.1.
Developers Letouzey whose estimated the same after repeated testing of source code snippets, all of a sudden the accuser was pilloried Rajlich - Rybka Developer of the accused with his accusations against Strelka quasi itself.
Meanwhile Rybka program became modified by the competitor Houdini under pressure, a program that is deemed to still play more - but also as possibly plagiarized: Houdini is said to have served the competitor Ippolit, and that in turn is based on Rybka - and thus on Fruit 2.1 ? The question is not decided in the rankings of the top spot Houdini has therefore not been able to take.
Many small clones?
Letouzey middle of the last decade had set one of the last major milestones: With Fruit, he ventured a slightly different approach in the style of play that made the program temporarily to play one of the strongest in the world.
Fruit of the open source engine has been running the core of many chess programs based on it - what's under the GPL okay, as long as the resulting program is then re-released under the GPL. Matthias Wüllenweber: "The power of open source is that all improvements are again exposed to the quality of the software increases continuously Rybka is not open source.."
Rajlich, so Letouzeys appeal to the ICGA, had violated this rule, Rybka should therefore be classified as Fruit-clone. Whether that happens will be decided with the "Clone and Derivative Investigation Panel" means a committee of the ICGA, which was founded only last month in response to the bizarre programmers Dispute: From now on, the computer makes World Cup organizers, a commission, whose sole purpose is to clarify allegations of cloning is.
That might be worth it obviously.
URL:
*
Http://www.spiegel.de/netzwelt/gadgets/0, 1518,748616,00. Html
MORE ON THE INTERNET
* ICGA International Computer Games Association
http://ilk.uvt.nl/icga/
* Computer chess program Rybka
http://www.rybkachess.com/deutsch/index.php
* Fruit Engine
http://www.fruitchess.com/indexde.htm
* Homepage ChessBase
http://www.chessbase.de
SPIEGEL ONLINE is not liable
for the content of external internet sites.
© SPIEGEL ONLINE 2011
All rights reserved
Reproduction only with the permission of SPIEGELnet GmbH
Listen
Read phonetically
Dictionary - View detailed dictionary
Google Translate for my:SearchesVideosEmailPhoneChatBusiness
About Google TranslateTurn off instant translationPrivacyHelp
©2010Business ToolsTranslator ToolkitAbout Google TranslateBlogPrivacyHelp
►
