I just did put a few underlining of my own.
Sorry to add a new thread but I think that this post by Vas is very important, probably more than any other post so far on the subject.
Marc
Hi,
I got quite a few emails about this. Let me clarify a few points:
1) There is no doubt in my mind that the similarities(in executables and in output) between Rybka 1.0 and Strelka 1.8 are not a coincidence. This is in contrast to Strelka 1.0, where the similarities didn't seem to me to be excessive.
2) I have no idea how this was achieved. Whether something 'wrong' was done, whether Strelka author is just making fun of me and Rybka, whether Strelka 1.8 should be tested by CEGT, CCRL, etc - these are deeper questions. I am 100% sure that nobody has gotten a hold of my source code - this stays on just one machine which is not connected to the internet. This has been the case since well before Rybka 1.0.
3) This Strelka release does not bother me personally, partly because I don't really know what was done, but mainly because these similarities are with Rybka 1.0, which is ancient, incomplete, prototype-level code. If the similarities were with Rybka 2.3.2a, and if the playing-strength was on par with Rybka 2.3.2a, then Convekta would need to investigate this and consider their options.
4) From the point of view of the computer chess community, this is an issue which is in my view worth investigating. I'd like to thank the guys who have contributed to this - it is important. Writing something of Strelka 1.8 strength is a significant effort which requires multiple thousand man hours of development and testing. In light of this, we should have a procedure which is fair to the Strelka author and also fair to other aspiring engine authors. It's good to have an environment which is conducive to productive contributions from everybody.
Vas