Firebird 1.0 beta w32. Vs. Rybka 4 w32

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

kingliveson

Re: Firebird 1.0 beta w32. Vs. Rybka 4 w32

Post by kingliveson »

gerold wrote:
Carlos777 wrote:
gerold wrote: P.S. Is there any engine that you see as being stronger
then Houdini at this time.
Can run a match for you if you like.
Thanks for the offer, Gerold! You are very kind.

Hmm, probably IvanHoe B63Mod12c is the strongest right now (by a little margin though.) I'd like to see it against DR 4 and also I'd like to see how Fire 1.31 and Houdini 1.01 do against DR 4 too.

Best,
Carlos
I did test with IvanHoe 10c and it was about the same as the
earlier version of Houdini.
With other test i have seen there is little difference between
IH 10c and 12c.

I do not like how the Rybka forum has talked about Harry.
IMO he has the strongest and most stable books. He
is also very honest man. Until He is cleared of any wrong
doing or i get an email from one i sent to Vas i will not
test any Rybka products.

Best,
Gerold.
Gerold, I agree with you regarding crucifixion of an accused without evidence. However, I respectfully disagree with you not testing Rybka product because of Felix Kling's childish antics. Vas should not be held responsible for behaviors of others. As for "group think" on rybka forum, it does amaze me at times. Here is an example of making shit up without any evidence to back it up:
A poster from Rybka Forums wrote: Generally speaking, programmers will place encrypted or unencrypted information in the files or even in the executable to identify the particular original source in case there is a leak. I recall that Vas did this with Rybka 3, too. This was obviously done here, and it was by this that Felix @ Co. identified Harry as the leak with 100% certainty.
But of course, what is 100% certain about the above statement is bullshit. I bought Rybka 4 from chessok which comes with 10 binaries. The leaked package has 4 binaries. If you do a checksum (md5sum) to compare the 4 binaries, they match. As an engineer, you and I know that shifting a single bit (01 to 10) changes integrity of the data.

Code: Select all

cd5a29e2d86e4360f987e8473425f420 *Deep Rybka 4 SSE42 w32.exe
4b12d5cd1a328c3d1e01d8630f3d9381 *Deep Rybka 4 SSE42 x64.exe
7aa5c8f355175411525a5a78fd766143 *Deep Rybka 4 w32.exe
0d347db5d5bfd1ab3825e61b576628e1 *Deep Rybka 4 x64.exe
txt file with 01 and 10 strings
96a3be3cf272e017046d1b2674a52bd3 *01
d3d9446802a44259755d38e6d163e820 *10

That being said, the accused might well be guilty as charged. There is just not enough evidence or info available publicly to make a reasonable conclusion -- therefor, presumption of innocence must prevail.
kingliveson

Re: Firebird 1.0 beta w32. Vs. Rybka 4 w32

Post by kingliveson »

Carlos777 wrote:
gerold wrote: P.S. Is there any engine that you see as being stronger
then Houdini at this time.
Can run a match for you if you like.
Thanks for the offer, Gerold! You are very kind.

Hmm, probably IvanHoe B63Mod12c is the strongest right now (by a little margin though.) I'd like to see it against DR 4 and also I'd like to see how Fire 1.31 and Houdini 1.01 do against DR 4 too.

Best,
Carlos
I think if you get Fire to not play choosing Knights over Bishops, you may be surprised how strong it plays.