Why Is Houdini Not Entered In The WCCC??

[Robert Houdart]

I'm not a chess engine expert (I hardly even know C/C++), but I'd like to see the eval, as it's quite different from Ippo*'s eval. I want to know if you just do some math on it to make it more pessimistic like a lot of people think, or if it really is different.

Peter

Didn't you write: "You don't have to publish the whole source. Just small parts of it." ? .
The evaluation function is not a "small part" of the engine, it currently contains over 2500 lines of code and is at the core of Houdini's strength.

You can probably understand that I cannot publish this code, but to give you a point of comparison, the functional equivalent in the current Ivanhoe sources (see evaluation.c) contains only about 1500 lines.

Robert

[notyetagm]

What I don't understand is why not make the simple statement:

I took the ippolit sources, studied them, and made changes to produce what we have now.

Because the simple fact is that this is not an accurate description of how Houdini was created.

What I don't understand is why you insist on my making a false statement.
The word "Inquisition" comes to mind, I'm really glad we no longer live in 16th century...

Robert

Nonsense this is not an inquisition merely a request for the truth.

No, the way I see it is this IS an inquisition by Rybka fanboys who cannot accept the following facts:

1) Houdart's Houdini is stronger than Rybka 4 (not by much though)
2) Rybka 1 was a clone of Fruit (read thesource code comparison and weep)

[Dr. Ivannik]

I sit here in my Moscow apartment sipping my Stolichnaya vodka and sweltering in the heat as I analyze a old Rudolf Spielmann game full of speculative sacrifices. I have engines in front of me that can tactically get to the heart of the chess problem and show me resources that my comrades before me could not have seen years ago. I feel fortunate to have these chess engines of today and do not care where or who or what they are. As I read these posts I wonder if anyone uses these engines for their original purpose? All this silly debate as to their legality and on and on I wonder why does one care?? Enjoy all the engines and yes I'll have another on the rocks.

Thank you

Dr. Ivannik

[kranium]

I sit here in my Moscow apartment sipping my Stolichnaya vodka and sweltering in the heat as I analyze a old Rudolf Spielmann game full of speculative sacrifices. I have engines in front of me that can tactically get to the heart of the chess problem and show me resources that my comrades before me could not have seen years ago. I feel fortunate to have these chess engines of today and do not care where or who or what they are. As I read these posts I wonder if anyone uses these engines for their original purpose? All this silly debate as to their legality and on and on I wonder why does one care?? Enjoy all the engines and yes I'll have another on the rocks.

Thank you

Dr. Ivannik

thank-you Dr. Ivannik!
best post i've read in ages!
i agree 100%

except for one very important thing:
i prefer Jack Daniels...(w/ cola or orange juice and lots of ice)



PS-
don't forget to stay indoors...in case you haven't looked out the window in awhile, in addition to a severe heat wave...
apparently there's thick black forest-fire smoke covering the city!

(no wonder you're drinking!)

[Peter C]

I'm not a chess engine expert (I hardly even know C/C++), but I'd like to see the eval, as it's quite different from Ippo*'s eval. I want to know if you just do some math on it to make it more pessimistic like a lot of people think, or if it really is different.

Peter

Didn't you write: "You don't have to publish the whole source. Just small parts of it." ? .
The evaluation function is not a "small part" of the engine, it currently contains over 2500 lines of code and is at the core of Houdini's strength.

You can probably understand that I cannot publish this code, but to give you a point of comparison, the functional equivalent in the current Ivanhoe sources (see evaluation.c) contains only about 1500 lines.

Robert

Shows how much I know about chess programming.
2500 lines? In one file? Java programmers are much more civilized.....
While I do understand that you don't want to publish all that, I think if Houdini really isn't based on Ippolit you should publish at least large parts of it.

Peter

[BB+]

Another example is that Naum never competed in the WCCC, even though it was #2 in a few rating lists for awhile.

[hyatt]

There's lots of food for thought. But from my experience, authors that have _really_ spent any time writing a chess engine are equally as interested in competing in these events to see how they stack up against the rest. It's simply human nature to want to compete. When someone doesn't, it is one of those things that makes you go "hmmmmm"...

food for thought regards...

Bob

[Carl Bicknell]

Robert,
Two questions:

There are probably quite a few people who would offer to write you a pretty decent book, perhaps even for free. If you got a decent bookmaker onside Houdini would be formidable in competitive chess. What do you think?

Secondly, related perhaps, what about going commercial?

[Gerold]

He could rewrite or improve the Houdini and include a short book with the next version.
Maybe in 90 days we will see a new Houdini. Hope it stays free for awhile. .

Best,
Gerold.

[Robert Houdart]

@Carl,
Going commercial is an option to fund at least partially the hardware and the expenses of a WCCC participation.
The question is which option would be preferred by the users: a free engine that hasn't participated in the WCCC or a $40 engine that has participated in the WCCC? Would make an interesting poll...

@Gerold,
I could surely include a short book with the next release, but so many good general purpose books are available that this may not be very useful.

Robert