OpenChess

Ippolit has been public now since May 2010...more than 1.5 years.
the codebase has been translated into plain and well-known chess programming terms...

Why aren't there more very strong programs...(i.e. at the level of Fire, IvanHoe, Rybka4)?
Several (many) programs: Stockfish, Sjeng, Critter, and others have copied ideas/techniques from Ippolit... and some have clearly benefited from these published/notable 'computer chess programming' and theory advancements...

but what's the problem...i.e. why isn't Stockfish as strong as IvanHoe or Rybka 4>
seems that only Rybka is keeping up with the Ippolit progress...

Clearly the tech leader, the persistent innovation is via IvanHoe... ?
almost every week new, exciting, and innovative ideas emerging...!

Take Stockfish for example: obviously, getting a C++ programmer to simply re-write (optimize) Glaurung's 'cowboy-style' existing code to modern C++ standards isn't enough...although Marco's efforts here seem certainly to have been worthwhile...
and, clearly, getting a layman to meticulously optimize search and eval parameters can help...but only to a certain degree,
although Jouna certainly seems to have had some success in this area...

I ask: how hard is it to utilize ideas/code presented in Ippolit/RobboLito/IvanHoe/(Rybka??)
Is simply it beyond the grasp of of the Komodo/Stockfish, and most programmers ?

The reason is that the author of Ippolit is Vasik Rajlich. Nobody, except him, is not able to fully understand the ideas implemented by the author.

jury_osipov wrote:The reason is that the author of Ippolit is Vasik Rajlich. Nobody, except him, is not able to fully understand the ideas implemented by the author.

You may be right that Ippolit true author is Vas. It could be that Computer Chess world was not getting enough buzz, there was no competition against Rybka, and the best way to generate publicity is to create rivalry. So when Rybka 4 was released everyone hurried up and got it to see how it faired against Ippolit (IvanHoe). Are we being played as pawns?

jury_osipov wrote:The reason is that the author of Ippolit is Vasik Rajlich. Nobody, except him, is not able to fully understand the ideas implemented by the author.

Sorry...good try but that's wrong.
how then to explain Houdini/IvanHoe/Fire, etc. being so much better than R3...?
because no one other than Vas fully understands the ideas implemented?

kingliveson wrote:

jury_osipov wrote:The reason is that the author of Ippolit is Vasik Rajlich. Nobody, except him, is not able to fully understand the ideas implemented by the author.

You may be right that Ippolit true author is Vas. It could be that Computer Chess world was not getting enough buzz, there was no competition against Rybka, and the best way to generate publicity is to create rivalry. So when Rybka 4 was released everyone hurried up and got it to see how it faired against Ippolit (IvanHoe). Are we being played as pawns?

yes, absolutely...
we're all a bunch of idiots and he's an absolute genius...
manipulating us in a grand scheme like a master puppeteer uses his marionettes!

kranium wrote:
yes, absolutely...

No need to thank me. I've doing my part contributing since the grand deception of 1947.

kingliveson wrote:

kranium wrote:
yes, absolutely...

No need to thank me. I've doing my part contributing since the grand deception of 1947.

for the sake of accuracy:
i didn't thank you.

need I for some reason?
1947?

Earth to Major Tom...
Can you hear me Major Tom?
Here am I sitting in a tin can...

I believe you've nailed the real reason, already -- Vas is a professional chess programmer, not an amateur. He and his team, have more time and resources at his disposal, than any amateur team has. (Hyatt might be equal in this, however)

Since his income is tied to Rybka, Vas is also highly motivated to work on his engine, and be sure every part of it (opening book, etc), is taken care of. Multi PV on multiple cores, may be a problem, but the most important thing for his is:

1) Winning important tournaments

2) Marketing those wins, into sales.

Software gains have been rather slow in computer chess, historically. Hardware gains have far surpassed software improvements. After the first year or two, I believe most CC authors would be glad to have a 20 point elo improvement, in a year's time.

That's where I believe we are today - the low hanging fruit (no pun intended), has been picked, already. What remains is much harder to find, and you have to be willing to climb much higher in the trees, to reach it.

Dave Mitchell wrote:I believe you've nailed the real reason, already -- Vas is a professional chess programmer, not an amateur. He and his team, have more time and resources at his disposal, than any amateur team has. (Hyatt might be equal in this, however)

Since his income is tied to Rybka, Vas is also highly motivated to work on his engine, and be sure every part of it (opening book, etc), is taken care of. Multi PV on multiple cores, may be a problem, but the most important thing for his is:

1) Winning important tournaments

2) Marketing those wins, into sales.

Software gains have been rather slow in computer chess, historically. Hardware gains have far surpassed software improvements. After the first year or two, I believe most CC authors would be glad to have a 20 point elo improvement, in a year's time.

That's where I believe we are today - the low hanging fruit (no pun intended), has been picked, already. What remains is much harder to find, and you have to be willing to climb much higher in the trees, to reach it.

well i have to agree with you there Dave...

getting a very small ELO improvement at such a high level seems like squeezing water from a stone,
very very difficult indeed...but probably (theoretically) possible with enough effort!

I for one, am trying....
many many others choose to stand on the sidelines offering nothing, simply judging and complaining...
(see CCC talkchess.com)
but beware!
this site might constitute a severe health risk for anyone even slightly open-minded!...

kranium wrote: 1947?

Did I say 1947?! I meant the gay deception of 1935.

Anyways, I would say the publication of Ippolit was huge. Look at how many engines took a massive jump in its wake: Critter, Houdini, Komodo, Rybka, Sjeng, Stockfish, and so on. I personally have gone through the Ippolit source code forwards and backwards and learned many things.