Komodo 8 Chessbase

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Paloma
Posts: 857
Joined: Thu Dec 25, 2008 8:07 pm

Komodo 8 Chessbase

Post by Paloma » Wed Aug 27, 2014 7:12 pm


Modern Times
Posts: 2417
Joined: Thu Jun 07, 2012 9:02 pm

Re: Komodo 8 Chessbase

Post by Modern Times » Wed Aug 27, 2014 7:53 pm

Well spotted ! Nothing in the chessbase shop so far.

A big surprise, but a good call I think to release through them. They handle all the support, which takes that burden away from Mark and Larry so they can concentrate on product improvement.

User avatar
Rebel
Posts: 4787
Joined: Thu Aug 18, 2011 10:04 am

Re: Komodo 8 Chessbase

Post by Rebel » Wed Aug 27, 2014 11:08 pm

Just when I thought I had seen everything.

Komodo is a great program but leans heavily on scrutinized Ippolit, the legacy of Vas. Komodo author, member and driving force behind the ICGA fiasco now profiting.

Real life wins from Hollywood script writers.

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Komodo 8 Chessbase

Post by Sedat Canbaz » Thu Aug 28, 2014 12:09 am

I wonder, is this version of Komodo stronger than Stockfish ?

lkaufman
Posts: 3741
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: Komodo 8 Chessbase

Post by lkaufman » Thu Aug 28, 2014 1:45 am

Rebel wrote:Just when I thought I had seen everything.

Komodo is a great program but leans heavily on scrutinized Ippolit, the legacy of Vas. Komodo author, member and driving force behind the ICGA fiasco now profiting.

Real life wins from Hollywood script writers.
There is certainly a fair amount of similarity in the evaluation functions of Rybka 3, Ippolit, and Komodo, but as the primary "evaluation guy" of both Rybka 3 and Komodo I can hardly be faulted for using many of my own ideas in both Rybka 3 and in Komodo. Vas and Don wrote the actual eval code in Rybka and Komodo respectively (excluding what Mark has added in the past year) and the result is quite different code, while the eval weights were tuned independently so they are not similar.
Regarding the search I can say that although we (Don and I) did study both the Ippolit and Glaurung (later Stockfish) code quite a bit, we ended up keeping few of the Ippolit ideas that are not also in Glaurung or Stockfish or older engines like Fruit. I think Komodo has more original search ideas than any other top engine. If I want to give credit to other programs to open-source engines from which we got ideas, I would put Glaurung ahead of Ippolit or Rybka, and of course those engines got many ideas from Fruit and from older engines. Even to this day I still don't understand why Ippolit and derivatives are so much stronger in bullet chess than in slow chess.
In conclusion, I think it would be quite just for the Komodo team to benefit from ideas used in Rybka 3 (even if they are few in number), since I was half of each team! But frankly, neither Mark nor I makes more than a small fraction of what we would made doing other work (programming in his case, chess teaching/writing in mine); Komodo is not a good project financially for either of us. We just enjoy working on it.

shrapnel
Posts: 1245
Joined: Fri Nov 02, 2012 8:43 am
Location: New Delhi, India

Re: Komodo 8 Chessbase

Post by shrapnel » Thu Aug 28, 2014 2:56 am

When is the Komodo 8 standalone Engine being released ?
i7 5960X @ 4.1 Ghz, 64 GB G.Skill RipJaws RAM, Twin Asus ROG Strix OC 11 GB Geforce 2080 Tis

lkaufman
Posts: 3741
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: Komodo 8 Chessbase

Post by lkaufman » Thu Aug 28, 2014 3:53 am

I can't answer that yet, I actually don't know myself as it depends on other people. We'll have more to say about it soon.

Larry

User avatar
Rebel
Posts: 4787
Joined: Thu Aug 18, 2011 10:04 am

Re: Komodo 8 Chessbase

Post by Rebel » Thu Aug 28, 2014 5:48 am

lkaufman wrote:
Rebel wrote:Just when I thought I had seen everything.

Komodo is a great program but leans heavily on scrutinized Ippolit, the legacy of Vas. Komodo author, member and driving force behind the ICGA fiasco now profiting.

Real life wins from Hollywood script writers.
There is certainly a fair amount of similarity in the evaluation functions of Rybka 3, Ippolit, and Komodo, but as the primary "evaluation guy" of both Rybka 3 and Komodo I can hardly be faulted for using many of my own ideas in both Rybka 3 and in Komodo. Vas and Don wrote the actual eval code in Rybka and Komodo respectively (excluding what Mark has added in the past year) and the result is quite different code, while the eval weights were tuned independently so they are not similar.
Regarding the search I can say that although we (Don and I) did study both the Ippolit and Glaurung (later Stockfish) code quite a bit, we ended up keeping few of the Ippolit ideas that are not also in Glaurung or Stockfish or older engines like Fruit. I think Komodo has more original search ideas than any other top engine. If I want to give credit to other programs to open-source engines from which we got ideas, I would put Glaurung ahead of Ippolit or Rybka, and of course those engines got many ideas from Fruit and from older engines. Even to this day I still don't understand why Ippolit and derivatives are so much stronger in bullet chess than in slow chess.
In conclusion, I think it would be quite just for the Komodo team to benefit from ideas used in Rybka 3 (even if they are few in number), since I was half of each team! But frankly, neither Mark nor I makes more than a small fraction of what we would made doing other work (programming in his case, chess teaching/writing in mine); Komodo is not a good project financially for either of us. We just enjoy working on it.
The irony of ruining Vas' business (by Lefler) and then take from Vas 400 elo legacy and make money out of it, now notable via Chessbase is unbearable.

Don is one case, Lefler is another case of a different magnitude due to his leading role in the ICGA fiasco.

Without Ippolit and friends Rybka 4.1 would still top the rating lists.

pohl4711
Posts: 1105
Joined: Sat Sep 03, 2011 5:25 am
Location: Berlin, Germany
Contact:

Re: Komodo 8 Chessbase

Post by pohl4711 » Thu Aug 28, 2014 6:51 am

lkaufman wrote:I can't answer that yet, I actually don't know myself as it depends on other people. We'll have more to say about it soon.

Larry
What I (and all other testers) need, is a UCI-Komodo 8, which runs on Shredder GUI, LittleBlitzer, cutechess...and not a ChessBase-native-version, which only works with the FritzGUI.

Stefan

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Komodo 8 Chessbase

Post by Sedat Canbaz » Thu Aug 28, 2014 7:23 am

Larry,

Can you inform me please:

Is this version of Komodo stronger than latest releases of Stockfish ?

Post Reply