Guess who made it to the New York Times?

[Ted Summers]

I saw there were no comments appended to the NYT blog entry yet, so I submitted this for moderator approval. Hopefully it will appear later this morning & give readers who are new to computer chess some additional context ...



As a long-time chess player & computer geek, I've really been enjoying this event for the last few days. This event doesn't have the man versus machine overtones of the 2003 Kasparov-DeepJunior matchup, but it's still interesting to watch this battle of the eight best available chess engines play out.

Houdini is remarkable in that it is a free chess engine developed by one individual (Robert Houdart) as a hobby, but it's also remarkable that the other free engines are doing so well. Any one of these commercial or free chess engines running on even modest PC hardware can now beat any Grandmaster alive.

On various internet chess forums, some of the commercial developers like to cry foul when a free engine beats their for-pay creations, insinuating that stolen intellectual property is the basis of of the free engine's success. But most commercial developers also admit that they freely use ideas and even snippets of code from existing open-source chess engines, and offer no hard proof that developers of free chess engines are doing anything different.

Kudos have to be given to Martin Thoresen, a programmer and computer chess afficionado who set up the TCEC website and the hardware to allow these engines to compete. There are a number of established chess-engine rating sites that will play many blitz games (chess played with very fast time restrictions) to evaluate new engines, but Mr. Thoresen's TCEC site is using a single very high-end PC (a six-core Intel Core i7 980x overclocked to 4303 MHz) and long time controls (initial 40 moves in 100 minutes, per engine). That's unique, and attracting a fair amount of attention. Using one very fast PC to play the games one-at-a-time also adds a bit of suspense to the proceedings, and allows computer chess geeks to chat during the live play on sites like Chessdom/Chessbomb mentioned in this NYTimes blog entry (http://livechess.chessdom.com/site).

Anyone interested in playing chess against one of these free engines can download a free GUI (a graphic interface that displays the chessboard and interacts with the user) and any of the free engines, all of which now conform to a standard protocol to talk to chess GUIs. I'd recommend getting the free gui Arena (http://www.playwitharena.com) which comes with several basic engines built in. After playing with the basic engines for a while, you can download and install Houdini (http://www.cruxis.com/chess/houdini.htm) or Stockfish (http://www.stockfishchess.com).

Have fun watching the competition and playing against these inspiring creations!

- Jim in Austin, TX

+1000

[Ted Summers]

There was a link on the Rybka Chess Forum that discussed the New York Times article, but guess what?

http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=20339

Access denied

[Zapped]

The comments I posted earlier made it thru moderation onto the NY Times blog posting as comment #3. Unfortunately, comments #2 and #4 refer to the reverse-engineering controversy instead of the merits of the TCEC contest itself.

As a VLSI designer whose entire undergraduate education was based on studying the techniques of prior art, all arranged and systematized in textbooks and in scholarly journals, I can't for the life of me understand why the study of previous open-source code and refinement of earlier ideas is such a sticking point in these forums.

[Uly]

Unfortunately, comments #2 and #4 refer to the reverse-engineering controversy instead of the merits of the TCEC contest itself.

I find such comments fortunate, people should know the truth, otherwise they may get the idea that engines like Houdini are original, something that is still not known.

[Gino]

Unfortunately, comments #2 and #4 refer to the reverse-engineering controversy instead of the merits of the TCEC contest itself.

I find such comments fortunate, people should know the truth, otherwise they may get the idea that engines like Houdini are original, something that is still not known.

It's quite unfortunate to ignore what the author says about his own engine.

[Dave Mitchell]

The comments I posted earlier made it thru moderation onto the NY Times blog posting as comment #3. Unfortunately, comments #2 and #4 refer to the reverse-engineering controversy instead of the merits of the TCEC contest itself.

As a VLSI designer whose entire undergraduate education was based on studying the techniques of prior art, all arranged and systematized in textbooks and in scholarly journals, I can't for the life of me understand why the study of previous open-source code and refinement of earlier ideas is such a sticking point in these forums.

How can you assert this?

It's *absolutely shameful* the way Houdini, Stockfish, etc., have taken ideas and prior published data, to create better programs.

Tomorrow, I'm asking all the airplane designers to quit stealing the idea of shaped wings, to create lift for their planes. Clearly, they have NOT paid royalties to the birds, and the poor birds have had no advocate for their case, in court.

Ruffled feathers regards.