Rybka is a clone & has been banned from WCCC..?

Originally posted by Darax The Good
A computer program does the testing rather than a person.

Edit: if Rybka's big addition is automated testing, then it is highly likely than he trains Rybka against other engines. This is more of a machine learning process than automated testing. If my assumption is correct, then Rybka is going to end up looking to play like the engines it is trai ...[text shortened]... I mean look,it has links and you can buy stuff, clearly you've stolen the code from amazon.

hahhahhah, priceless! now we have a guy defending the guy who cheated on computer chess! you can't make stuff like this up!!! 😀DD

It would be interesting to know if rybka took actual code or just it's logic re-coded (the evidence is that it took actual logic and details not just ideas). If they copy/pasted some of the code and modified it for rybka, the program is a derivate of those 2 open source programs. Both licenses require that any derivate program has to make the sources available. Fruit and crafty authors could sue rybka and get the sources public and available for anyone to modify.
Fruit license is GPL 2. You are mostly free to do what you like but it requires that any derivate code is free to read, modify and distribute.
Crafty license is more restrictive and does not allow derivates to enter chess tournaments or have a name of their own. Crafty is open source but not free.

Hmm, I wonder if Chessbase News or the USCF will be publishing similar reports.

Originally posted by Mad Rook
Hmm, I wonder if Chessbase News or the USCF will be publishing similar reports.

The USCF is sending someone to dig up Bobby Fischer. A shovel ready project indeed.

Originally posted by ketchuplover
The USCF is sending someone to dig up Bobby Fischer. A shovel ready project indeed.

Nah, the guys running the USCF would argue over what kind of shovel to use. They'd sue each other over it, and it would be tied up in the courts forever. The job would never get done.

I assume they found large numbers of situations where Rybka is making the same bad moves as Fruit, and/or Crafty.

Rather than simply pointing out that they have the same good moves, the same bad moves would be a very big red flag.

Apparently there is some code in Crafty that does nothing.
It just sits there and is totally random.

(perhaps planted as an anti-plagiarism device)

This same code had been found in Rybka. The only explantion
is that the code was copied from Crafty.

It's an old stunt to catch plagiarism.
You put in a deliberate error and if it crops again in the stolen piece
then there is your proof. (I do it all the time with the RHP blog) 😉

This makes perfect sense: At the cheating championship the best cheater has been disqualified for cheating.

They shrank down Kasparov and cloned him and then hid them inside the gubbins. An old trick but a good one.

Apparently a sneeze is what got them found out.

If you ever doubt your own software, a little pepper is often all you need to diagnose it.

Originally posted by greenpawn34
Apparently there is some code in Crafty that does nothing.
It just sits there and is totally random.

(perhaps planted as an anti-plagiarism device)

This same code had been found in Rybka. The only explantion
is that the code was copied from Crafty.

It's an old stunt to catch plagiarism.
You put in a deliberate error and if it crops again in the stolen piece
then there is your proof. (I do it all the time with the RHP blog) 😉

That's a nice conspiracy theory, but not really true. Bob Hyatt has explained the reasons why the obsolete code was left in Crafty, and it had nothing to do with trying to trap plagiarists. (Basically just mistakes in leaving in old, obsolete code.)

By the way, the alleged code copying from Crafty wasn't used as direct proof of VR breaking ICGA rules, since the version of Rybka that allegedly copied Crafty code was not used in any ICGA tournaments. It was simply mentioned to establish a pattern of behavior. What caused the ban ruling was the alleged copying of Fruit code.

Originally posted by Daobobo
I assume they found large numbers of situations where Rybka is making the same bad moves as Fruit, and/or Crafty.

Rather than simply pointing out that they have the same good moves, the same bad moves would be a very big red flag.

If anyone noticed similarities in good or bad moves between the programs, I can't remember reading about it. As far as I know, it's always been about the similarity of code. I think what elevated suspicions was when Strelka came out. Some people, I think including VR, thought Strelka was an illegal clone of Rybka. Then someone started comparing the disassembled code of Strelka with Fruit code, and then the alleged similarities were noticed. And it just snowballed from there.

Oh, and to clarify one point by the OP - The ICGA ruling does not accuse Rybka of being a clone. The ruling accused Rybka of being a derivative program of Fruit, which would violate ICGA rules.

Rybka merely taking one of its five a day.

Hi Mad Rook.

It is a nice conspiracy theory, let's keep it going.

I picked up that gem from the Rybka forum and twisted it a bit.

Although companies do use it to trap plagiarists.
Map makers are the prime example where they put in false
rivers or even made up towns.
Programmers started using this method and it gave birth to 'Easter Eggs'.

Hey Greenerpawn I think you are right.

Instead of using pepper I left one of these Mephisto computers
in a bucket of water overnight and now it does not work.

I think I've drowned my wee shrunken down GM. Where can I get another?

Still nothing from Chessbase News. Maybe they're taking extra time to do some really in-depth, hardcore reporting. Or maybe they just haven't caught wind of the story yet. 😉

Anyway, it's now hitting the mainstream press - The Washington Times, the New York Times, and Fox News, along with some smaller news outlets.