Shark is probably Rybka 3 Clone

[tano-urayoan]

Has anyone thought of the possibility that if you have "rybka.exe" and you rename it in windows to "othername.exe" it will be reported as another engine? It could be a kid having a bit of a laugh.

Mike Scheidl already explain this in his posts. Have you read them?

[lonola]

Mike Scheidl already explain this in his posts. Have you read them?

Sorry, I didn't read his version but consider mine an abdridged version of the truth for non-computer chess geeks.

[bob]

I'm not worried about the distribution issue, I'm more concerned with playing in a tournament and having more than one copy of Rybka 3, more than one copy of Stockfish, etc... That is a real problem and is happening regularly.

Do you have any examples? I know a few cases in the distant past, but I don't recall any recent examples of something like this.

The most recent online tournament had obvious clones. I won't say any more at the moment, but there are some obvious cases.

[bob]

Mike Scheidl already explain this in his posts. Have you read them?

Sorry, I didn't read his version but consider mine an abdridged version of the truth for non-computer chess geeks.

This is not the kind of "cloning" I am talking about. I am specifically talking about online events like the last ACCA. There are two types of problems.

(1) legit author writes a program, and everyone knows about it. But in online events he uses a _different_ program to play, one he did not write.

(2) illegit "authors" just copy a program, perhaps use a debugger to change the character strings internally so that it appears to be a different program, and then participate in tournaments with that program.

Both happen _far_ more often than you would guess.

[Tord Romstad]

This is not the kind of "cloning" I am talking about. I am specifically talking about online events like the last ACCA. There are two types of problems.

(1) legit author writes a program, and everyone knows about it. But in online events he uses a _different_ program to play, one he did not write.

(2) illegit "authors" just copy a program, perhaps use a debugger to change the character strings internally so that it appears to be a different program, and then participate in tournaments with that program.

I hope you are planning to eventually show the evidence you have for this claim. Based on the result list of the ACCA tournament, it seems highly unlikely to me that anything like this has happened.

Number 2 above can be almost completely ruled out, as all programs with the possible exception of HfC (which I've never heard about, but it didn't perform well) are written by experienced and respected authors.

This leaves number 1. There is one clear example of an author who did use a different program from the one everyone knows: Allard used Spark instead of Bright. I see no reason to believe that Allard did not write Spark, and it doesn't make much sense to "cheat" by using an engine clearly weaker than the one you are known for writing. It's possible in theory that some of the other participants weren't who they pretended to be, but it doesn't seem likely, considering that almost all the authors are very well known and respected, and no program scored much better than expected.

[Rolf]

it doesn't make much sense to "cheat" by using an engine clearly weaker than the one you are known for writing. It's possible in theory that some of the other participants weren't who they pretended to be, but it doesn't seem likely, considering that almost all the authors are very well known and respected, and no program scored much better than expected.

Sorry, but that argument or assumption is faulty because the strength of the engine is of less importance if a little fake on book basis is being planned. Motivation of clearly weaker contenders than the known number one, experimenting with all kind of confusion to extract half a point here or there. So, here it's absolutely not the question if the author was capable of writing the program but the cheat is the confusional attempt as such. Just IMO as a lay in computerchess.

Rolf

[Allard Siemelink]

Actually, I don't think the latest Spark is (much) weaker than Bright.
After all, IIRC, it did manage to beat Crafty in ACCA

I don't assume Bob would suspect Spark.
Nor do I suspect any of the other entrants.
Therefore I wonder what the fuzz is all about?

There is one clear example of an author who did use a different program from the one everyone knows: Allard used Spark instead of Bright. I see no reason to believe that Allard did not write Spark, and it doesn't make much sense to "cheat" by using an engine clearly weaker than the one you are known for writing.

[Allard Siemelink]

As I was not aware of an appearantly wellknown clone engine called Shark, I made the same mistake thinking this was about my new engine Spark...

For my next engine (if any), I shall remind myself to think up a name and investigate it before tying myself to some tournament's deadline...

I thought initially that he talks about some known engine because there is an engine with a similiar name.

see the following post:
http://talkchess.com/forum/viewtopic.ph ... 37&start=0