I NEED HELP With This Question

[geots]

It concerns Strelka 5.1 and Strelka 5.5. They are both in some rating lists. I wonder how Strelka 5.5 is being tested in general. Surely I am hoping they are testing it the way it was released.

That was one of the most crude and rude things I have ever seen a programmer do when Richard Vida went into the Strelka 5.5 engine, removed the database of moves in it- then told people that version was ready to use now. No way he had a right to do that- not IMO.

Surely not one person is testing it with the database of moves removed, I would hope. I have made my feelings known, but I am not interested in a debate over it. Just rather how most people seem to be testing it, and how- with or w/o the moves database in it.



Thanks,

george

[rvida]

Just rather how most people seem to be testing it.

Does not matter in engine-engine matches.

The built-in DB contains some well known test positions which you have almost zero chance to encounter in real games.

[Matthias Gemuh]

Just rather how most people seem to be testing it.

Does not matter in engine-engine matches.

The built-in DB contains some well known test positions which you have almost zero chance to encounter in real games.

Thanks for taking out the DB.
How does one know whether a given Strelka55 is the bugfixed version or not ?

[gleperlier]

In fact for me it's quite simple :

- If Strelka author wants to have a DB for engine vs engine : No problem at all. It's like own book or something equivalent.

- For test suits, it's obvious that these databases are just cheating, so we have to disable.

Cheers,

Gab

[geots]

In fact for me it's quite simple :

- If Strelka author wants to have a DB for engine vs engine : No problem at all. It's like own book or something equivalent.

- For test suits, it's obvious that these databases are just cheating, so we have to disable.

Cheers,

Gab

Gab, let me explain to you if you don't mind how this works. Or should. It is the same as an author having an engine that has an opening book built into the engine. So a generic book cannot be used with it. So the testing group cannot test it. At that point they get in touch with the author of the program- and explain all that to him. THEN it is HIS choice as to whether or not to disable his own book- depending on how bad or whether or not he wants it removed. The user should never take it upon himself to do this himself- w/o the author's permission. Others may say they don't agree with me- but it is just a common, sometimes unspoken, act of respect and courtesy.


Best,

george


PS: There is a huge difference in going into an engine's parameters or config. file and disabling an opening book. That and going into his engine to disable things are as different as day and night.

[Dan Honeycutt]

Gab, let me explain to you if you don't mind how this works. Or should. It is the same as an author having an engine that has an opening book built into the engine. So a generic book cannot be used with it. So the testing group cannot test it.

Why can't they test it?

If the gui is using a book it will put the engine in force and send it the book moves. When those run out the engine is on it's own but it might play a couple more moves out of its own book. Somebody might say the engine has an unfair advantage but I assume you would not agree based on your statement above :

Surely not one person is testing it with the database of moves removed, I would hope.

Best
Dan H.

[rvida]

How does one know whether a given Strelka55 is the bugfixed version or not?

Try this position:

rn1qr1k1/1p2bppp/p3p3/3pP3/P2P1B2/2RB1Q1P/1P3PP1/R5K1 w - - bm Bxh7+

original version shows Bxh7 as best move at depth 6 almost instantly even though the score is negative
version with disabled database finds it _much_ later (> 30 seconds)

[IWB]

Just rather how most people seem to be testing it.

Does not matter in engine-engine matches.

The built-in DB contains some well known test positions which you have almost zero chance to encounter in real games.

Hello Richard,

I would be more interested in that database than the engine. I will not test the rip off engine but I would like to compare that little database with my own test set. Is there a possibility to get it in a readable format?

Thanks
Ingo

[Sedat Canbaz]

Just rather how most people seem to be testing it.

Does not matter in engine-engine matches.

The built-in DB contains some well known test positions which you have almost zero chance to encounter in real games.

Dear Richard,

It seems, you hacked Strelka 5.5 or maybe you have permission by Jury Osipov ?!

As far as i know, Strelka 5.5 was released as closed source engine

So i wonder, have you looked in other chess engines too,i mean about other engines include similar database ?

Btw,i wonder too...how many chess programmers (exception you) are looking (coping,learning,deleting,renaming...) in the others engine.exe files work ?

All this work is legal ?


Best,
Sedat

[Matthias Gemuh]

How does one know whether a given Strelka55 is the bugfixed version or not?

Try this position:

rn1qr1k1/1p2bppp/p3p3/3pP3/P2P1B2/2RB1Q1P/1P3PP1/R5K1 w - - bm Bxh7+

original version shows Bxh7 as best move at depth 6 almost instantly even though the score is negative
version with disabled database finds it _much_ later (> 30 seconds)

Thanks !
My copy still needs fixing.