Sting SF 1.0 is out

[rodolfoleoni]

Hi Marek,

I ask Jim Ablett to compile my project and to make it available for all (thanks in advance!):

Unfortunately I have an agreement with the Stockfish team only to compile the official Stockfish version,
but I wish you much success with your project.

Jim.

It's not easy to understand why this agreement could involve the Sting engine. Author has been respectful of glp rules and it's not an "unofficial" Stockfish as many others we recently saw. It's a legal derivative (as well as Tinapa). A fork, not a clone.

I hope Stockfish team will allow you to compile it.

We asked Jim to not compile sources directly downloaded from our repo to avoid fostering random bianaries of a still in development software. But we have nothing against derived engines of official release: sources based on an officially released version, like Stockfish 2.1.1, and then further modified by the author and, of course, with the name changed. I think Sting belongs to this second group.

Clear and honest from you. As always.

[Graham Banks]

But we have nothing against derived engines of official release: sources based on an officially released version, like Stockfish 2.1.1, and then further modified by the author and, of course, with the name changed. I think Sting belongs to this second group.

Marco - are there any other Stockfish based derivatives which you'd consider to be in this second group?

[MikeGL]

Hello Mike:

I have problem connecting to .pl domains I guess.
Can we grab it in mediafire or some other sites with different URL without .pl

It seems that this file contains the sources. Mediafire link:

http://www.mediafire.com/?jf1omcgkejq1b4d

Enjoy! And thanks to Marek for his effort. Of course, thank you very much to SF team for programming an excellent engine such as StockFish is.

Regards from Spain.

Ajedrecista.

Thanks Jesús

[mcostalba]

Marco - are there any other Stockfish based derivatives which you'd consider to be in this second group?

I don't keep the list of SF derived (although perhaps I should), anyhow the little "rule" (an hint actually) I gave above should be self explanatory and everybody could judge for themself, in case you have some doubts on a particular engine you can open a discussion here on the forum: it would be instructive, mainly to clarify once more the GPL concepts with practical examples.

[Ajedrecista]

Hello:

I have found this info in Chess2U forum:

http://www.chess2u.com/t4988-sting-sf-1 ... able#28947

sting_SF_10_w32.rar

SF 2.1.1 won 290.5 - 209.5 (+168 -87 = 245). I did my calculations and (more less) I get (with 2-sigma confidence ~ 95.45% confidence) a rating difference of +56.8 ± 22.2 ~ [+34.6, +79] in favour of SF 2.1.1.

I do not know if in that test SF 2.1.1 was w32 or x64; I have the same doubt with Sting SF 111208 (again I do not know if it is the same executable of the download link I post).

Regards from Spain.

Ajedrecista.

[Michel]

Hello:

I have found this info in Chess2U forum:

http://www.chess2u.com/t4988-sting-sf-1 ... able#28947

sting_SF_10_w32.rar

SF 2.1.1 won 290.5 - 209.5 (+168 -87 = 245). I did my calculations and (more less) I get (with 2-sigma confidence ~ 95.45% confidence) a rating difference of +56.8 ± 22.2 ~ [+34.6, +79] in favour of SF 2.1.1.

I do not know if in that test SF 2.1.1 was w32 or x64; I have the same doubt with Sting SF 111208 (again I do not know if it is the same executable of the download link I post).

Regards from Spain.

Ajedrecista.

Well even if it costs elo, if this "sacrifice detection" really works it might be
useful for analysis. So far I do not understand the code (or the underlying principles).

[lech]

Hello:

I have found this info in Chess2U forum:

http://www.chess2u.com/t4988-sting-sf-1 ... able#28947

sting_SF_10_w32.rar

SF 2.1.1 won 290.5 - 209.5 (+168 -87 = 245). I did my calculations and (more less) I get (with 2-sigma confidence ~ 95.45% confidence) a rating difference of +56.8 ± 22.2 ~ [+34.6, +79] in favour of SF 2.1.1.

I do not know if in that test SF 2.1.1 was w32 or x64; I have the same doubt with Sting SF 111208 (again I do not know if it is the same executable of the download link I post).

Regards from Spain.

Ajedrecista.

Well even if it costs elo, if this "sacrifice detection" really works it might be
useful for analysis. So far I do not understand the code (or the underlying principles).

Of course, it is a very good result. My proposition is, to try some positions.
It may help us to know usefulness of my idea too.
I have not a computer to work at this idea (it works in my head mainly), but it will change soon. Then my work will be faster and better than now.
It is a time to prove that machines may play chess like human without A.I.

[MikeGL]

Hesús,

Seems like that compilation is strange.
It has no option to modify the Hash. It has only a default hash of 5MB.
Engine keeps on crashing in Arena.

Also, if I ran that specific Sting1.0 compilation in command prompt and typed uci, it displays the option Hash but it cannot be set to anything.
I typed: setoption name Hash 512

But it is claiming no such option.
Anyways I'll attempt my rusty and lousy compilation skills with that source code.

[Graham Banks]

[Eelco de Groot]

Very nice logo Graham! This is what I associated the name with:



The Stingray submarine, from the Supermarionation series by Jerry Anderson. "Marina, Marina!"

Eelco