Stockfish 1.4, the final weapon

[Christopher Conkie]

As the Glaurung site is well known, would it be possible to distribute Stockfish from this site too?

kind regards
Bernhard

He could even rename it Glaurung and all would be well.



Christopher

[Christopher Conkie]

As the Glaurung site is well known, would it be possible to distribute Stockfish from this site too?

kind regards
Bernhard

He could even rename it Glaurung and all would be well.



Christopher

Surely he doesn't have the right to do that.

I'm not sure Gabor since he does not completely own it anymore. Maybe he will have to ask the co-"authors". Anyway for us Stockfish 1.4 is already Glaurung 2.7.

Those nice people provided source you know

We like that name better.



Christopher

[mcostalba]

Hello Tord,

Do you maybe know if the Chess960 bugfix could also be used for the Stockfish 1.3 code? Or was it a specific problem with Stockfish 1.4.? I have not yet tested if everything works in Chess960 with the older version, I was hoping that you guys would find a solution to the problem

Regards, Eelco

The Chess960 bug is due to a silly parsing bug in uci.cpp

This is the patch that fixes it.

Code: Select all

        if (token == "value") 
       { 
+             // Reads until end of line and left trim white space 
+            getline(uip, token);
+            token.erase(0, token.find_first_not_of(" \n\r\t")); 
+
-           getline(uip, token); // reads until end of line
        set_option_value(name, token); 
        } else

The bug causes the uci option Chess960 to be ignored even when is set to true, so this actually disables Chess960 support.

Marco

[zamar]

Anyway for us Stockfish 1.4 is already Glaurung 2.7.

Well,

as Tord has stated here many times recently: Stockfish is the only Glaurung version that is under active development. So your your opinion that Stockfish 1.4 = Glaurung 2.7 is very close to the truth actually I see nothing wrong with that.

And I share your view that it must not be forgotten that >90% of the code is written by Tord, so he absolutely deserves the most credit. But it must also not be forgotten that around ~5-10% of the code is written by Marco, who seems also to be very talented person and has made many great findings to improve the program's strength.

For me, it is an honour to work with these two experts

[Eelco de Groot]

Hello Tord,

Do you maybe know if the Chess960 bugfix could also be used for the Stockfish 1.3 code? Or was it a specific problem with Stockfish 1.4.? I have not yet tested if everything works in Chess960 with the older version, I was hoping that you guys would find a solution to the problem

Regards, Eelco

The Chess960 bug is due to a silly parsing bug in uci.cpp

This is the patch that fixes it.

Code: Select all

        if (token == "value") 
       { 
+             // Reads until end of line and left trim white space 
+            getline(uip, token);
+            token.erase(0, token.find_first_not_of(" \n\r\t")); 
+
-           getline(uip, token); // reads until end of line
        set_option_value(name, token); 
        } else

The bug causes the uci option Chess960 to be ignored even when is set to true, so this actually disables Chess960 support.

Marco

Many thanks Marco

I will go try this bugfix out. Glad your team found that, I hope now Ancalagon can play Fischer Random too
So my analysis posted was just with the regular code and Ancalagon did not see any castling rights, because the engine thought it was regular chess.

It is by the way very well possible to check the first, opening positions this way, I collected quite a few files with Pro Deo on Chess960 opening positions at a time. Pro Deo of course does not play chess960 at all but doing castling in the first few moves is usually not recommended anyway, because some of these positions are so sharp, that castling wastes valuable tempo.

Thanks to Tord and Joona too!
Eelco

[Spock]

Yes, I view Stockfish as a positive step in the development and evolution of Glaurung. On his own, I think Tord pretty much stated that there would be no new versions of Glaurung itself and he had stopped work on it. Now there is a team of people working on it to move it forward and improve it, and that is a good thing and has actually kept Tord on board as well.

[mcostalba]

I will go try this bugfix out. Glad your team found that,

Credits to have found the bug should go to Tord, Credit to have intorduced it before 1.3 goes to me

[Christopher Conkie]

Anyway for us Stockfish 1.4 is already Glaurung 2.7.

Well,

as Tord has stated here many times recently: Stockfish is the only Glaurung version that is under active development. So your your opinion that Stockfish 1.4 = Glaurung 2.7 is very close to the truth actually I see nothing wrong with that.

And I share your view that it must not be forgotten that >90% of the code is written by Tord, so he absolutely deserves the most credit. But it must also not be forgotten that around ~5-10% of the code is written by Marco, who seems also to be very talented person and has made many great findings to improve the program's strength.

For me, it is an honour to work with these two experts

There may yet be hope of a compromise.

Tord has said here.......

http://lefounumerique.xooit.com/t511-St ... 4-2766.htm

......that Stockfish is Glaurung.

"It's still the same program. The only things which are different compared to before are the name and the fact that we are now three developers instead of just one."

So why does it need a new name?

All that remains is for him to put it on his site, rename it accordingly and then we can all wish you good luck.

We cannot have the Fruity business again.......it's way too much hassle.

Christopher

[mcostalba]

There may yet be hope of a compromise.

A compromise for who ?

There is no need of any compromise among our three because we are already aligned and well focused on what to do and how to do it.

If you mean a compromise with your wishes then I´m sorry for you but your wishes count less then zero to me.

I really don´t understund what is your goal. To divide our 3 ? To try to slow down SF development ? to make our group to split ?

I really don´t understund, but any is your target, is a bad one and, FWIW, is deemed to fail at 100%.

It is better you redirect your envy in someting more positive and useful for everybody... if you are able to do that, but I have big doubts about it.

[Christopher Conkie]

There may yet be hope of a compromise.

A compromise for who ?

There is no need of any compromise among our three because we are already aligned and well focused on what to do and how to do it.

If you mean a compromise with your wishes then I´m sorry for you but your wishes count less then zero to me.

I really don´t understund what is your goal. To divide our 3 ? To try to slow down SF development ? to make our group to split ?

I really don´t understund, but any is your target, is a bad one and, FWIW, is deemed to fail at 100%.

It is better you redirect your envy in someting more positive and useful for everybody... if you are able to do that, but I have big doubts about it.

I see you are unable to answer what is here......

http://lefounumerique.xooit.com/t511-St ... 4-2766.htm

You should be happy to work on a program called Glaurung.

Christopher