Ares 1.005 available

[fern]

Charles, I would love -and perhaps others too- a 32 bit version. I have not a 64 bit PC.
Yes, I am out of tune, but...


Fern

[CRoberson]

Fern,

For you, I'll do it. It may be a day or two.

Charles

[CRoberson]

This is one of the best news recently. Thank you, Charles.

I recall you used to test NoonianChess. I have a new version of that too. It needs more testing and I am still modifying it.

Uh, that was a looong time ago, Charles. Which does not mean I would not test it again. What is the download link?

BTW, is Ares a fork of Noonian or an independent program?

It was just a moment ago.
I don't have a download link yet. Still working on it.
Ares is an independent program. A complete rewrite with different data structures and a different language. I coded Noonian in C and Ares in C++.

After the 2002 ICGA WCCC, I did quite a bit of research on somethings I thought about at the tournament. I planned for these ideas to make their way into Noonian, but then I decided to start over. Over the last decade, my mind just kept wondering what would those ideas have done to Noonian.

[fern]

Thank you Charles.
I will contribute, of course
I am surte your qares will be or already is even more entetaining than my human like CSTAL I and II

Fern

[CRoberson]

Thank you Charles.
I will contribute, of course
I am surte your qares will be or already is even more entetaining than my human like CSTAL I and II

Fern

If you don't like the playing style, let me know. I tried to improve certain evaluation terms without too much change to the style. So, I am not committed to keeping it exactly like this. This isn't about getting to the top of CCRL, CEGT ... It is about having a virtual human opponent.

I was up late 2 nights ago a played it while practically asleep. I thought I was going to win for a moment, but the moment didn't last.

[pgn]
[Event "Human Test Game"]
[Site ""]
[Date "2014.05.21"]
[Round "?"]
[White "Charles Roberson"]
[Black "Ares 1.005 64 bit"]
[Result "0-1"]

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 b5 7.Bb3 e6 8.O-O b4
9.Nce2 Nxe4 10.Re1 Bb7 11.Ng3 Nxg3 12.hxg3 e5 13.Nf3 Be7 14.Bd5 Bxd5 15.
Qxd5 Nd7 16.Bd2 Rb8 17.c3 bxc3 18.Bxc3 O-O 19.Ba5 Nf6 20.Qd2 Qd7 21.Rac1
Rfc8 22.Rxc8+ Rxc8 23.Rc1 Rxc1+ 24.Qxc1 Qa4 25.Qc8+ Bf8 26.Bd8 Qxa2 27.Be7
Qb1+ 28.Kh2 Ng4+ 29.Kh3 Nxf2+ 30.Kh4 Qh1+ 31.Kg5 Qh6+ 32.Kf5 Qg6# 0-1
[/pgn]

A 32bit version will run at half the speed, thus losing about 100 Elo and it could lose more than that if your computer is old and slow.

[fern]

My PC is not that slow. In any case 100 elo pounts less means nothing to a program that anyway shpould be 500 points stronger than me, at the very least.

Fern

[carldaman]

Thank you Charles.
I will contribute, of course
I am surte your qares will be or already is even more entetaining than my human like CSTAL I and II

Fern

If you don't like the playing style, let me know. I tried to improve certain evaluation terms without too much change to the style. So, I am not committed to keeping it exactly like this. This isn't about getting to the top of CCRL, CEGT ... It is about having a virtual human opponent.

Finally, a developer after my own heart!

Thanks a lot, Charles! Your work is appreciated.

Best regards,
Carl

[JVMerlino]

After a long break, I've been working on Ares again. The new version is improved and is released as a full 64 bit binary instead of 32 bit for Ares 1.004. My tests indicate about a 200 Elo improvement over Ares 1.004.

I have a website for it. The licensing is at the bottom and is simple. You can use it, but not copy, reverse engineer, modify or any other copy method. I don't allow redistribution either.

The site is http://aigames.net/AresChess1-005.html

There are two verison:
1) 64 bit with popcnt
2) Plain 64 bit.

If anybody needs a 32 bit version, let me know. If anybody needs a Linux version, let me know.

Charles

Congratulations!

jm

[CRoberson]

Ok, Ferdinand notice some issues which led me to notice an extra one.

So, new versions have been uploaded to aigames.net. I left the version number the same due to only a one character change in the code.

1) The trailing a1a1 in the pv has been removed and I tested analyze mode in Arena 3.5. It works fine now.
2) The UCI_ShowCurrLine issue is not an issue. Years ago I turned off showing current line until 10 ply. Basically, it is a way to speed up analysis mode on fast computers. All that extra info in the early plies seems a waste.

3) THIS IS THE BIG ONE! ! ! !
When installing Ares in Arena do NOT let Arena auto-detect the protocol. Force the UCI protocol yourself at installation time. Ares supports a minimal amount of the xboard protcol for my purposes. Arena's auto-detection sends the xboard command and if the proper response comes, then it assumes xboard. If there isn't a response, it tries uci. With Ares, it falls into the xboard protocol. I tested this in detail with Arena 3.5 and it works fine when UCI is the forced protocol.

Fern,
I'll get to the 32 bit version in a day or two.

[CRoberson]

Must issue with Arena 3.5.

Also, the initial setting to UCI doesn't persist past closing Arena. You must go into Engines->Manage->Details and then force UCI for Type. If you don't Arena 3.5 will set it to Autodetect.