Arasan 24.1

[Frank Quisinsky]

Hi Jon,

after I start your engine, the engine asks for hash size normally.
When I start Arasan 24.1, your engine does not ask for hash size.

Maybe this has something to do with it.

Here ucidebug ...
Type ucidebug in the Shredder GUI ...

Best
Frank

Code: Select all

### 24.: to Akimbo 0.8.0 NN (0): quit

### 25.: to Arasan 24.1 NN (0): uci

### 26.: to Arasan 24.1 NN (0): debug on

### from Arasan 24.1 NN (0): Arasan 24.1 Copyright 1994-2024 by Jon Dart. All Rights Reserved.

##### Unknown token: Arasan
### from Arasan 24.1 NN (0): id name Arasan 24.1

### from Arasan 24.1 NN (0): id author Jon Dart

### from Arasan 24.1 NN (0): option name Hash type spin default 0 min 4 max 64000

### illegal option string: name Hash type spin default 0 min 4 max 64000
### from Arasan 24.1 NN (0): option name Ponder type check default true

### from Arasan 24.1 NN (0): option name Contempt type spin default 0 min -200 max 200

### from Arasan 24.1 NN (0): option name Use tablebases type check default true

### from Arasan 24.1 NN (0): option name SyzygyTbPath type string default syzygy

### from Arasan 24.1 NN (0): option name SyzygyUse50MoveRule type check default true

### from Arasan 24.1 NN (0): option name SyzygyProbeDepth type spin default 4 min 0 max 64

### from Arasan 24.1 NN (0): option name MultiPV type spin default 1 min 1 max 10

### from Arasan 24.1 NN (0): option name OwnBook type check default false

### from Arasan 24.1 NN (0): option name BookPath type string default book.bin

### from Arasan 24.1 NN (0): option name Favor frequent book moves type spin default 50 min 0 max 100

### from Arasan 24.1 NN (0): option name Favor best book moves type spin default 50 min 0 max 100

### from Arasan 24.1 NN (0): option name Favor high-weighted book moves type spin default 100 min 0 max 100

### from Arasan 24.1 NN (0): option name Randomize book moves type spin default 50 min 0 max 100

### from Arasan 24.1 NN (0): option name Threads type spin default 1 min 1 max 256

### from Arasan 24.1 NN (0): option name UCI_LimitStrength type check default false

### from Arasan 24.1 NN (0): option name UCI_Elo type spin default 3450 min 1000 max 3450

### from Arasan 24.1 NN (0): option name Use NNUE type check default true

### from Arasan 24.1 NN (0): option name NNUE file type string default arasan-d9-SFv4-20231228.nnue

### from Arasan 24.1 NN (0): option name Move overhead type spin default 30 min 0 max 1000

### from Arasan 24.1 NN (0): uciok

setoption name Ponder value false

### 27.: to Arasan 24.1 NN (0): setoption name Ponder value false

### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

setoption name OwnBook value false

out of do_all_pending, list size=0
### 28.: to Arasan 24.1 NN (0): setoption name OwnBook value false

setoption name UCI_LimitStrength value false

### 29.: to Arasan 24.1 NN (0): setoption name UCI_LimitStrength value false

### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name Ponder value false

pending command(a): setoption name Ponder value false
### from Arasan 24.1 NN (0): info string do_command: setoption name Ponder value false

do_command: setoption name Ponder value false
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name OwnBook value false

pending command(a): setoption name OwnBook value false
### from Arasan 24.1 NN (0): info string do_command: setoption name OwnBook value false

do_command: setoption name OwnBook value false
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name UCI_LimitStrength value false

pending command(a): setoption name UCI_LimitStrength value false
### from Arasan 24.1 NN (0): info string do_command: setoption name UCI_LimitStrength value false

do_command: setoption name UCI_LimitStrength value false
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
setoption name SyzygyTbPath value c:\Chess\_syzygybases-5

### 30.: to Arasan 24.1 NN (0): setoption name SyzygyTbPath value c:\Chess\_syzygybases-5

### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name SyzygyTbPath value c:\Chess\_syzygybases-5

pending command(a): setoption name SyzygyTbPath value c:\Chess\_syzygybases-5
### from Arasan 24.1 NN (0): info string do_command: setoption name SyzygyTbPath value c:\Chess\_syzygybases-5

do_command: setoption name SyzygyTbPath value c:\Chess\_syzygybases-5
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
setoption name SyzygyProbeDepth value 1

### 31.: to Arasan 24.1 NN (0): setoption name SyzygyProbeDepth value 1

### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name SyzygyProbeDepth value 1

pending command(a): setoption name SyzygyProbeDepth value 1
### from Arasan 24.1 NN (0): info string do_command: setoption name SyzygyProbeDepth value 1

do_command: setoption name SyzygyProbeDepth value 1
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
setoption name BookPath value <empty>

### 32.: to Arasan 24.1 NN (0): setoption name BookPath value <empty>

### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name BookPath value <empty>

pending command(a): setoption name BookPath value <empty>
### from Arasan 24.1 NN (0): info string do_command: setoption name BookPath value <empty>

do_command: setoption name BookPath value <empty>
setoption name Move overhead value 100

### 33.: to Arasan 24.1 NN (0): setoption name Move overhead value 100

### 34.: to Arasan 24.1 NN (0): isready

### from Arasan 24.1 NN (0): info string found 5-man Syzygy tablebases in directory c:\Chess\_syzygybases-5

found 5-man Syzygy tablebases in directory c:\Chess\_syzygybases-5
### from Arasan 24.1 NN (0): info string loaded network from file arasan-d9-SFv4-20231228.nnue

loaded network from file arasan-d9-SFv4-20231228.nnue
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): setoption name Move overhead value 100

pending command(a): setoption name Move overhead value 100
### from Arasan 24.1 NN (0): info string do_command: setoption name Move overhead value 100

do_command: setoption name Move overhead value 100
### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
### from Arasan 24.1 NN (0): info string in do_all_pending

in do_all_pending
### from Arasan 24.1 NN (0): info string pending command(a): isready

pending command(a): isready
### from Arasan 24.1 NN (0): info string do_command: isready

do_command: isready
### from Arasan 24.1 NN (0): readyok

### from Arasan 24.1 NN (0): info string out of do_all_pending, list size=0

out of do_all_pending, list size=0
Engine: Arasan 24.1 NN (0 MB)
by Jon Dart

[peter]

When I start Arasan 24.1, your engine does not ask for hash size.

When I start Arasan 24.1 in Shredder (13) GUI, hash is to be set the same way as that works with other UCI engines and Shredder.
Are you sure, you have installed Arasan as UCI (not winboard)- engine?
Just in case, regards

[Frank Quisinsky]

No understanding for your question.
I added ucidebug ...
Sorry!

[Frank Quisinsky]

Hm, I think I found the mistake!
If I delete the rc file from the Arasan 24.1 directory, Arasan asked for hash and load the hash-tables.
Depth seems normaly, all works fine.

Never I delete the rc file for Arasan versions with an uci configuration.
In Arasan 24.0 directory the rc file is inside and all works great.
Looking in differences (v24.0 / v24.1) in rc file but can find nothing.

So, it have to do with the rc file.

[peter]

No understanding for your question.
I added ucidebug ...
Sorry!

In winboard- protocol hash is set by arasan.rc, not by GUI (as that is the case in UCI- installation). In Shredder you can install Arasan as UCI- engine or as winboard- engine, so my suspicion was, you could have installed as winboard- engine maybe.
Sorry for interference, just wanted to tell I cannot reproduce your problem with my installation of 24.1 in Shredder (13)- GUI. Hash is to be set by GUI normally by me (like with other UCI- engines).

Edit: and I haven't deleted arasan.rc, (just reading not until now your follow- up posting you added while I was answering your latest but one still, our typewritings got crossed).
Let's wait for Jon's anwer, regards

[Frank Quisinsky]

Hi Peter,

after all I know from Arasan ... different settings are only possible with the rc file, not with UCI options.
I have that all the years in my heat, so I have the rc file all the time in Arasan*s directory.

Yes, let us wait of the answer by Jon.
But the problem is solved for me.
I can test the new version 24.1 ...

Thanks and Best
Frank

[Peter Berger]

Hi Peter,

after all I know from Arasan ... different settings are only possible with the rc file, not with UCI options.
I have that all the years in my heat, so I have the rc file all the time in Arasan*s directory.

Yes, let us wait of the answer by Jon.
But the problem is solved for me.
I can test the new version 24.1 ...

Thanks and Best
Frank

Arasan is a nice chess program. But the setup is kind of complicated. I couldnt even edit the arasan.rc file as it was write-protected on a windows system in a way that was not easy to overcome. Also the syzygy files are at a modern named place that I can't give the path to.

[jdart]

The intent is that UCI or CECP options override the arasan.rc settings. If that's not happening in CECP mode, that's a bug.

Command-line options override UCI or CECP options: for example -H 4G will set the hash size to 4G.

--Jon

[jdart]

Arasan is a nice chess program. But the setup is kind of complicated. I couldnt even edit the arasan.rc file as it was write-protected on a windows system in a way that was not easy to overcome. Also the syzygy files are at a modern named place that I can't give the path to.

Can you give some more information, especially about the second issue? What was the pathname?

[Frank Quisinsky]

Hi Jon,

please do not misunderstand ...
Or the next programmer is annoyed about Quisinsky (no interest on it).

1. Playing style for Arasan is just great (since around three-five years all the improvements --- playing style view ---) are wonderful.

2. What you do for openings is just great ... Arasan have a lot of understandings for openings, can be see after book moves with your eval. I like that a lot. That is highest TOP-Level niveau!!

3. Move-average, endgames, attacking games (fast wins) and the late and complicated mid-game (so many own ideas Arasan have for make in bad position the draw) ... all is wonderful.

Enough ...
You know all of this better as each other here.

But the rc file is complelty out, sorry!
The good and old winboard times, thinking about my Phase X in Gambit-Soft times I explain Winboard, the site had 3 million hits in only one year. But all this is history.

You are the top-programmer ... but maybe it make more sense to add the opening-book features you have, the learn-features you have, log file features as UCI options in Arasan. For every new release I have to check your rc.file. Means, have Jon changed here a bit and I must thinking for each Arasan configuration again and again.

Arasan today is must have and I like your engine a lot.
Good tested and optimated over many years by yourself ... and all this in each details.
But the rc file looks complicated and I think completly out today.
Yesterday I am looking 5 hours how I can solve the problem I have.
Checked my problem with 4 others GUIs and so one.

Best
Frank