TalkChess.com

Arasan 20.4 is available from https://www.arasanchess.org.

I am putting this out a little sooner than I had planned because it fixes a number of bugs that were in 20.3.

Changes in Arasan 20.4:
1) Fix Linux/Mac Makefile so PGO works with clang.
2) Code cleanup. Replace NULL with nullptr in most places (except some C library calls).
3) Fix serious bug in Syzygy tb code for handling cursed losses.
4) Fix "ecocoder" target in Makefile and fix ecocoder code (sync with changes to Notation).
5) Updates to "pgnselect" utility program.
6) A few fixes in the tuner program. Remove ordinal objective (was not working properly).
7) Add -d option to test command so can do fixed depth. Profile uses fixed depth now.
8) Some updates to the label_positions.py tool.
9) Re-tune parameters with larger traning set and more frequent PV updates.
10) Bug fixes in "easy move" code.
11) Some minor changes/fixes to root move ordering.
12) Bug fixes for UCI multi-pv mode.
13) Add UCI contempt option.

This version is about +16 ELO stronger than 20.3 in my testing, on an arbitrary scale where current versions about ELO 2400. So maybe a bit more on a CEGT or CCRL scale.

--Jon

Nice! Thanx Jon. Too bad my current tournament started a week ago with 20.3 already. Will have to wait till february for the next run and will include 20.4

jdart wrote:Arasan 20.4 is available from https://www.arasanchess.org.

I am putting this out a little sooner than I had planned because it fixes a number of bugs that were in 20.3.

Changes in Arasan 20.4:
1) Fix Linux/Mac Makefile so PGO works with clang.
2) Code cleanup. Replace NULL with nullptr in most places (except some C library calls).
3) Fix serious bug in Syzygy tb code for handling cursed losses.
4) Fix "ecocoder" target in Makefile and fix ecocoder code (sync with changes to Notation).
5) Updates to "pgnselect" utility program.
6) A few fixes in the tuner program. Remove ordinal objective (was not working properly).
7) Add -d option to test command so can do fixed depth. Profile uses fixed depth now.
8) Some updates to the label_positions.py tool.
9) Re-tune parameters with larger traning set and more frequent PV updates.
10) Bug fixes in "easy move" code.
11) Some minor changes/fixes to root move ordering.
12) Bug fixes for UCI multi-pv mode.
13) Add UCI contempt option.

This version is about +16 ELO stronger than 20.3 in my testing, on an arbitrary scale where current versions about ELO 2400. So maybe a bit more on a CEGT or CCRL scale.

--Jon

Maybe someone will be interested. Here is my compilation of Arasan on android (arm7, arm64v8):
https://app.box.com/s/5vcpcdis72z0dbvf5a3ihkuuw9yvsudv

Good news. Thanks Jon

Till now I played 36 games with Arasan 20.4. Based on this small number of games it scores at least 80 elopoints more than 20.3.

One thing amazes me. It shows no use of Syzygy bases.

attakinski wrote:It shows no use of Syzygy bases.

As I have noted earlier, generally it is not sufficient to configure Syzygy base access in the arasan.rc file, unless you are running Arasan standalone from the command-line. Otherwise you need to set the Syzygy path in whatever GUI you are using. If you do this, it should access the tablebases.

Can you tell me what UI you were using and if the TB options are set in the UI?

--Jon

I use CB/Fritz GUI.

Arasan 20.3 and 19.4 have no problem with the use of syzygy under the Fritz GUI.
Besides that: the syzygy bases are used without problem by at least 30 other programs on my PC.
Fire, Raubfish, Stockfish, Vajolet, Fruit reloaded, Pedone, Laser, Sugar, Chiron, Fizbo and so on.

Forgot to mention: the syzygy bases are set under the GUI.

I have looked at this a little bit and I think it is broken under Fritz. In fact I don't think it ever worked; it should not have worked in version 20.3 either. It works in Arena, because Arena sends the UCI commands in a different order.

I should be able to have a fix fairly soon.

--Jon

jdart wrote:I have looked at this a little bit and I think it is broken under Fritz. In fact I don't think it ever worked; it should not have worked in version 20.3 either. It works in Arena, because Arena sends the UCI commands in a different order.

I should be able to have a fix fairly soon.

--Jon

Hi Jon,
a short test inside Fritz 16 GUI shows tbs are used (SYZYGY)

Neue Partie Line
4k3/4p3/8/8/3p4/8/4N3/4KB2 w - - 0 1

Analysis by Arasan 20.4 x64 1CPU:

1.Se2xd4
Weiß setzt Matt: +- (#30) Tiefe: 13 00:00:00 171kN, tb=3
1.Se2xd4

(of Course other engines show here more tbs)