Halogen 8 time management problem

[Angle]

I am impressed by the playing style of new Halogen 8 but I have a problem with its time management. On my Nitro AN515-52 Intel(R) Core(TM) i7-8750H with Arena 3.5.1 it losses on time after 65-70 moves using incremental TC while with tournament TC 40/2 it leaves approx. 3 seconds for the last 5 moves which leads to blunders. The situation is similar both for x64 and x64-avx2 versions. Anyone can help?

[Kieren Pearson]

Hi Angle, Halogen author here. Can you give me a little more specific information about the time losses? What was the time control used? Halogen 8 did well in the 40/15 leader boards this week but I won't be surprised if its not great at repeating time controls because I don't test on repeating controls. The time losses on the incremental TC is a big issue though.

[Gabor Szots]

My additional information is that using 2+1 under the Shredder 13 GUI there was not one single time loss out of 512 games. Haven't yet tried under Arena.

[Angle]

Hi Kieren! At 20s+3s TC Halogen used about 3.1 s per move losing оn time at move 65 to 70. At 40/2 TC it remained with about 3s at move 35 which forced him play almost instantly next 5 moves.

[Gabor Szots]

Hi Kieren! At 20s+3s TC Halogen used about 3.1 s per move losing оn time at move 65 to 70. At 40/2 TC it remained with about 3s at move 35 which forced him play almost instantly next 5 moves.

By any chance did you not install it as a WinBoard engine?

[Kieren Pearson]

I've not been able to reproduce the 20s+3s TC the time losses. I just today wrote and tested some patches pertaining to time control and time losses that should fix the issue. If you are using Halogen for enjoyment and are able to compile it yourself feel free to run the latest dev build on the master branch. Halogen is mostly tested under incremental time controls and I do understand does not manage time well in repeating time controls. I will aim to fix this in the next few days.

[Angle]

Hi Kieren! At 20s+3s TC Halogen used about 3.1 s per move losing оn time at move 65 to 70. At 40/2 TC it remained with about 3s at move 35 which forced him play almost instantly next 5 moves.

By any chance did you not install it as a WinBoard engine?

I installed it as a UCI engine. It doesn't work as a WinBoard.

[Angle]

I've not been able to reproduce the 20s+3s TC the time losses. I just today wrote and tested some patches pertaining to time control and time losses that should fix the issue. If you are using Halogen for enjoyment and are able to compile it yourself feel free to run the latest dev build on the master branch. Halogen is mostly tested under incremental time controls and I do understand does not manage time well in repeating time controls. I will aim to fix this in the next few days.

Thank you! I will try to compile it myself.

[Angle]

I downloaded Halogen compiles from the Chess Engines Diary:
https://chessengines.blogspot.com/2020/ ... -nnue.html
and now have no time losses using avx2 version. The problem of allocating too little time for the last few moves at ТС 40/2 still remains but there are no time losses in this case either.

I am now running a big tournament with various versions of Beef, Cheese, Counter, Francesca, Fruit Reloaded, Koivisto, Mr Bob, Orion, Rodent, Seer, Tucano, and Weiss. I will add Halogen to this torrney and report you the result by the end of this week.

[Guenther]

I downloaded Halogen compiles from the Chess Engines Diary:
https://chessengines.blogspot.com/2020/ ... -nnue.html
and now have no time losses using avx2 version. The problem of allocating too little time for the last few moves at ТС 40/2 still remains but there are no time losses in this case either.

I am now running a big tournament with various versions of Beef, Cheese, Counter, Francesca, Fruit Reloaded, Koivisto, Mr Bob, Orion, Rodent, Seer, Tucano, and Weiss. I will add Halogen to this torrney and report you the result by the end of this week.

Ouch, do you realize that this site just recycles otherwise free downloads and opensource projects just for making click money,
by re-uploading them to file hosters?

Actually (as expected) the binary is an identical copy of the avx2 compilation as from Halogens github release site! ( I did a file comparison)
What compilation did you use before? May be you used the nopopcount one?
https://github.com/KierenP/Halogen/releases

Code: Select all

Halogen8-x32.exe
848 KB
Halogen8-x64-AVX2.exe
961 KB
Halogen8-x64-nopopcnt.exe
966 KB
Halogen8-x64.exe
963 KB