Ponder 'on' in LittleBlitzer

[asanjuan]

I've recently discovered LittleBlitzer, is a great tool for testing, but I found the "ponder" check disabled in the tournament options.
I find this option very interesting to measure how well implemented is the hash table.
I'm currently using version 2.72. Is there any version with this option enabled? If not: When we'll find a new version?

Thanks in advance.

[Ajedrecista]

¡Hola Alberto!

I have just tried a thing with LittleBlitzer 2.5 (which should not be very different to 2.72 version): it is related with Engines.lbe file. What I do is the following:

a) I open a Notepad and I write the parameters of the engines (I learned that from Readme.txt file and some looks to ImmortalChess Forum); here is an example:

Code: Select all

Engine=Z:\Descargas\Test\a.exe
NumThreads=1
useSleepingThreads=false
Ponder=on
Hash=32
Engine=Z:\Descargas\Test\b.exe
NumThreads=1
useSleepingThreads=false
Ponder=on
Hash=32

b) Here is where I add 'Ponder on'. Then, clicking in 'File\Save as...' you change the name to "Engines.lbe" (please do not forget the " " for changing the extension of the file: from *.txt to *.lbe) and save it in LB (LittleBlitzer) folder. Well, obviously you have to change the paths of the engines (I tried two renamed StockFishs for simplifying the paths).

I try this Engines.lbe file with the Task Manager opened, and each engine used more than 25% (more than one core) in an Intel i5-760 (4 cores), so it should works fine. If anyone has a better idea, please post it. ¡Buena suerte!

Regards from Spain.

Ajedrecista.

[nthom]

The GUI ponder on feature was never implemented, sorry Modifying the engines file is indeed the way to do it.

[asanjuan]

The GUI ponder on feature was never implemented, sorry Smile Modifying the engines file is indeed the way to do it.

But the GUI must send a "go ponder" command to the engine, doesn't it? So modifying the engines file won't work.
Ok, I'll try with a greater time control to test my TT.

I supose i have to wait until version 3.0
Thank you Nathan.
(and of course, thank you for your tool)

[nthom]

Ah yes you are correct... the engine in question must not be following the UCI standard properly. Pondering was never a big priority as the aim was to get many threads running at once to use all CPUs.

Also sorry for the lack of new versions lately, I have been side tracked with another project