New Clubfoot and Bitfoot release builds available

[zd3nik]

Clubfoot-1.0.048995d and Bitfoot-1.0.65acfcb binaries available.

https://github.com/zd3nik/Clubfoot/rele ... .0.048995d
https://github.com/zd3nik/Bitfoot/relea ... .0.65acfcb

Mostly search fixes and enhancements. Also some minor eval tweaks.

Enjoy.

[Graham Banks]

Clubfoot-1.0.048995d and Bitfoot-1.0.65acfcb binaries available.

https://github.com/zd3nik/Clubfoot/rele ... .0.048995d
https://github.com/zd3nik/Bitfoot/relea ... .0.65acfcb

Mostly search fixes and enhancements. Also some minor eval tweaks.

Enjoy.

Hi Shawn,

roughly how strong is Clubfoot in CCRL terms?

Graham.

[zd3nik]

Hi Shawn,

roughly how strong is Clubfoot in CCRL terms?

Graham.

I have no idea, but probably not very strong.

Bitfoot is much stronger, though still relatively weak in comparison to today's top engines. It's good enough to beat SOS consistently and good enough to give Fruit 2.1 a good game - up until the endgame that is.

How does an engine get on the CCRL?

[Graham Banks]

Hi Shawn,

roughly how strong is Clubfoot in CCRL terms?

Graham.

I have no idea, but probably not very strong.

Bitfoot is much stronger, though still relatively weak in comparison to today's top engines. It's good enough to beat SOS consistently and good enough to give Fruit 2.1 a good game - up until the endgame that is.

How does an engine get on the CCRL?

I'll take a look at Bigfoot for CCRL 40/40.

Although we test a much wider strength range in CCRL 40/4, at present I usually restrict my 40/40 testing to engines that are 2000+ on that list.

Cheers,
Graham.

[Gurcan Uckardes]

On CCRL scale, both should be 2000+. Previous versions on Android already 1900 and 2100 roughly. On CCRL 150 to 200 elo more may be expected.
Above numbers concern previous versions compiled for arm5 i've tested. Newest ones should add some more elo's by definition.

[Graham Banks]

Hi Shawn,

roughly how strong is Clubfoot in CCRL terms?

Graham.

I have no idea, but probably not very strong.

Bitfoot is much stronger, though still relatively weak in comparison to today's top engines. It's good enough to beat SOS consistently and good enough to give Fruit 2.1 a good game - up until the endgame that is.

How does an engine get on the CCRL?

Could you please provide a non-popcount exe as well?
I'm hoping to start a gauntlet for CCRL 40/40, but can't unless I can get such a compile.

Graham.

[zd3nik]

Could you please provide a non-popcount exe as well?
I'm hoping to start a gauntlet for CCRL 40/40, but can't unless I can get such a compile.

Graham.

Sorry about the slow response.

I'll have to dig up some old code to see what I was doing before using pop_count (probably some bit twidling code I found somewhere online). It might be a few days before I get to it. How soon do you need it?

[Graham Banks]

Could you please provide a non-popcount exe as well?
I'm hoping to start a gauntlet for CCRL 40/40, but can't unless I can get such a compile.

Graham.

Sorry about the slow response.

I'll have to dig up some old code to see what I was doing before using pop_count (probably some bit twidling code I found somewhere online). It might be a few days before I get to it. How soon do you need it?

No hurry. I can wait.

[zd3nik]

Do you only want popcount calls removed? Or do you want _BitScanForward64, _BitScanReverse64, __builtin_ctzll, etc removed as well?

Also, I'm curious: why do you need versions without popcount in the first place? Are you running the tournaments on old hardware?

[Graham Banks]

Do you only want popcount calls removed? Or do you want _BitScanForward64, _BitScanReverse64, __builtin_ctzll, etc removed as well?

No idea. Perhaps another programmer might answer.

Also, I'm curious: why do you need versions without popcount in the first place? Are you running the tournaments on old hardware?

Sadly yes.
I'm running a Q6600 and a Q8200 on which I do my gauntlets.
I use the i5 for tournaments.