A complete 2000 lines of code engine

[xr_a_y]

Ok thanks, I'll fix that !

But I'm not sure what wouldn't work following protocol 1 spec ?

[Werner Taelemans]

But I'm not sure what wouldn't work following protocol 1 spec ?

Probably:

version 1:
MOVE

version 2:
usermove MOVE

[xr_a_y]

ok indeed ....

I pushed a dirty patch ... https://github.com/tryingsomestuff/Mini ... 67fa10245c

I'll have to clean it up better some day.

Thanks again.

[xr_a_y]

Recent (too small) progress

Code: Select all

Rank Name                          Elo     +/-   Games   Score   Draws
   1 minic_0.55_linux_x64_avx2_bmi2      53      13    1546   57.5%   41.7%
   2 minic_0.53_linux_x64_avx2_bmi2      39      13    1547   55.7%   40.9%
   3 minic_0.50_linux_x64_avx2_bmi2      27      13    1547   53.9%   40.7%
   4 minic_0.47_linux_x64_avx2_bmi2      -7      14    1547   49.0%   37.8%
   5 minic_0.43_linux_x64_see4.2    -116      14    1547   33.9%   36.5%

3867 of 10000 games finished.

[Raphexon]

Still awesome that even with so few lines, it still would have been the strongest engine back in the late 90s/early 2000s.

[xr_a_y]

Well, I must admit this is just developing for fun (re)using what pioneers already tested back then and what is already working well on top engines. The real great part in the process is learning about chess, chess programming, re-discovering things by myself and sharing with the great chess programming community. I am still amazed about how kind everybody is here, spending time sharing, testing, searching and teaching.

[xr_a_y]

Minic 0.60 will be released tonight with +50elo thanks to TT use inside qsearch

[xr_a_y]

Minic is now better than fruit2.1

Code: Select all

Score of fruit_21 vs Minic: 33 - 65 - 34 [0.379]
Elo difference: -85.94 +/- 52.46

[silentshark]

Minic 0.60 will be released tonight with +50elo thanks to TT use inside qsearch

A +50 elo tweak is always nice, well done.

For my engine, probing the TT table in qsearch (but not storing in TT in qsearch) was the best. How about for you?

[xr_a_y]

Using TT in qsearch was useless in Minic with a very simple eval. It became good to store and probe in qsearch when eval started to be the bottleneck.

Just probing TT in qsearch seems strange to me, this will only be good when in check I guess... I store things in qsearch with depth=-1. I guess the important thing is to get eval score for free, I don't expect a lot from TT cut-off.