so my vacances are over and i have time for you guys
@tumbum (coin issue) its a bug with hrtbuddy
@lsqlhl5862
https://www.thebuddyforum.com/heart...custom-deck-silverfish-ai-82.html#post1667060
@Jack you find the exe somewhere on my github page, to use it copy it to the rigth place (Data folder) and change this line frome false to true:
this.useExternalProcess = true;
and the speed is in the most cases the same, simply because your tested boards are to small
which number related to = Boards calculated each depth level .?
there is not such an setting (only in the second turn simulation:
int nextturnsimMaxWidth = 10
which number related to = Childs board kept for next depth level search .?
int mxwde = 3000;
which number related to = Boards kept for enemy turn simulation .?
there is not such an setting! only a number for on how much boards are the second turn simulation is performed int twotsamount = 256; (if you use silver.exe set it to 1000 and above)
@ik777 the placement setting is only recomended with the exe
@jacks the second turn simulation uses (up to) 32 cores.
@kaime jeah, because silver.rush have his own settings. just search for this a second time: settings#####################
@Vector did you tried the mulligan-stuff i posted? (see link above)
@jacks i have to lower the settings in the nozdurm case
@runittwice you have to activate it in source and copy it into the right folder(see above)
and please dont put a empty test.exe in the same folder as the silver.exe! (it have to contain a board to simulate (the stuff you should post in an correct bugreport!)
@NaughtbutToLongName i look at this can you post a uai-log?
@jacks for the delay: i created a special version, with some log for the timings.will post it later!
edit use this one and send me crawler + uai log:
https://drive.google.com/file/d/0BxhxijYPL-BGMGxWVWhfdG42Yzg/view?usp=sharing
and read
https://www.thebuddyforum.com/heart...custom-deck-silverfish-ai-83.html#post1668004 for your settings.
@all thanks for your reports, im going to look at it soon.