hgm wrote:No progress with your Octopus project, yet?
SMIRF is outdated, nevertheless I am trying to extinguish some bad properties there.
In general my motivation to do chess programming has been violated severely. Thus my investigations in chess programming have been downgraded massivly and mostly are focussed elsewhere than in practice. E.g. I have found out (merely for myself, but not only me) null move heuristic to be an inconsistent idea, as you can experience in Zugzwang positions. First I tried to repair that by introducing a special answer null move (not simply allowing double null moves), which is working fine and has solved the Zugzwang problem. But null move pruning still remains the wrong approach.
The reviewed SMIRF thus will no longer use that null move heuristic. Instead this has been replaced by parts of a better idea. Unfortunately this now is somehow in concurrence to the existing internal intelligence feed back method. Thus the benefit of this exchange is not that big and partially covered by that in big parts probably now redundant old SMIRF basic idea. Anyway abandoning null move pruning here has not been a lost.
The follow up project to SMIRF has the name Octopus (which will change soon into a shorter four letter name). It is growing very slowly. And it is done creating a monochrome approach, completely symmetric for both parties Black and White, which only are occurring at the interface to the outer world.
I mentioned both themes partially within a German language forum. There has been rarely interest. Thus I better keep those things private and slow.
For your event I will then release a renewed SMIRF bonus engine.
Reinhard.
