Hi all,
well, the first news items of note that I don't know exactly what I did, but
at version 3.0.0 Freecell Solver has become a bit faster than before. Using "-
l by" and "./Tatzer -l p4b", it can solve the MS 32,000 in 93.8695681095123s
(340.89 boards per second).
Another item of note is that I added support for Google's dense_hash_map (see
http://code.google.com/p/google-sparsehash/ ) for both the positions/states
storage and the columns/stacks storage. This was inspired by this hash
benchmark:
http://incise.org/hash-table-benchmarks.html
Google's hash performed worse than fcs_hash.c (the fc-solve custom hash) in
the standard fc-solve benchmark, and there can be several factors: the C++->C
wrappers, the sub-optimal hash function, the fact the hash itself is less
optimal, etc. I may try to experiment a bit with it.
Otherwise I fixed a crash where if a soft_thread->name was uninitialised (and
thus NULL) then Freecell Solver would crash when compiling the prelude.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
Understand what Open Source is - http://shlom.in/oss-fs
God considered inflicting XSLT as the tenth plague of Egypt, but then
decided against it because he thought it would be too evil.
Please reply to list if it's a mailing list post - http://shlom.in/reply .
Received on Wed Jun 09 2010 - 06:19:42 IDT