Hi all,
here are some updates for what is new on the Freecell Solver trunk since the
release of version 4.2.0:
1. I noticed a bug in the JavaScript version where running "-l ve -mi 120000"
on MS deal No. 124 resulted in an intractable state instead of a solved state
with the command line version. I first blamed the clang/Emscripten compiler,
but after I was able to get it to work here on Mageia Linux (see
https://bugs.mageia.org/show_bug.cgi?id=18138 ), it turned out that a
clang-compiled fc-solve executable did not exhibit the problem either.
I eventually discovered that it happened if I limit the iterations and
gradually increment it by steps of 1,000 (or so), and found out that the
problem was in the handling of the flares. So I mad the handling less error
prone and wrote a test for it - in python - and committed the change and now
the JS version is fine:
Relevant commit -
https://github.com/shlomif/fc-solve/commit/70a0a780cf87f7f6c0791f630bf339ac99086094 .
2. I have applied some optimisations to the fcs_hash code.
3. I replaced the command line processing radix tree with the output of gperf
(see
http://www.ibm.com/developerworks/library/l-gperf/index.html ), which made
things somewhat faster.
4. Many small cleanups , optimisations and refactorings.
----
Right now the test suite runs at 19 seconds which is lower than what I recall
of it running at 20,21 or 22 seconds and formerly 24 seconds.
Furthermore, I hit a benchmark of
https://github.com/shlomif/fc-solve/commit/e46fd6ded4bd2fb5d58835a8174a70a1498bd9dc
- 8.3072s - or 3852.09119973118 deals per second.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
Stop Using MSIE - http://www.shlomifish.org/no-ie/
Chuck Norris knows what the gender of Great A’Tuin, the Discworld world turtle,
is.
— http://www.shlomifish.org/humour/bits/facts/Chuck-Norris/
Please reply to list if it's a mailing list post - http://shlom.in/reply .
Received on Tue Apr 26 2016 - 09:37:39 IDT