The fix I made earlier for this relied on a callback that was skipped in
release builds. Now always take the path that involves making the
callback when one is provided. Also remove an optimization that was
trying to eliminate possible moves based on scores prior to doing the
more expensive full check. In 2018 I prefer simplicity, and can make the
remaining code faster if that's required.
I *think* the reason I'm occasionally seeing toasts about not finding a
move is that when the engine's interrupted by there being a UI event in
the queue that error is posted. Instead try posting only when at the end
of the search nothing's been found.
where you do "hint", then "prev hint" and get told there are no moves
found only to have the next "prev hint" succeed. This is a hack, but
the right fix is eluding me, and will certainly be riskier.
if all tiles are used (ending the game and triggering penalties for
other players.) Changes are protected by compile-time flags currently
not turned on.
forward only) but disabled at compile time. Idea's to have a dict
browser. There was some simple refactoring in common code Android
uses, and that tests fine.
one and changing the site that used it to use the other. Not visible
outside of common, but should make it easier to harvest a list of all
words for one move or the entire game.
player. Works for gtk client. Compiles for Android but there's no UI
yet to specify more than one dict. Management of dupicate dicts
without duplicating memory -- refcounting -- will be up to the
platforms.
implemented (when not smart) as trying to match the human's score to a
per-robot value between 1 and 100 that gives the percentage of best
moves to store before picking randomly from among them. So a 1 means
save only the best move and always pick it; 100 means save all the
best moves (how many are saved is compile-time configurable) and pick
one of them. Because it's settable per-robot a smarter robot can be
played against a dumber one (though I may choose not to make it
settable per-robot on shipping versions.)
possible moves in either order both within a cached subset and when
building cached subsets. Still a bit buggy (shows the same move twice
when moving backwards and reaches the top-scoring move) and not well
tested.
deal with the output by removing params where possible and elsewhere
by adding XP_UNUSED macro wrapping __attribute__((unused)). There
should be NO change in function in spite of the large number of files.
