[gnugo-devel] experimental influence cleanup

[Arend Bayer]

Would it be consensus to remove the --experimental-influence command
line/configure option?
I care most about cleaning that ugly Y/FY classification mess from the
non-influence pattern databases, to which out tuning didn't pay attention
anyway (since --experimental-influence got default). Nuking those, it
wouldn't make much sense to keep the old value_territory function either.

(I would leave the F/FY stuff in the influence-specific pattern databases
however -- this would then continue to indicate patterns that should only
be used in either pre-owl or post-owl influence computations. I.e.,
pre-owl influence computation would remain untouched as before.)

Arend