for benchmarking. gpu (cuda/thrust) version appears to be 50x faster.

fixed typo in unit efficiency curve

workspace update

schwefel problem: add broken termination condition

hydro problem objective function adjustments

local-maxima strategy to invert byte distributions

added null-analyser

gitignore Debug and Release folders

makelist tweaks

breakout util functions

file relocations

added convergence testing

added missing NDEBUG compile flag, removed unnecessary EXE_LINK flag additions

hierarchical makelist structure

documentation additions