Is there publicly available information on what makes k/q interpreters so much faster? Or is that part of the secret sauce that the companies maintaining those interpreters sell? I assume it's the latter, but hope for the former, since good descriptions of the technical work that goes into high-performance systems usually makes for fascinating reading.
