Indeed, I thought that was probably what you were talking about. It's somewhat a mystery (at least to Lisp programmers) why they wouldn't just port it to a faster Lisp (any major commercial Lisp, for example - doing such ports should be relatively trivial), but that may have to do with the perception of Lisp programmers being hard to find. You'd think a company would rather undergo that than the massive technical problems involved in reinventing a lot of Lisp (especially stuff like Graham's which (I would assume based on his books) uses a lot of very Lispy constructs), in another language, but it's their choice, I suppose