> It seems that all abstractions, from Pascal or Lisp virtual machines to C to higher and higher level languages to the ones popular today, are all descended fr

> It seems that all abstractions, from Pascal or Lisp virtual machines to C to higher and higher level languages to the ones popular today, are all descended from the search for a way to deal with that initial lack of protocol (rule) to get hardware makers to use the same instruction set and thereby let programmers use the same assembly language. Maybe I am misunderstanding this. But Lisp and C derives from very different views of the underlying machine. It is rather hard to unify a lisp machine (or other lambda calculus machine) instruction set with the instruction set assumed by a language like C. Even assuming extensibility of the instruction set, constructing a machine that can execute both the "usual" instruction set and lambda calculus efficiently is very very hard. So this may not really be practical.