Tuesday, November 8, 2011

Open, extensible composition models


Did VPRI just enter the FEXPR game?
The evaluator corrects several semantic inadequacies of LISP (described by Stoyan [5]) and is metacircular (written in the language it evaluates). The language provides:
  • lists and atomic values, including symbols
  • primitive functions called SUBRs
  • symbolic functions (closures) called EXPRs
  • a FIXED object that encapsulates another applicable value and prevents argument evaluation
  • predicates to discriminate between the above types
  • built-in SUBRs to access the contents of these values
  • a way to call the primitive behaviour of a SUBR
  • a quotation mechanism to prevent evaluation of literals
(my emphasis)
More later. HT @CraigOverend.

3 comments:

John Shutt said...

Yes. Yes, they did. Apparently implicit-evaluation fexprs (a sad choice), but definitely fexprs.

Manuel Simoni said...

Amazing. Seeing that they don't mention Kernel, maybe we need to propagandize 'em.

John Shutt said...

Hm. No <expr> access to dynamic environment, it seems. Going to have study this in-depth to really grok what they're getting at — the efficiency results may be of interest in themselves.