On Sat, 11 Jul 2020 21:52:17 -0400 John Cowan <co...@ccil.org> wrote: > On Sat, Jul 11, 2020 at 7:10 PM Chris Vine <vine35792...@gmail.com> wrote: > > So when you said "continuable and non-continuable in Scheme are exactly > > like CL [signal] and error (and its variants)" are you then saying that > > SIGNAL == continuable (ie raise-continuable in R6RS) and ERROR == > > non-continuable (ie raise in R6RS)? > > That is what I meant. > > > In what way do you say that continuable exceptions are not in some sense > > analogous to common lisp restarts (noting the "in some sense"? > > Well, okay, but in *what* sense? You made the claim in the first place; I > think the burden of persuasion lies with you.
I am not sure that burdens is the best way of looking at it. But my point of analogy was that with either restarts or continuable exceptions, the stack is not unwound to the dynamic context in which with-exception-handler was called, which is the norm in conventional stack unwinding exception handling systems. Control returns to (for continuable exceptions) the point at which the exception was raised or (for restarts) the relevant restart.
