Re: continuation efficiency

[Thomas Bushnell, BSG]

Marius Vollmer <address@hidden> writes:

> It uses a different algorithm.  

The code you posted for call/ec is a wrapper around call/cc.  Asking
about *guile*, if it's a wrapper around call/cc, it can't be faster
than call/cc, unless guile does fancy continuation optimizations (which
I was told it doesn't).

So again: you said that:

1) In guile, call/ec might be a primitive, but in fact is not.
2) In guile, call/ec is faster than call/cc.

How can those two statements be true?