[plt-scheme] SRFI-4 and FFI

On Aug 30, John Kozak wrote:
> A suggestion: I think it would be better if SRFI-4 vectors were
> layered on top of cvectors: tidier, easy genericity when needed.  Is
> there a reason I'm missing why this wasn't done?

It's a very similar implementation, but IIRC the extra field in
cvectors (the type field) made it difficult to share the
implementation.


> I knocked up a quick implementation of SRFI-4 vectors as cvectors
> and discovered that foreign.ss provides all the SRFI-4 primitives.
> Presumably this is a bug, as the code seems to be trying not to do
> this?

No, the foreign library implements both.  cvectors are more generic
(obviously) but just a bit slower since each one comes with a fixed
type.

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                  http://www.barzilay.org/                 Maze is Life!