>reflected in the type of atoms.  Of course, to preserve type safety,
>property keys need to be typed, so that when we recover a data from a
>property list, its type is known.

Yeah, this was the part that I was missing (that you needed to have the preallocated property around to pass in).

So, in lisp, I assume I don't need to pass in this other piece of data since it's dynamically typed.

However, with this, it seems like I need to pass in the same property instance to "get" that I gave to "put" (so that the closure sets the right thing), so if I want somebody else to be able to get the properties I set, then I have to pass them all my property instances as well, right?  Which sort of causes the same heterogeneity problem again.  ;)  Or am I missing something?

I guess you could preallocate properties for the basic types in the atoms.mli file so everybody could use them, but you'd still have to have a set that your app uses for your own types.

Still, it's definitely a cute trick!

Somebody needs to make a website with "wacky but potentially useful functional patterns", since there's no way I'd have thought of this, or combinators, or whatever.  :)

Chris

-------------------
To unsubscribe, mail caml-list-request@inria.fr.  Archives: http://caml.inria.fr