- From: Michael Dyck <jmdyck@ibiblio.org>
- Date: Sun, 28 Feb 2016 00:10:30 -0500
- To: "Robie, Jonathan" <jonathan.robie@emc.com>, "public-xsl-query@w3.org" <public-xsl-query@w3.org>
In the editor's draft, the sentence in question now reads:
If F is a partially applied function,
the implementation of F is called, supplying
the value of each nonlocal variable binding
as the argument of the corresponding parameter.
The change of "fixed position" to "nonlocal variable binding"
addresses 2 of my 3 original points.
However, none of the other points I raised has been addressed. To recap:
(1)
We can, and should (and used to!) fully define the semantics of invoking
a partially applied function derived from a user-defined function
or anonymous function. This text does not.
(2)
It's unclear what "the corresponding parameter" means. A function F has
parameters and nonlocal variable bindings, but they are disjoint by
definition, so there's no "correspondence".
(3)
A partially applied function has no "underlying function".
Conceivably, we could specify a PAF that way, but we don't.
(4)
The CR *does* say what happens when you invoke a partially applied function.
-Michael
Received on Sunday, 28 February 2016 05:10:58 UTC