11-165
To: J3
From: Nick Maclaren
Subject: Interop TR: first part of notes on type(*)
Date: 2011 June 07
Reference: N1854


There has been a great deal of cross-purposes discussion on many aspects
that come down to people not closely involved in this group
misunderstanding the purpose and intent of type(*).  Upon looking
further, there is a factual error in the current extended notes.  I
think that we should spell out the situation rather more bluntly.

The error is that point (2) in A.1.1 on page 31 states that it is an
interoperable dynamic type, which is not the meaning of interoperable
type used in the main standard.



Edits to N1854:
---------------

[31:12] replace "interoperable dynamic type for which" by "possibly
unspecified dynamic type but which" and after "with the entity" append
"to the companion processor".

[31:20+] Add a new paragraph:

"Because the purpose of assumed-type is to allow the companion processor
to bypass some of the strictness of the typing in the Fortran standard,
it is not generally a suitable type for use within a Fortran program and
no facilities have been provided to make it more useful for that.
Fortran programs wanting a similar functionality should use CLASS(*) and
not TYPE(*).