J3/00-134

Date:		29th February 2000
To:		J3
From:		Malcolm Cohen
Subject:	Issue 215

1. Introduction

Two papers (from separate meetings) edited the same lines in the document.

For this reason the editor omitted the note inserted by 99-186r1 which said
  "The above rules guarantee that there can be no run-time type errors
   through argument association."
I agree that this is confusingly worded; what it means is that type errors
in argument association can be detected by static analysis - i.e. type
correctness is not data-dependent.  Although this is an important design
issue, perhaps we can pass over it in silence in the standard itself.

The editor also notes
  "A sentence about the dummy argument assuming its type from the actual seems
   to have disappeared somewhere.  I'm not sure now whether some other paper
   said to delete it or whether I just did so by accident."
Either way, it should be re-inserted for now (another paper may attempt some
simplification in this area).

2. Edits to 00-007

[257:9] Replace "INTENT(OUT" with "INTENT(OUT)".
{Fix typo.}

[257:13+] Insert new paragraph
          "A polymorphic dummy argument assumes its dynamic type from the
           dynamic type of the associated actual argument.  When execution of
	   the procedure terminates, if a polymorphic dummy argument does not
           have INTENT(IN) and is a pointer or is allocatable, the dynamic type
	   of the associated actual argument becomes the same as that of the
	   dummy argument."
{Reinsert missing text.}

[257:14-28] Delete the J3 note.

===END