J3/00-334

Date:     12 Dec 2000
To:       J3
From:     Richard Maine
Subject:  Bindname (issues 94, 156, and 232)

This paper resolves the issues relating to BINDNAME=.

Issue 156 was resolved by paper 99-271r1, and is indicated as
resolved in 00-011r3.  It was accidentally left in the 007.
There was not an intent to reraise the issue, so this paper
deletes it from the 007.

Issue 94 is addressed by note 12.41.

There is a good chance that the edits below may end up getting
subsumed by a subsequent paper that re-organizes the material
relating to BIND.  But that paper is a bit more ambitious and
relates to things in addition to BINDNAME.  In the meantime,
and in case the other paper doesn't get done or passed, this
paper at least tries to fix the problems mentioned in the
unresolved issues.

Edits to 00-007r3.

[266:9-29] Delete issue 232.

[272:15] Replace this para with

  "A BINDNAME= specifier for a procedure is a processor-dependent
   specification of a name and mechanism by which the procedure
   may be invoked by a companion processor.  The valid values for
   and interpretation of the character expression in a BINDNAME=
   specifier are processor dependent."

[272:37+]  Add a new para to Note 12.41

  "Another possible processor choice is that the BINDNAME=
   specifier has no meaning.  In this case, it is recommended
   that the processor generate a warning diagnostic if the
   specifier is used."

[272:16-27]  Delete issue 156.

[391:11-13] Delete issue 94.

[401:41] replace this para with

  "A BINDNAME= specifier for a variable or common block is a
  processor-dependent specification of a name by which that
  entity may be known to a companion processor.  The valid values
  for and interpretation of the character expression in a
  BINDNAME= specifier are processor dependent."