J3/01-218
To: J3
From: Malcolm Cohen
Subject: Unresolved issue 332
Date: 21st May 2001

1. Introduction

Issue 332 says
  "Shouldn't R823 have a constraint like the 2nd one after R818?  I don't
see
   how one can argue that the one on R818 applies."

Since R823 invokes R818 itself, the one on R818 ought to apply.

On the other hand, the BNF is worded in a slightly confusing way here.
In particular, the constraint does not apply if you miss the
<associate-name>
syntax out!  This is a mistake which I'll fix, at the same time as making it
a little more transparent about what constraints apply where.

2. Edits to 01-007r1

[154:39] Replace "[ <associate-name> => ] <selector>" with "( <selection>
)".
[154:39+] Insert
"R817a <selection> <<is>> <associate-name>
                   <<or>> <association>
 Constraint: The <associate-name> shall be a polymorphic variable."
[154:40-] Move R823 from [156:12] to here (rule for <association>).
{Amalgamate the "name=>selector" syntax.}

[154:32] Delete constraint.
{This is now already enforced by the syntax rules.}

[155:17-18] Delete "if specified ... <selector>".
{We now always have an associate-name.}

[156:12] We moved this to [154:40-] already.

[156:13-16] Delete J3 note 332.

===END