From Ian Medland<368218CF.D840403D@s-and-t.com>
(snip)
>What is at issue is the requirements that an object reference variable
>hold at all times a valid reference.
>
>It could be argued that the variable in the application can not hold a
>reference to an object that is not in scope (hence it can only hold
>references to objects in the application or current scene) - to do
>otherwise is not valid.
>
>Otherwise it could be suggested that references are "dumb". They are
>just placeholders. As long as the reference is syntacticaly correct this
>is enough. IF the reference is then used to target a non existant object
>- well that's a different matter altogether.
>
>Can people confirm the validity of this code or argue against it please
>(the more votes the better).
I think this issue had already discussed before and the latter opinion
is right.
>I'm sorry that some of us have gone round this before, but in the spirit
>of MHEG I am trying to get the group consensus. It has been suggested
>that the spec is not suitably strict about this.
It is sure there is no clear declaration in the spec, but basically
the action which included such a invalid object reference should ignore.
>Oh, and have a great Christmas too :-)
Merry Christmas and Happy New Year!
-- Yoshiharu
Sony Corporation IT Labs Network Lab. BDS Gp.