(j3.2006) a question on allowing internal procedures as procedure targets

[Michael Ingrassia]

Yes, using the definition of "host instance" works better than treating
host instance as English words meaning the same thing as host subprogram,
thanks.   Making it a defined term as you suggest sounds like a good idea.


Point noted that we only say that scopes have hosts.
I may have just read 16.5.1.4p6 
"If an external or dummy procedure with an implicit interface is accessed
via host association, then it shall have the EXTERNAL attribute in the host
scoping unit" where it seems obvious now that "in the host scoping unit"
does not mean that the external or dummy procedure has a host scoping unit,
but it didn't yesterday.

By the way NOTE 12.19 has a reference to "host procedure scoping unit"
that should probably be a reference to the defined term "host scoping
unit". 
	--Michael I.