Forward referencing problem

[Fredrik Jonsson]

On Tuesday, February 17, 1998 5:07 PM, S.Lo@orl.co.uk [SMTP:S.Lo@orl.co.uk] 
wrote:
> The compilation fails not because of circular reference but because the
> interface B is not defined when the stub is compiled.

So, does that imply that as soon as I have a circular relation ship
I'm forced to compile those interfaces into a single translation unit?

(I wish there were IDL semantics to do both forwarding and some
indirect implementation inclusion where the inclusion is only affecting
the generated skeleton file)

--Fredrik Jonsson
Ericsson Austria AG