Re: ORBit order dependence within idl files?

[Todd Graham Lewis <tlewis mindspring net>]

On 23 Jul 1998, Raja R Harinath wrote:

> > I'm presuming that since this came from an OMG spec that it's legal
> > CORBA, but it's impossible for me to tell.  
> 
> But a random fragment may not be a full-fledged IDL definition.

Yeah, my piecing the various definitions scattered throughout the
spec might have been an imperfect process...

> This just requires a forward declaration of one of the interfaces.  Just
> add
> 
> 	interface SubtransactionAwareResource;
> 
> at the beginning. 

In private mail, Elliot:

1) made me aware of forward declarations;
2) made me aware the ORBit does not support them;
3) reminded me of his exceptional amenability to patches.  8^)

> With this, the messages are:
> 
>   sejong $ /opt/Orbix_2.2MT/bin/idl bar.idl
>   6:(semantic): Identifier `Inactive' not found 
>   6:(semantic): Identifier `NotSubtransaction' not found 
>   10:(semantic): Identifier `Resource' not found 
> 
>   sejong $ idl bar.idl                           #mico
>   bar.idl:6: scoped name 'Inactive' not defined
> 
>   sejong ~/z/z/orbit > orbit-idl -d2 bar.idl
>   bar.idl:6 `Inactive' undeclared identifier
> 
> (These are of course, real errors in the fragment that you posted).

Sure.  Here's a much cleaned-up version to assist those working on
ORBit:

	interface Foo;
	interface Bar;

	interface Foo{
		void function_the_first(in Bar arg);
	};

	interface Bar{
		void function_the_second(in Foo arg);
	};

and here's what it gives me:

** ERROR **: file orbit-c-header.c: line 539 (cbe_hdr_do_forward_dcl): "!"Not yet implemented""

Happy coding; I'm going to bed.

--
Todd Graham Lewis                                     (800) 719-4664, x2804
******Linux******         MindSpring Enterprises      tlewis@mindspring.net