[omniORB] Using #include in IDL files

[Nathan Banks]

>

I have encountered this problem (scratch that) inconvenience as well. I have
threatened myself to resort to an 'intermediate' language to specify my structures and
enumerations. I would feed these specifications to a simple filter or compiler which
would emit both C++ and IDL equivalences including some translation logic to assist in
the movement of data from C++ types to
IDL generated types and vice versus. I am not talking about anything spectacular here
- just a simple perl or python script. Anything would be acceptable just to eliminate
the redundancy and minimize writing of the 'boring'  logic which copies my legacy
class attributes to the IDL interface class attributes and back.

While on the subject: To whom do I voice my request that IDL enumeration members
should have an optional author specified value?  When I attempt to model in IDL my
legacy telecom backends' use of value specified enumerations, I get frustrated because
IDL will not allow me to specify a value for an enumeration.  I suppose a major intent
of IDL is 'brevity of expresssion' and I should 'hold the party line' and keep on
writing my 'boring' legacy enumeration to IDL enumeration translation logic.

Here's to CORBA and all they who build and sail her mighty ships on the Information
Ocean!


Cheers,
Nathan Banks
Telecommunications Engineer & CORBA Enthusiast