[omniORB] DynAny and typecode aliases

[Stephen Crawley]

Duncan,

I've submitted a CORBA revision issue related to this problem.

For what it is worth, the consensus so far on the <org_revision@omg.org>
mailing list is that:

  1)  The intent of the CORBA 2.2 & 2.3 DynAny specification is that 
      aliases should not be stripped.

  2)  Even if the text doesn't say this, there are sufficient hints
      in the spec to imply this.

  3)  A DynAny implementation that does strip out aliases is broken.

  4)  However, the spec does need clarifying to spell this out.

I understand that this work will be done in the current RTF cycle.

-- Steve