Re: ORbit2 naming server modifies IORs

[Michael Meeks <michael ximian com>]

Hi Ansgar,

On Thu, 2004-05-13 at 11:38 +0200, Ansgar Radermacher wrote:
> I used the ORBit2 (version 2.10.1) name server within a small benchmark 
> (with multiple ORBs). Strangely, the execution time of omniORB for the 
> transfer of large character sequences differed considerably depending on 
> whether I used the orbit name server or the name server of another ORB.
> I found out, that the orbit name server (started with the option 
> -ORBIIOPIPv4=1) modifies the code-set related part of the object 
> reference. This probably causes omniORB to perform code translation. Is 
> this a bug or am I doing something wrong?

	Oh - well, I can quite believe it. We deliberately discard code-sets,
it's true - since I have no deep desire to get into the knotty problems
of code-set translation; and the world should all use UTF-8 ;->

	I guess (given that we are not going to support any code-sets except
UTF-8 any time soon), it's prolly a lesser evil to push UTF-8 where a
different code-set is expected, and not mangle their IOR;

	Fix much appreciated to:

	ORBit2/src/orb/orb-core/iop-profiles.c
	 (IOP_TAG_CODE_SETS_marshal, / _demarshal):

	HTH,

		Michael.

-- 
 michael ximian com  <><, Pseudo Engineer, itinerant idiot