ORBit IOR not understood by other CORBA

[Schultz Peter <Peter Schultz hvr1 siemens de>]

Hello,

We use ORBit 0.5.6 under HP-UX, C language mapping and try to communicate
with a CORBA 2.3 from Sun Java 1.3, running under Windows NT 4.0. The IOR is
written as string to a file.
Server and client running on HP-UX: ok.
Server and client running on NT: ok.
Server running on NT, client on HP-UX: ok.
Server running on HP-UX, client on NT: "invalid IOR".

At XEROX ILU I found an "ILU IOR Parser"
(http://www.parc.xerox.com/istl/projects/ILU/parseIOR/), which tells me
"Error parsing IOR (status 1) ... error: No object for IOR" when working
with an ORBit IOR-string.
The ORBit program "ior-decode" (HP-UX) does understand the IOR string, but
it tells: "Profile type: TAG_ORBIT_SPECIFIC", in contrast to "Profile type:
TAG_INTERNET_IOP" when decoding the Java IOR-string.

Is this the problem? When the answer is "yes", how can I tell ORBit to use
"Profile type: TAG_INTERNET_IOP" ?

Examples:
ORBit:
IOR:0003aa200000001349444c3a43616c63756c61746f723a312e30000100000001badfaeca
000000480001001b000000202f746d702f6f726269742d70726f73756d2f6f72622d39333030
31343036330000000000000000180000000040a1aed32961a41c0000000158a25903074aee37
Java:
IOR:000000000000001749444c3a48656c6c6f4170702f48656c6c6f3a312e30000000000001
000000000000005800010100000000103132392e3130332e3139382e31343300049b00000000
0018afabcafe00000002d09fedfc000000080000000000000000000000010000000100000014
0000000000010020000000000001010000000000

Thanks.