[omniORB] CosNaming.idl

[Olaf Meding]

> ... original one and a backwards-compatible extended one ...

Which one is omniORB using?

Couldn't find these two CosNaming.idl files on the OMG web site.  Would
someone be able to point me to an authoritive source?

Looked at versions of this file from VisiBroker for Windows, VisiBroker for
Tornado, omniORB, OOC ORBacus, mico.  The all look different.  Most have
some comments about special IDL compiler options at the top of the file.
Some use pragma's, some use #if's.  So yes, maybe the content that counts
might be same but that is hard to tell for me w/o going thru line by line.

Seems to me that Vertel is making the customer going thru a lot of pain just
to be able to connect to a non-Vertel naming service.


Anyway, thanks much for your help.

Olaf

-----Original Message-----
From: dpg1@uk.research.att.com [mailto:dpg1@uk.research.att.com]On
Behalf Of Duncan Grisby
Sent: Friday, February 01, 2002 3:41 AM
To: Olaf Meding
Cc: omniORB Mailing List
Subject: Re: [omniORB] CosNaming.idl


On Thursday 31 January, "Olaf Meding" wrote:

> One of our ORB vendors (Vertel eORB) claims that we need to IDL compile
the
> CosNaming.idl file of whatever naming service we use (OpenORB in our
case).
> I do not understand why I would or should have to do this.  Aren't the
> naming service functions specified by the CORBA standard?  I do not
compile
> said  file with omniORB, yet have no problem connecting to the OpenORB
> naming service.
>
> Even if I did have to compile a CosNaming.idl file, shouldn't that file be
> supplied by OMG rather then by each ORB (naming service) vendor?

Yes, the CosNaming.idl file used by all ORBs should be the same as the
one distributed by the OMG. There are actually two different versions
of it, the original one and a backwards-compatible extended one.

In omniORB, we choose to include the stubs for the naming service in
the ORB library, since it is such a commonly used service, but there
is no requirement in the specification to do that. It is therefore
valid for Vertel to claim that you must compile CosNaming.idl for
yourself, but it is not true that it differs from ORB to ORB.

Cheers,

Duncan.

--
 -- Duncan Grisby  \  Research Engineer  --
  -- AT&T Laboratories Cambridge          --
   -- http://www.uk.research.att.com/~dpg1 --