Re: smilint remembers named number in OID value assignment (fwd)

["David T. Perkins" <dperkins@dsperkins.com>]

HI,

Mike, your intepretation is correct. That is, any label used
on a "named number" subidentifier in an OID value is local.

Doing it differently (assigning semantics) has been done incorrectly 
by many MIB compilers over the years. However, they are incorrect.
You quoted the ASN.1 spec, and the meaning has not been modified
in anyway by SNMP's SMIv1 or SMIv2.

On Mon, 29 Nov 2004, C. M. Heard wrote:
> Greetings,
> 
> Learned comments on the following would be appreciated.
> 
> Thanks,
> 
> Mike Heard
> 
> ---------- Forwarded message ----------
> Date: Mon, 29 Nov 2004 08:56:11 +0100
> From: Juergen Schoenwaelder <j.schoenwaelder@iu-bremen.de>
> To: C. M. Heard <heard@pobox.com>
> Cc: Frank Strauss <strauss@ibr.cs.tu-bs.de>
> Subject: Re: smilint remembers named number in OID value assignment
> 
> On Sat, Nov 27, 2004 at 09:22:03AM -0800, C. M. Heard wrote:
>  
> > In parsing the attached MIB module smilint appears treat the
> > following expressions as defining the symbol snmpDot3MauMgt:
> > 
> >      ianaMauTC MODULE-IDENTITY  
> >        [ ... ]
> >        ::= { mib-2 snmpDot3MauMgt(26) 7 } -- mauMod+1
> > 
> >                 .
> >                 .
> >                 .
> > 
> >      dot3MauType OBJECT IDENTIFIER ::= { mib-2 snmpDot3MauMgt(26) 4 }
> > 
> > and as a result complains about the second one (see attached e-mail
> > robot output).
> > 
> > As I understand it, the names associated with these "named number"
> > constructions are supposed to be local to the OID assignment and not
> > known outside that scope -- at least that's how it works in ASN.1,
> > from which these constructs were inherited.  In effect, the name
> > attached to the number in these constructs is nothing more than a
> > comment.
> 
> Frankly, I do not really know what the rule is here. If the name is
> truely local to the assignment, then it has no purpose at all since
> you hardly can reference the name. Obviously, libsmi does not treat
> this as purely local. I guess the best thing would be to raise this
> question on the mibs mailing list to see whether there is concensus
> how this is actually supposed to work.
> 
> Regarding the MIB that triggered your email: I think this MIB module
> should be registered directly below mib-2 anyway.
>  
> /js
> 
> -- 
> Juergen Schoenwaelder		    International University Bremen
> <http://www.eecs.iu-bremen.de/>	    P.O. Box 750 561, 28725 Bremen, Germany
> 
> 
>