[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: objectIdentifierMatch on ambiguous name



At 11:52 AM 2/12/2003, Hallvard B Furuseth wrote:
>Kurt D. Zeilenga writes:
>>At 05:29 PM 2/11/2003, I wrote:
>>>        Implementations SHOULD avoid the use of registered
>>>        short names (descriptors) or, when not available,
>>>        an object identifiers in numeric form, to avoid
>>>        ambiguity.
>
>I prefer my own text.

I think I was addressing a different issue.  I think
something more is needed.  Likely [Syntaxes, 4.3.26] needs
to say something, as well as [Models].

[Syntaxes] likely should say that the <descr> form may
only be used in contexts which specifically detail which
short names are applicable.  As this is a common issue with
the use of the <oid> production, I think it might actually be
appropriate to say something about it in the [Models, 1.3].

>Yours doesn't imply that
>  attribute type ( ... NAME 'oid'
>                   EQUALITY objectIdentifierMatch
>                   SYNTAX 1.3.6.1.4.1.1466.115.121.1.38 ) ; OID
>from the start of this thread may not contain short names,
>like you and others have said in previous articles.

Well, actually, if 'oid' was an operational attribute it
could be defined such that short names could be used.  For
userApplications attributes (excepting objectClass), short names
cannot be used as the server has no knowledge of the user
application semantics.

(I'm going to review past posts before offerring any
additional suggestions.)

Kurt