RE: Attribute Name Length Bounds

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

At 02:44 PM 6/13/2003, Chris Apple wrote:
>Are you saying that you believe that SHOULD (rather than MUST)
>is the appropriate strength of requirement to use so that
>implementers can limit their scope of applicability as they
>see fit?

No.   I am saying that, in my opinion, no RFC 2119 imperative
is warranted here.

Kurt


>Chris Apple - Principal Architect
>
>DSI Consulting, Inc.
>
>mailto:capple@dsi-consulting.net
>
>http://www.dsi-consulting.com
>
>-----Original Message-----
>From: owner-ietf-ldapbis@OpenLDAP.org
>[mailto:owner-ietf-ldapbis@OpenLDAP.org] On Behalf Of Kurt D. Zeilenga
>Sent: Friday, June 13, 2003 3:04 PM
>To: ietf-ldapbis@OpenLDAP.org
>Subject: Re: Attribute Name Length Bounds
>
>
>I have a few (personal) thoughts on this issue...
>
>>From a protocol point-of-view, interoperability can be said to
>be demonstrated if a presented overly-long attribute descriptor is
>treated as unrecognized type.  Likewise for attribute options and
>other such protocol tokens.  Additionally, no harm (to protocol
>peers, security, the Internet) is done by an implementation which
>simply treats overly long (by whatever definition) as unrecognized.
>Hence, I don't think an RFC 2119 imperative is not warranted here.
>
>This is not to say that an implementation which imposes significant
>limits on the length of attribute descriptors, options, and other
>such things has not limited its applicability.  It just to say its
>not a protocol interoperability issue.   Implementations should be
>free to limit their applicability.
>
>There are many issues that a technical specification detailing
>how to support "open-ended sets of attributes" likely would
>need to address.  I think these issues can be left to a future
>document.
>
>Kurt