Re: draft-ietf-ldapext-locate-04.txt

["RL 'Bob' Morgan" <rlmorgan@washington.edu>]

> I have a question on Mapping Algorithm in Section 3.  The second paragraph 
> of this section has this sentence:
> 
> "An RDN is able to be converted if it (1) consists of a single 
> AttributeTypeAndValue".
> 
> Is this restriction necessary.  If an RDN is multi-valued, why should the 
> mapping break?  Shouldn't things work out OK if the RDN is something like:
> 
> "dc = foo + ou = bar"?

Well, we *could* permit this, of course.  This would require deciding
whether "dc=foo + dc=bar" is OK, and how to deal with ordering (since an
RDN is a SET of AVAs, which means it's unordered, right?).  So, more work,
more potential confusion, presumably little constituency for actually
using multi-AVA RDNs, especially multi-AVA RDNs one of whose components is
DC=.  Easier to say "don't do that", IMHO.  Do you think this would
actually be useful in practice?

 - RL "Bob"