Re: dITStructureRules

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

At 02:40 PM 2/24/2004, Steven Legg wrote:
>In the Adacel directory server we disallow ambiguous name forms in the
>schema. An update that attempts to add an ambiguous name form to a
>subschema subentry fails with a constraintViolation error.
>
>To the best of my knowledge, X.500 doesn't say what to do with ambiguous
>name forms. We chose to disallow them.

I don't recall any specific statement, but doesn't X.501
contain a general statement that implies DSA are to ensure
that the subschema is, at least, self consistent.  Anyways,
folks who think some sort of specific statement is needed in
[Models] should offer text for the WG to consider.

Regards, Kurt


>Regards,
>Steven
>
>Jim Sermersheim wrote:
>>It would be interesting to know what some X.500 vendors currently do. This seems to beg for some kind of 'ambiguousStructuralRule' error.
>> 
>>Jim
>> >>> "Steve McLain" <smclain@novell.com> 2/22/04 8:28:10 PM >>>
>>What is the desired/proper behavior if when adding a new entry, there are two or more possible dITStructureRules that could equally be the governingStructureRule?  For exampe:
>>NameForm1   mandatory cn, optional ou, namedObjectClass orgPerson NameForm2   mandatory cn, optional sn, namedObjectClass orgPerson
>> 
>>StructureRule5
>>     NameForm     nameForm1
>>     SuperiorRule   StructureRule4
>> 
>>StructureRule6
>>      NameForm     nameForm2
>>      SuperiorRule   StructureRule4
>> 
>>StructureRule7
>>       NameForm    nameform2
>>       SuperiorRule  Structurerule6
>> 
>>Attempting to add a orgPerson object with the rdn of "cn=ldapuser" to a parent object that has a governingStructureRule
>>of StructureRule4.
>> 
>>What dITStructureRule is set for the governingStructureRule or should this error?
>>I have looked through X.500 and the current model draft, I have not found the answer yet.
>> 
>>Steve McLain