[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
Re: objectclass attribute
At 04:17 PM 3/22/00 +0100, Mikael Grehn wrote:
>Hi again!
>When I test my hacked OpenLDAP server and the backend linked to a
>Btrieve database I use several Ldap clients, for example
>'ldapsearch.exe' and swix.exe clients.
>I might even try it with our friend Ströders client later on. :-)
>In my response to the client I have a linked list of attributes (among
>other things) starting with the objectclass attributes.
>Question 1:
> I need to have USERSECURITYINFORMATION as an attribute (customer
>requires an unique personal code
> for each of his employers in the database). I first added
>objectclass=top and objectclass=newpilotperson
> of each person record returned to LDAP client, then I add attributes
>cn=<person name>,mail=<person email>
> and USERSECURITYINFORMATION=<personal unique code>.
>
> The USERSECURITYINFORMATION attribute doesnt show on Client side and
>I guess that is
> becourse I havent included an objectclassattribute that contain this
>attribute (cn and mail are in
> newpilotperson attribute).
> Do I have to find an objectclass that contain a suitable attribute
>for this unique code(for example
> USERSECURITYINFORMATION) and include this in my linked list to be
>able to include this
> attribute later on (in the list) to the client?
In general, server generated entries should be consistent with
schema.
>Question 2:
> If I wish to add (in slapd.oc.conf) this USERSECURITYINFORMATION
>attribute directly in the
> objectclass newpilotperson 'struct', is this possible?...
Possible? yes Appropriate? no The schema items provided are
not yours to modify.
>and even
>more interesting; How can the client know
> my objectclass struct(after I change it/them for example)?
Use an auxiliary objectclass to augment provided schema....
such as uid (userid) and uidObject.
>Question 3:
> Does it matter if I start the linked list with the objectclass
>attributes or that 'actual' attributes themselves?
The returned list of attributes is unorderred sequence.