*MACE-Dir Conference Call*
April 29, 2002

*Participants*

Keith Hazelton -- Wisconsin (chair)
Rob Banz -- UMBC
Tom Barton -- Memphis
Brendan Bellina -- Notre Dame
Paul Hill -- MIT
Bob Morgan -- Washington
Steve Olshanksy -- Internet2
Todd Piket -- MTU
Nate Klingenstein -- Internet2 (scribe)

*Discussion*

Metadirectory Document Revisions

        Some time was spent going back over the Metadirectories document
to refine it further in preparation for the NMI.  The group briefly considered
whether DNs should be indexed, but Rob and Todd agreed that indexing
should be determined by applications and not by structural, schema issues.
The decision was to drop a recommendation on this after verifying with
Michael that this was acceptable.
        Brendan added a glossary which is intended to encompass only the
new terminology or other words such as "group" which have a specific meaning.
The group's general opinion is that if broader definitions are needed, the
reader should be directed towards another document where this has been
done.  This is similar for other parts of the document which may need
greater detail, e.g. pointing towards the Best Practices in Directory
Groups document for a deeper discussion.  Storing the terminology in a
centralized place also limits proliferation of text and divergence in
terminology between MACE-Dir documents.

Miscellany

        There was also a conversation about how to package and distribute
country names, and whether a country name attribute should be put in
eduPerson.  There were concerns that existing standards might not be in a given
object class, but nobody wanted to tackle a problem of this magnitude. One
reason that a short country name attribute didn't appeal is that it seemed
simpler to the group to make an attribute for country code rather than some
display name, allowing the interface to translate that to the local form of
country name that is appropriate for the user.
        The group wanted to follow the general philosophy of using existing
attributes which are correct and defining attributes that are needed, 
rather than redefining or augmenting existing attributes with new meanings
or values.  There is an existing location auxiliary object class in x.521,
but nobody in the group was familiar with it.