Core User Directory Project Working Party
Minutes of the Core Directory Project Working Party - Meeting 9
There was a meeting at 3 pm on Friday, 27 Feb, in the OeRC Conference Room
Present: Mike Fraser (chair), Paul Jeffreys, Lou Burnard, Tony Brett, Jonathan Ward, Ilana Veitch, Beth Crutch, Ann Gavin, Neil Jefferies, Maureen MacNaboe, Jim Davies, Alana Davies, Rob Hebron, and Miranda Llewellyn (minutes).
Apologies: Lynn Waddington, Emma Potts, Chris Cattermole, Tom Payne, Tim Shaw, Anne Bowtel, Tom Jones, Mirjam Siderius, Mike Heaney, Dave Price (Neil Jefferies attending for Dave Price).
Minutes of the last meeting (27 January) and matters arising
The minutes were confirmed as an accurate record. The majority of the actions had been completed but one was carried forward: Lou Burnard confirmed that he would follow up with Alana Davies (this happened after the meeting). Neil Jefferies had been invited to the Pilot meeting (but could not make it) and was kept up to date on developments. Mike requested that reports of Rob Hebron's work be added to the public website.
Issues arising from OUCS internal meeting of 13 February 2009
The group was concerned about the absence of agreed University policy regarding data management. Decisions about how and on what criteria new attributes should be included in (or removed from) an operational CUD needed to be made in the future. Mike commented that he would like to see examples of how this might be addressed as he felt it lay outside of the project's remit. Beth Crutch asked whether the University would be constructing an Information Strategy (similar to the ICT Strategy). Paul Jeffreys confirmed that he had suggested this in the past but it had not been agreed. Roger Boning is currently looking into it.
Lou reported on the secondary affiliation problem: when a new
status is added sometimes (but not always) it replaces the
previous status. The group discussed the authoritative sources
which the CUD would rely on. Mike confirmed that the group had
previously agreed to include metadata for every piece of data in
the CUD.
Action: Lou to provide a draft list of
attributes.
Lou confirmed that the interface shown at the last meeting was merely for demonstration purposes and did not represent the final interface. There will also be an LDAP interface.
- the data attribute contains information about a person who is associated with the University
- there is a use case in which more than one data consumer requires access to the same data attribute
- there is a use case in which some central University system makes available the attribute, having obtained it from some other data source
The group discussed whether these criteria were the only ones: for example, some attributes were included solely because they made reconciliation of data easier. Neil Jefferies noted that meeting the criteria did not necessarily mean the data would be included in the CUD.
The group discussed the need for documentation of the meaning of the data attributes. Should this be provided for all potential users, or should it simply reflect the meaning of the attributes for the data provider? It was suggested that it was best to keep things simple: rather than repeat definitions which already existed, a link could be provided.
The group discussed waifs and strays. Mike confirmed that the group had decided previously that the CUD would be the authoritative source for the identifier; waifs and strays would need a separate service.
Report from Rob Hebron, to include demonstration
Rob Hebron reported that every component of the Pilot was now in place and some outgoing interfaces were complete. His next steps were to include the remaining data which is within scope, to review and enhance where required, and to provide documentation. Mike thanked Rob for the presentation and the work completed.