April 4 - 6, 2001
Voorburg, The Netherlands
(Recorder: Wendy Treadwell)
April 4, 2001
Present: William Block, Ann Green, Emiel Kaper, Jean-Pierre Kent, Jostein Ryssevik, Wendy Treadwell
Meeting Goal: Working compromise for expansion of DDI DTD to cover aggregate data
All agreed that we must have a functional compromise ready for approval at the June meeting of the DDI. All members are committed to leaving this three day meeting with an agreement on the approach and basic element expansion set. The agreement will be presented at the informal meeting of DDI group at the IASSIST Conference in May for additional comment and questions.
Both FASTER and Minnesota are committed to building systems on the DDI and recognize the need to solve the problem of dealing with aggregate data in a manner that meets the needs of both groups.
A working compromise was defined to be an agreement which met the list of "Criteria for acceptable model" (see attached).
Clarification of terms in the criteria list took place. It was agreed that semantics were playing a major role in both understanding the problems and agreeing on the solutions. In particular, the concepts of VAR, variable matrix, cube, vector and matrix dimension were expanded upon in order to identify the specific areas of misunderstanding and/or disagreement.
It was agreed that the second and third days of the meeting would be limited to William Block, Emiel Kaper and Wendy Treadwell who would work out the design details.
April 5, 2001
Present: William Block, Emiel Kaper and Wendy Treadwell
Discussion of concepts continued and a general approach for combining the two approaches which met the needs of both groups as defined by the criteria was accomplished. It was agreed that a clear separation between description of the logical structure of the aggregate data matrix/cube and the physical storage structure(s) needed to be maintained. The remainder of the session was spent on detailing the additional elements and attributes required to describe the logical structure of the data (section 4.0).
April 6, 2001
Present: William Block, Emiel Kaper and Wendy Treadwell
Discussion of the various form of physical structures for storing or creating aggregate data took place. Specific unique characteristics that required description or specific infuriation required to manipulate these structures was identified.
The general model for additional elements and attributes required to describe the various physical structures of aggregate data was completed. Wendy Treadwell will take these notes and complete the recommended revisions to the schema for presentation to the DDI in May.