Caveat Lector

(James Sanborn, Jim Tuttle)

Early planning: set up committee
NOT (at least at first): distributed committee structure, open submission, ‘IR’
INSTEAD: library-managed collections, building block for campus partnership, learning opportunity

Building blocks: technical-report print collection (as circulation declined, these started disappearing, nobody was taking care of them), special-collections resource centers, GIS data

Repository plan
- Target ‘research’ collections first: tech reports, ETDs, faculty publications/citations
- Each collection as own project
- Pursue common technological solutions insofar possible, but don’t shrink away from heterogeneous architecture

Tech reports
- Lightly-customized DSpace
- Library harvested: local metadata DB, scripted ingest-object creation, batch ingest (logical; this is the smart way to do it)
- Mix of ongoing submission by institute/departmental personnel and library capture

ETDs
- not live yet
- Hybrid between DSpace and ETD-DB
- ETD-DB for submission/approval/management
- Direct database extract for DSpace ingest-object creation
- Scheduled batch-ingest process
- DSpace considerations: metadata mapping, author browse (need to exclude contributor.advisor), some interface changes

Faculty publications
- Originally Access DB with ColdFusion UI (ugh!); bare citations, not all full-text
- Rebuilt into Oracle/PHP (to hang onto authority-control capability, do OpenURL, “vita-like” cite display, full-text or submission link)
- Full-text stored in DSpace

(Interesting setup here. They take on individual projects, make them work, and then move on to open it up to the rest of the university. Maybe that’s a better answer to the chicken-and-egg content problem?)