This Regional Design Workshop took place at Duke University, in Durham, NC, on December 15 – 16, 2008.

Photographs from this workshop are located at http://www.flickr.com/photos/oleproject/sets/72157611288014373/ with the tags oleproject and durham and duke.

Materials referenced in these notes are linked directly in the notes, and/or available on the Resources from Duke page linked here.

Day 1

Jean Ferguson welcomed the 58 people who attended this workshop.

Libraries represented included:

• Auburn University
• Campbell College
• College Center for Library Automation (CCLA)
• Duke University
• Elon University
• Emory University
• Florida Center for Library Automation (FCLA)
• Forsyth County Public Library
• Georgia Tech
• George Washington
• Johnson C. Smith University
• Meredith College
• Mississippi State University
• North Carolina State University
• North Carolina Central
• University of Alabama
• University of Central Florida
• University of Georgia Libraries
• University of Maryland Libraries
• University of North Carolina at Chapel Hill
• University of Virginia
• Wake Forest University
• Winston Salem State University

Library departments or roles represented included:

• Acquisitions and Collection Development
• Cataloging
• Circulation, Reserves, and Interlibrary Loan
• Electronic Resources Management
• Serials
• Systems

ILS’s in use at these libraries are

• Aleph from Ex Libris
• SirsiDynix
• Innovative Interfaces, Inc.
• Voyager from Ex Libris, formerly from Endeavor

Lynne O’Brien then gave a presentation about the overall goals, activities and timeline of the OLE Project. Questions after the presentation included:

• What will be the time line for the build project?
• Governance Issues.  What would be the business plan for the build project?  How to you ensure a balance between a monolithic design and rapid testing.  Project must be built in a modular way so it can be tested as it is built, rapid proto-typing.  Design must be tested as it is planned.
• How will the project work with other campus IT systems?  Blackboard etc.
• Who would pay for the build?  Would libraries contribute developers? How would it be paid for?  Who would pay for the development?  Would it come through a grant or libraries or some mix?

For the rest of the morning, we identified key workflows and business processes in our libraries. We discussed the current labels we apply to our activities (such as Cataloging, Circulation) and tried to organize our workflows into more generic headings. One challenge was distinguishing between activities that are the core business practices of the library, versus processes which support those core business practices. For example, providing access to our materials is a core business function of the library, while running internal reports is a process in support of our work, not a core library business function. Similarly, authentication supports library workflows, but is not a core business within the library. Gary Johnson, Assistant Director for Library Software Operations at the College Center for Library Automation (CCLA) developed an excellent summary of the revised categories and key activities within those categories are shown in the linked document http://www.archive.org/details/LibraryProcesses

We identified a number of issues or trends that the OLE Project needs to consider as it continues with its work. For example:

• Determine how much of ILS is federated
• Broadening data types sources to include GIS
• Watch out just becoming bibliographic centers
• Increase in E Materials
• Importance of user authentication
• Relationship between items in different systems
• Variety of discovery places
• Ability to recombine things
• Need better access to data at all stages
• Need more connections between data sources
• Need to push our data more places and the ability to do so
• Ability to comply with complex licenses
• Increase in serialized items
• Need more ability for inter-operability
• Relationships between items in different systems
• Multiple user starting points.
• Need better access to all
• Ensuring that our materials are findable wherever the user starts from
• Being able to comply with complexity of licencing with different materials
• Increase in serialized items and decrease of items purchased outright
• New pressures on preservation
• Growing demand for individual records
• Broadening data types/sources

Bob Trotter, IT Manager for the University of Georgia Library, gave a 4/6 presentation (an idea we borrowed from the Bamboo Project (http://projectbamboo.uchicago.edu/what-foursix). He gave a 4 minute talk on the various consortia to which libraries in Georgia belong, and the issues that arise as a result of libraries being part of multiple consortia. The audience then had six minutes to ask questions.

Questions focused on

• A library – or a library user, can have multiple roles
• Questions regarding multiple authorities and licensing issues
• Creating data
• Transactions

John Little gave a brief introduction to Service Oriented Architecture (SOA) and Business Process Modeling (BPM).  We then broke into small groups and practiced modeling the process of “checking out a book.” Our large group review of the small groups’ work showed us that we can find general elements that we all agree on. The main differences in how we mapped this process had to do with the level of detail deriving from our individual libraries’ policies and procedures.

Questions raised included:

• What level of webservices are you going to define in the process?
• Talk between services
• The Build group will have to answer that question but it must be as Universal as possible
• SOA is a model – style should not be dependent of the type of interface
• SOAP is one way to implement SOA.  In an application based interface, have you chosen this protocol or any other?
• That will also be part of the build process.

We wrapped up the day with an open discussion or what worked well, and what could have been improved about the day’s workshop.

Evening:

Duke’s University Librarian, Deborah Jakubs, hosted a dinner for directors from nearby libraries. Libraries represented were: Duke, Belmont Abbey College, Meredith College, Elon University, Johnson C. Smith University, Winston-Salem State University, UNC Chapel Hill, NCWC, Triangle Research Libraries Network, and the University of VA. The purpose of the dinner was to introduce the OLE Project, encourage participation in the project and get feedback about the project’s goals and activities from directors of a variety of library types: large and small, public and private, using a variety of ILS’s and a variety of support models.

Questions included:

• When will the OLE Project have software ready for libraries to use?
• How can small libraries, which do not have programmers and whose staff members have multiple roles within their libraries, plan for the use of open-source library automation systems?
• What reactions have we had from vendors?
• How does this project relate to work being done by OCLC?
• Are we in touch with other library projects, such as the eXtensible Catalog project?

Day 2

Two of the individuals attending this workshop already had done business process modeling for Technical Services workflows at Emory University, so we started the day with brief presentations by Lauren Corbett from Wake Forest University (formerly from Emory University) and by Lola Halpin from Emory University.

Next, we applied what we learned yesterday about BPM by breaking into small groups, with each group mapping two specific processes identified in yesterday’s work.

Will Sexton, Metadata Analyst/Programmer from Duke’s Perkins Library gave a 4/6 presentation about the Duke Library’s recent work in defining needs and hiring programmers to develop enhanced metadata tools to support our digitization programs.

Questions for Will included:

• Why did we not use Open Archive’s tool?
• Will the tool developed at Duke be available as open source?
• What metadata schema(s) is Duke Supporting?
• Will it be available as OS?
• Define scope of what the developers will do?  We will approach through SOA methodology.
• Developer has been using Flash to build wizards???
• All agreed that a major issue with Cataloging is the integration of authorities.
• How will it be displayed? The idea is not to intergrate metadata with a specific front end.

After lunch, John Little reported back from a lunch discussion with workshop attendees who focus on library systems. Some of ideas that group discussed were:

• Need for well developed API and/or the extensive ability to export, import, and move data between services.  The need to interoperate with other enterprise systems is key
• The ability to have OPT-IN or OPT-OUT sharing of books lists and circ data profiles
• The ability to default to generalized book circulation history
• The ability to display connections between objects (titles, etc.).  E.g. this book circulated with this other book.
• Are our users ready for an environment to opt in or opt out for data collection which would facilitate developing such connections and recommendations?
• Difficulty in sharing user data in a mobile society.  How does the data get shared, by user, institution, etc .
• The ability to ingest and export data using various data models
• Identify Management is important although there was varying discussion as to whether this would be part of the SOA services or part of another enterprise units interoperable services
• Object and management and holdings management are very important
• Non disruptive upgrades are very important
• Federating – managing resources in more than just OLE.
• Need to connect metadata in a variety of formats.
• (These topics are from the notes <http://www.flickr.com/photos/oleproject/3115310171/>)

We concluded Day 2 by talking about “blue sky” ideas for the ILS.

Here are some of the ideas small groups brought back to the full group:

• Only store locally what we need to store locally; instead, link out to reliable sources.
• Good interoperability with external systems. (financial, vendor, patron databases, etc.)
• Improve staff interfaces. Consider “far out” things like using voice activation, eye movement tracking to speed data entry and modification.
• More automated processes.
• Library owned resources easily discoverable from non-Library search tools (e.g., Google, Yahoo).
• Modular architecture. Able to change on the fly as work flows change.
• Self documenting for data changes – automatically records changes you make.
• Electronic Resource Management within the ILS.
• System would automatically keep track of when an e-version of a book becomes available, find it, create a record for it, and put it in the system.
• Expose all data to harvesters.
• Import data from all sorts of places, and do so automatically wherever possible.
• Handles licensing and rights information automatically.
• Facilitate management of links between records.
• Build in workflow for contracts.
• Have a built-in resolver.
• More web interfaces integrating staff and public views.
• Tools staff could leverage, reporting on data sets that cross boundaries, moduls
• Operating system independent (as opposed to the web based?)
• Granularity with permission?
• Want to easily know relationships, links between records and hierarchies.
• Want to break apart records into elemental units.
• A data base model that enforces data base rules.  Automatically.
• All information could be accessed in a multitude of ways in different moduls – ingests all data formats.
• ARL magic stats button(is that the Undo button?)
• Reserves – issues everywhere.
• Intelligent subject search – more in line with natural language
• While it should be easy for users to get what is available also want to brand our resources – (libraries need to do some marketing)
• More robust API access to all information (not just bibliographic, but holdings, patron data, system functions,)
• Audit trails for all record types
• Better ERM functionality within ILS
• Expose to our users to rights they do and don’t have.
• Spell Check
• Distinguish between access and ownership
• Tracking rights for a variety of different formats.  (Digitize a video that we already own), investigate copyright clearance, Universal proxy controls.

Positive feedback from the meeting included:

• Process modeling was informative
• The lunch for IT people
• Ability to work with same functional group
• Blue sky
• Organization was good
• 4×6 presentations were very good

Suggestions for improvement or changes were:

• Blue sky would have been better if groups were told in advance that they should develop one major concept.
• Wanted functional areas to meet regarding Bluesky
• More time in breakout sessions
• Acoustics
• Duke Signage – hard to find Perkins Library
• Needed Coffee in the morning