DSpace UG Meeting July 2005 Summary

The DSpace Federation 2nd User Group meeting was held in Cambridge, UK on 2005-07-07 and 2005-07-08. The programme and presentations are available at DSpace@Cambridge at https://www.dspace.cam.ac.uk/handle/1810/206459.

Page seeded by obert Tansley - please add/amend.

reakout groups

The outcomes of the breakout groups, listed below, are a mix of discussion topics, identified requirements and priorities and recommended courses of action.

Digital Preservation

• Need to articulate DSpace's current capabilities and develop/document a set of best practices relating to digital preservation
• May be best to move as much of the digital preservation process as possible to IT departments – backup, replication etc.

  Extending metadata support

• Clearly need more than DC
• We need better ways to describe bitstreams themselves
• METS appears to be short-term resolution

  Access control

• Improving the UI seems biggest requirement
• Other potential areas of improvement include timed policies, locking out metadata/part of metadata, allowing submitters to adjust access
• Various parts of the system not using access control uniformly (e.g. OAI-PMH allows all metadata to be exposed)
• Need to establish a relationship between policies and licenses
• Shibboleth integration (apparently being done by a group in Australia)

  Network interfaces

• Comments of various approaches, exposure of object model via WebDAV etc
• General consensus to keep it simple, release early to allow people to experiment

  Statistics and logs

• Three types of statistic identified:
• usage/downloads – for advocacy etc.
• archival – e.g. number of objects, storage quotas
• administrative – e.g. number of users, how long workflow steps are taking
• Two main approaches – live vs. periodic generation
• Periodic seems fragile in the long term – formats change, logs become another object you have to manage/preserve
• Log tables in D may be easier to manage, provide live feedback
• live generation useful especially with respect to usage logs; could even be OAI-PMH harvested

  UI and publishing frameworks

• Top current user requirements seem to be:
• searching by document type (image, thesis etc)
• arranging by topic
• arranging for a particular audience (e.g. undergrads vs postgrads vs professors)
• Need to be able to tag refereed material
• Migration and conversion on ingest – help educate users about the best formats
• Usable rights management
• Search and retrieve across multiple systems
• Discussion of customisable display – configuration, i18n, themes

  ich media and complex object support

• small group – 3 members, only one with DSpace experience
  (mention of framework 6 project – missed the name)
• equirements discussed:
• IP issues
• many want to destroy physical originals and preserve the digital – physical are expensive to preserve
• Need interfaces with other systems
• Need to scale DSpace up in this area
• Need to support the various metadata standards in this field, e.g. SMPTE, MPEG7
• Need surrogates, viewing aids
• There are opportunities below the bitstream level – e.g. segments of video, keyframes
• DSpace UI seems more academic than public-facing

  Community discussion 2 - communication and processes

General consensus that great progress had been made in the move to open source in the last year. Much more development activity happening now than a year ago.
People from smaller institutions asked how they can become involved. Contributions do not need to be in terms of development – deployment experience, advocacy, testing are among other valuable contributions.
On the development front, if the architecture work was divided into clearly defined and specified steps, people could become involved more easily. The difficult bit is making the decisions required to divide the work this way. In this regard, the community is looking to the committers to provide some leadership in this area. People desire a visible road map, and maybe a public "to do" list. (oth of which already exist in basic form.)
DSpace.org site appears geared to academics, though DSpace has broader appeal. Perhaps it needs to be restructured?
Although there is the Wiki and the mailing lists, there still seem to be communication issues.

TODO (notes from MacKenzie Smith at MIT)

• close down the current set of Special Interest Group listservs (dspace-I, dspace-theses, dspace-publishing, dspace-LO, dspace-EM, dspace-datasets, dspace-preservation) but keep the dspace-general list and the two technical mailing lists on SourceForge (dspace-tech and dspace-devel)
• create specialized, topical mailing lists of specific technical topics for defined periods of time, as needed (for example, to continue OF discussions from the user group meeting)
• overhaul the current DSpace Federation website to represent all the stakeholders in the community
• add sections to the wiki for DSpace systems administrators, other specialized types of developers or users
• plan user group meetings for 2006 and beyond, with locations in different parts of the world so that DSpace developers/users can attend meetings closer to their home at least once in awhile (e.g. alternate between North American venues and European, South American, Asian, and other venues). Ideally there would be two user group meetings per year, but since they are sponsored and organized by voluteers that may not prove practical
• consider starting a DSpace newsletter to be mailed to subscribers or posted on the website with contributions from community members
• with the committer group, develop a detailed oad Map for development that the community can volunteer to help with
• begin work on a DSpace Service Framework that will point to a Service Oriented Architecture for future releases of the system

  Community discussion 3 - Governance

Proposal by Julie Walker for determining governance structure generally accepted/approved by those present
Short discussion on what it means to be a "member of the DSpace community" – should there be any "sign up" process? Suggestion for initial lightweight process (e.g. using Wiki)
equirements for governance identified during discussion:

• Enabling communication within DSpace community
• Ensure the community is healthy
• Conflict resolution - strategic rather than technical
• Ensuring sustainability
• Infrastructure (e.g. Wiki, mailing lists) currently a 'volunteer' effort – enable sharing this load
• External communication:
• Advocacy, "marketing"
• epresentation in other fora
• epresentation in standards bodies

• Wasn't "Target for funding / donations" mentioned as well? JimDowning
• A few photos (UserGroupMeetingPhotos2005) of the festivities, courtesy of Peter Walgemoed from Carelliance in the Netherlands