ADL takes first step to repository profile
With the publication of a report on "Emerging and Enabling Technologies for the Design of Learning Object Repositories", ADL is taking the first tentative steps to designing a learning object repository application profile to complement its existing learning object reference model, SCORM.
One of the big promises of learning objects are their ability to be re-used outside the context they where first designed for. A lot of debate has therefore focussed on things like package formats and the ideal level of granularity or generality of a learning object. But none of this will matter a lot when there are no agreed means of finding, gathering, retrieving and storing learning objects. Hence the general interest in repositories that will do all of these things. That ADL is surveying standards and specifications in this area is significant beyond what they found; it indicates that the technology is both reaching maturity and in pretty urgent need to be standardised.
As Paul Jesukiewicz, Director of the ADL Co-Laboratory in Alexandria, notes, "We are in about the same place today with repositories as we were three years ago with LMSs. We look forward to continuing the investigation into repositories and hope to continue as a catalyst in bringing the various communities together."
Going by ADL's past practice with SCORM, that would mean that after the report that has just been published, a process of selection between the current technologies will take place. One or two of the current repository specs will then be anointed as the basis of an ADL application profile-- an agreed interpretation and implementation of the chosen wider specs. ADL would be hoping that the resulting repository profile will prove to be as popular as SCORM, thereby facilitating the much hoped for wide learning object economy.
As always with any technology or its sponsors, one size may not fit all (very well). That is, even if and when ADL gets to publish any repository application profile, it is most likely to focus on specific SCORM needs rather than any other type of learning object. For example, the metadata handled by the profile could be limited to those bits of metadata that are compulsory in SCORM now. This would make sense, and doesn't preclude vendors from supporting all sorts of other fun repository stuff alongside ADL's profile. It also means that such a profile is unlikely to be some sort of silver bullet.
Before any of that happens, however, ADL needs to be satisfied that the existing specs are sufficiently mature. No use making a profile of a spec that fails to take off. As such, the current report does not commit to any existing spec. Instead, it concludes with a longish list of potential pitfals and holes. Some of the uncertainties about enabling technologies have already been partially resolved (XML schema, for example, is a full W3C recommendation, SOAP a candidate), but other risks like undeclared patents popping up in supposedly open standards are quite real. Another very pertinent issue is the fact that there are several methods for providing unique identifiers for learning objects (e.g., GUID, UUID, DOI, URN, URI).
A very detailed overview of the current learning object respository specs is available in the Emerging and Enabling Technologies for the Design of Learning Object Repositories Report (907 Kb, MS Word doc) from the ADL website.