It was gratifying to see at the NSDL Registry stakeholders meeting yesterday afternoon, that our initial feeling that this was the RIGHT time for a metadata registry was confirmed.

It wasn’t the greatest time for a meeting–5:00 on the second (very long) day of the NSDL Annual Meeting–but the attendance was good and the discussion active. Stuart and I presented the plan for the project, showed a bit about what the use case and functional requirements contained (begging piteously for feedback), and asked some fundamental questions:

Question 1: What kind of services will registry maintainers require to continue to use the Registry?

As we’ve mentioned before–we think this is a critical question to answer at this design stage, not at the end of the project (where it might traditionally be asked). Some of the things we’ve thought of are in the nature of input and output mechanisms, which would allow a maintainer to use the Registry as the main place where their controlled vocabulary is managed, with output options to cataloging applications or other places where the vocabulary drives services. We know this means that we need to work closely with our stakeholder community as we design and prototype, and we’re commited to doing just that. We’re asking for volunteers first–arm twisting is the backup strategy!

Question 2: What additional value can the Registry provide for those who use registered vocabularies?

The corollary to the first question is what kinds of services do vocabulary USERS need? Obviously, those in search of a vocabulary to use for a specific purpose might be pursuaded to come search or browse the registry. What might pursuade them to consider one possibility over another? Would the knowledge that Project A was using a particular vocabulary be pursuasive? If so, how can we encourage projects to register use, as well as ownership or availability, as a way to encourage re-use and collaboration?

Once a vocabulary was chosen, a user might be offered the option of registering the fact that they were intending to use a particular vocabulary, and be offered as well some possibilities for updating or notification services around that vocabulary’s certain changes. Depending on what they might do or wish to do to update their cataloging or search applications and/or instance data, there might be a number of possibilities. They might want email notification of specific changes or additions to terms, or notifications of new files available for download or harvest. Some tool maintainers might want to schedule interactions with the registry for updates to multiple on a regular basis. We’re hoping that our survey of our stakeholders will fill in our understanding in this area.

Question 3: How can the Registry become an integral part of the work of NSDL Projects?

This is, of course, the big question–one that the stakeholders can’t answer but the Registry Project must, if its work is to become the important infrastructure component we think it should be. But the time is right for this project–the stakeholders didn’t need explanations of why registration of controlled vocabularies was an important idea, and it was clear that they were eager to work with us to make it happen. What could be better?

Watch this space as we struggle with these questions (and certainly, if you have ideas, tell us about them!)