Well, the first public draft of the straw man data definitions document has been published (Version 1.6 of Data Definitions). It's rather longer than I thought it would be and will probably get longer, because there are some bits currently left out.
Thank you to all members of the working group who have replied (so far). I hope it was a good idea to have this semi-formal arrangement, rather than just releasing the documents into the wild. The concept is to permit some in-depth discussion within a smaller group than the CourseDataStage1 mailing list, thus hopefully saving time for those who aren't particularly interested in a lot of detail on the data definitions and vocabularies topics. It seems to be working so far.
Current issues include:
  • How to handle special characters; UTF-8 encoding and character entities the suggested solution
  • Images; do we want them? If so, what sizes and formats?
  • Cost; is structure required for this element?
  • Subjects and categories; a thorny problem, probably best addressed through the vocabulary work. My own preference would be to recommend that everyone uses the learndirect classification system (LDCS), then we'd get consistency - and we could map other things to it (JACS, SSA, sector specific lists) for those that don't want to use LDCS directly. However, others may have alternative views.
  • Namespaces; I'd resisted mentioning this issue initially, but Craig's quite right to say that guidance would be a good idea, so we've made a start. How much more detail will we need? Time will tell.
  • Reliable feed location stuff; again, I'm uncertain whether this really about data definitions (IMHO it's certainly not), but we have a doc that guidance about this issue could sit in.
  • Identifiers and URIs; important discussion here about whether we really need to use URIs in XCRI-CAP identifiers - we have separate URL elements in any case, so maybe the requirement in the standard to use URIs should be lifted? More collective thought needed I feel.
  • Target audience; I think this was perhaps an omission from XCRI-CAP 1.2. It's also being discussed in relation to the Learning Resources Metadata Initiative (LRMI), a much wider international initiative - see http://wiki.creativecommons.org/LRMI/Properties/Version_0.7 if you're interested.
  • Multiple providers; some worms in that one too.
I think we have a good start. Things never stand still though - I have the Vocab Framework doc to finish off, and I'd very much like to publish a documented XCRI-CAP 1.2 doc - currently in draft and subject to slight modification to meet the data definitions.