Out of the MARC frying pan
I’ve been lurking on the RDA mailing list for the last week, and one thing stands out: some people in the cataloging community are having a hard time letting go of MARC. MARC is a syntax that they have spent years using in puzzling out the nuances of describing catalog items, and the thought of moving to a much more malleable and parsable representation presents something of a “threat or menace” choice to them.
From my point of view as a complete outsider to the library community, it’s clear that evangelists like Diane Hillman and Karen Coyle have a long way to go to make this type of person comfortable. While there appears in some circles to be excitement in the announcement of the teaming-up of the RDA and DCMI communities to push the standard in the direction of greater relevance to a broader community, I worry that this well-intended effort might fall victim to the technology adoption problems that befell the RDF standard over the last decade. That is, that premature design commitments that fail to take hard-won implementational experience into account provide those resisting change with an excuse to reject new technology.
Case in point: some of the initial work on this synthesis leverages the DCMI Abstract Model, which makes explicit use of blank nodes to represent aggregates of related information, e.g., statements of responsibility. The traditional, largely academic RDF community is quite comfortable with this method of representation, but virtually everyone (including my company) who has been involved in the implementation of deployed commercial systems using RDF thinks blank nodes are a bad idea. By assuming blank nodes as part of the standard, the RDA/DC/RDF synthesis may be painting itself into a corner that it won’t be able to get out of. It’s hard to tell from my distant perspective whether this is a real problem, as the multiple layer of draft documents about RDA and FRBR are all over the map, but to me it is cause for concern.