[open-bibliography] FRBR examples

Thu May 27 11:26:50 UTC 2010

Christopher Gutteridge wrote:
<snip>
Thanks, Karen. This pretty much rubs in that there is no "quick start" 
for this model, and the person creating records is expected to 
understand the details of the data model. This may work well with fully 
trained library staff, but in the author-deposits world I work in, 
that's impractical.

Part of the goal of open bibliographic data is that it should be 
reasonably easy for people to create and consume (and discover, but 
that's a separate discussion).

This model seems to describe the 'truth' better than some, but does it 
facilitate or hinder creation and consumption?
</snip>

I will point out again, that even if we make the effort to somehow create a coherent FRBR datamodel and manage to get most of our records into it, I still have not seen any advantage it will have over what we have now. I still do not see how it will add a single access point that we do not have presently, how it will help metadata creators (professional or amateur), how it will increase productivity, or anything at all. Karen's examples show the enormous difficulties we face to even retain the capabilities we currently have, even in the extremely simple examples she uses. I fear matters could become far, far more complex when applied in a comprehensive way to the real world.

Putting records into a more accessible XML format or RDF format would be a good first step, but it seems increasingly clear that FRBR is strictly an idealistic intellectual model that falls apart in the practice of the real world, while practical implementation appears more and more unwise.

Perhaps one way of beginning to approach the problem would be to clear the mind of all FRBR, RDA, AACR2, MARC and everything else, and ask: if there were no other formats or records of any sort, what would you do? This would be to approach the problem strictly as a user, which is I think the most important task, and focus on the needs of users. Only then can you begin to build tools that attempt to fulfill these needs, and then decide how best to incorporate the past records with the future records, and so on.

Jim Weinheimer