I really enjoyed the post by Katie Howard // Metadata LS566 regarding "Dummy Proof Data." I could completely picture the office with papers stacked everywhere as she tried to impose order; I was amused that she "digs" metadata while using wells as her example; and I quite like the idea of the dummy-proof filing scheme. What I would like to know, though, is whether it was dummy-proof for the next person in her job. Seeing Katie's passion for organization, I'm going to guess that it was, but that, in my opinion, would be the ultimate dummy-proof filing system.
It's easy when assigning metadata to go to so much effort to make everything findable by the most inexperienced searcher that we create a system that is impossible to maintain. This can take many forms, including duplicating the data (and the work) in too many places that must be separately maintained, or spending so much time creating access points that we can't keep up with the work, or developing an arcane, undocumented workflow that can't be understood by others.
This tendency can be balanced by asking questions at the outset like "If we need to make a change, how can we make sure it happens in all places?" and "How far do we really need to extend the Dublin Core for this project?" and the old standard "If you win the lottery tomorrow, can someone else step in and do the work?"
So, bring on the dummy-proof discovery schemes, but let's not forget to dummy-proof it for ourselves as well!
Excellent!
ReplyDelete