Library standard insularity
The RDF wars continue; Google has placed a bet on microdata instead. If these catch on, RDF’s ongoing effort to tell the world that it is too a web standard really truly may well be scotched for good.
This prospect bothers the Loon not in the slightest; RDF bulldozed her beloved topic maps (which are easier to think about, easier to build, and vastly less hacky) for absolutely zero benefit as best she can tell. RDF has been cruising for a bruising as long as it’s existed.
Metadata and data-format standards go through cycles, something like this:
- We must standardize everything! EVERYTHING! And then the world will be perfect!
- Look at this standard! It is insanely complicated and no one understands it! We will simplify it, go back to basics! And then the world will be perfect!
- Look at this standard! It doesn’t cover special cases! We must make it extensible! And then the world will be perfect!
- Look at this standard! No two people use it the same way!
And ’round the cycle goes. With microdata, we seem somewhere between 2 and 3.
Accepting that this cycle is eternal and unstoppable, it strikes the Loon that libraries miss out on opportunities to inject library-created standards and data into the cycle. She saw the exception that proves the rule over on Threepress Consulting’s blog: using MARC relator codes for granular contributor types in ebook metadata.
The Loon thinks that perfectly delightful as well as useful. She wishes such cross-fertilization happened more often, and she fears it can’t as long as libraries remain inwardly focused, working only on their own bibliographic data, letting the outside world pass by.
After reading this post, I cannot stop think that, if we take library standards (MARC formats and cataloging codes) as a part of metadata standard, how can we interpret the development of these standards using your theory?
Interesting question! The best answer the Loon can muster is that libraries seem to have gotten themselves semi-permanently stuck somewhere between 4 and 1. No two libraries use MARC plus AACR2 in exactly the same way (for a constellation of reasons that most certainly include “stunningly bad tools”), and part of what RDA is trying (not notably successfully yet, but there is time) to do is (re-?)standardize quite a few bits of the implicit cataloging data model.
Admittedly, some of the motivation for RDA has more than a sprinkle of 2 in it as well…
The Loon wouldn’t call this a “theory,” by the way; it’s not nearly dignified enough. It’s only a jocularly-intended description of a pattern the Loon sees in the development of some web and library standards.