I was surprised by the responses I received from the initial posting in this thread on this list, so I drew back, reconsidered, and talked some things over with some experts. It turns out that in my initial post, I said nothing wrong: all I did was lay out some facts. I did later make an error about Lucene and ISO2709, but that doesn’t change the point I was trying to make and in fact, it makes formats even less important. Still, I appreciate this being pointed out because I am always learning.
That initial post about Bibframe and linked data was aimed at catalogers. Many catalogers have gone through huge changes in almost every way, practically all negative, over the last 15 or 20 years. Over and over, catalogers have read and been told that the “textual strings” they create are obsolete, along with their format. So, after everything they have gone through, there finally seems to be some hope: many catalogers now expect marvelous and wonderful advances from Bibframe. Such an expectation should not be so strange–it makes sense that if the obsolete format changes, then one of the major problems should be solved–and many hope that once the format is changed, libraries can then begin to create some wonderful things, based on better formats and of course, linked data.
My post (originally in Autocat) was to point out that these expectations are unrealistic. A developer who understands library data and MARCXML (and there are many of them) can do anything they want with those records now. RDF/Bibframe offers nothing that they couldn’t have done before. RDF/Bibframe is primarily for the non-library community, who do not want to delve into the MARCXML structures and–it is said–will be able to work with our data much more easily than the ways we offer today.
So, if libraries want to create something with their own data, they have already been able to do it for a long time. We can all see and use other websites created from bibliographic data that is worse than ours. RDF/Bibframe will not allow libraries any capabilities in addition to what they have already had for a long, long time, so people should not expect lots of new, wonderful changes just because of Bibframe.
Then I mentioned that if the underlying purpose is linked data, what is needed is not so much the formats, but the links. We have been able to add links for a long time, but it is really unclear what to link to. Links into id.loc.gov which contains the UF, BT, RT, NT, Notes and different cross-references for various types of names is not what gets people excited when they imagine linked data. VIAF also doesn’t seem to fit the bill. This is no criticism of those two projects, just a simple observation. So, to make either of those tools more useful for the public will demand a lot of thinking (I hope this will include at least some amount of market research among the public) along with a lot of development. That translates into lots more time and lots more money.
Finally, you can be “all dressed up but there’s nowhere to go.” In other words, there can be beautiful RDF; it can be on the web and available in ways incredibly easy to search and use, but there is no guarantee that anybody will want it.
In a later post I mentioned that data is not all equal: that bibliographic information should be seen as directional information, similar to traffic signs that lead people to the real information (content) they want. This goes almost without saying among catalogers since they work intimately with the real data that is contained in the books, serials, scores, and so on. I ventured that tools that may work rather well on content probably do not work so well on the directional information catalogers make and therefore, it is important to distinguish and understand the differences. I personally believe that both types of information are important, but I know that others have different opinions.
These observations are simple statements of fact, it seems to me. People can make their own inferences from these facts of course, but I want to emphasize once again that I am for the RDF and linked data projects, although the main problem remains, that is, of making something that the public wants.
James Weinheimer email@example.com
First Thus http://blog.jweinheimer.net
First Thus Facebook Page https://www.facebook.com/FirstThus
Personal Facebook Page https://www.facebook.com/james.weinheimer.35 Google+ https://plus.google.com/u/0/+JamesWeinheimer
Cooperative Cataloging Rules http://sites.google.com/site/opencatalogingrules/
Cataloging Matters Podcasts http://blog.jweinheimer.net/cataloging-matters-podcasts The Library Herald http://libnews.jweinheimer.net/