On 22/02/2012 22:52, Jonathan Rochkind wrote:
I actually don't think it's neccesarily 'marc', although marc is a terrible terrible format we should be working to abolish. But, hey, MarcXML is XML, everyone loves XML, or at least is okay with it, right? And anyone can already turn any Marc into MarcXML in one line of library code in almost any programming language there is; and many many of our systems will already provide MarcXML out of the box anyway.
So what's the problem?
It's _not_ the surface representation of the data, in Marc, or XML, or even RDF. Taking the exact data we've got and translating it to one of those formats is not going to help -- becuase after all, we already CAN do that, so what's the barrier?
The barrier is the data itself. Is muddled, ambiguous, unclear.
This is all correct and I won't argue with any of it, but this raises an issue that I have with Linked Data: I don't think other formats are much better from ours. Karen mentioned the FaBIO system that has really bizarre (at least for me) works, expressions, etc. The (poetically named!) "Thing" from schema.org with item types: article, blog, book, recipe, TVSeason, etc. is tough to reconcile with others. I honestly do not think these other projects will change any of their practices just for us.
Therefore, in the aggregate, the linked data universe seems--to this cataloger--to be chaos. While I can imagine a solution, and you probably can too, such a solution would make reconciling all of the different MARC formats in the world seem like child's play.
This is not saying that we should not be aiming for linked data. Doing it with identifiers would be better (maybe) than what we have now. The biggest obstacle of entering the linked data world, in my opinion, is to interoperate with what is already there, and will be there. This will be exceedingly difficult. While FRBR might (maybe) make it easier, it is not worth waiting for it, especially considering the effort that will be demanded from already-pressed catalogers now. I would much prefer to put the onus on the already-pressed IT staff. :-)
As a corollary to this, I mention something hesitantly. I don't know how useful putting up records for our printed materials will really be. The public is evolving toward digital resources. Yes, I understand that someone searching for a book from a cell phone with a GPS chip will, in the correct system, readily find that they can get a copy of a book they want from a local library only 10 km away. Some few may actually decide to go there to get the book, but many more would probably be interested to know that they could download a copy to their phone or tablet.
This is why I mentioned in my paper in Buenos Aires the NPTEL free online courses that lots of people would really and truly find useful. There are so many of these sorts of resources that it is absolutely astounding! Unfortunately (I am definitely a book lover!) these are some of the directions I think we will have to take if we are to make a real difference to the public.