On 22/08/2012 23:46, Kevin M Randall wrote:
But until we do have some mechanism for dynamically keeping descriptions current, the notes that you say are “completely useless” are absolutely essential. I certainly agree with you that the traditional methods of creating metadata are not adequate for handling the universe of online resources. But that does not mean that we shouldn’t still have standards that will allow “traditional” metadata and “created-in-an-as-yet-unknown-method-and-system” metadata to be able to interoperate. Developing those standards is what we’re trying to do with RDA. Hopefully there will be ways to harvest data from the resources themselves, and map them to the data definitions in RDA, to get them into our discovery tools. But until we reach that goal, we still need to be creating traditional records, and we need to know what it is that the records are describing. With a “Description based on” note, there is a clue to what was described, and when.
This was my point: that catalogers need to find methods to make sure that the records describe something that actually exists! While I have little argument against placing a note detailing when the description was made, I doubt how important it is since it only shows people (librarians, because the public will not really understand it) more precisely how obsolete the information is. “Oh! This record is five [ten, twenty, etc.] years out of date.” Therefore, it is a rather sad bit of information, but from another viewpoint, it is the easiest part of the record! Still, the mere fact that catalogers will be spending their time creating records that have a very high probability of being obsolete in x number of weeks or months or years does not bode well for how administrators will view the value of catalog records. And they are the ones who must be convinced–not catalogers or even the public.
Several years ago, I wrote an article where I tried to deal with this issue: http://www.emeraldinsight.com/journals.htm?articleid=1668881&show=pdf and I think it is still not too bad of an idea. If I had the chance to do over again, I would improve the article because I tried to write for both an IT audience and a cataloging audience at the same time and I don’t believe it worked all that well. Still the basic idea still is worth a try (I think), where embedded metadata would be linked to separate metadata records in catalogs and spiders would keep the two in sync. There was a major role for selectors, responsibilities for web masters and for catalogers too.
But novel ideas are needed more than ever, I think that is becoming clear enough to all.