On 12/06/2011 20:53, Edna Stirrat wrote:
<snip>This could work in an entire variety of ways, and each library/consortium/other grouping of metadata creators can implement their own solutions. One thing that is important to keep in mind however, is that linked-type technology is the normal way computer systems and the internet works. The traditional library method of transferring an *entire record* is actually what is weird today. The way browsers work is that it downloads all kinds of files from different places and the browser brings them all together to create *what appears* to be a single web page. So, any web page you see can easily be made up of dozens of files, or more, behind the scenes.
Will a linked record created with FRBR and RDA still be shareable between vendor to library or will the links break?
How would the sharing of a linked record actually work in practice?
For example - a bibliographic record that contained information of Harry Potter and all its associated manifestations, works, subjects, concepts etc, etc.
The files themselves can be held anywhere: locally, or quite literally anywhere on the web. What is different with this linked-data movement is that these files are linked semantically, i.e. based on their meaning and this requires another format that allows meaning, in this case XML, but the fundamental technology itself is nothing new and has been around since the beginning of computers. Therefore, I can envision a system that, when you want to download a record, saves all related records along with it: all the authority records (wouldn't that be nice to have right now?) and *if* FRBR is implemented, all associated entity records as well. At the same time, the system could check the dates of each version of each of the authority/entity records to see if they need updating, and if they do, the new versions overwrite the old versions. Or, the system could simply link to the related records on the internet.
Much of this, as so often, comes down to a matter of *money* and depends on how open all of this data will be, or if it will be proprietary and libraries will have to pay for it. For instance, assuming that full FRBR is implemented: libraries are now expected to pay for access to the rules, will those same libraries also have to pay for access to the entity records, much as libraries have to pay to download OCLC records? So, let's imagine that a cataloger wants to download a manifestation record from a library that allows this for free--will the work and expression records also be available for free? If not, there will be much less reason to take the manifestation record in the first place.
The issue of "rights" becomes much more complicated the more we break up a record into various types of entities. I could even imagine that libraries could be expected to pay for "relation" information separately!
Still, the actual work of the cataloger will not necessarily have to change that much, even if full FRBR is implemented. Or it can. It depends on what the powers-that-be decide, and I would hope that library traditions of openness prevail, while catalogers themselves would be among the most important of those "powers-that-be."