- From: Thomas Baker <tbaker@tbaker.de>
- Date: Sun, 13 Mar 2011 19:37:45 -0400
- To: public-lld@w3.org
Karen:
> 1. I _hate_ blank nodes ... A blank node says I have data about this thing,
What I find frustrating in FRBR is the need for a blank node where you have
no properties... in THEORY, every M manifests at least one E, which in turn
expresses one W. If you have M properties (say a title, a publisher, a date)
and W properties (an author and some subjects), FRBR implies that you must
create an E to get from the M to the W -- even if you have no E properties.
In this case you don't know enough about your something to give it
properties appropriate to the E entity.
Someone on the public list (I'm thinking it was Ross) said essentially that
if you find yourself going through absurd gyrations to fulfill a model,
especially if those gyrations don't appear to enhance your functionality,
maybe your model doesn't fit your reality.
How about if we let FRBR be a model for library cataloging systems --
systems that will only be used by a relatively small group of people with
particular logon privileges, and come up with an entirely different model
for the public face of library data? Well, maybe not entirely different, but
one based on what we want to with the data in an open world rather than the
closed cataloging world.
--
Tom Baker <tbaker@tbaker.de>
Received on Sunday, 13 March 2011 23:38:25 UTC