Fellow Metalogers,
I am hopeful that David Bearman is right when he says:
"the design principle that supports both [simple and complex
resource description] can be quite simple."
I believe that our original intent (simple, author-generated
description) has expanded to include the possibility for rather more
sophisticated applications as well that might well involve the creation
of DC records by more experienced information managers.
A project report at the Bonn Metadata workshop last week provides a
good example of this: Dublin Core-style records for mathematical
preprints embedded in HTML according to the syntax convention
elaborated at the W3C workshop. The project staff indicated they would
convert their field names as soon as there is clear indication that DC
elements are stable.
The trick is to have an upgrade path that supports a range of
description, from simple author-generated records to more formal
records embellished with SCHEMEs, TYPEs, and perhaps other parameters
such as the flags that the User Guide proposed.
The User Guide controversies of the past couple of weeks result from
an attempt to achieve this range of expression, but I think
the guide got too deep too fast, and the simplicity we've all been
trying to preserve suffered somewhat.
I would like for us to step back a bit and separate the new issues
raised in the Guide from the Guide itself, fold in the issues raised in
the Image Metadata workshop, and see where we stand.
I'm putting together a list of Open Issues which I will post to the list
tomorrow probably. Mail me with your favorite if you'd like.
Meanwhile, my fortune cookie at lunch read:
You can depend on the trust of the collective.
Lucky numbers 2, 3, 29, 39, 40, 41
stu (who thinks he might buy a lottery ticket on the way home tonight)
|