Ricky,
I understand what you are getting but my concern is that it may be
unintentionally over complicating DC. This may not seem possible from the
perspective of a person who is a cataloguer or used to MARC - for them the
DC is way too simplistic already. This does not hold for the museum users
and record creators who are not cataloguers. What we found with our SGML
work, and again in the Z39.50 implementers testbed was that tagging
structures and methods we thought were dead simple in the committee room
turned out to be devilishly complicated to get real people to use in real
applications. Most of them would say basic DC is pretty complicated.
The requirement for users to enter two sets of data - info about
the original and then about the surrogate - for a single record sets off
warning bells. There might be a simpler solution, what you are proposing
might be the best solution that has to wait for practical tools and sw to
be made for it to work, or implementers might say it's no problem. I don't
know. But I am unsure enough that I'd like to see the issue aired more
broadly in the user and implementer community.
Maybe with DC-5 focusing on implementation issues we might get an
implementers testbed organized that would work on the concrete aspects of
this issue.
Best, JP
At 5:11 PM 9/23/97 -0700, Ricky Erway wrote:
>REPLY TO 09/23/97 13:38 FROM [log in to unmask] "John Perkins": Re:
>Resource Types discussion
>
>John,
>
>We (who want to use DC for all kinds of resources, web-based and
>not), in fact, assume there IS a better, richer set of metadata in
>some form other than DC metadata. We are merely trying to make sure
>DC can accommodate a brief description of non-web-based resources
>(that ultimately points to more detailed information), so that people
>looking for stuff on the web are reminded that there are lots of
>other information resources, too. Some are available via FTP,
>Gopher, and OPACs, and sometimes you gotta go look at old-fashioned,
>analog things.
>
>So we don't expect the DC to do all the heavy lifting -- just the
>descriptions that help researchers find things. Our discussion about
>the original/surrogate issue is only to ensure that, at least, the
>original CAN be described.
>
>Ricky
>
>To: [log in to unmask]
>cc: [log in to unmask]
