[stdsreg] Metadata Standard v2.5

[Bob Hager <BHager@ansi.org>]

The link to v2.5 of the metadata standard definition and a summary of
changes made since v2.0 (two separate documents) are now available from this
page:

http://www.ansi.org/reports/master.asp?room=70

Changes since v2.0 were made based on discussion and minutes from 6/4/02
meeting, comments made since, and my own additional recommendations.  The
overall goal in making these changes was to continue to simplify and
consolidate elements, while also harmonizing with Dublin Core v1.1 as much
as possible.  In v2.5, there are several new elements, several elements
eliminated, and several existing elements that have been re-named and/or
re-defined.  So it's worth reviewing every element and every attribute.

Following are outstanding issues that still need discussion:

+  DC 1.1 mapping: are there appropriate mappings for Designation and
SDOInfo?

+  Archival records:  are we intending to encourage the inclusion of
archival records in submissions using this metadata standard?  If so, we
need to consider the inclusion of the "Is Replaced By" element and the
addition of "Withdrawn" as a Current Status option.

+  Language:  We now have a provision for representing different languages,
both of the content itself (through the Language element) and of the
metadata (through the Language attribute).   Martin raised the point of
language metadata matching the language of the content:

"Also note that the name and acronym can be dependent on the language being
used (see ISO re French and English versions). Should we mandate that the
English version of the name be specified (which mitigates against national
standards) or allow both national and English variants to be stated? (In
other words, allow this field to be repeated.)"

Presumably, the language of the content would match the language of the
metadata and if different language versions of a given standard are
presented as separate documents, each document would carry a separate
metadata record in the corresponding language.  How would multi-language
standards be represented?  Since such a standard would be represented by a
single record, perhaps the Language element could contain more than one
value and English could be the default metadata language.

As discussed, the next step is to expand the best practice guidance and to
develop an XML representation.

Looking forward to your further input.

Bob Hager