Hi, Corey--

I think my issue is more that I want to use a LOGICAL structMap TYPE (for a
performance that may occupy portions of several tapes) rather than PHYSICAL
(parent divs for the tapes themselves and nested divs for the portions of
each tape on which various performances reside). My problem is, there is no
LABEL attribute (nor can I link descriptive metadata) to identify just the
area where a particular performance and a particular tape overlap.

I might be able to accomplish this with multiple structMaps (logical and
physical), and divide the tape physical divs into nested divs for the
performance pieces on each tape. Then, when a performance spans two tapes,
I could have nested physical divs on each tape point to the same
descriptive metadata as the logical div for the performance itself. But
this seems unnecessarily complicated, especially when a LABEL attribute for
the area element (or the ability to link descriptive metadata to it) would
meet my need to identify each portion of the performance.

Well, I suppose I *could* use the area's ADMID attribute to link to
dc:identifier "administrative" metadata, in order to label it--that would
be an interesting kludge...

--Eric

At 05:18 PM 2004-08-19, you wrote:
>Hi Eric,
>
>I'm fond of the nested div approach because it enables you to link
>metadata, be it descriptive or administrative, to that particular part of
>the tape.  Also here at LC we are not putting "descriptive" labels in the
>structMap but we link to descriptive metadata in the dmdSec.  We only use
>the LABEL attribute of a div node for "structural" labels.
>
>Corey Keith
>Library of Congress
>
> >>> [log in to unmask] 8/19/2004 4:51:58 PM >>>
>I'm using METS to represent video performances that we have accessioned
>into our library from NYU's performing arts school. Each catalogued
>"object" is a performance that may be one of several on a particular tape,
>may span several tapes, may start at the end of one tape and continue onto
>the beginning of the next tape, etc.
>
>Rather than create div elements for each tape, part of a tape, or parts of
>several tapes on which a particular performance resides, I am trying to use
>the seq and area elements with BEGIN and END attributes to indicate where
>each piece of the performance begins and ends on each tape. Things are
>working out fine, except that I need a LABEL attribute for the AREA
>element, to indicate what part of each tape is being utilized for the
>particular performance being represented.
>
>Does this look like a good approach, and would it be worthwhile for METS to
>add a LABEL attribute to the area e! lement? Or, am I making things
>unnecessarily hard for myself, and should use nested divs to indicate the
>parts of tapes on which each performance resides?
>
>Thoughts or suggestions appreciated,
>
>--Eric

=======================
Eric L. Stedfeld
[log in to unmask]
NYU Bobst Library
70 Wash. Sq. South
New York, NY 10012
212 995-3545