Leonard Will wrote:

> (That is the theory; does anyone have software in which this type of
> structure has been implemented fully and effectively for use in
> practice?)

We have been developing a relational data structure over the last ten
years which uses a the following basic parent-child (whole-part, if you
prefer) relationships:

DONOR/VENOR/SOURCE <--> ACCESSION <--> OBJECT <--> PART <--> FRAGMENT

We have run into an interesting problem.  Our collections span fine (and
not so fine) art, furniture, housewares, archaeological specimens,
archives, books, small (and not so small) watercraft, vehicals, etc.,
collected over a century.  Now the problem.

Which end do you start from?

When I did archaeology I naturally assumed you began with fragments, now
as a registrar I have learned that you start with sources, the curators
begin with objects (whatever they are - a subject for another
discussion).  I guess what I am saying is that how you impliment the
database functionally really depends on the worldview.  I have developed
very successful applications from a variety of museums and
archaeological projects using this logic and essentailly the same data
structure, it was the business - functional model that really made how
the database operated different.

Richard Gerrard