Re: Representing distinct item states

Rob sez:

> This wouldn't do, because when the data is collected
> into a unified store for query, there would be no way
> of telling with which state each dc:title is associated.
> (It's by virtue of the statements being in separate files
> that one can tell right now.)

JSE: I'm confused. I thought that one of the innovations of the Harmony model
was that it was able to capture events by certain actors or roles. In the
example that Rob gives, an aspect of the Item is modified by some actor --- it
doesn't matter whether this change happens because the original accessor
mis-types the title or the title changes for editorial reasons. We need to be
able to capture the action (the property changed) and the context for that
change.

A week or so ago I suggested that we need to create data models for curatorial
activities, to better capture the sematics of "actions" that take place
against items in a collection. It was suggested that this might be beyond the
scope of current history system, but really, only use cases will tell --- use
cases that drive the collection of change data, and use cases that exercise
that data (incl. queries across it, and decisions or actions taken based upon
it, like a roll-back of some previous activity). Examination of such cases
would ditact the granularity of metadata that we need e.g. for actions against
properties, like a title change...

John

Received on Saturday, 17 May 2003 08:40:13 UTC