Re: More History comments

Hi Rob,

I'm not sure if you ever got an answer to your specific question:

>9/ What is part of a single situation?  How do events that affect objects 
>contained within other objects affect the containers?

With your example of adding a bitstream to a bundle (e.g. supplying a
missing TIFF image from a scanned document)

I think you have to look at this from the functional (archivist) point of 
view:
when is a change significant from a collection management point of view,
and thus would be nice to record in the history system? So in your example,
from a purely practical point of view, what's changed that the history system
might want to know about?

1. The bundle changed, as you said (BND 1:2).

2. Has its containing ITEM1 changed in any important way? Assuming we
want to expose all the versions of an item/bundle that have ever existed,
then yes, ITEM1 now has two versions available instead of one, so we record
ITEM1:2.

If, on the other hand, we will *not* show all versions of a an item/bundle
that have ever existed but only the last version, then this is not a 
significant
change to ITEM1 (i.e. it still contains one bundle).

3. Has the collection changed in any significant way? No. It still has an
ITEM1 and the fact that ITEM1 has changed is not significant to the
management of that collection.

[NOTE: Our faculty advisory board are split on this question, but those
with the best sense of history for the scholarly record think that we should
expose all the versions of the item/bundle and default references to the
latest version.]

So in my opinion the problem is not as exponential as you worry it is...
we can bound the problem for any given resource change pretty easily.

This is, of course, assuming we're asking the question in the context of
OAIS rather than the more general "what should you record in a general
system to track changes of any resource over time", in which case you'd
have to take the more conservative approach that any state change to
any element of the system effects the entire system... like an ecosystem.
Ugh.

Received on Monday, 26 May 2003 12:26:30 UTC