On Feb 22, 2006, at 4:46 PM, Riley, Jenn wrote:
> 1) Every single file being transported (even the ones though of as
> descriptive or technical metadata) has an entry in the <fileSec>,
> allowing us to indicate checksum, etc. These are grouped by type of
> file.

Nothing particularly wrong with this, although if you're doing it  
because this allows you to record a checksum for the metadata record,
I think it points to a problem in the greater metadata world.  Many  
of the extension
schema available to use with METS don't include a facility to allow  
you to check
the integrity/authenticity of the metadata record.  This conflicts  
with several of
the writings on metadata that we've all grown to love over the past  
several years.
<indecs>, for example, specified that the author of any item of  
metadata should
be *securely* identified.  The OAIS reference model states that the  
information for an information package consists of both the data  
object itself
and the representation information, and also says that fixity  
information "provides
the data integrity checks or validation/verification keys used to  
ensure that the
particular content information object has not been altered in an  
manner," which pretty clearly implies that your fixity information  
needs to cover
some or all of your technical metadata, as well as your actual data.

METS at the moment assumes that an extension schema will provide its own
facilities for carrying the fixity information necessary to insure  
that the record
is intact.  The board might want to reconsider that assumption.  It  
would be possible
to add some elements to the mdSecType definition to allow for a  
checksum and/or
digital signature to be recorded for any piece of metadata included  
in a METS
object.  This would allow for recording fixity information for  
metadata when
the metadata schema involved does not support it.

An alternative would be to a PREMIS record within the METS object for
each metadata record for which you needed to record fixity information
(in theory, all of them).  That would accomplish the same thing without
involving changes to METS.

Jerome McDonough, Asst. Professor
Graduate School of Library & Information Science
University of Illinois, Urbana-Champaign
501 E. Daniel Street, Room 202
Champaign, IL 61820
(217) 244-5916
[log in to unmask]