To turn this question around 180 degrees, where should the preservation
record of the METS file itself go in the METS file? Could one reserve
an <amdSec> to use for the history of the METS file, separate from the
<amdSec> for the objects that the METS file describes? Is there a
recommended best practice for additional information about the METS
document itself?
-----Original Message-----
From: Metadata Encoding and Transmission Standard [mailto:[log in to unmask]]
On Behalf Of Jerome McDonough
Sent: Monday, August 30, 2004 3:15 PM
To: [log in to unmask]
Subject: Re: [METS] Agent role in metsHdr
I should really not answer e-mail before I've had my fifth
cup of coffee.
So, I'll try to clarify my take on this, and if anyone else on the
editorial board wants to jump in, they may feel free. The metsHdr
really is intended to be about the METS document itself, and not the
complete digital object. As noted in the original question, the METS
schema provides slots for recording descriptive and administrative
metadata for the digital files included in the digital object; the
metsHdr should in theory be used only for CREATOR/EDITOR/IPOWNER etc.
with regards to the METS document itself.
That being said, it's fairly obvious that someone entrusted with
preservation responsibilities for the METS document is probably also
going to have preservation responsibilities for the rest of the files
composing the complete digital object. If an organization wants to use
such roles as PRESERVATIONIST, DISSEMINATOR, ARCHIVIST, etc. as a
shorthand way of indicating that someone has responsibility for both the
METS file and associated files in a digital object, I don't think anyone
is going to morally censure them for it. Moreover, some folks may be
using the <FContent> element to embed the content bitstreams inside of
the METS document itself, in which case saying someone has preservation
responsibility for the METS file would of necessity imply preservation
responsibility for the embedded bitstreams. So, the metsHdr
documentation stating that the <agent>s should be responsible for the
"METS document" is actually more inclusive than it might sound on its
face.
But using the metsHdr to indicate someone has roles/responsibilities for
digital content files in an object when they do *NOT* have
responsibility for the METS file itself would be, to my mind, a mis-use
of the specification. If someone has an agent role with respect to
digital files that they do not also play with respect to the METS file,
that should be recorded somewhere in a dmdSec or an amdSec with the
<file> element(s) linked to the information using the DMDID and ADMID
attributes. So, officially, <metsHdr> should be used to record metadata
for the METS file itself, and if you want to use <metsHdr> to record
information about agent roles for the METS file *and* associated data
files, probably no one will give you grief, although it would be a
*very* good idea to indicate that in the <note> element for that agent.
But if someone has a role with respect to content files that they do not
have with respect to the METS file itself, you should keep that out of
the <metsHdr>.
Does that help any, or do I need a sixth cup of coffee?
On Aug 30, 2004, at 2:19 PM, Karen Coyle wrote:
> OK, then I need a bit more of an explanation of the other agent roles
> in the metsHdr, in particular:
>
> ARCHIVIST
> PRESERVATION
> CUSTODIAN
>
> It seems to me that the inherent meaning of these implies not just the
> METS document but also the digital resources. I can see where CREATOR
> and EDITOR can refer to the METS document alone, but with the others
> I'm having a hard time giving them that semantic.
>
> kc
>
> On Mon, 2004-08-30 at 08:15, Jerome McDonough wrote:
>> On Aug 30, 2004, at 10:48 AM, Evan Owens wrote:
>>
>>> I read the documentation to say that the agent roles in the
>>> <metsHdr> are roles in connection with the METS file, not the
>>> digital object that the METS file documents. Thus for example agent
>>> role "IPOWNER" means IP
>>> owner of the METS file, not the digital object described by METS
>>> file.
>>> Information about the IP owner for the digital object would go in
the
>>> rights MD section of the administrative MD.
>>>
>>> Is that a correct reading of the standard?
>>>
>>
>> Yes.
>>
>>
>> Jerome McDonough
>> Digital Library Development Team Leader
>> Elmer Bobst Library, New York University
>> 70 Washington Square South
>> New York, NY 10012
>> (212) 998-2425
> [log in to unmask]
> --
> -------------------------------------
> Karen Coyle
> Digital Library Specialist
> http://www.kcoyle.net
> Ph: 510-540-7596 Fax: 510-848-3913
> --------------------------------------
>
Jerome McDonough
Digital Library Development Team Leader
Elmer Bobst Library, New York University
70 Washington Square South
New York, NY 10012
(212) 998-2425
[log in to unmask]
|