On Tue, 9 Dec 2003, Marie-Louise Ayres wrote:

> 2.  LC colleagues - any timelines on updated MODS2MARCXML and MARCXML2MODS
> stylesheets becoming available?

We posted the new schema yesterday; expect to have MARCXML2MODS very soon,
and MODS2MARCXML a little later.

> 3.  I've noticed a discrepancy between the MODS2MARC mapping for
> <originInfo><dateCaptured> and the Guidelines.  Mapping reads:
> <dateCaptured> with encoding="iso8601" 033 ind1=0 ind2=0 $a <dateCaptured>
> with point="start" and 033 ind1=2 ind2=0 first $a encoding="iso8601"
> <dateCaptured> with point="end" and 033 ind1=2 ind2=0 second $a
> encoding="iso8601"
> with the ind2=0 - Capture, defined as:
> The recording of sound, the filming of visual images, the making or
> producing of an item, including the creation of an item not involving
> recording or filming.
> However, the MODS guidelines state:
> <dateCaptured>
> "dateCaptured" contains the date on which the resource was digitized or a
> subsequent snapshot was taken. This is particularly useful for Web resources
> because of their changeability. This could be a date that the resource was
> captured or could also be the date of creation.
> I'd just like to check that <dateCaptured> is the most appropriate place to
> store information about recording dates.

The guidelines generally add to the MARC descriptions. So the text
about date captured being a digitization date is of particular
relevance to digital resources, but is in addition to the basic MARC
definitions to clarify its use. I did check with someone dealing with
recorded sound here at LC, and it is appropriate to include recording
dates under dateCaptured, whether MODS or MARC 033. Other dates relevant
may be dateIssued (260$c in MARC), which may or may not be the same as
date captured/date recorded.

> 4.  I'd really like some advice on <relatedItem type="host">, especially as
> I see there are some additions there, and there are some special issues as
> we'll be converting these records to MARCXML (and MARC doesn't handle these
> thiings particularly well!).
> We have some sample records output from ScreenSound Australia's MAVIS
> database, which incorporates the concepts of 'cover' and 'track' level
> records.  I've included the following coding:
> <relatedItem type="host">
>                 <titleInfo>
>                         <title>Josie and Me; Pagan Love Song</title>
>                 </titleInfo>
>                 <physicalDescription>1, Shellac disc (78s, transcription),
> 10" Standard Play</physicalDescription>
>                 <identifier type="local">395552</identifier>
> </relatedItem>
> inside a MODS record for one of the components (which includes a URL to the
> online sound recording etc.).  This will all map to various 773 sub-fields,
> but looking at the changes to V3.0, I can see quite a lot of 'part' coding.
> Any advice on how these might be used, and what MARC fields and sub-fields
> they might map to?

If you have physical description of the component (the track), it would go
in <physicalDescription> of that MODS instance and map to 300. It is not
necessary to include in <part> unless it is parsing out pieces of a
citation, or the location within a host which would otherwise not be
included elsewhere. But I'm not sure exactly what data you are proposing
putting in part, so an example would be useful. What you've done above
looks fine to me. It is questionable whether or not you need that
<physicalDescription> under relatedItem because if you had a record for
the host, it would also be included in that record.

> 5.  In V3.0, we now have
> <place>
> <placeTerm type="code" or "text" authority="">
> Could you advise whether <placeTerm> is required (as namePart is required),
> or whether it is acceptable to simply have <place>Australia</place>

You would need the subelement placeTerm; <place> is now a wrapper element
and a value wouldn't go in it.

> 6.  Now that MODS allows us to include a URL for an online resource in
> <identifier type="uri>  AND
> <location><url>
> are others planning to include this information in both places and only
> index one?

If it is strictly a location (which may or may not be persistent)
providing access to the resource, record in <location><url>. What we are
doing at LC is using the URL in location, and if it has been assigned a
handle (which we do for digital resources), including the "raw" handle in
<identifier>. If a URI is identified as a persistent one it can be
recorded in both places.

Clearly we need some guidelines here.

> 7.  My colleagues in the Australian National Bibliographic Database wondered
> whether any other organisations had raised the needs for a MODS equivalent
> to MARC 008 character position 28 to represent government publications?
> Would you use
> <genre authority="marc">government publication</genre>
> But in this case, there is no way to indicate the jurisdictonal level.  This
> is an issue for us as we are using MODS and our new Harvester system to pass
> metadata bout online government publications through to the NBD.

I had suggested earlier that we include the values from 008/28, but hadn't
decided whether to mix them with the MARC genre list (which means they
will be at a different level of granularity than other terms on that
list) or to make a separate identifiable list. Any other thoughts on that?

Hope this helps.


 > > Thanks
> Marie-Louise
> Dr Marie-Louise Ayres
> Project Manager, Music Australia
> National Library of Australia
> Parkes  ACT  2600
> Email: [log in to unmask]
> Phone: +61 2 6262 1536
> Fax:      +61 2 6273 5081