On Tue, 23 Sep 1997, Gary Malet wrote:
> Re: Bibliographies
>
> (A) <META NAME="DC.type" CONTENT="data.Structured-Text">
>
> This doesn't represent a bibliography. Will DC constrain us to a registry?
I was just going by Roy Tennant's structuralist list at
<URL:http://sunsite.berkeley.edu/Metadata/structuralist.html> which is the
current "pulse" of consensus from the DC community. It might be an idea
to suggest that Roy add a Text.Bibliography type for these?
> Is there a reason you might want:
>
> (B) <META NAME="DC.type" CONTENT="(SCHEME=MedMetadata)
> data.Structured-Text.bibliography">
>
> This might allow a selection of "Structured-Text" from a global search
> engine pick list or satisfy the sensibilities of disciplined cataloguers?
If you're using your own scheme you could use anything for the content as
its explicitly not fitting in with the standard DC scheme. So you could
just use:
<META NAME="DC.type" CONTENT="(SCHEME=MedMetadata) bibliography">
However by putting in the data.Structured-Text and dropping the local
scheme you'll get support DC scheme-aware indexing and processing systems.
Its your call on whether that is important to you or not.
> Is there an objection to:
>
> (C) <META NAME="DC.type" CONTENT="(SCHEME=MedMetadata) data.bibliography">
>
If you want you're own local scheme that mainstream DC aware tools will
ignore, then that's fine.
> Would you have us revise the 8 million records in Medline with metadata
> elements, ex. "Structured-Text", that don't add anything for us? Again,
> in terms of resource discovery and efficiency (C) is best. Am I missing
> something?
If you've got 8 million records and you've got your own tools that are
used more or less exclusively by your user community then you can probably
got with your local scheme. Hell, with 8 million records you've probably
got more data than all the other current DC users put together so you can
do anything you like! :-)
> Please advise what the DC WG recommends for DC syntax with an established
> metadata set:
>
> (A) (B) or (C)
In your case I'd say (C) probably does make sense, although you might want
to stick an (A) (or Text.Bibliography if the group agrees that is a
reasonable structural resource type) in *as well* for the DC aware
"mainstream" tools when they appear in force.
Tatty bye,
Jim'll
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Jon "Jim'll" Knight, Researcher, Sysop and General Dogsbody, Dept. Computer
Studies, Loughborough University of Technology, Leics., ENGLAND. LE11 3TU.
* I've found I now dream in Perl. More worryingly, I enjoy those dreams. *
|