[uf-new] Dublin Core (was: hAudio FN or Title)

Andy Mabbett andy at pigsonthewing.org.uk
Fri Feb 1 12:39:10 PST 2008

In message <C3C88F71.9B977%tantek at cs.stanford.edu>, Tantek Çelik
<tantek at cs.stanford.edu> writes

>>> The main disagreement seemed to be in DC's choice of class names...
>That's only one problem with DC.

You fail to explain why you think DC's class (sic) names are a problem.

>The other problem is that DC itself is more theoretical rather than based on
>any actual content publishing research/behaviors.

On the contrary: DC is based on a deep understanding of the metadata
published by the type of organisations for which (and by whom) it was
intiially designed.

>Anyone that wants to look at re-using DC should instead look into
>helping move the citation microformat effort forward, which has documented
>DC as one of many previous formats that relate to citations.

DC is not only for citations.

>DC by itself is not the answer.

Before you can assert that, you should, at the least, state which
question you think applies.


My above comments not withstanding, more effort towards completing the
'citation' work (and that for several other of the much-needed, pending,
microformats) would be a good thing. If this community does not do it,
some other group probably will.

>>> * It re-uses a vocabulary that is largely accepted in the web semantics
>>>   community.
>It's been mostly used to publish hidden metadata in pages that is either
>ignored or polluted.

If so, a facility for using it on "visible" metadata would be an
improvement, would it not?

>  It's not really got much support of tools that support
>it and do something useful with it

There *is* support and there *are* tools, not least in the fields for
which it was intended. It is even government-mandated in some quarters.

>- mostly academic projects.

That's not necessarily a bad thing (after all, HTML was first designed
for academic projects!).

>I just wrote up this process FAQ entry regarding "re-use whole-sale"

There is a requirement on the wiki that opinions are marked up as such -
see point 4 on:



Folks new to DC might like to read:


and note in particular the distinctions between "simple" and "qualified"
DC, and that the former has just 12 properties:

           1. Title
           2. Creator
           3. Subject
           4. Description
           5. Publisher
           6. Contributor
           7. Date
           8. Type
           9. Format
          10. Identifier
          11. Source
          12. Language
          13. Relation
          14. Coverage
          15. Rights

(By way of illustration, Qualified DC takes a property, such as "date",
from Simple DC and makes properties like "date.created" and

A method of applying DC using HTML class names on published data would
be 'A Good Thing', and could be used alongside existing microformats.

For example:

        <div class="haudio">
           <span class="audio-title">DigitalPlanet Podcast</span>
           <abbr class="published" title="20071029">29 Oct 07</abbr>

could, using a wrapper class to encompass the item to which the DC
metadata applies, become:

        <div class="haudio dc-wrapper">
           <span class="audio-title dc-title">
              DigitalPlanet <span class="dc-type">Podcast</span>
           <abbr class="published dc-date-created" title="2007-10-29">
              29 Oct 07

which could then be parsed by DC consumers, without the need for such
consumers to be updated each time a new microformat emerges.

Note, for example, the use of:

        class="published dc-date-created"

in an hAudio; whereas an hReview would have:

        class="dtreviewed dc-date-created"

and an hListing might use:

        class="dtlisted dc-date-created"

Whether such usage is called a microformat, "POSH" or some other term is
really a matter of bike-shed colouration.

Andy Mabbett

More information about the microformats-new mailing list