On Tue, 15 Jul 1997, Jordan Reiter wrote:
> At 10:51 PM -0000 7/14/97, Andrew Daviel wrote:
> >If we have FORMAT to say what kind of representation is being used, maybe
> >we should keep TYPE for a more generic description ?
> >
> >TYPE=PORTRAIT
> > FORMAT=IMAGE/JPEG - a photograph
> > FORMAT=VRML - a 3D head
> >
> >TYPE=REPORT
> > FORMAT=TEXT - a written report
> > FORMAT=AUDIO - a verbal report
>
> I'm pretty sure that in metadata, resource-type refers to what you call
> "TYPE", while in fact the term "MEDIA" refers to what you call "format".
> In web implementation, for example, it might look like:
> <META NAME="DC.Resource-Type" CONTENT="document.essay" MEDIA="audio/aiff">
That looks awfully like HTML to me ... and in the 4.0 draft, we have:
(http://www.w3.org/TR/WD-html40/struct/global.html#h-7.1.3.2)
ELEMENT META
i18n lang, dir
http-equiv NAME #IMPLIED
name NAME #IMPLIED
content CDATA #REQUIRED
scheme CDATA #IMPLIED
.. no MEDIA.
In DC we currently have a TYPE element and a FORMAT element
("assigned from enumerated lists such as .. MIME types")
(so maybe I should have put world/vrml, audio/basic etc.)
.. however, in CSS we do have
ELEMENT STYLE
i18n lang, dir
type CDATA #REQUIRED
media CDATA #IMPLIED
title CDATA #IMPLIED
(http://www.w3.org/TR/WD-html40/present/styles.html)
but that's HTML, not DC.
Anyhow, my basic premise was that if FORMAT tells you whether
a resource is text, audio, video etc. then TYPE doesn't need subtypes
giving the same information.
Andrew Daviel
|