"John A. Kunze" wrote:
> Comments by the end of April
I am concerned about how qualifiers are illustrated in the examples in
the draft.
My concerns are partly about internationalization (I want to see
authoritative examples), and partly about the use of the dot-notation
(I don't like it and don't consider the issue sufficiently resolved
to be used in examples).
1. Internationalization
-----------------------
The 15 elements of the DC element set has been stable since 1996, and
the authoritative RFC (Weibel 1998) does not specify qualifiers or any
other method to qualify semantic elements.
Section 7.4.4 in HTML 4.0 (Raggett et al 1998), however, specifies
that a META-tag may be used to embed metadata in an HTML-page. It goes
on to specify the use if qualifiers, referring to them as attributes.
The HTML. 4.0 Specification makes particular mention of the "scheme"
attribute, but it also refers briefly to two other attributes "lang"
and "dir" defined elsewhere.
For some reason, the "charset" attribute is not mentioned in section
7.4.4, but it makes sense to include it with the other two (as is also
implied in section 5.1 in the HTML 4.0 Specification).
This gives the HTML-metatag the following four attributes "inherited"
from HTML 4.0:
- Scheme: This qualifier the context the property should be
interpreted in.
- Lang: This qualifier specifies the language by means of an by means
of an RFC 1766 (i.e. iso 639) tag of the element value of the property
field.
- Dir: Direction of text (RTL or LTR).
- Charset: Character set used to describe value of the property.
I think all these four makes a lot of sense, and all of the should be
exemplified in a section 7 in [DCHTML]. The present draft shows the
use of "scheme" and "lang", but not "dir" and "charset". I also think
that it would be nice to have an example showing how to encode
non-ASCII characters (which in HTML I believe is based upon using SGML
entity names (as defined in HTML 2.0 and annex D of ISO 8879).
I.e. I propose that that the document includes authoritative
examples to show how to deal with non-ASCII character sets.
By this I mean that examples such as the following should be added
to [DCHTML]:
<META NAME="DC.Title" LANG="de" CHARSET="iso-8859-1" DIR="ltr"
CONTENT="Römische Elegien">
<META NAME="DC.Title" LANG="he" CHARSET="iso-8859-8" DIR="rtl"
CONTENT="ìàøùé
õøàî íåìù">
2. The dot-notation
-------------------
Some of the examples i DCHTML shows the use of dot-notation to
refine DC properties (e.g.: "DC.Date.Available").
My notion is that it is _not_ a good idea at the present time to
"endorse" this idea by means of example. As far as I've understood the
DC Initiative, this notation is still controversial (and personally, I
think that this is a bad idea). The dot-notation isn't really a
qualifier. Instead, it is a mechanism for refining a property.
Refinement is a dangerous item, as it will allow some people to use
the DC to create encyclopaedias rather than metadata for resource
discovery. Somebody will want to do exactly that, trying to map the
entire world into the DC in an almost Borgesian fashion. This does not
work. It also violates the KISS principle (Keep It Simple & Stable)
that I believe is all important if a standard are to be successful.
References
----------
Weibel, Stuart, Renato Ianella and Warwick Cathro: The 4th Dublin Core
Metadata Workshop Report; D-Lib Magazine, June 1997,
<http://www.dlib.org/dlib/june97/metadata/06weibel.html>
Weibel S., J. Kunze, C. Lagoste, M. Wolf: Dublin Core Metadata for
Resource Discovery; RFC2413, September 1998,
<http://ftp.univ-lyon1.fr/mirrors/rfc-processed/24xx/2413.gz>
--
- gisle hannemyr ( [log in to unmask] - http://home.sol.no/home/gisle/ )
------------------------------------------------------------------------
"Use the Source, Luke. Use the Source." -- apologies to Obi-Wan Kenobi
------------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|