Folks,
Some quick comments about URNs in HTML documents.
1) People print out HTML pages. This loses any implicit or hidden
information. Therefore, it is sometimes important to put URNs as text
strings in the body of documents.
2) Programs process HTML pages, e.g., for indexing. Therefore, it is
important to put URNs in standard places using agreed syntax.
3) People use hyperlinks from HTML pages. Therefore it is reasonable to
use URNs in hotlinks (hrefs).
The conculsion is that URNs, like URLs before them, will be be used in
various contexts and that flexibility is a virtue.
As an example of how these things can work, we have developed browser
extensions for resolving handles (which are a class of URN). The
extensions are available at http://www.handle.net/ for downloading. Once
an extension has been installed, the browser will resolve automaticaaly any
string of the form "hdl:..." wherever it might find a URL, including hrefs
in HTML documents. If the handle resolves to a URL the browser then
accesses the URL automatically.
The handle system is used for DOIs. It is a trivial change to have the
browsers also recognize strings that begin "doi:" or indeed "urn:".
Bill
=====================================
William Y. Arms
Vice President
Corporation for National Research Initiatives
[log in to unmask]
http://www.cnri.reston.va.us/
http://www.dlib.org/
tel: (703) 620-8990
fax: (703) 620-0913
=====================================
|