OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   FPIs as locators in XML-links ?

[ Lists Home | Date Index | Thread Index ]
  • From: "Alain Michard" <Alain.Michard@inria.fr>
  • To: <xml-dev@ic.ac.uk>
  • Date: Fri, 8 Aug 1997 13:58:10 +0200

I may miss something (and in that case thanks for the help!) , but I
feel that the current specification of locator in XML-LINK don't open an
easy way to avoid including URLs in XML documents.
It is true that a URL can be considered as a unique id of one physical copy
of a  resource, but this id is transient : machines may be changed, and
"publishers" (any entity putting content on the Web) may decide to migrate a
document repository from one place to another.
If locators in XML-LINKS are URLs, this implies that in case of change of
URL, many authors
-a- should ideally be notified in some way of the change if it is relevant
for them (ie: if they have in their own documents links pointing to
resources which have changed of URL)
-b- have to retrieve all the document they have published which contain a
link to the modified URL;
-c- have to edit all these documents.
That's in fact exactly the situation with the HTML-based Web.

I feel that the SGML practice to use  Public Identifiers and to store
mappings of PUBLIC identifiers to SYSTEM identifiers in a Catalogue file
facilitates greatly the management of large collection of documents:
- a "publisher" may distribute updates of his public catalogue to the
community with which he shares a number of resources;
- the catalogue is the only file you need to edit when any Public ID gets
associated to a new physical resource;
- in case of mirror copies of Web sites, the catalogue may be an easy mean
to impose to your browser to look for a given document at a given site,
without having to specify it at each traversal of a link;

Moreover, including URLs in XML documents appears contradictory to the
general phylosophy of SGML, which I guess could be resumed as "Ensure
long-term life of documents".

I would be very interested to read some comments from SGML experts on the
list, to help me understand the reasons why the XML draft specs
exclude -so far- using Public IDs in links.

Best Regards

Alain Michard
Mediaculture - Direction du Développement
INRIA  -  Domaine de Voluceau
BP 105
F-78153 Le Chesnay Cedex - France
Tel: +33 1 3963 5472   Fax: +33 1 3963 5114









xml-dev: A list for W3C XML Developers
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To unsubscribe, send to majordomo@ic.ac.uk the following message;
unsubscribe xml-dev
List coordinator, Henry Rzepa (rzepa@ic.ac.uk)





 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS