W3C home > Mailing lists > Public > www-tag@w3.org > April 2008

Creating Catalog of W3C's XML Schemata

From: Ted Guild <ted@w3.org>
Date: Tue, 29 Apr 2008 11:09:40 -0400
To: www-tag@w3.org
Message-ID: <nnbk5igofyj.fsf@dev-null.guilds.net>


Last Summer I wrote [1] asking the W3C TAG to consider writing a Best
Practices document on caching XML schemata.  The TAG agreed it was an
important issue and part of a bigger concern [2][3] on the scalability
of popular resources.

The W3C Systems Team posted an article [4] a couple months ago on our
DTD/schemata traffic which drew a fair amount of attention to this
continuing problem along with a few useful comments.  One suggestion,
reiterated by Daniel Veillard former W3C staff, ongoing participant of
XML Core activity and creator of libxml, was that W3C should make
available a XML Catalog of it's various schemata [5].  Libxml has
utilized a catalog from it's inception [6], is widely deployed on
various UNIXes, including Linux distributions and MacOS, and often the
catalog is in a separate software package than the library and
utilities (eg [7]).  This topic came up as part of a recent Staff
Project Review where Tim suggested the question of whether or not to
make a catalog available of W3C's various DTD/schemata should be
brought to the TAG to decide.  The Systems and Communications Teams
would discuss further how to carry it out should the decision be

W3C could also work to persuade various Operating System providers to
make libxml or a similar XML Catalog model a default part of their OS
in a defined location (system environment variable or predefined OS
specific default) for other XML processing software and libraries to
reference.  Also they should have reasonably frequent updates part of
their overall OS' automated update strategy and define policies
whether libraries and utilities are able to update or contribute to
the catalog as well, treating it like a caching catalog.  Whether
static and only updated routinely by OS maintainers or the online
resources' http headers are more regularly checked to fetch newer
versions the catalog could be basically considered a http cache from
the architectural point of view.

[1] http://lists.w3.org/Archives/Public/www-tag/2007Jul/0200.html
[2] http://lists.w3.org/Archives/Public/www-tag/2007Aug/0032.html
[3] http://www.w3.org/2001/tag/group/track/issues/58
[4] http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic
[5] http://www.oasis-open.org/committees/download.php/14809/xml-catalogs.html
[6] http://xmlsoft.org/catalog.html
[7] http://packages.debian.org/etch/all/xml-core/filelist

Ted Guild <ted@w3.org>
W3C Systems Team
Received on Tuesday, 29 April 2008 15:11:39 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:56:21 UTC