The W3C has been created 10 years ago and has most part of its content on the Web. To be able to understand the W3C environment, we have to go through a number of facts and figures:
In May 2005,
One of the main reproaches made to W3C is usually is lack of usability in terms of coherency in navigation and Web design of its pages (not particulary appealing). These are valid comments though we can balance them with other facts.
Though we have strict rules, called the pubrules, for the space where the W3C recommendations are kept. It is a consistent information space. We have also a kind of very basic templating system, called ,new, based on a script which can help a consistent design across a section of the Web site. For example the QA Team is using this system.
We can also argue that there is a lack of use of the full potential of HTML, XHTML semantics to improve the indexing of Web pages and built useful indexes.
Though this depends on the section of the W3C Web site. For example, the QA Web site has a lot of metadata included in its headers.
We could consider that there are too many editors or webmasters, or we could consider that all people have the possiblity to edit this information space and that is a benefit for the whole W3C.
We can certainly improve and we have to do it.
How many Web sites with 875,000 Web pages and ten years old meet these facts:
Every pages are archived and historically kept in a simple incremental archiving system (CVS). We can access to any revisions of the pages in the past.
Almost all pages are accessible by people with various disabilities and by various devices (mobile phones, desktop browsers, braille readers, vocal synthetizers, etc.)
Almost all Web pages are valid with regards to the technologies they are using, mainly XHTML, HTML and CSS. We are using different techniques (Markup validator, CSS Validator) to ensure this level of quality that we will present later in this paper.
HTTP is one of the forgotten children of the Web, because it's not directly viewable. For example, Someone will not notice he/she has received an old version of a page, because a content location parameter in HTTP has not been correctly implemented in a browser or not well defined on a server (See CHIPs and CUAP for implementation problems). We do care of HTTP on our Web servers.
And finally maybe one of the most simple facts: Stable URIs! Cool URIs don't change. ever which has a huge usability cost for the Web community. Stable URIs!!!
No future? No we have a past, a very important one almost as old as the Web and we are still able to rely on these documents which contains the knowledge of hundreds of participants. We have improved and developed the techniques which help us to keep a very high level of quality, even if not sensational visually. The understanding of the technology is shared at different level accross the Team but any person who knows how to write in a real Web authoring tool (HTTP PUT) can start to edit documents right away without having to be trained specifically.
The W3C system team has developped an important number of tools to help people editing the Web site. There are called the ,tools (pronounced comma tools). At the end of any URLs on the server, you can add a comma and hit the enter key, you will get a page with a number of tools like
Let's say, One wants to create a very simple template where the user will be able to enter the title of the document and the name of the author. An HTML authoring tool is the only thing, the template editor will need. A few
%%variable%% parameters at the appropriate places in the document are enough to define the template, which is saved in a
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> <head> <meta http-equiv="content-type" content="text/html; charset=utf-8" /> <title>%%Document_Title%% - QA @ W3C</title> <meta name="DC.Creator" content="%%Author%%" /> <body> <h1>%%Document_Title%%</h1> <address>Author: %%Author%%</address> </body> </html>
Each time someone will put ,new at the end of an URI, he/she will have the possibility to choose a template and finally edit the forms for the value of Author and Document_Title, in this case, and save the document which is now ready to be filled with content.
Caveat: Do not help once the page has been already written, though there is a possibility to define a system where you could watch URIs and apply transformation to this URI when the content has been modified. It's a project to explore and use the full notion of URI being just a pointer to a ressource.
As we have seen most of the content at W3C Web site is composed of static Web pages. But This information is structured with elements and class attributes in XHTML files, as you would structure content in a database. XHTML files are XML files and so can be manipulated by XSLT and XPath for example.
This article is the companion guide of a Talk given at the WWW 2004 Conference in New-York on May 19, 2004.