Web 2.1: Making it whole again

Attended the W3C Technical Plenary, in Cannes, France. W3C has played the United Nations of Web standards for a decade now. In that period they have created 85 recommendations (finished specifications). In the works are 2 proposed recommendations, 28 candidate recommendations, and 111 working drafts, 25 of which are in last call (actually a few more, but I dropped some of the more process-oriented documents from the list).

The World Wide Web comprised three standards initially, URL, HTTP, and HTML. The URL, the location of a resource, was the crucial one. It created one standardised way of describing an address. The HTTP transport protocol allowed any machine to pick up or send to that resource, and HTML like this included the URLs as hypertext links. WWW built upon existing Internet standards as well as predecessors like Gopher, FTP, and Enriched Text.

Even so going from just three to a couple hundred standards, not including programming languages and frameworks that are standardised elsewhere, has made the web undertaking less seamless. Every standard has been created to solve a particular problem, or in some cases to cause a particular problem, but often with less interest in making it a well-functioning whole. W3C rules states that new standards should build upon existing ones. This is good in that you can reuse what you have already got and bad in that each new specification will be on top of a huge stack of existing ones, so new specs are going from baroque to rococco before they even get started.

One of the terms that cropped up last year was Web 2.0 (for which we can blame O’Reilly). If there is a Web 2.0 it certainly is in first working draft. However the interesting, in fact the necessary, project is to make all these standards work with each other. Not only “Make it work” but “Make it work well”.

This will be slow, methodical work with little hype attached to it and it will take time. Realistically it will take another decade. Most, if not all, of the hundred-some recommendations and recommendation-wannabes will have to be reworked. This process, analogous to the development of CSS 2.0 and CSS 2.1, will not add much features to speak of but it will add quality and ease. Call it Web 2.1, the interoperable web. The process has started, and it is causing friction of the good kind. For instance SVG has been used for specific purposes in the past, now that it is getting used for the broader web audience issues that didn’t matter matter before matter now. One of the integration tools is the Compound Document Framework that actually tries to integrate different formats into one. This will not solve the problems but will show where most of them are.

In the five years I have been involved with standards I have seen hundreds of demos on how well a particular method solves a particular problem. Now what remains is the problem of how to solve that particular method.


Your email address will not be published. Required fields are marked *