The Open Group and the Web Standards Project duly kicked off a partnership attacking what they say is the unfulfilled promises of browser manufacturers to support W3C standards, some of which are over two years old now (CI No 3,508). Vendor response to the so-called Open Browser initiative was tepid or non-existent, yet the two claim the extra testing and workaround development required to support the way browsers implement W3C standards differently, adds an estimated 25% to the total cost of web site development. Their goal is to encourage consistent implementation of W3C specifications. They are to provide test suites that enable developers to test content and client display environments. HTML 3.2 is supported well in most browsers, the two say. However no browser, they claim, would currently pass tests created for HTML 4.0 and advanced presentation features such as cascading style sheets (CSS) and document object model (DOM) more often referred to as Dynamic HTML (a vendor buzzword to describe many of these types of features). Vendors need to be convinced it’s a good idea to support these features consistently. After all users are not going to upgrade web server software and hardware if they cannot take advantage of new W3C functionality. Moreover as the browser becomes less relevant to vendors’ business models the need to lock users into specific implementations of new functionality should, in theory, decline. The Open Group already has a suite of tests that evaluates conformance to HTML 3.2. It’s looking for partners to help define, finance and develop test suites for HTML 4.0 and DHTML, including CSS 1.0. Its charter doesn’t enable to it go out and proactively test browsers and publish results. That’s what its partner, the 4,800-member Web Standards Project will do. The Web Standards Project is also seeking to hear from users that have experienced specific problems stemming from inconsistent W3C specification implementations. Ultimately it would like to create tests for client environments and for the web developer tools used to create web pages that incorporate HTML code generators.
Paying the price
Is there support from the content/browser community and W3C? The Open Group is a member of W3C. W3C people are on the Web Standards Project’s board. W3C says it supports any effort to promote its work and good implementation practices. Parsing bad HTML code is the worst thing that browsers do, W3C says. The browser vendors were lenient to begin with and now they are paying the price, it says. But the group also reminds us it makes recommendations. It does not set legally-binding standards (although the International Standards Organization is currently creating standards around HTML 4.0). W3C does not have an enforcement mechanism and does not proactively test for the correct implementation of its recommendations. However it does offer suites for testing certain functions. It does have a test for CSS 2.0 – the current implementation – and freely admits support for CSS 1.0 was not good. W3C also has a validator for HTML 4.0. W3C is not including any more presentation features in HTML – those have become CSS and the other DHTML functions – so there’s no visual testing component part of the validator. It’s also a created an Amaya browser as an example implementation of its recommendations but emphasizes this is not a commercial endeavor. It’s also working on a guide for vendors of user agents, a document which will promote accessibility and good implementation practices. It has no formal links with The Open Group or Web Standards Project on the Open Browser initiative.
Why are we paying?
Netscape Communications Corp said that while it supports efforts to promote good programming practices and the development and display of portable content, it has reservations about the Open Browser initiative. To begin with why are we paying all this money to W3C if they are not setting standards? it wonders. Not only is it fearful of propping up a dysfunctional standards group (The Open Group), it thinks it might end up having to support multiple standards. That’s because W3C’s recommendations would be subject to a variety of interpretations if the effects of their implementation were to be measured and compared. There would be differences of opinion about what behavior W3C specifies. Publicly Microsoft still refuses to comment, despite repeated approaches, although privately it has asked the two groups to show it commercial examples of the kinds of problems they are talking about. It was supposedly less than interested in examples that were provided. The two groups showed some W3C pages using DHTML viewed in Netscape that looked terrible, and equally terrible using Internet Explorer.