Weblogs: Web Standards

WHAT about Internet Explorer?

Thursday, June 10, 2004

The W3C workshop into Web applications and compound documents has sparked off quite a discussion in the web community. The aim is to have a consistent foundation for building web applications - this seems to be the next evolution of the web.

Evolution or revolution?

Already interested parties are fracturing. The big companies have essentially given up on the browser prefering to adopt a platform founded on XForms and SVG. They've effectively killed any growth of the browser. They are looking at a clean slate, which is more revolutionary than building upon the web in its current form.

A group of developers headed by Opera and the Mozilla Foundation have coalesced into a group called Web Hypertext Application Technology (WHAT) to establish a specification around enhancing the current browser technologies to add in the extra bits for web applications to be comfortably feasible.

XForms and SVG

The big companies are all focusing on XHTML, XForms and SVG as a basis for web applications, yet prefer a profile or basic subsets. Of course, this raises the concern that if subsets and profiles are used, isn't that a clear sign that these specifications are already too unwieldly for normal use?

The driver toward XForms is the political point that there were more implementations of XForms on its first day as a recommendation than all the other recommendations on their first days. For some reason this has been translated as being an indication of high adoption. Yet XForms is hardly visibly on the Web, and no sign of real-world support.

SVG only really has a web presence in the Adobe plugin in IE6. Hardly the basis of wide adoption considering the high preponderance of Flash usage in the same browser as well as other browsers. I'm only aware of one major site using SVG, and that's Jonathan Chetwynd's peepo which is a resource for learning disabled people.

It looks like the consortium of companies is more interested in building a new independant infrastructure over the Internet.

Evolving the browser

The Mozilla Foundation and Opera initiative - WHAT - is guided by the veritable whos-who of browser development. Ian Hickson, David Hyatt, Brendan Eich, Håkon Wium Lie, David Baron - all well recognised, highly respected individuals.

The commitment to the browser platform is critical. The Web is an everyday resource. To start a new infrastructure that is independant of the browser is a clean slate start from scratch. And without an existing base of users to build from that will drastically reduce the speed of adoption.

The work that WHAT have undertaken is greatly needed now. The main objective - I would guess - is to have something implementable, up and running as well as stable enough well before Longhorn is in a position to influence direction. Longhorn is expecting to ship in 2006, so a feasible web application approach needs to be mature in 2007.

Internet Explorer 6

From the looks of it, Internet Explorer 6 has no future within Microsoft whatsoever. It is a dead browser with a 90% usage rate. Over the course of the next two years that share will gradually drop as people start flocking to better alternatives in Mozilla, Firefox, Safari and Opera - mainly because of features they will be offering.

Internet Explorer 6, as an end-of-line product, will not be able to keep up with the innovations introduced by the other browsers - notably XHTML, XHTML2, CSS Level 2.1 and Level 3. Even the introduction of RSS Aggregators won't happen inside IE6.

It looks like other browser vendors are stuck with supporting the competitor IE6. Because of its massive user-base, innovations introduced by other browsers cannot have traction on the web unless there is support or compatability with Internet Explorer 6. Bugwise compatability with IE6 is essential for web applications to gain any support for the next few years at least.

Longhorn

The alternative is worse. Wait for Longhorn, which offers XAML and Avalon - a closed locked-in solution for web applications which may, or may not, deliver web applications in 2007.

The web needs a web applications standard / guideline now. Waiting for a proprietary, patent-encumbered solution is not a viable alternative for the Web. Building a brand-new solution using XForms and SVG is not viable for the web - the delivery scales and uptake will be overshadowed by a pending Longhorn.

If Longhorn and the W3C/SVG/XForms approach compete head-on-head, Longhorn will win eventually - marketing money beats sense everyday in the real world. The only decent option is to have something open-standards based working on the web before Longhorn is launched.

My perspective

I am in two minds about browser-based web applications. On one hand I realise the importance of evolving within the current browser, so web applications written using today's technology of HTML, CSS and Javascript is something I understand. I also much like Mozilla's approach to web applications, leveraging the use of XUL, XPCOM and XBL. Sadly, seeing XUL work in IE6 is an extremely long-shot.

For browser-based web-applications, what would be great is for browsers like Opera to support the XmlHttpRequest object (Safari have recently done it) - that would be the first piece of the jigsaw puzzle.