Designs on the Web
Stuck in The Mud
So how have these standards moved on, and what should state-of-the-art web design be offering today? Visit www.w3c.org for an answer that’s as astonishing as it is depressing. With CSS1 released in 1996 and CSS2 in 1998, you might expect CSS7 to be nearing completion, but in fact there’s no clear date yet for CSS3 – and it’s the same story all round. W3C work on PNG has stopped, while the current SVG and SMIL standards are stuck at 1.1 and 2 respectively. It’s not just the web-authoring packages that have run out of steam but the underlying technologies themselves, and it’s not Macromedia and Adobe’s fault, nor the W3C’s. The fundamental problem that’s stalled development is one that none of them can do much about: browser support.
There’s no point in standards that run way ahead of practice, which is why CSS3 focuses on breaking down the specification into achievable and testable modules, while CSS 2.1 was content to set a minimum standard based largely on current browser practice. Equally, there’s no point in authoring applications adding functions the end user won’t be able to see. In fact, it’s worse: web authors can’t realistically take advantage of features like CSS-based positioning or SVG-based navigation until they’re sure the overwhelming majority of users will see the page as intended. Web practice can only move ahead at the speed of the slowest ship, so one browser can effectively halt the progress of all.
It’s a situation we’ve been in before when the original take-up of CSS1 was blighted by the half-baked support in Netscape Navigator 4, but who’s the culprit now? Look at a site such as http://nanobox.chipx86.com that compares current browsers’ CSS 2.1 compatibility and you’ll see that none of the main contenders gets a clean bill of health, but Firefox and especially Opera score relatively highly, while the browser with much the worst record is the most popular, Internet Explorer 6 (IE6). While you might consider designing a rich page that works in the dominant IE6 but falls over in little-used Opera, you’re certainly not going to do the reverse…
So why is IE6’s CSS support so poor and, come to that, why doesn’t it natively support PNG transparency, SMIL or SVG? The answer is simple: IE6 is ancient. It was first released in 2001 when technologies like SVG and SMIL were relatively new and seen as niche interests, so no-one expected instant browser support. By comparison, CSS was long-established, and there IE6 actually led the way, correcting IE5’s disastrous misreading of the CSS box model in standards-compliant mode and generally providing the best support then available. Indeed, it was IE6’s marketdominance that allowed CSS to finally take off. Now, though, IE6 is fast approaching its fourth birthday and expectations have moved on, while the program has become outdated.
The next obvious question is why hasn’t Microsoft released a new version? Having seen off Netscape and won the browser wars, why would it stop development? The answer, of course, is that it hasn’t. Bill Gates’ long-term aim, once he’d recognised and then seen off the threat that Netscape and Java represented, was to integrate the Internet – which is only a network after all – into the heart of Windows at OS level. Following the launch of IE6, the previously separate IE and Windows development teams were merged, and in May 2003 it was announced there wouldn’t be another standalone release of Internet Explorer; instead, all browser development was folded into Microsoft’s work on Longhorn, its next version of Windows.