HTML5 – reinventing the wobbling wheel?
Jan 10 2009
A recent post by Jeremy Keith inspired me to catch up with the latest experiences on real-life usage of HTML5 out there – and some related consideration whether it’d be about time to switch from XHTML to HTML5.
Your mileage may vary
Despite the over-a-decade-away “deadline” of HTML5, scheduled to have the state of a proposed recommendation by 2022 (according to Jeremy Keith that means “it’s been finished for ages and now it’s 100% supported in at least two implementations”), the working draft of HTML5 appears to be already pretty well-understood by the major browsers. Even for IE 6 and 7 the JavaScript community has already found some workarounds to make it work.
But reading, among others, Eric Meyer’s experiences from designing the AEA 2009 website in HTML5 it appears that applying the newly-won semantic possibilities come at a high price regarding browser incompatibilities and the related tweaking of Style Sheets that we just about got rid of thanks to the standard-support of the modern web browsers.
Take some, leave the rest
An Event Apart is not the only live experiment with HTML5 out there, and what seems to unite them all is that, in the end, the number of elements that can already safely be used is very limited. Sure, as Bruce Lawson proves with his HTML5 test page your browser – after some tweaking – will be able to display the page as intended, but as soon as a more advanced layout comes into play, the time spent on hacking and fixing the CSS is (in a production environment) better spent otherwise.
Nevertheless, these public experiments are a valuable contribution to not only the development and promotion of HTML5 – and semantic web technologies in general – as well as it’s support in browser software. They are hopefully able to create important momentum towards a more general adaptation of common semantic rules, even while we are still using either version 4.01 or one of the XHTML specifications: The use of HTML5′s terminology for the naming of markup elements could well be as valuable as the spreading use of microformats. Thinking of it as a web-wide POSH microformat, this would already enable an all new dimension in the evolving Semantic Web. Or as Jeremy Keith puts it:
“Even if you decide against writing in HTML5 itself, at least consider using those HTML5-inspired class names for your structural naming convention where appropriate.”
Is HTML5 the solution after all?
John Allsopp discussed what I consider an issue well worth contemplating about in a recent article on ALA. HTML5 will not only give us new ways of embedding semantic information into HTML documents, but at the same time restrict it again – on a higher level.
Personally, I find it an appealing idea to use standardized XHTML attributes for semantic markup, rather than re-inventing the (wobbling) wheel. The new elements and syntax of HTML5 introduce a very limited set of semantic tags and will cause major backward and, as John Allsopp predicts, forward compatibility issues for years to come (the experiments referred to earlier give a good taste of what to expect).
After all, using XHTML attributes for semantic markup is already an established technique; think microformats.
