I have been getting more and more excitable about the potential of HTML5 as a future web standard. Rather than throwing away existing HTML (as XHTML2 will), it attempts the more �CSS like� approach of evolution. It’s being specified extremely well, filling in some gaps in HTML4.

Now, this is a year or so away (and browser support may remain tetchy). Interesting now, though, is that the draft re-specification of some HTML elements (i, small) could be applied now. It offers some ‘better practice’ that could really improve the whole experience of developing for the web.

Some of this �better practice� is an attempt at a backwards compatible (and search engine compatible) version of XHTML 2’s new h element for headings. The reason is because having separate h1 through h6 is, frankly, a really stupid idea. It’s a stupid idea that has really hit home in the recent years of mass-adoption of the Content Management System. If I want a subheading in my blog entry, I have to remember which bloody level the headings appear at. If I ever change my template, so that all my headings end up one level higher, or lower, then I have to edit all my content. Bad Thing�.

XHTML2 and HTML5 both have a generic section element, in which you can place a generic heading and the level of that heading is implied by the depth of nesting. In XHTML2, this element is h, in HTML5 the proposal is to reuse h1 (and choosing not to use the other �hx� elements), thus ensuring that the element is stylable in older browsers. Clever, eh?

The headings idea is the one that’s inspired this post. I’m interested in exploring the arguments against adopting the new practice of only using h1. Is there anything truly �harmful� in adopting such future-practices on HTML4/XHTML1? I dunno�