All modern browsers support HTML5, although feature support is not necessarily complete.
HTML5 is a response to the observation that the HTML and XHTML in common use on the World Wide Web is a mixture of features introduced by various specifications, along with those introduced by software products such as web browsers, those established by common practice, together with many syntax errors in existing web documents.

I’d read that blog post of yours, and we are basically in agreement.

My main point is that one should consider the effects of a particular semantic value. The end result should be usability and accessibility. And perhaps parsability (by search bots and server side scripts). Without such effects semantics will stay an academic exercise, of no further benefit than a span-element.

When I read about the nav-element, it basically seems like something a user would like to be able to find quickly or quickly pass over to find more relevant information, i.e. it serves the same purpose as skip links do today, for vision impaired users. In addition to that, I think browsers could implement shortcuts (or gestures) that will enhance the user experience also for “normal” (or at least “power”) users. Ergo: The question “what kind of benefit will this bring to users of the site?” provides the best guidance about how to use a specific feature.

Today that basically means add ARIA-landmarks, continue to use skip-links. Add nav-elements for personal sites.

When ARIA-aware OS/browser/AT combinations has become the norm – and I’m not talking about fake implementations like in Safari 4 (that does not expose ARIA properly to the OS) – we can begin drop skip links. I suppose that is 2-3 years into the future, when today’s cutting edge browsers are being phased out. (I also predict faster uptake on new browsers in the future…) By then we have native support for HTML 5 structural elements in most “A-grade” browsers, I’d presume. That means phasing out a *few* ARIA features, or relegating them to those niches where HTML still is not enough.

In principle I prefer native elements to attributes, on the “less cruft” principle. Otherwise we could just have one element and code like this ( don’t know what HTML you’ll allow. I’m trying pre and code here):

<elem class="div">
<elem class="p"><elem class="em">No
way!</elem>, she said...</elem>..
</elem>


Parts of HTML 5 are usable today for corporate web sites though. How many sites do not add Flash through a JavaScript that actually inserts an embed-element? They are de facto using HTML 5, even though the doctype might say something else.

I’ll provide this link to a blog post of mine, as it is relevant to that last point.

http://itpastorn.blogspot.com/2009/06/validation-and-doctype-myths-and.html