The use and misuse of these elements have been discussed here fairly often with an especially vociferous contribution on the negative side from deathshadow. For now, I think he is mostly in the right, at least from a practical standpoint. In the long run, I think this structure will prove to be an improvement over the present loose structural ties among different sections (not the <section> element). For now, though, there are issues. I will quote Ian Hickson, editor of both the html5 and WHAT specifications, from a longish thread in the WHAT maillist: These new specs are apparently with us, whether we like them or not. That said, we need to be aware of current issues and work around them while we learn to use them correctly. cheers, gary
To me it just pisses on the entire reason headings HAVE numbers -- maybe it's my background from writing technical documents, where heading depths serve a purpose in creating a logical document structure -- maybe it's because TBL created them in HTML originally to serve that same purpose (to the point he pretty much assumed they didn't need explaining to his original audience of people writing reports, technical documents, project papers, etc, etc...). A top level heading -- H1 -- is the heading under which EVERYTHING else on the page is a subsection. It's that simple -- it's why in professional writing you would only ever HAVE one H1... H2's are the start of subsections of the H1's,... H3's are the start of subsections of the H2's and so-forth. They divide the page into sections with a heirarchy, fanning you like a tree regardless of how they are nested in other tags. It's why using a H4 or H5 or H5 before you even HAVE a H3 on the page is gibberish halfwit bull -- no matter how many developers still do it. Likewise a horizontal rule-type element most always is a divider where a heading is unwanted/unwarranted -- like before citing your references or a headerless footer. It's moronically simple and I don't understand how throwing more tags at it is going to fix people not having the brain cells to rub together to use them properly. ... and that's WHY SECTION, ARTICLE, NAV, and most of the rest of the HTML 5 halfwit bull is little more than code bloat, encouraging people to put pointless allegedly semantic containers in their markup for no good reason. RESETTING the count to H1 so you have multiples of them on the page just makes heading navigation gibberish (though admittedly the ONE browser to get that right doesn't do it any more since they turned it into pathetic crippleware), and to what end? Satiating the needs of data scraping asshats that most site owners try to keep out? It's the same idiocy as microformats, aria roles, and a whole host of other stuff that on websites for visitors serves no legitimate purpose. What ever happened to "write for the user, not the engine?" Naturally, don't even get me STARTED about ASIDE, PROGRESS, CANVAS... and I like CANVAS, it just has no business as a scripting only element HAVING a tag. That's NOSCRIPT's job! Of course, there's also XHTML 2.0, which was doomed to failure from the start BECAUSE of idiocy like this, letting people just make up their own tags, and on the whole being as much of a train wreck of a "specification" as HTML 5. Of course, I'm using the term "specification" in the loosest sense since having an engineering background, specifications should be authoritative, not documentative; which is why crap like allowing EMBED into the specification, creating pointless redundancies that are the opposite of what HTML 4 Strict was about like AUDIO and VIDEO (might as well bring APPLET, MENU and DIR back dumbasses), and the loosening of structural rules that can only serve the purpose of catering to the whims of the nudniks still vomiting up HTML 3.2 and slapping 4 tranny on it. Now they can slap 5 lip-service around the same full-assed coding practices, bloat it out even more with extra wrappers for nothing, and then slap each-other on the back over how "modern" they are. If your idea of "modern" is the worst of 1997 pre-Strict coding practices, re-introducing redundancies, presentational thinking and all the other things we were SUPPOSED to be getting away from. AND FOR WHAT? Data scraping scumbags, sleazeballs vomiting up broken outdated methodology and the W3C just shrugging it's shoulders and going "Oh well!"?!? If this is the people in charge of the standards idea of the "future", well... I guess they're just doomed to repeat the past. It's a bunch of grade A farm fresh manure I still cannot fathom how anyone can use by choice, much less defend except out of total ignorance. Here's hoping we have a HTML 6 that deprecates all these stupid halfwit redundancies and gets back to ACTUALLY using logical document structure to build a page... but no, given the path we're heading down with this manure I'm half expecting them to announce VERB, ADVERB, NOUN and PRONOUN tags any day now -- as that's how stupid this nonsense is getting. Sooner or later, you're gonna have to let your CDATA do it's blasted job! Or as Dan Schulz put it long ago, (and I parrot frequently) "When people can't be bothered to use the existing tags properly, throwing more tags at it is not the answer!"