As I’m coming to understand it, it looks, at least on the surface, that the motivation for <section> is to simply provide a grouping of heading+content that a machine could extract or manipulate. As it currently stands, content on the internet consists of any number of headings, paragraphs, lists, etc., in virtually random order (from a machine’s perspective). Say I wanted to programmatically pull out the second heading with its content from an article. Currently, this would be an audacious if not insurmountable task. But, if each logical division of a piece of content was wrapped in a <section> element, I could jump right into the DOM, and get what I wanted with something as simple as “getElementsByTagName(’section’)[1]” (`1` here representing the second element in the array).

<section> seems mostly if not entirely devoted to machines (i.e. web applications), only holding value to end-users by proxy of the greater abilities it gives to web applications.

I like the “natural heading” idea of a section to separate it from a div. Considering a homepage, you can have sections for “news”, “banners”, “focus items” etc etc. The danger of the section element is that people would obviously want to replace the div.

What I don’t like about this article is that the div is put forward as an element needed for styling or javascript. The div is (and always has been) a generic structural element, and should still be considered as such. It’s probably best to keep using divs to structure content within an article or section element. I think the meta data of an article (author, date, whatever other info) should still be wrapped in a div, as it makes very little sense as a standalone blob of information.

So on the homepage you could have a section class=”newsList”, containing a list of articles, containing a header with a heading and div for meta data.

On a sidenote, the whole mess seems a result of the lack of structural understanding. Divitis is not only a disease, the counterreaction is a disease by itself. At least tags like section and article might change this again.

It is true that not everyone will use HTML5 correctly, just like the majority of existing content won’t be updated to HTML5. However, disabled users are picky about the websites they use and if HTML5 proves to be as useful as we hope, they will most likely prefer to access sites written in it. The big boys like Amazon and eBay can implement it and make it massively easier to navigate these large, complicated websites. And for the other small sites that don’t, the screen readers can fall back to the technology that they’ve been using throughout this century.

In short, it is nothing but a benefit to the users at a small cost to the developers.

As for the element confusion, remember that Web Developers need to read manuals too. In the most part, section can be completely ignored and replaced with proper heading elements, the outliner will create the sections itself.

thanks for the comments, we did have the notes for what is/isn’t allowed in but I accidentally commented them out when making some changes recently, thye’re no restored.

Ta

Rich

r/e your blind friend: everyone is different. I have blind friends who scan a page using “L” to go to lists (eg, go round navigation etc), “T” to go to tables, etc. Most screenreader users rely on a proper hierarchy of headings to navigate a page with (See a video of a blind web dev doing this: http://www.youtube.com/watch?v=AmUPhEVWu_E). Having a “flat” structure makes this impossible.

The sectioning content and outlining algorithm simplifies the ability to properly structure headings, particularly with syndicated content.

@alohci is right: the outliner is not a validator. But I agree that “untitled section” isn’t helpful. There is a bug filed that the outliner should generate something meaningful – so when it encounters nav, which is sectioning content, it should show the top-level heading within the nav element (if there is one) or say “(navigation)” instead of “untitled section”. It’s perfectly legitimate not to have a heading in navigation, footer or aside.

In my diagnosis, the main cause of sectionpox (the new div-itis) is people want to put it around “main content”. I think that the best reason for adopting a main or content element is to stop people misusing section. But I doubt that the working group will sanction a “safety tag” as it were.

@Don asked “why? I can write scripts, algorithms and related stuff to apply to a div with class=”section” as easy as it is to a section-element . Sure, not everybody will use the same class-name, but then again, not everybody will use the section element correctly.”

On Saturday I was listening to Jamie Knight, a sighted web developer with autism explaining that he sometimes uses a screenreader when stressed as it’s easier than reading. On a blog post on my own site, I wrote

He made his own screenreader that listens out for certain ids or class names like “navigation”, “menu”, “sidebar” so that his screenreader can skip the peripheral information on repeated views.

With the “baked-in” elements like nav, footer etc, this would be achievable with much greater precision.

http://www.brucelawson.co.uk/2009/standards-next-cognition-and-accessibility/

Having real elements takes away the ambiguity of different naming conventions and means that programmatic user agents – like search engines or screenreaders – can “understand” content better.

@thinsoldier – you asked “yes or no”. No.
Most people would use sections a lot; to wrap up areas of code to disambiguate where sections start and end; to surround areas demarcated with headings inside articles that were to be syndicated, so that the sites that import that content can apply CSS that works for the importing sites’ outline and so the outlining algorithm works correctly.

Regarding non-blog sites, for example http://portal.opera.com or iGoogle http://www.google.com/ig, I would mark up each “lump” of content as a different section because, well, each is a different section.

Cool. Yea I wasn’t really thinking of the outliner as a validator but it’s still strange to “warn” me of an untitled section even though it’s really not warranted.

Thanks for the info.