17975 2012-07-18 07:26:56 +0000 what about if the <body> IS an article? 2012-12-30 02:29:23 +0000 1 1 1 Unclassified WHATWG HTML unspecified Other other RESOLVED FIXED P3 normal Unsorted 1 contributor ian giorgio.liscio ian jackalmage mike contributor oldest_to_newest 70411 0 contributor 2012-07-18 07:26:56 +0000 This was was cloned from bug 14540 as part of operation convergence. Originally filed: 2011-10-22 04:08:00 +0000 Original reporter: Giorgio <giorgio.liscio@email.it> ================================================================================ #0 Giorgio 2011-10-22 04:08:44 +0000 -------------------------------------------------------------------------------- hello, what about if *the entire <body> IS the article*, <body> <article></article> </body> I think that the difference between <nav|aside> and <article> is that nav and aside can't be primary content, the main section, of a document, article yes so maybe <article> should not be a sectioning element, or you should provide a way to mark the <body> as article p.s. I don't want to break up your hard work of years, I'm just submitting my opinion. I hope this helps! ================================================================================ #1 Tab Atkins Jr. 2011-10-22 04:25:58 +0000 -------------------------------------------------------------------------------- In this case, just don't add a wrapping <article> tag. ================================================================================ #2 Giorgio 2011-10-22 05:24:34 +0000 -------------------------------------------------------------------------------- so if I can not mark a page as article, makes <article> totally useless by the way, I've extended my thought *please read carefully because this can be interesting and my English is not so good* One of the problems of html4 was "site title" vs "document title" I think html5 should introduce something that resolves this issue. As I said, my english is not so good so I will try to explain with some code. the homepage: <body> <h1>bestSPORTWEAR.org</h1>...some nav... ... </body> product page: <body> <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> </body> now, users want to brand their sites with the same header of the homepage, so, according to html5 spec: product page: <body>  <aside><h1>bestSPORTWEAR.org</h1>...some nav...</aside>  <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> </body> but examining this: <body>  <aside><h1>bestSPORTWEAR.org</h1>...some nav...</aside> <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> <aside> <h1>How to wash the t-shirt</h1>  </aside> </body> seems that "bestSPORTWEAR.org" is at the same level of "how to wash the t-shirt" how bots should interpret this? and accessibility tools? it is nonsense. I examine another approach: <body> <h1>bestSPORTWEAR.org</h1> <section> <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> <aside> <h1>How to wash the t-shirt</h1>  </aside> </section> </body> now the outline is correct, but bots and accessibility tools can't distinguish the document title and the site title a new element can be the answer: (It's just an example) <body> <parent> <h1>bestSPORTWEAR.org</h1> </parent> <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> <aside> <h1>How to wash the t-shirt</h1>  </aside> </body> the outline will still be: bestSPORTWEAR.org A red NIKE t-shirt How to wash the t-shirt but the document title is now hyper-contextualized - bots can identify the entry point of a page (the document title) - bots can distinguish between the document context (the site or something in the site) and the document's related contents - users can easily provide meaningful outlines without doubts another solution can be an attribute, instead of introduce one element <body> <h1>bestSPORTWEAR.org</h1> <p></p> <article entrypoint> <p></p> <h1>A red NIKE t-shirt</h1> <p>A fantastic nike t-shirt with flames and skulls</p> <aside> <h1>How to wash the t-shirt</h1>  </aside> </article> </body> ================================================================================ #3 Giorgio 2011-10-22 05:33:30 +0000 -------------------------------------------------------------------------------- <body> <h1>excellentSPORTWEAR.org</h1> <nav> Tshirts, Tweatsuits, Weight-lifting tools </nav> <section> <h1>Tshirts</h1> <nav>Red ones, Blue ones, Lime ones</nav> <article entrypoint> <h1>Some cool NIKE red Tshirt</h1> <p>description</p> <aside> <h1>How to wash it</h1> </aside> </article> </section> </body> ================================================================================ #4 Tab Atkins Jr. 2011-10-24 02:09:31 +0000 -------------------------------------------------------------------------------- There's a good chance you don't need both of those <nav>s. (Rule of thumb: should it appear in a separate menu titled "Main Navigation for the Page"? If so, make it a <nav>; if not, don't.) Otherwise, this is a perfectly fine structure. There's no need to indicate the "entrypoint" - the outline is already fine as it is. ================================================================================ #5 Giorgio 2011-10-24 02:44:20 +0000 -------------------------------------------------------------------------------- > There's a good chance you don't need both of those <nav>s. (Rule of thumb: > should it appear in a separate menu titled "Main Navigation for the Page"? If > so, make it a <nav>; if not, don't.) > > Otherwise, this is a perfectly fine structure. There's no need to indicate the > "entrypoint" - the outline is already fine as it is. hi, please double check http://www.w3.org/Bugs/Public/show_bug.cgi?id=14540#c2 is fine for us, but with the point of view of an accessibility tool? the document is called "<h1>bestSPORTWEAR.org</h1>" but this is the site title, not the page title this is not so "hyper-textual" seems that a "document" is totally standalone, but it is not, it is placed in a context ( the site, the presentation, the chapter list) and there should be a way to distinguish context from content ================================================================================ #6 Tab Atkins Jr. 2011-10-24 06:28:49 +0000 -------------------------------------------------------------------------------- Since most sites on the internet have exactly that structure, I expect that screen-readers, bots, and other things that consume markup directly can handle that structure just fine. If they couldn't, most of the internet would be inaccessible. ================================================================================ #7 Ian 'Hixie' Hickson 2011-12-09 23:39:27 +0000 -------------------------------------------------------------------------------- I should add an example that explicitly covers the case of a one-article page vs a multiple-article page. ================================================================================ 80693 1 ian 2012-12-30 02:29:13 +0000 I've added some text to the spec that basically says that you are allowed to include the <article> in this case, but don't have to. The site-title vs document-title thing is a bit of a red herring in practice. There's no "site" concept in HTML, it's always about the document. It's just that there's a document title, and then a document subtitle. You can use a single mostly-page-wide <article> for that, or you can use <hgroup>, or <section>, or <h2>, or pretty much any other solution that makes sense. 80694 2 contributor 2012-12-30 02:29:23 +0000 Checked in as WHATWG revision r7608. Check-in comment: Elaborate on <article> in one-article documents, in case of doubt. http://html5.org/tools/web-apps-tracker?from=7607&to=7608