- From: Cameron Jones <cmhjones@gmail.com>
- Date: Mon, 10 Sep 2012 21:21:50 +0100
- To: Steve Faulkner <faulkner.steve@gmail.com>
- Cc: HTMLWG WG <public-html@w3.org>
On Mon, Sep 10, 2012 at 8:59 PM, Steve Faulkner <faulkner.steve@gmail.com> wrote: > hi Cameron, the same arguments you are using for <maincontent> could > be used for role=main Not so, as i said at the end of the last mail: "The great thing about ARIA is that it requires people to be explicit, and with the intention of providing accessibility" The intent is to provide accessibility, so they wouldn't misuse the annotation. If their intent is not to provide accessibility but just define some stylistic hook, they can just use a <div> and class names which are not overloaded. > >> i don't think that the majority of people read specifications, or pay >> attention to them even if they do. Far more people will use it more >> liberally than ever intended because it "makes sense to them". > >> I would expect to start to see a "main content" wrapper inside almost >> every element, essentially as a replacement for <div>. We don't need >> another meaningless <div>. >> > > considering the correct usage of ARIA role=main is so high some people > must have read the spec or read some source which correctly > interpreted the spec and people get the idea, its a simple one that is > hard to misinterpret kinda like <maincontent> The people who have implemented ARIA are not your standard HTML author. The reason most implement ARIA is because they are passionate about web accessibility or are mandated to by policy or regulation. Therefore your statistics are not representative and must be taken within that limited context. > > your statement 'Far more people will use it more liberally than ever > intended because it "makes sense to them"' is mere speculation it > would further discussion if you could provide some analogous data to > back up your statements (as I have). > Yes, it is speculation but it's hard to provide statistical evidence for something which doesn't exist. Instead, i provide insight into the nature of man born from experience and frustration :) Other than the meta-elements, there are few which impose the type of limited use you are suggesting. This, i postulate, is ripe for misuse. Since the benefit of the element is in its restrictive application, the overuse will negate any practical benefit. >> The great thing about ARIA is that it requires people to be explicit, >> and with the intention of providing accessibility. It won't be misused >> because that would defeat the point of specifying it in the first >> place. > > the same could be said of a HTML and the <maincontent> element: > > "It won't be misused because that would defeat the point of specifying > it in the first place." > > regards > SteveF > This could make perfect "sense" to someone who didn't read the spec, (substitute the example elements with any other): <p><span>First three words</span> <maincontent> the rest of the paragraph</maincontent><p> The intent here, which is the overall same intent as the whole reason for the element, is to avoid using a css class name and instead fallback on the inferred semantics of the element name. Thanks, Cameron Jones
Received on Monday, 10 September 2012 20:22:18 UTC