Re: Validity constraints on <section>

2013-10-17 13:40, Steve Faulkner wrote:
> the change does not require a heading for every section. that would be 
> MUST, what it says is SHOULD, but there may be cases , such as you 
> describe where a heading does not make sense.

A SHOULD requirement still requires authors to carefully consider the 
implications etc.

> The change is designed to encourage authors to think about whether it 
> is appropriate to use an unlabelled section (as it has stated in the 
> spec for some time that sections are for content that would be listed 
> in the document outline) take this example 
> (http://www.awardwinnersonly.com) which contains 393 section elements 
> (sourced from from a comment in this post 
> http://blog.paciellogroup.com/2013/10/using-html5-section-element/ ) 
> This is an example ,of many, where the section element is being used 
> contrary to the spec.

If the real problem is misuse of <section>, shouldn’t that be addressed 
directly, rather than saying something quite different (about things 
that SHOULD appear within a <section>)?

I don’t see how the example page violates the HTML5 CR. On a page that 
provides information about many books, each book can be regarded as a 
topic, so <section> looks quite OK to me (when I take the <section> 
element seriously in the first place).

Besides, although it *would* here make sense to turn the book titles 
(now classed <div>s) to <h3> elements, this is quite independent of the 
<section> issue. Would you say that it would make the use of <section> 
conforming here?

It would be a different matter if <section> were defined simply as a way 
of specifying the scope of a heading element (to be used when desired to 
make it explicit, or when the scope should differ from the scope 
inferred by certain rules).

-- 
Yucca, http://www.cs.tut.fi/~jkorpela/

Received on Thursday, 17 October 2013 11:31:35 UTC