Re: Validity constraints on <section>

I agree we should think about tightening up the conformance requirements.


>but I think it would break too much existing content

nothing would be broken, there is no change in implementation, some stuff
(which is crufty anyway) may become non conforming if run through a
validator.

a good place to start is by looking at how it is used currently and
identifying where it is used unnecessarily/poorly etc

as a start here are 252 HTML5 pages using the section element: CSS is
applied to the pages to display section use
http://www.html5accessibility.com/HTML5data/section/section.html

PS have plenty more where that came from :-)

with regards

--
SteveF
HTML 5.1 <http://www.w3.org/html/wg/drafts/html/master/>
<http://www.paciellogroup.com/resources/wat-ie-about.html>


On 21 March 2013 10:30, Robin Berjon <robin@w3.org> wrote:

> On 21/03/2013 11:18 , Jirka Kosek wrote:
>
>> On 21.3.2013 11:02, Robin Berjon wrote:
>>
>>> The specification does have some advice about only using <section> for
>>> content that is meant to appear in the document outline, but given that
>>> the outline doesn't show up anywhere, that's not something that's ever
>>> likely to stop this drift.
>>>
>>
>> I always though that introducing semantic elements like <section> will
>> not be very useful for HTML. They will be misused as any other HTML
>> element. HTML is not rigid and semantic format like DocBook or DITA.
>>
>
> I beg to differ :)
>
>     http://alistapart.com/**comments/semanticsinhtml5#**325554<http://alistapart.com/comments/semanticsinhtml5#325554>
>
>
>  I've therefore been wondering: would it make sense to make section
>>> invalid if it does not have heading content as its direct children?
>>>
>>
>> We can even make this more strict and require heading content to be
>> first child of section.
>>
>
> I thought of that, but I think it would break too much existing content
> that is pretty legit, of the form:
>
> <section>
>   <div class='date'>1977-03-15</div>
>   <h2>Blah</h2>
>   ...
> </section>
>
> I don't think we need to constrain things more than needed.
>
>
>  Put
>>> differently, what are the use cases for a headless section?
>>>
>>
>> More sexy <div>? :-)
>>
>
> Stick to your <div>s, I'll bring the sexy.
>
>
> --
> Robin Berjon - http://berjon.com/ - @robinberjon
>
>