[html5] Using <section> and <h1> ? Theoretical?

[Jukka K. Korpela]

2014-05-15 10:06, Bruce Lawson wrote:
> On 14 May 2014 20:29, Jukka K. Korpela <jukka.k.korpela at kolumbus.fi> wrote:
>>> But yes, to support tools that don't yet support <section>, one should use
>>> <h2>-<h6>. That's why the spec allows this. It was part of the original
>>> design of <section> and friends, almost 10 years ago now. :-)
>> And <h2> and friends have worked well even longer. The question is really
>> why the use of <h1>, relying on support to <section>, is allowed at all (and
>> often seen as recommended and “more logical”).
> Is it often "recommended and seen as more logical"?

It seems to me that advocates <section><h1>... favor it as more logical. At
http://www.whatwg.org/specs/web-apps/current-work/multipage/sections.html#the-h1,-h2,-h3,-h4,-h5,-and-h6-elements
the example is first shown with the classic style that uses <h1>, <h2>, 
<h3>, then in the new style, and last in a mixed style “for 
compatibility with legacy tools”.

I think this suggests that sections can be implied or explicitly marked 
up with <section> elements, but if you choose the latter, then the 
primary approach is to use <h1> inside a section independently of 
nesting level (and, presumably, to make *all* of your sections 
explicitly marked up). Then the use of <h2> etc. is just allowed “for 
compatibility”.

>   In the book I
> co-wrote, and every other book or presentation I've seen, the use of
> <h1>..<h6> + sectioning elements is recommended.

People who write books tend to be more practical than people who write 
specifications and draft specifications.

>   I know of no-one who
> recommends the use of h1 all the time; even if the presenter/ author
> doesn't care about lack of AT support for the outlining algorithm, it
> makes writing x-browser CSS to ensure that all "logical" <h2> look the
> same truly evil.

The issue is why <section><h1>... is used at all, given the practical 
problems with it and lack of benefits except assumed ease of editing in 
some cases.

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