W3C home > Mailing lists > Public > public-html@w3.org > September 2007

Re: [html] Semantics of "aside", "header", and "footer"

From: Robert Burns <rob@robburns.com>
Date: Thu, 6 Sep 2007 17:34:49 -0500
Message-Id: <7E71FFF9-C5AC-454F-A248-AD11AD7CEF71@robburns.com>
Cc: public-html@w3.org
To: Jens Meiert <jens.meiert@erde3.com>

Hi Jens,

On Sep 6, 2007, at 12:03 PM, Jens Meiert wrote:

> (Apologies for recent WG discussion and collaboration absence; I  
> moved.)
> I need to bring up the "aside" [1], "header" [2], and "footer" [3]  
> elements again, no matter that they've been discussed a few times  
> yet. I'm still not convinced that their names are very appropriate  
> as they seem to be too "presentational" and almost meaningless.

While I agree these elements have quite presentational names, I think  
they also signify a common semantic that has some strong  
presentational implications. That is, for screen presentation, we  
quite often see a document broken down into a header, footer,  
sidebars (aside) and contents (section or article). I"d like to see  
some of this handled more from the CSS  side, but I'm not sure if  
we're there yet or if it will ever eliminate the need to markup these  
important pieces of the document.

> I'm very interested in how you judge that and whether more WG  
> members see those as an issue, and beside asking for your pick, I  
> may illustrate the problem by at the same time proposing alternate  
> names and slightly different/complementary semantics like:
> * "auxiliary"/"aux" or "supplement" instead of "aside"; quite more  
> generic and probably the most debatable suggestions here, but also  
> less related to the element's presentation.
> * "identity" instead of "header"; I know that more people gripe  
> when it comes to "header" and "footer" (yet with IDs and classes of  
> the same names which seems reasonable to me), and observing people  
> who look for alternatives and also doing a "heuristic analysis" of  
> current use, "identity" seems to be a descriptive and in many cases  
> sufficient alternative for "header".
> * "about" instead of "footer"; again, it's just a hasty suggestion,  
> but the situation is similar to "header", and "about" already  
> appears to be an improvement, too.

My view is that simply changing the names of the elements cannot   
alone  change them into non-presentational elements. More  
importantly, I think we need to think about the semantic contents of  
these parts of the page more systematically. For example the  
auxiliary columns of a multi-column layout often does information  
auxiliary information or site-wide or collection wide information as  
the name you chose suggests. However, it still represents an  
important semantically distinct section of the document, whatever we  
call it.

For the header and the footer I think more needs to be done than  
simply changing the names of the elements to be more semantic.  
Instead the focus should be on providing better semantics of the  
types of metadata that might end up in the header or footer of a  
page. We might want to look at Atom and RSS to see if there's any  
other metadata elements or attributes that might b e worth defining  
for HTML. Or alternatively should we just try to encourage the use of  
Atom to meet such needs of authors?

Some possible suggestions:
   adding a 'copyright' or a 'rights' element to express the dates/ 
years and the copyrights ownership of document
   adding a 'pubdate' element that shares many of the properties of  
the 'time' element but means specifically the publication date for a  
   adding 'scope' attributes to the 'link' and 'meta' elements to be  
able to set an IDREF that scopes those pieces of metadata.
   adding a 'h' or 'heading' element to specify a generic heading  
for a document, section or subsection.
   adding a 'subh' element to be included specifically within the  
'heading' or, 'h' or 'h1'  'h6' element's content model and the subh  
content model itself.

Even with these suggestions I think it might make sense to still  
define 'header' and 'footer' as regions of a document when presented  
as a web page. As currently defined, the  'header' element seems to  
be confused a bit  with a 'heading' element as those terms are  
commonly used. In other words a 'header' is a presentational region  
of a page, while a 'heading' is a title or subtitle for a section or  
subsection. While those headings often get displayed in headers they  
are semantically distinct.

Providing thorough mechanisms to express the metadata of a document  
helps to define those metadata semantics that might be presented in  
the header and the footer of a page. However, we should keep those  
concepts separate. Once CSS2 catches on, I think we'll have the  
presentational mechanisms we need to present that information in the  
header and footer of a page on screen (as well as in print). For the  
HTML WG we should focus on those distinct semantics and not just in  
name only (such as renaming the 'footer' to 'about' without more  
narrowly defining what 'about' should contain).

Some other metadata semantics defined in Atom that we might want to  
consider bringing to HTML5[1]:

    author/contributor (including: EMail, URI and Name; perhaps more  
fully specifying structure for the ADDRESS element)
    updated (date)
    published (date)

There are other pieces of metadata in atom, but they seem less  
relevant to HTML.

Take care,

[1]: <http://tools.ietf.org/html/rfc4287>

original footnotes:
> [1] http://www.w3.org/html/wg/html5/#aside
> [2] http://www.w3.org/html/wg/html5/#header
> [3] http://www.w3.org/html/wg/html5/#footer
Received on Thursday, 6 September 2007 22:35:02 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:26 UTC