W3C home > Mailing lists > Public > public-comments-wcag20@w3.org > February 2009

Re: Include H69 as technique for 2.4.6

From: Sailesh Panchang <spanchang02@yahoo.com>
Date: Fri, 27 Feb 2009 21:16:27 -0800 (PST)
To: Loretta Guarino Reid <lorettaguarino@google.com>
Cc: public-comments-wcag20@w3.org
Message-ID: <736670.82871.qm@web37901.mail.mud.yahoo.com>
I think this is a broader issue affecting SC 1.3.1, 2.4.4 and 2.4.6. Also overlap between H42 and H69.
Pardon the length.
 
1. SC 1.3.1 refers to H42 and not H69. I note this confusion on your part too.
In an earlier  comment-email sent some months ago, I had noted this duplication between H42 and H69  and said that they should be merged.

2. It is incorrect to contend that headings should be introduced to aid navigation. 
Screen readers' implementation  a method of providing  direct access to headers work when they are properly marked up.
Headings markup  is not meant to aid navigation per se. 
Headings should exist in a properly structured document. One cannot introduce headings where none are needed merely to aid navigation.
If headings are absent and sections or structure  or topic is conveyed only by layout and presentation
then hidden headings may be introduced to expose this structure  to screen reader users.  
Screen readers can programatically determine this markup if available and then provide access by quick navigation.
The HTML4 specs do not refer to headings as a navigation aid (that is the role of the user agent / AT) but only says, 
"A heading element briefly describes the topic of the section it introduces."[1]
Headings need to be  marked up because they are headings that convey semantic structure.
So I content that SC 2.4.4 should also refer to H42. (As stated above, merge H42 and H69)

3. One needs to have  a properly marked up heading and it should be concise, descriptive and serve as topic / section identifier at the same time.
These aspects cannot be separated: SC 2.4.6- use descriptive  headings, and 
SC 1.3.1- convey info and relation between heading and related content.    

A cryptic / badly designed  but properly marked up heading does not convey info though it may convey some relation. So it fails 1.3.1. It fails 2.4.6 because it is not  a good heading / not "descriptive".  
A heading text  without proper markup is not programatically identifiable. 
So SC 2.4.6 should also refer to H42

[1]http://www.w3.org/TR/REC-html40/struct/global.html#edef-H1

Sailesh Panchang
www.deque.com
Tel 571-344-1765


--- On Fri, 2/27/09, Loretta Guarino Reid <lorettaguarino@google.com> wrote:

> From: Loretta Guarino Reid <lorettaguarino@google.com>
> Subject: Re: Include H69 as technique for 2.4.6
> To: "Sailesh Panchang" <spanchang02@yahoo.com>
> Cc: public-comments-wcag20@w3.org
> Date: Friday, February 27, 2009, 5:19 PM
> On Fri, Dec 5, 2008 at 11:47 AM, Sailesh Panchang
> <spanchang02@yahoo.com> wrote:
> >
> > SC 2.4.6 lists a general technique (G130) but omits to
> list
> > H69: Providing heading elements at the beginning of
> each section of content
> > as a sufficient technique.
> >
> > Also F2 may be listed as a failure for 2.4.6.
> Currently no failures are listed.
> > Sailesh Panchang
> > www.deque.com
> >
> 
> We have included H69 as a related technique to G130, but
> since it is
> primarily about adding headings to facilitate navigation
> and not about
> whether they describe topic or purpose, we have not
> included it under
> 2.4.6. Similarly, we have not added F2 because it is not a
> failure of
> SC 2.4.6, but a failure of SC 1.3.1. Headings can be
> properly
> descriptive (meeting SC 2.4.6) but still improperly marked
> up (failing
> SC 1.3.1). There are no common failures listed for SC 2.4.6
> because
> there were no special cases to call out, apart from just
> failing to
> meet the success criterion because a heading is not
> suitably
> descriptive.


      
Received on Saturday, 28 February 2009 05:17:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:11:11 UTC