Re: Tagging question...

I must admit that I think characterising some people (I am not entirely clear
whom, but suspect it might be me ;-) as quasi-religious and fanatical seems a
bit unkind.

In fact, it also detracts from the truth of the matter. Although I think it
is important to write markup as well as possible, so that information is
available for those who can take advatnage of it, and so there is some
incentive for developers to bother providing accessibility improvements in
their products, I am also very strongly of the opinion that real-world
realities must be recognised if accessibility is going to be a possibility
for the web at large.

In the particular example of dividing sections, I would do something a bit
different...

For starters, I would use a div element to mark out the sections. This is the
one way HTML provides of grouping a page into semantic chunks. And it allows
a title attribute to describe the purpose of the div for readers (and if
applicable a class attribute to describe it in machine-processing terms -
e.g. to enable styling).

Personally I would also give it borders using CSS - people without CSS
capable browsers presumably haven't felt the need for my fancy styling, and I
haven't felt the need to complicate my life by trying to hack it into
existence - legacy browsers render content fine, if a little plain.

However, I might also be inclined to add an hr element. This means that
anyone will in fact get a standard horizontal rule, saving me the need to
even think about what kind of alt might be required. 

For list items I would again use a list, styled with CSS so that anyone who
doesn't see the need for CSS styling gets a list item in a familiar format
and I don't have to worry about what is a useful alt. (Using a variety of
speech systems quickly makes it clear that choosing the "correct alt text for
markup" is best left to the user interface, rather than the author. (See also
accesskey...) One of the things I find annoying about the legacy web is
listst that aren't - non-web programs such as word processors have been able
to deal with lists as a piece of structure for about a decade, so I don't see
why I shouldn't expect the same of a language and environment designed around
the idea of structured information.

just my 2c worth

Charles McCN

On Wed, 7 Jun 2000, Gregory J. Rosmaita wrote:

  aloha, y'all!
  
  i agree with karl and charles -- where the markup exists, use it -- 
  particularly when the markup is endowed with attributes that increase 
  accessibility, by providing contextual/semantic information about the 
  element or the segment of the document in which it is used...
  
  for example, prior to HTML4, i had to use an ugly hack in order to express 
  logical divisions in documents, since the HR element in HTML 3.2 did not 
  provide for alternative content, as it now does, courtesy of the title 
  attribute...
  
  so, instead of using a kludge such as:
  
  <img src="redline.gif" alt="Please Read The Fine Print (At Least Once!)">
  
  i could use the title element, in conjunction with the HR element, thus:
  
  <hr title="Please Read The Fine Print (At Least Once!)">
  
  or
  
  <hr title="End Terminal Index/Begin Validation Information">
  
  which, incidentally, has the added benefit of not increasing the weight of 
  the page for those with graphics turned on...
  
  of course, the problem is that most adaptive technologies do not recognize 
  (or offer their users the option of exposing) the title attribute for the 
  horizontal rule element, whereas they do recognize the alt attribute...
  
  which puts the page author between a rock and a hard place, which i think 
  is bruce's point...  what is more important -- ensuring that content is as 
  accessible as possible now, using work-arounds based on legacy technology, 
  or pushing page authors to implement markup languages according to the 
  specifications (and, of course, to check not only for accessibility, but 
  validity) while concurrently advocating that mainstream user agent 
  manufacturers expose the extended semantic information (such as title and 
  label and legend et. al.) so that adaptive technology developers can 
  program their products to convey such information to the end user, which 
  is, after all, the point of the exercise...
  
  of course, individual users should be able to configure their AT so as to 
  provide them with as much or as little extended semantic information as 
  that individual desires -- i personally would prefer an on-the-fly toggle 
  for such a setting, for in some cases, i'd want to know what the horizontal 
  rules signify, in others, i really don't care...
  
  if you want to trace this skein back to 1997, start at Gregg Vanderheiden's 
  post, with the subject line "Title on Horizontal Rule"
  <http://lists.w3.org/Archives/Public/w3c-wai-ig/1997OctDec/0056.html>
  my comments at the time, archived at:
  <http://lists.w3.org/Archives/Public/w3c-wai-ig/1997OctDec/0061.html>
    follow:
  
  quote
  i too support a TITLE argument for horizontal rules, provided that the
  explanation of how to use the TITLE appropriately is also included in
  any discussion of using an ALT argument to signify a graphical line
  break, so as to avoid meaningless ALT/TITLE texts such as:
  
           horizontal rule
           straight line
           a line incorporating the olympic colors
           squiggly decorative line
  
  and the like...
  
  when i have had occasion to use graphical dividers (i.e. red or green
  bars), i have opted to either use the ALT tag either to mimic Lynx's
  rendering of an HR
  
  <p align=center><img src="redline.gif" height=4 width=600
  alt="_________________________________________________________________"></p>
  
  if the division is a purely aesthetic one...  otherwise, i have used ALT
  text in conjunction with a graphical rule/divider precisely because there
  currently is no TITLE argument for the HR which would enable me to indicate
  the purpose of the content that follows the divider, as in the following
  example:
  
  <p align=center><img src="redline.gif" height=4 width=600
  alt="The Fine Print: Please Read This Disclaimer at Least Once"></p>
  
  or, when rendering an entire book, where it is often important to
  communicate the actual pagination of the printed edition, i have resorted to
  using the ALT kludge:
  
  <p align=center><img src="thinline.gif" height=1 width=600
  alt="Begin Printed Page 86"></p>
  
  if the TITLE argument was adapted, it would then be possible to use the
  less intrusive HR to indicate the printed edition's page breaks...
  unquote
  
  gregory.
  
  
  --------------------------------------------------------
  He that lives on Hope, dies farting
        -- Benjamin Franklin, Poor Richard's Almanack, 1763
  --------------------------------------------------------
  Gregory J. Rosmaita <unagi69@concentric.net>
      WebMaster and Minister of Propaganda, VICUG NYC
           <http://www.hicom.net/~oedipus/vicug/index.html>
  --------------------------------------------------------
  

--
Charles McCathieNevile    mailto:charles@w3.org    phone: +61 (0) 409 134 136
W3C Web Accessibility Initiative                      http://www.w3.org/WAI
Location: I-cubed, 110 Victoria Street, Carlton VIC 3053
Postal: GPO Box 2476V, Melbourne 3001,  Australia

Received on Wednesday, 7 June 2000 07:15:23 UTC