W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2000

RE: FW: Re[2]: Database-driven Web pages

From: Charles McCathieNevile <charles@w3.org>
Date: Sun, 29 Oct 2000 14:30:12 -0500 (EST)
To: Charles Oppermann <Charles@coppersoftware.com>
cc: basr-l@trace.wisc.edu, w3c-wai-ig@w3.org, arron@gwmicro.com, lynx-dev@sig.net
Message-ID: <Pine.LNX.4.21.0010291427581.15188-100000@tux.w3.org>
Charles Opperman is right on the money as I see it. As to Dave's original
request for database-backed software to ensure that there is alternative
text, it is a requirement covered by several different checkpoints in the
Authoring Tool Accessibility Guidelines, so getting increased conformance to
those will hopefully improve matters considerably. 

Cheers

Charles McCathieNevile

On Fri, 27 Oct 2000, Charles Oppermann wrote:

  The item you're referring to is a image representing a button with words
  "Post a comment" in the graphic.
  
  It's some god-awful HTML.  I don't blame the authors for having long
  URLs - that's necessary for this kind of stuff with the current
  technology, but not having a ALT attribute is the first mistake.  I'd
  also like to see a TITLE attribute on the <A> element - not as a
  substitute for the ALT though. 
  
  So, in psuedo-code...
  
  <A TITLE="Add your own comments on this story" HREF=....>
  <IMG ALT="Post a comment">
  </A>
  
  -----Original Message-----
  From: Lloyd G. Rasmussen [mailto:lras@loc.gov]
  Sent: Friday, October 27, 2000 1:26 PM
  To: basr-l@trace.wisc.edu
  Cc: w3c-wai-ig@w3.org; arron@gwmicro.com; lynx-dev@sig.net
  Subject: Re: FW: Re[2]: Database-driven Web pages
  
  
  I want to comment on the second issue, long URLs generated by automated
  software. 
  
  I find Cnet News  <news.cnet.com> to be quite usable with Window-Eyes
  and
  Lynx.  But they have a feature whereby you can read and comment on
  messages
  submitted by readers of the current article.  Most of its output is just
  fine, except for one button which allows you to submit a comment, which
  has
  no title or alt.  In the article at
  
  http://news.cnet.com/news/0-1005-200-3310477.html?tag=st.ne.1002.lthd.10
  05-2
  00-3310477 
  about Congress wanting to crack down on hackers, the offending link in
  the
  "message boards" section of this story reads
  <a
  href="/news/reviewentry/1,10799,0-1005-602-3310477,00.html?tag=st.ne.ni.
  ucwy
  tnav.wytfm"><img src="/Images/News/Graphics/UserReviews/button_21.gif"
  width="78" height="34" align="right" border="0"></a>
  in which Window-Eyes reads the unintelligible URL 
  news/reviewentry/1,10799,0-1005-602-3310477,00.html?tag=st.ne.ni.ucwytna
  v.wy
  tfm
  and Lynx instead parses the name of the image and provides 
  [button_21] as the name of the link.  If this image had an alt or a
  title,
  both programs would use it.  I never until now tried to find out where
  this
  link takes you, but using Window-Eyes to read the long URL in the midst
  of
  intelligible text is disruptive enough that I sometimes stop the
  automatic
  reading of the story, skip that line, then resume the read-to-end
  function.
     
  
  Databases should produce alt or title attributes for all clickable
  images,
  period.  It would be even better if they meant something in the current
  context.  
  
  At 12:21 PM 10/27/00 -0400, you wrote:
  >If anyone has any thoughts/comments regarding the below messages,
  please
  >send them along.
  >
  >Don Barrett
  >
  >
  >
  >-----Original Message-----
  >From: MARC FINK [mailto:MFINK@email.usps.gov]
  >Sent: Friday, October 27, 2000 11:35 AM
  >To: Don_Barrett; pat.sheehan; Paul_Schafer; Bruce_Bailey;
  >shelia_hamblin; Michele_Zozom; Elaine_Goheen; Ron_Luycx
  >Cc: Gerald_Malitz
  >Subject: Re[2]: Database-driven Web pages
  >
  >
  >
  >     Thanks, Don,
  >
  >     I'm especially interested in what approaches folks have used to
  solve
  >     the following challenges:
  >
  >     1. When generating a linearized table dynamically, what have you
  found
  >     to be a useful method for generating cell coordinates--in other
  words,
  >     how do you get the column and row headers associated with each
  cell
  >     member? Judging from Don's comments below, perhaps it would be
  >     possible to pass a row header variable to each cell in addition to
  the
  >     column header.
  >
  >     2. What work-around has anyone found for dealing with session and
  >     variable ID's which are passed from one page to the next in Web
  >     applications like e-commerce shopping carts? As you know, these
  can be
  >     very long strings which convey little meaning to screen readers.
  >
  >     These are sticky technical issues, but I see them as unavoidable
  >     considering how important database-driven Web content is. This is
  >     especially true when building pages that serve a search/query
  function
  >     to access large data/information resources in an efficient and
  >     meaningful way.
  >
  >     Sincerely,
  >     Marc Fink
  >
  >
  >______________________________ Reply Separator
  >_________________________________
  >Subject: RE: Database-driven Web pages
  >Author:  Don_Barrett@ed.gov at INTERNET
  >Date:    10/27/00 8:41 AM
  >
  >
  >I am forwarding this message to a number of individuals on our team and
  on
  >the Web Development team in the hopes that we can garner some
  additional
  >interest/information on the problems raised in Mark's message below.
  If
  >anyone has any thoughts on how we might proceed in framing this
  discussion,
  >please share your thoughts.
  >
  >For example, here at Education, we have had a high degree of success
  with
  >Cold Fusion, in that the resulting interface is HTML, and all query and
  >script strings are handled without interfering with the interface.  We
  did
  >have a few applications which involved the use of complex tables which
  >needed column header identifiers to be read along with cell contents by
  the
  >screen reader in order to make the table intelligible.  This was
  handled by
  >the contractor by modifying the template so that header information
  would be
  >dynamically generated in the table in each cell.  However, this
  solution is
  >not documented as far as I know, and it should be so others facing the
  same
  >problem can rely on this experience for an easy solution.
  >
  >Thoughts?
  >
  >Don
  >
  >
  >-----Original Message-----
  >From: MARC FINK [mailto:MFINK@email.usps.gov]
  >Sent: Thursday, October 26, 2000 5:31 PM
  >To: Don_Barrett; pat.sheehan
  >Subject: Database-driven Web pages
  >
  >
  >
  >     Dear Pat:
  >
  >     I recently had the pleasure of meeting and talking with you at the
  >     IDEAS 2000 conference. (I work with the 508 team at the United
  States
  >     Postal Service.) I talked with you specifically about
  >     database-generated Web pages and other interactive pages such as
  form
  >     elements and links which are query strings. We agreed that these
  are
  >     areas of great concern but have not really been dealt with
  in-depth in
  >     the present guidelines.
  >
  >     At the time I offered to pursue this issue further with you. Don
  >     Barrett, who works periodically here at Postal Headquarters,
  expressed
  >     a similar interest in studying this area in more detail.
  >
  >     One goal of our mutual collaboration could be to share knowledge
  of
  >     different and successful approaches we know of--hopefully saving
  >     everyone time and money in the long run.
  >
  >     Let me know what you think, and thank you for your time.
  >
  >     Marc Fink
  >     Web Project Manager
  >     508 Team
  >     USPS/LittonPRC
  >     202 268-4716
  >
  >
  Braille is the digital divide.
  Lloyd Rasmussen, Senior Staff Engineer
  National Library Service f/t Blind and Physically Handicapped
  Library of Congress    (202) 707-0535  <lras@loc.gov>
  <http://www.loc.gov/nls>
  HOME:  <lras@sprynet.com>       <http://lras.home.sprynet.com>
  
  

-- 
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, Australia
September - November 2000: 
W3C INRIA, 2004 Route des Lucioles, BP 93, 06902 Sophia Antipolis Cedex, France
Received on Sunday, 29 October 2000 14:30:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:13:50 GMT