RE: Dynamically generated page content tests

An additional point is that querying a database to find out if there is a
particular piece of information (a field) in each record is trivially easy.
Assessing whether that information is appropriate is a bit more difficult,
but the results can also be recorded.

The EARL language - http://www.w3.org/2001/03/earl (but there has been
new work done this week so that page needs to be updated) - being developed
by the evaluation and repair tools group - http://www.w3.org/WAI/ER - is one
way of recording this information, and there are tools that implement this
approach so that it is also available from the database. That means that it
is possible to query for things that have not yet been checked - again
reducing the workload to the bare essentials.

I would encourage this group to work closely with the Authoring Tool group
- http://www.w3.org/WAI/AU - on this topic. That group is developing
techniques for evaluating authoring tools, and techniques that can be used
by authoring tools for evaluating content, as part of the Techniques for
Authoring Tool Accessibility - http://www.w3.org/WAI/AU/ATAG-TECHS - which
should contain useful information. There have also been discussions in that
group which might be worth reading before repeating them.

Cheers

Charles

On Thu, 27 Jun 2002, Alan Chuter wrote:

  I'm sorry I missed last night's meeting.

  It's rather late in the day to add to this, but it occurs to me that while
  this section allows that it may be impractical to test all the permutations
  of pages generated dynamically from a database, and suggests testing
  templates separately from content, perhaps we should consider testing the
  content independently of the templates.

  I suggest declaring in an authoring tool whether or not a field can contain
  mark-up or not. If it can't (like date or customer name for example) then
  there's no need to test it. If it can (like a news story), it would be
  feasible to create a test viewer to systematically test all the content in
  that field (in every row), at least those aspects for which the
  accessibility is independent of the template. Differentiating the content on
  markup would reduce the work load enormously.

  I don't think that it's justifiable to slacken the rules for testing marked
  up content just because it comes out of a database. After all, using the
  dynamic architecture is a way to make more efficient use of developer
  resources; some of those resources should be used to check the content more
  thoroughly.

  I realize that authoring tools probably don't support this but it could be
  included described in an abstract way.
    a.. Distinguish between fields containing marked up content and those with
  plain text only.
    b.. Fields that contain only plain text can be assumed not to cause
  accessibility problems
    c.. Fields containing marked up content may cause accessibility problems
  and should be tested:
      a.. Evaluate the relationship between the content and the enclosing
  template (ie, navigation, formatting), and investigate any accessibility
  issues
      b.. Check all the content for accessibility problems in a test page
  I also suggest moving the first point ("if all dynamic content cannot be
  evaluated... test the output"), to the end, as its a catch-all condition to
  be used as a last resort.

  Alan Chuter
  achuter@teleservicios.com
  Fundosa Teleservicios (ONCE Foundation), Madrid, Spain
  ONCE (Spanish National Organisation of the Blind)


  -----Mensaje original-----
  De: w3c-wai-eo-request@w3.org [mailto:w3c-wai-eo-request@w3.org]En nombre de
  Chuck Letourneau
  Enviado el: miercoles, 26 de junio de 2002 18:16
  Para: Harvey Bingham; EOWG
  Asunto: Re: Dynamically generated page content tests


  Hi Harvey,

  I have attempted to incorporate (most of) your suggestions into the
  Evaluating Web Sites for Accessibility document.  See Section 4 at
  http://www.starlingweb.com/wai/eval2.htm#dynamic

  I added them as sub bullets under (I think) the appropriate bullets.
  However, I am not entirely sure this level of detail is appropriate for this
  document.  Perhaps the group could discuss this on the Wednesday or Friday
  call.

  Regards,
  Chuck


   At 14/06/02 09:58 AM, Harvey Bingham wrote:


  Here are some thoughts on assessment of dynamically generated page content:

  1. Does the content of every dynamically generated page have consistent
  content
  with respect to accessibility? If so a single sample can suffice. If not,
  seek
  consistent classes that represent these differences for testing?

  2. Does generated content drawn from database meet accessibility
  requirements for
  all images, alt texts, and if needed the longdesc? [require these in
  database.]

  3. Does the generated tab order from the template allow getting to the
  generated text
  content effectively?

  4. Do generated data tables have accessibility aids: captions, ids on th
  header cells
  and axis idrefs in th data cells?

  5. If generated video, is it captioned?

  6. If generated audio narrative, is textual equivalent available?

  Regards/Harvey Bingham


-- 
Charles McCathieNevile    http://www.w3.org/People/Charles  phone: +61 409 134 136
W3C Web Accessibility Initiative     http://www.w3.org/WAI  fax: +33 4 92 38 78 22
Location: 21 Mitchell street FOOTSCRAY Vic 3011, Australia
(or W3C INRIA, Route des Lucioles, BP 93, 06902 Sophia Antipolis Cedex, France)

Received on Thursday, 27 June 2002 06:13:57 UTC