Web Accessibility: Will WCAG 2.0 Better Meet Today

1
Web Accessibility Will WCAG 2.0 Better Meet
Todays Challenges?Experiences Of WCAG 1.0

• Brian Kelly
• UK Web Focus
• UKOLN
• University of Bath
• UK

2
Contents

• Introduction
• What's Happening?
• Survey of UK University Home Pages
• Reports From Other Sectors
• Typical Problems
• Conclusions

3
UK University Home Pages

• In Sept 2003 survey of accessibility of 160 UK
  University entry points carried out
• Used Bobby (to report on problems which an
  automated tool could spot)
• How many WAI AA pages were found?
• The survey found
• Only four entry points complied with AA
• One was a JavaScripted page so isn't accessible
• The UK HE community is generally aware of and
  supportive of WAI issues, uses email lists to
  discuss issues and share solutions (esp. in light
  of legislation introduced in Sept 2002). So why
  this low figure?

What's Happening?
See lthttp//www.ariadne.ac.uk/issue33/web-watch/gt
4
Scottish Political Parties

• Survey of the accessibility of 8 parties standing
  in May 2003 Scottish Parliamentary elections
  carried out (by David Martin Sloan)
• Four parties' home page failed Cynthia Says test
  and manual testing found that all have
  accessibility problems across the Web sites
• missing ALT tags, contrasts, graphical
  navigation, poorly implemented frames,
  non-compliant HTML, PDF files,
• A number of political parties pledged support for
  accessibility, the Web sites had been developed
  for the election and had a high profile. So why
  the poor findings?

What's Happening?
See lthttp//www.dmag.org.uk/election/gt
5
RNIB Web Site

• Bobby was used on 7 May 2003 to test the RNIB
  home page at lthttp//www.rnib.org.uk/gt
• Two priority 2 errors were found
• Is the RNIB home page really inaccessible?

What's Happening?
Similar findings have been reported for other
high-profile accessibility organisations
6
The Context
Concerns

• One University Web manager, following survey
  publication, said
• "I too have been struggling with just how
  rigorously the WAI guidelines should be
  implemented I certainly aspire to comply as
  full as I can with the WAI guidelines but "
• Some guidelines are too theoretical
• I will have a pragmatic approach
• Will use tables for positioning
• Will not associate form controls for search boxes
• Will not necessarily nest headers correctly

7
Specific Problems
Concerns

• Typical problems reported by Bobby's automated
  testing
• Missing ALT text
• Missing DOCTYPEs
• Use of absolute positioning
• Repeated link phrases
• The justifications for these requirements is
  well-known
• They could be fixed easily for an entry point
• But
• What about workflow issues
• What about tools used today
• Are there usability issues?

8
MS Office Case Study

• A typical organisation (including universities)
• Has significant investment in Microsoft Office
  products
• Has conservative users who typically won't
  appreciate new tools being forced on them)
• In MS Word / PowerPoint
• How many users will know how to add ALT text to
  images?
• How many would use this option if they knew about
  it?

Typical Problems
If PowerPoint presentations are held on the Web
primarily for file delivery with little
expectation of use by others should (a) effort be
spend on ALT tags, (b) do as at present or (c)
remove files from Web site?
9
Using A Text Editor

• Many experienced Web authors / software
  developers may use a text editor in preference to
  a HTML authoring tool (I use HTML-kit)
• This should be more usable these days (just
  create simple HTML elements, and leave formatting
  to a CSS file)
• But
• Isn't it too difficult to maintain ids for cell
  elements in complex tables
• Isn't it worse to get ids wrong than not have
  them?

Typical Problems
Should the WAI guidelines be explicit on this
point? How will users of text editors react?
10
Large Web Sites

• A typical university Web site
• Has devolved authorship
• Uses a wide range of technologies, applications,
  etc.)
• Has hundreds of thousands of Web resources
• Differing perceptions
• Web teams would like to install centralised
  Content Management Systems to help apply
  consistent best practices
• Users typically don't like central service
  departments and want to manage their own
  resources, use their own favourite applications,
  etc.

Typical Problems
11
WAI Compliance Levels

• Is it unreasonable to regard
• WAI A Good effort
• WAI AA Even better
• WAI AAA Top of the class
• But
• Is this really the case?
• Aren't some of the AA and AAA requirements based
  on assumptions of how the Web will be in the
  future?

Typical Problems
12
Too Theoretical?

• Are some WAI guidelines too theoretical?

Typical Problems
13.2 Provide metadata to add semantic information
to pages and sites. Priority 2 For example,
use RDF (RDF) to indicate the document's
author, the type of content, etc.

• Some questions
• How many use RDF today?
• Isn't RDF an unproven technology which is
  currently of research interest?
• Isn't this using WAI as a mechanism to promote a
  favoured W3C format?
• If I can't / won't do this, will other Priority
  2 requirements be ignored?

13
Too Theoretical?

• Have some WAI techniques not being used
  sufficiently to expect widespread use?

Typical Problems
1.1 Provide a text equivalent for every non-text
element (e.g., via "alt", "longdesc", or

• But
• longdescr not supported in widely used browsers
• There is little implementation experience
• Should the file be text, HTML, (it's not
  defined)
• How will the information be rendered?
• Should I provide navigation to the original
  document?
• What about the management of the content?
• If it's not widely used, can we implement a
  better solution (e.g. based on XLink)

14
Best Practices Or Today's Practices?
XML CSS SMIL SVG RDF

• Does WAI
• Act as an evangelist for emerging W3C
  technologies?
• Assume that the W3C philosophy is true ("by
  following these guidelines content developers can
  create pages that degrade gracefully ")
• Address real world concerns in an environment of
  broken browsers, commercially driven interests,
  proprietary formats,

Typical Problems
G6 Ensure that pages are accessible even when
newer technologies .. not supported
If I use SMIL, how do I dumb things down to HTML?
15
Cost Of Web Accessibility
MYTH 2 Accessible Web authoring is expensive
and time-consuming MYTH 3 Web accessibility is
too difficult for the average web designer
Typical Problems
http//aware.hwg.org/why/myths.htmlm2

• But doesn't
• 2 ignores the workflow issues
• 2 ignores the documented costs of providing and
  maintaining metadata (an ALT tag is metadata)
• 3 ignores the real world difficulties of, say,
  deploying CSS

It is acknowledged that this is not from WAI
Wouldn't it be better to be open about the costs
in order to gain acceptance? We don't pretend
that safety in cars, providing fire safety in
building, etc. is cheap.
16
Cost Of Web Accessibility
p font-size 12px ///a body p font-size
x-small voice-family "\"\"" voice-family
inherit font-size small htmlgtbody p
font-size small / /

• Diveintoaccessibility.org provides valuable
  advice on making Web sites accessible.
• But look at what it describes

1. First, we're defining an absolute size (12px)
   for every ltpgt. All browsers apply this style
2. Then we include the odd-looking comment "///".
   Due to bugs in Netscape 4, everything between
   this comment and the following one will be
   ignored. That's right, all the following styles
   will only be applied in non-Netscape-4 browsers.
3. Immediately after the odd-looking comment, we
   include an empty rule "a ". Opera 5 for Mac is
   buggy and ignores this rule (and only this rule).
   It applies everything else.

17
Conclusions

• To conclude
• Public sector bodies who want to provide
  accessible Web sites seem to find it difficult to
  do so, even on individual high-profile pages
• The WCAG 1.0 guidelines appear to promote
  little-deployed emerging W3C technologies
• It appears to be difficult / expensive to produce
  richly functional accessible e-learning
  resources

Or is this taking the WAI WCAG guidelines too
literally? Don't the guidelines do a good enough
job in the majority of cases, and to highlight
exceptional cases or esoteric aspects is to
undermine the valuable work that WAI is doing
(and provide a loophole for avoidance)?