Professional Documents
Culture Documents
CGS2871
6 August 2003
In the brave new XML world, accessibility as defined in the W3C Web Accessibility Initiative
(WAI) Should be universal. The whole point of XML is to make documents available in a platformindependent form that is evolution proof, based on text and descriptive tags so that anyone can look at the
file and figure out what the document is and what the pieces are. The WAI is designed to extend that
transparency to everyone, so that browsers, authoring tools, and the entire Web experience is equally
enjoyable or useful for persons with visual, motor, or cognitive disabilities. But that doesnt mean your
troubles are over. Its unlikely that the world will have universal access to XML-capable browsers anytime
soon. So part of your job is going to be thinking of ways to make your XML content visible to the large
numbers of persons without direct XML support.
In a modern context, this means that youll have to take into account display devices with many levels of
display capability. Devices may include:
Text-only or aural displays with a single keyboard.
Web-TVs with only a pointing device.
High-resolution graphics terminals with keyboard plus a graphics tablet plus whatever.
Internet appliances, devices intended to address a few specialized needs.
Text-only includes both terminals, such as those used by persons with vision disabilities, and
PDAs or even cell phones.
This wide range of potential output devices means that either youll have to supply documents dynamically,
based on sniffing browsers capability, or provide lowest-common-denominator documents using interim
technologies. Those technologies will include XHTML along with the capability to embed (or transclude)
XML or other content along with XSL style information to allow dynamic modification of the incoming
data stream by SML-aware devices. Like scripts today, browsers not capable of supporting advanced
capabilities will ignore them.
With similar techniques, users could use the NET as easily with low vision or no vision as others can with
20/20 eyesight. If you cant distinguish blue from green, why shouldnt your computer know this and
compensate in appropriate ways? With CSS of XSL style sheets, the compensation is trivial as long as the
underlying meaning of every data items is encoded in the page.
All these things are possible. The technology exists today. The means includes universal use of XML in
Web pages. For example, its quite common now for Web documents authors to use visual clues, like color,
font, or point size, to highlight meaning in particular data fields. Since HTML doesnt allow you to easily
identify meaningful distinctions between types of data, the highlighting mechanism tend to be coded
directly into the page. Just as calendars are customarily printed with weekend and holiday dates in red,
which is meaningless unless you already know the red-letter-day conventions, HTML authors tried to
duplicate the conventions of their national languages by brute force rather than letting meaning flow from
the tags themselves as is possible in XML.
So an HTML author might naively encode information too specifically and too uninformatively like this:
<p>Click on the red text to select a link an explanation<a href=URL><font colorred>here</font></a></p>
An XML author, on the other hand, should be aware that not everyone will have a mouse so click is a
poor description of selecting a link, and using the word here to describe a link doesnt give much of a
clue to what is on the other end when presented as an isolated list of links, as some aural browsers do
automatically. So in XML you might want to do something more like this:
<paragraph>Select the following link to see a<a href=URL>further explanation of </link>of this
subject.</paragraph>
Note that this XML version doesnt assume anything about how links are highlighted or displayed and
offers the phrase further explanations to describe what one finds on the other end of the link. In a longer
example, you might want the link descriptions to be even more formal, since they may be presented
separately as an outline as mentioned above.
There are other types of accessibility which include:
The Human Case For Accessibility
The Business Case For Accessibility
The Legal Case For Accessibility
Using Links Accessibility
Using Frames Accessibility
Making An XML Document Accessible
Unless youre very careful in your testing, your site may attract a complaint from a disabled user who feels,
quite justly, aggrieved and doesnt hesitate to let you know it. Pay close attention. If there is anything you
dont understand, reply with a polite question showing willingness to learn. Disabled individuals have a lot
more experience dealing with their disability than you. Quite often, they are rich sources of information of
how to easily and inexpensively deal with the issue they raise. Also consider that theyve done you a
tremendous favor. Many of your potential customers, perhaps hundreds, may have clicked on through
rather than take the time to pen a note. Or perhaps they couldnt find anyone to send the note to. Make sure
you have clearly labeled contact information on your splash page.
Although all sites should maintain a webmaster@[sitename] email identity to simplify life for persons with
visual disabilities who may need assistance, its surprising how many sites make it difficult to contact a
human being. Some use forms located on a page which may itself be inaccessible, which is, quite frankly,
unconscionable.
Some things have changed, albeit with little affect on the end result:
Usability problem: Users can't read the text
Typical reason 1997: Poor contrast (e.g. black text on dark red background)
Typical reason 2003: Text too small
Excuse 1997: We must use the corporate colours!
Excuse 2003: We need to fit more on the screen!
1998 saw growing excitement around the potential of Intranets and many organizations had begun projects
to move their internal administrative systems into web-based interfaces. The foreseen advantage was often
cited as "it will make them much simpler - users just need to click on things!". Sadly, the fact that "clicking
on things" was about the only thing standard web interfaces could do was overlooked, and initial attempts
at moving complex applications to the web resulted in incredibly slow and cumbersome user interfaces.
1999 and the onset of the new millennium saw a trend towards the merger of web-based and 'traditional'
software interfaces. Not only were software applications 'borrowing' ideas from the web, but the better
web-based applications were also beginning to follow the interface design guidelines of software
applications.
During 2000 most Internet companies more or less reached their peak in terms of size. Larger companies
had by this stage taken on usability specialists, and several had usability/UI groups. However, success was
very mixed - our evaluations revealed that the same company could produce different web sites with wildly
varying usability. This can be explained in part by the rapid rate of growth and mergers of such companies
- the understanding of the importance of usability and user-centred design varied greatly from one group to
another.
2001-2003 - Reality check for the web?
With the decline and fall of many 'internet agencies', much of the hype surrounding the web also subdued.
Greater emphasis has begun to be placed on designing web sites to meet business and user needs, rather
than to provide all the latest functions regardless of how (or if) they might be used.
Awareness of the importance of good usability is now relatively widespread. The problem, however, is that
knowledge concerning how to reach good usability - through user-centered design - is still relatively poor
(see our "Lack of customer focus in e-commerce" editorial).
Phillips, Lee Anne. Special Edition Using XML. Que Corporation, 201 W. 103rd Street,
Indianapolis, Indiana 46290, August 2000.
Whitehand, Richard. Usability Partners.
http://www.usabilitypartners.se/news/2003/editorial06.shtml, SWEDEN, June 2003.