- From: David Woolley <david@djwhome.demon.co.uk>
- Date: Wed, 31 Oct 2001 07:13:34 +0000 (GMT)
- To: w3c-wai-ig@w3.org
> > Please excuse my ignorance, I lack a programing background and thought > someone on this list might have an answer. > > The first word processor I used required the insertion of codes to > produce various effects - bold, italic words, etc. This is one of the great problems for HTML, users try to use it that way rather than introducing codes to indicate structural elements. > This concept, if technically feasible, would allow the person creating > the content to control what text/graphics would be read by the screen > reader. This can be done (the standard is almost 4 years old) using CSS2: @media tactile, audio, tty { /* braille, direct speech and indirect speech using text only and simple screen reader */ .eyecandy { display: none } /* Do not render any content of the eyecandy class */ } then putting class="eyecandy" in the start tags of elements containing such noise components (possibly adding DIV or SPAN elements for this purpose). Note that display: none supresses everything subordinate to the element to which it is applied. However it requires the use of the latest browsers with tools that are not simple screen readers, but rather expensive accessibility aids that interface with the browser to obtain information about the structure of the document. Many blind users would claim not to be able to afford such tools, certainly those in Access Technology's client groups wouldn't be able to afford them. General purpose screen readers will see the results of interpreting the document with a media type of visual, and many who use them with text only browsers will find that they don't implement style sheets (support for style sheets needs large internal changes on projects maintained by a handful of people in spare time, and most styling cannot be represented in text only, anyway). Moreover, the singular failure of authors to use alt attributes, which can go a long way towards de-cluttering the text, or to linearise their text into proper reading order (which will still be necessary) does not bode well for them marking up HTML in such a way. > All a person would have to do would be to insert a specific start and > stop command to instruct screen reading software to skip the content > between these commands. Note that HTML requires that structures nest, even if early commercial browsers encouraged "tag soup" coding in which they didn't nest.
Received on Wednesday, 31 October 2001 02:13:40 UTC