- From: Geoff Deering <geoff@deering.id.au>
- Date: Mon, 09 Jan 2006 23:20:20 +1100
- To: Léonie Watson <lw@nomensa.com>, w3c-wai-ig@w3.org
Léonie Watson wrote: >Geoff Deering wrote: > >"I'd like to know if this means the user agent is correctly implementing the guidelines/specifications or not? So is it incorrect to markup words using <span> like this?" > > None of the screen readers currently on the market have a particularly accurate grasp of valid code. Most make decisions on behaviour based on their own interpretation of the code, rather than the accepted standard interpretation. > > For example, the Jaws screen reader will determine table type by the number of rows and columns the table contains. If the table contains more than one row or column, it is determined to be a data table. This decision is based on nothing more, the inclusion of relevant data table code, caption, associative markup etc is not taken into account. > > Providing the code used is valid, then it should be used and with luck or perhaps some prodding, the assistive UA vendors will consider following the published standards. > >Regards, >Tink. > > I have to admit your approach is the better to follow, and I admire that. But that leaves me saying to everyone who wants to know and learn about web accessibility; "You need to become educated in the standards and how to implement them... then you also need to learn all the bugs and foibles of the user agents and make sure you avoid implementing practices that clash with their bugs, even though they may be correct implementations". That's a huge ask. I'd really like to see WaSP get into a few of these companies, because their ability to fix their applications, in most part is just plain no good enough by anyone standard. ------------ Geoff Deering
Received on Monday, 9 January 2006 12:20:35 UTC