To answer the question, JAWS is the most widely used screen reader by a long way in the English speaking world and some other markets, and anecdotal evidence suggests that it is invariably used without any relevant changes to the configuration settings. I hesitate to call it a standard because its own behaviour changes from version to version, although some aspects are very consistent. Nevertheless, it is the only sensible reference point for discussion.
I would add the following points: 1. Some aspects of design affect screen reader users irrespective of the screen reader they use. Designers should be aware of these issues, as they should be aware of issues facing other user groups. 2. Few aspects of behaviour are the same with all screen readers. For instance most 'professional' products such as JAWS announce semantic information such as headings and lists, but some common ones including VoiceOver (built into Mac OS X 10.4) do not. Some do not announce tables, JAWS announces some tables and FireVox announces all tables. In the absence of <label> elements, JAWS 'intelligently' associates nearby text with form controls but other products do not. JavaScript support varies greatly. And so on. 3. It is dangerous to specifically design for a particular product because even the behaviour of one product varies from version to version. Not only do features get added, but existing behaviours change. And if you install a screen reader on a different browser it will behave differently due to differences in the way it interacts with the browser's DOM and the varying levels of MSAA support in browsers. 4. Some aspects of customisation reflect the user's preferences. The punctuation verbosity level (none, some, most or all) would be an example. The user adjusts this at their own risk, and the designer does not need to take it into account. The language and synthesizer voice would be others. 5. You cannot rely on users changing the configuration options even when it becomes easy to do so. Skill levels can be very high (our trainers are awesome) but average skill levels are very low. When you consider that most fully-able users don't even know you can change the font size, it's unreasonable to expect screen reader users to be able to understand the consequences of the hundreds (really!) of configuration options available to them. Before worrying about the minutiae of screen reader behaviour I think that designers should: a. Code to standards (not a problem for subscribers to this list). b. Understand the users and their needs. This is a big problem because few designers get the opportunity to see their designs used by any kind of users, disabled or otherwise. Hands up anyone who has done any user testing this year. Or ever. Steve -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Stuart Foulstone Sent: 25 May 2007 22:53 To: wsg@webstandardsgroup.org Subject: RE: [WSG] screen readers & repeated legends (was "dl v table for form layout") Hi, Does the ability for the user of a screenreader to customise "at leasst partially resolve the problem" or should we design for the default screenreader (which would mean Jaws presumably, since it seemms to be the most commonly used)? If we then design to this "standard", we should then at least have a starting point for further constructive criticism. ******************************************************************* List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] *******************************************************************