- From: David Poehlman <poehlman1@comcast.net>
- Date: Sat, 19 Oct 2002 07:31:42 -0400
- To: "David Woolley" <david@djwhome.demon.co.uk>, <w3c-wai-ig@w3.org>
and even those do some processing of their input although they generally do not process what comes to but rather what cdomes from the system if that makes any sense. boy would I love to see dynamic real time ocr of the screen with some intelegence behind it and configurability of the output. ----- Original Message ----- From: "David Woolley" <david@djwhome.demon.co.uk> To: <w3c-wai-ig@w3.org> Sent: Saturday, October 19, 2002 5:32 AM Subject: Re: LONGDESC > described below is because, in the utopian sense of the word, a screen > reader simply reads what is rendered on the screen, and does not perform In practice, what people call screen readers are not pure screen readers - in fact, with GUI type operating systems, such could only be created by using optical character recognition techniques. At the very least, they would need to invoke copy and paste capabilities in the application (or in the application's run time system - e.g. Visual Basic), and, in the case of the most often quoted of "screen readers" for web pages, they use MSAA, Microsoft's Accessiblity API, to interface with accessibility support features in the basic browser. Real screen readers tend to be used with tools like Lynx, although they may be quite common amongst those who pay for their own AT.
Received on Saturday, 19 October 2002 07:31:49 UTC