- From: Ian B. Jacobs <ij@w3.org>
- Date: Thu, 08 Aug 2002 13:12:53 -0400
- To: w3c-wai-ua@w3.org
Hello, For today's teleconference, I'd like to refer to this proposed revision to checkpoint 6.4. This revision takes into account various comments people have made. - Ian <new> Checkpoint 6.4 Programmatic access to information about rendered content. (P1) Provision 1: Make available bounding dimensions and coordinates of rendered graphical objects. Question: Are these coordinates with respect to the viewport or the screen boundaries? Rich has suggested that coordinates with respect to screen boundaries are necessary. But are these available, e.g., in javascript? If not, can one calculate "absolute" coordinates given relative coordinates and viewport location (available from the operating system)? Provision 2: Provide access to the following information about each piece of rendered text: a. font-family b. font-size c. foreground and background color Provision 3: As part of satisfying the provisions of this checkpoint, implement at least one API according to the API cascade described in provision two of checkpoint 6.3. Note(1): User agents should also provide information about the correspondence (in both directions) between graphical objects and their source in the document object. Note(2): The user agent should also provide access to information about each graphical object, such as its role, a text description and other useful information in describing the state and content of the object. </new> For the Techniques Document: The first provision of this checkpoint refers to what is actually rendered on the screen. In CSS, this means "actual values" rather than "computed values". Note, however, that in the DOM Level 2 Style Sheet specification, the CSS module does not provide access to actual values, only read-only access to computed values. -- Ian Jacobs (ij@w3.org) http://www.w3.org/People/Jacobs Tel: +1 718 260-9447
Received on Thursday, 8 August 2002 13:16:09 UTC