The InterNet Aural Accessibility Protocol
(IAAP)

The purpose of the InterNet Aural Accessibility Protocol (IAAP) is to enhance the accessibility of the World Wide Web to blind and visually impaired individuals using graphical browsers in conjunction with screen-readers, refreshable braille displays, and/or screen magnification software. Patches based on the IAAP would:

1) Allow the user to activate a numbered links option, which would cause the browser to associate a bracketed number with every hyperlink. Hyperlinks could then be selected simply by typing the number associated with the desired link and pressing <ENTER>

2) Communicate the ALT text correctly and uniformly, as outlined in the W3C's HTML specifications. This includes the use of ALT-text in the <AREA> portion of <MAP ...> definitions.

3) Facilitate the feeding of the ALT-text to a screen-reader and/or screen-magnification program when image loading is turned on--a feature which would greatly benefit low vision users who want/need aural feedback, but have enough visual acuity to perceive the images. While this is a commendable feature of MSAA (Microsoft Active Accessibility), MSAA is not a viable option for anyone: a) not operating in a 32 bit environment; b) not using a MSAA-capable screen access solution; and c) not using a MSAA-compliant browser.

4) A de-tablization option, which would force the browser to degrade tablized content by forcing a <BR> whenever a </TD> is encountered. This, in conjunction with Point 3 and the method of using the tools provided by HTML 3.2 to convey cellular structure outlined in the document entitled Constructing Aurally Comprehensible Tables Using HTML 3.2 will preserve the cellular structure of table-ized content in an aurally intelligible manner.

5) Generation of pseudo-ALT tags for graphically defined links that lack ALT-text in a manner similar to that employed by Lynx, which generates the pseudo-ALT [LINK] whenever it encounters an un-ALT-texted graphically defined hyperlink. This would make it possible for the speech-user to not only ascertain the existence of a graphically defined link, but would facilitate the navigation of a page (such as http://www.sony.com) which consists solely of un-ALT-texted, graphically defined hyperlinks

6) The generation of text-based output whenever a <MAP> ... </MAP> container is encountered. Whenever a <MAP ...> element is encountered, the browser, if so set, would automatically generate an ordered list, using the HREF of each AREA as the hyperlink text. If the AREA definition contains an ALT argument, the ALT-text will comprise the hyperlink text; if it does not, the hyperlink text will contain the resolved URLs to which the AREA HREFs point.

The patch should use the TITLE attribute of the IMG element, or the TITLE attribute of the MAP, if either is present in the markup, as the title and main header of the menu. Otherwise, it will use the ALT attribute of the IMG element as the title of the menu. If neither TITLE nor ALT attributes are present in the markup, the browser will create and use a [USEMAP] pseudo-ALT. if the MAPs are not in the same document as the IMG elements, the browser will fetch the document which contains the referenced MAP, and locate it based on its NAME or ID attribute. All MAPs should also be cached, so that they need not be retrieved repeatedly when referenced in different documents.

7) The use of frames (simultaneously displayed, independently scrolling windows) in HTML documents presents the blind/VI user with a multitude of access problems. Given the ever-increasing popularity of framed sites, therefore, it is essential that the following options be included in all graphical browsers:

7A) A no-frames option, which allows the user to disable the rendering of frames. With the browser set to "No Frames" (or "Framed Pages Off"), the user would be presented with the content of the <NOFRAMES></NOFRAMES> section of the document being rendered. If the document source does not contain a NOFRAMES section, the browser will present the user with a vertically aligned (one hyperlink per line) list of labeled links. The text of each link should be predicated on the NAME argument used in the FRAME definition, as in the following example, where underlined text represents the hyperlink text:

      FRAME 1: nav
      FRAME 2: main
      FRAME 3: fb

(Optimally, the NAME arguments utilized by the author of the document source should be more substantive than those used above for the purpose of example. Unfortunately, the vast majority of NAME arguments utilized by HTML authors are cryptically incomprehensible.)

7B) With "Display Frames" activated, the browser must generate an on-screen announcement/status-line warning which informs the user:

      a] that he or she is in a framed document
      b] in which frame the cursor is currently located (i.e. Frame X of Y)

Frame identification should be predicated on the NAME argument used in the FRAME definition.

7C) With "Display Frames" activated, the browser must allow the user to maximize and minimize individual frames.

All of the above suggestions should apply equally to hypertext documents generated by authoring/conversion utilities and by dynamic page generators, such as Cold Fusion and Active Server Pages (ASP). All such output should be filtered through an HTML validator/parser, in order to ensure that its output contains valid HTML that any browser, regardless of platform, is capable of rendering.


drafted by Gregory J. Rosmaita, unagi69@concentric.net
WebMaster & Executive Committee Member, VICUG NYC


Return to the IAAP Index
Go to the IAAP First Draft
Return to the top of this page


W3C Validated HTML 3.2!       Keep the Web Accessible!       Best Viewed With ANY Browser!


this page created September 11, 1997
hyperlinks last updated December 8, 1998