Updated 2007.04.15
When you buy the book, you get the entire text (but no graphics) on the included CD-ROM, along with a few extras, like fonts and utilities. The following files provide the original copy (with almost no graphics, save for Chapter 13), corrected as of August 2005.
In autumn 2005, I made every known correction to these texts. The corrections were limited to copy-errors and mistakes in code samples. The 2002-era advice is left intact for historical accuracy. Some of the methods espoused in the book are now outdated, and it may be of interest to future historians of Web accessibility to read what we thought would work in that bygone era.
Some of the advice in the book is now outdated for developers in the 21st century. Most importantly, the use of tables for layout can almost never be justified anymore. When I wrote the book, it was just barely possible to defend tables for layout (if you used exactly one table) because browser support for CSS was so poor. That isn’t the case anymore and you shouldn’t follow the book when it advises the use of layout tables.
As this site is a repository for the text of a book published at a certain historical moment, this is not the place to correct or update the book based on knowledge or facts that did not exist at the time I wrote it. You may view the advice to use tables for layout, and a few other things, as artifacts of a sort.
The Brewer palette: Colour combinations for colourblind people
These files contain the full text of the printed book, plus hyperlinks. My book designer Marc Sullivan and I made the command decision late in the process to include none of the 50 or more images used in the printed book. Why?
Because, as explained in the Colophon, producing HTML and Quark Xpress versions of the book added weeks to the production time. Stated another way, providing two formats ate two weeks out of my life. Manually converting Quark graphics boxes plus their associated cutlines (captions) into HTML, then inserting same with appropriate links in just the right places in each chapter, would have added yet another week.
And it still would have been inadequate. All you would have gotten is a simple alt text and title for each graphic rather than the full, rich verbal or textual description you could expect from a trained talking-book transcriber. (Guidelines exist for such conversions of illustrations to voice or text. It’s comparable to audio description of TV, film, and video.)
Consequently, the entire “Special ‘Advertising’ Supplement” allied with Chapter 6, “The image problem,” is not provided here, since it’s nothing but graphics and cutlines. I regret that I have but one life to give to accessibility; unneeded efforts must be eliminated.
Why didn’t we use PDF? Because Quark Xpress cannot save a tagged PDF, and those are the only kind with a reasonable chance of being read sensibly with a screen reader.
While all the foregoing may seem vaguely hypocritical – why isn’t a book on accessibility itself accessible? – the fact remains that books are old media, and old media must be made accessible outside themselves. New(er) media, like television, film, video, and indeed the Web, can carry along their own access features. So in fact, nothing more than what we’re offering could reasonably be expected. The way to make books accessible is to create alternate-media formats (these files are an attempt at creating such formats); books cannot carry along their own accessibility features.
For a later project, however, I returned to the book and added illustrations for Chapter 13, “Multimedia.”
All files included here that I created myself use valid HTML and stylesheets. (Note to strict readers: Using an image as a header, as the chapters do for typographic reasons, is perfectly valid as long as alt and title are provided, which they are. The book itself explains why.)
Note that all HTML files use extensive metadata (information about information), including full Dublin Core encoding and an enormous range of Next, Previous, Chapter, Section, and other attributes of the link element. The Dublin Core information won’t be useful to anything but some kind of future database software or aggregator, but the link-element additions are immediately practical for anyone who uses a browser like Mozilla, Netscape 6 or later, Opera, iCab, or Lynx, all of which give access to those navigational elements, which let you can move from chapter to chapter and section to section. Have fun with them. (Hint for Mozilla and Netscape users: You must turn on the Site Navigation Bar in the View menu’s Show/Hide submenu. In iCab, the Standard Links toolbar must be active. In Opera, reveal the Link Bar. Lynx shows you the navigation elements automatically.)