- From: David Woolley <david@djwhome.demon.co.uk>
- Date: Sat, 11 Aug 2001 10:53:57 +0100 (BST)
- To: w3c-wai-ig@w3.org
Joe Clark wrote: > The following is plain text; if your user agent tries to interpet it as HTML, it is broken. > <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" > "http://www.w3.org/TR/2000/REC-xhtml1-20000126/DTD/xhtml1-transitional.dtd"> > <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> > <head> > <meta http-equiv="content-type" content="text/html; > charset=iso-8859-1" /> > <title>WCAG 2 simplified by Joe Clark</title> Don't use initials here; titles are seen out of context, e.g. in bookmarks and search engine results. > </head> > <body> > <div id="toc"> > <h1>WCAG 2 simplified by <a href="http://joeclark.org/access/" Historically, at least, this should have had a rev="made" attribute, but, in spite of talks of the semantic web, and their ability to do the job done by meta better, rel and rev are more or less dead. > title="Joe Clark: Media Access">Joe Clark</a></h1> > > <h2><a accesskey="c" id="contents" name="contents">Table of > Contents</a></h2> > <ul> > <li> > <a href="#presentation">Guideline 1 - > Presentation. Design content that allows presentation according to This seems to have lost meaning to the extent that almost everyone could claim compliance. The following individual points to clarify, but they seem not to include the important one of using designs which will survive if all the presentation is overridden by the user. > the user’s needs and preferences</a> ^^^^^^^ This does not allow graceful fallback to older technology. There are still a lot of browsers in the field that do not understand Unicode, and even some Unicode aware ones may not know how to map this into available fonts. > <ul> > <li> > <a href="#natural-lang">1.4 > Identify the primary language of text and text equivalents and all > changes in language.</a> > > <ul> Although this appears not to be part of the published version, this still appears to be an abuse of lists for presentational purposes. > <li><strong>Note</strong>: I dropped <cite>human</cite> and ^^^^^^^^^^^^^^^^^^ This IS an abuse of cite for presentational purposes; subsequent uses not explicitly noted. > <cite>natural</cite> completely. Neither is necessary, frankly.</li> > </ul> > > </li> > <li> > <a href="#use-style">1.5 Keep > content and structure separate from presentation.</a> > <ul> > <li><strong>Note</strong>: “Separate content and ^^^^^^ This should be <q> for modern browsers or " for older browsers. In this context, I would go for " - similar points not explicitly noted. > structure” can be read as a noun phrase, with > <cite>separate</cite> as an adjective. This way it’s > unambiguous.</li> > </ul> > > </li> > </ul> > </li> > <li> > <a href="#navigation">Guideline 2 - > Interaction. Design content that allows interaction according to the > user’s needs and preferences</a> I think users should be plural, i.e. users'. > <ul> > <li> > <a > href="#consistent-behaviors">2.1 Handle input errors, such as > misspellings.</a> This can be overdone resulting in garbage in, correctly spelled garbage out. > </li> > <li> > <a > href="#consistent-responses">2.2 Provide consistent and predictable > responses to user actions.</a> > </li> > <li> > <a > href="#extreme-context-change">2.3 Give users control of mechanisms > that cause extreme changes in context.</a><ul> > <li><strong>Note</strong>: This is an example of a guideline > that cannot be simplified further because it requires a fuller > explanation just to understand the main point.</li> The average commercial web page author isn't going to realise that this refers to what he does every day! > </ul> > > </li> > <li> > <a > href="#avoid-interfering">2.4 When content requires a timed response, I think this has lost all meaning. The only obvious literal cases I can think of are where an e-commerce session gets timed out after about 20 minutes, or maybe a pre-"e-commerce" timeout is exceeded, e.g. a hold on an airline seat. I think, though, that it is actually talking about "push" technology, where there is no explicit response expected from the user at all, and the only implicit response is to finish reading the page. > <li> > <a href="#avoid-flicker">2.6 > Avoid causing the screen to flicker.</a> The only real cases I've seen of this would not have been obvious to the author (alternative dark and white scan lines on an interlaced display). We seem to have lost any requirements on distracting animation entirely. > </li> > </ul> > </li> > <li> > <a href="#comprehension">Guideline 3. > Comprehension: Make it as easy as possible to use and understand</a> "possible" is carrying a lot of responsibility here. The lawyers would want "reasonably possible", to avoid forcing people to go extremes, and authors could say the requirement is so subjective that they had done what was possible anyway. I'd consider turning it round and say "Make it no more difficult to use or understand than is strictly necessary for the purpose it serves" (even then the lawyers might want a "reasonably"). This also acknowledges that different purposes require different levels of prior knowledge. > <ul> > <li> > <a href="#consistency">3.1 > Use consistent presentation.</a> I don't think I would understand what this means if I didn't have prior knowledge of the issues. "presentation" is something of a technical term that is somewhat beyond the level of knowledge of most authors, who cut and paste or use WYSIWYG tools. They are doing it, but they don't need to name it. This also demonstrates the limits on making content easy to understand at all levels, without bloating it with explanations. The whole clause is vague. What you are really saying is make presentation a function of the structural component being presented and only introduce the designer's whim at the site or sub-site level. > </li> > <li> > <a > href="#use-style-to-emphasize">3.2 Emphasize structure through > presentation, positioning, and labels.</a> Some aspects of presentation should have adequate default handling by the browser; identifying it explicitly seems to put a responsibility on authors to always override the default styling (which actually conflicts with what I think item 1 is trying to say). Labels is the only aspect of this where the typical WYSIWYG page designer might feel they are failing. > </li> > <li> > <a > href="#clear-and-simple">3.3 Write as simply as possible in a way > that remains appropriate for the site’s content.</a> > <ul> > <li><strong>Note</strong>: This is the <cite>only</cite> > rendition that makes sense.</li> > </ul> > > </li> > <li> > <a > href="#supplement-text">3.4 Wherever possible, use a wide range of > modes of expression.</a> At face value, this conflicts with the write simply requirement. I think it means use text, images, sound etc., but it appears to mean use a wide range of different writing styles in the text. > <ul> > <li><strong>Note</strong>: This too is the <cite>only</cite> > rendition that makes sense.</li> > </ul> > <li> > <a > href="#wcag-possible-lang">4.1 Choose technologies that support the > use of these guidelines.</a> Potential for islands of compatibility. > </li> > <li> > <a > href="#use-markup-correctly">4.2 Use technologies according to > specification.</a> Too weak. The abuse of cite in this text, doesn't violate the objective part of the specification, but is still incorrect use of markup. There are objective parts of the specification, subjective parts, and the spirit. One is aiming for the spirit. <q> is a particular problem as it is not covered by objective parts of the specification and there are good reasons for avoiding its use with the current state of browsers. > </li> > <li> > <a > href="#AT-compatible-UI">4.3 Design user interfaces so they are > compatible with assistive technology.</a> I think I would want more specific points at this level. People may not realise that this covers the use of large fonts and non-standard colour schemes. These aren't properly covered by the next point as they are not turning off the features but changing the defaults. > </li> > <li> > <a > href="#new-tech-degrade">4.4 Design content so that, when > presentation effects are turned off or not supported, the content is > still usable.</a> > </li> > </ul>
Received on Saturday, 11 August 2001 05:59:12 UTC