Sam-I-Am's Builder Blog

Work type stuff - handy urls and notes on the trials and tribulations of a web builder.

Monday, August 28, 2006

Re: Has accessibility been taken too far?

(Jeff Croft posted this provocative article which seemed to tap a common feeling that accessibility is a pain in the ass, strictly optional and web designers should be cut some slack)

If you wade through the slop of the first round of comments to this post, there's actually some reasonable debate that follows. Jeff came out saying he wanted to provoke discussion and (eventually) seems to have done so. It would be nice to think he was playing devil's advocate, but that's probably too generous. He does however seem to withdraw some of the more provocative statements and wind up saying you do what you can, when you can, and dont beat yourself up too badly about it.

It seems to me that accessibility is being treated as one big lump you have to swallow. Particularly in the article heading "Has accessibility been taken too far". What does that even mean? From where I sit, in practice it hasnt actually budged much in the last 5 years, though an awareness of what you could do if you cared to might have improved.

There are some aspects of making an accessible design that present real difficulties to a designer, and some that do not. Making layout and content scale and flow sensibly across a useful range of font-size and effective window width is tricky. Making complex forms accessible can also add significant time to a project. But using semantic markup, and good page structure are not really hard at all.

So it seems there might be a useful distinction to be made between "not-meaningless" design, and "accessible" design. Where the former just implies the application of common sense and basic good practices, and the latter actually includes specific accommodations for some particular minority group/environment/technology.

Some is better than none. And if the brow-beating that Jeff refers to is real, it might be counter-productive. What is critical is awareness. And his post and lots of the comments that follow demonstrate some big gaps. There's a difference between a site looking/sounding crummy and being actually broken. Designers, developers and content authors need a better understanding of the impact of their decisions. Creating valid xhtml is useful, but not critical to accessibility. Css layout too - nice to have. Good alt tags? Only strictly /necessary/ in some cases, though without them it might be confusing and exasperating.

On a typical web project, that a site launches at all is usually a major accomplishment. You have to keep it simple, dont sweat the small stuff, and so on to get there. Accessibilty competes with a host of other requirements for attention. Here's my scale of 0:10 for accessibility: 0: not published at all, in any format. You just had to be there 5: published widely in an available format: perhaps a magazine, or a completely inaccessible web format like .gif or a downloadable wordperfect document. At least I might hear about it, and get someone to help me read it 8: Published in semantic, sensible html. But there's no alt tags, and no form field labels 10: All the above, plus all our favorite shortcuts and conventions that make quickly grokking the content a breeze in every conceivable browser, screen-reader, device and context.

If you are a brow-beater (and I'd guess anyone with an interest in accessibility has been guilty at some point), this might be a good perspective to keep.

Stepping back a little, awareness of accessibility on the web does seem to have grown to the point that it is one of the criteria I hear being used when assessing quality. And this might be a simpler way to think of it. If a site blows up in IE 5, is mute or unintelligible in Jaws, invisible to googlebot and strains the eyes on a projector - maybe its just plain bad. When "Good" includes being accessible, and inaccessible is "Bad", I think accessibility on the web has finally arrived.

Thursday, August 17, 2006

"Surveying OS Ajax Toolkits" article on infoworld

This is well worth a read. Unlike most reviews I've seen, this author obviously spent sometime with each of the libraries he includes - enough to get a meaningful impression of the strengths and weaknesses. For me, he's right on the money with Dojo, YUI, Rico, Atlas. I differ a little on GWT, but in truth it sounds like he spent more time with it than I did. Being a dojo guy at present, I think there's lots in there he missed or didnt mention, but the overviews are fair IMO. Oddly, he didnt review Prototype/Scriptaculous at all.