accessible design without drama

with Matt May (is it joe clark who always calls him MCMay?) — who must be related distantly to Greyson.

the drama in the field gets in our way. (hmmmmm….)

starting with high-level stuff, and then into concrete.

most of the audience is people for whom accessibility is part of a larger job.

break down hope at the beginning to bring it back later.

why are all these standards competing with each other? why isn’t there this accessibility thing that we can drop a cup of into our product?

we can’t guarantee that every message we create is going to be understood by the total set of all human beings. some people are good at convincing a broad range of people. there’s a point where you can’t get your understanding out into the rest of the world. not a cognitive disability: just a core tenent of human communication!

by creating standards, not to make it perfect. the messages that can be communicated can be percieved & understood by the widest range of people possible.

2nd disturbing thing: instead of thinking about people, we think about the standards, checkpoints, etc. keeps up from our full capability. the map is not the territory.

standards exist to be codified, policy-compatible. boil a grand concept into measurable criteria. can’t guarantee that it’s better. we have limitations, because we have to get something. (“if men were angels” and all that.)

folks “subject” to 508 et al start to think about the checklists and make something aweful.

why can’t standards be more specific? limitations of communication. when talking to people who don’t have inner motivations. rolls his eyes dramatically — when he mentions accessibility to designers, first comment is always “oh, alt text for people who are blind?”

508, in adapting an earlier version of wcag 1.0, dropped off lots of things in favor of simplicity.

looking past the documents & criteria (created so they could be measured) and focusing on how people use stuff.

simpsons reference: bart gets elephant. DJ jobs will be replaced by DJ-bot. “Don’t praise the machine”

if standards are an approximation; validation tools are then an “artschool knockoff” of that! “I passed Bobby” is NOT GOOD ENOUGH. the understanding stops at getting the icon; but at the bottom it says check this, this, and this. the machine can’t do that. “that’s why we have English teachers, right?” individuals need to look at what they’ve done.

“paraphrasing and getting it mixed up with with einstein.” (simple but no simpler)

was working on a meta-tool to mashup all the other tools, because of all the discrepancies.

best result: minor improvement. worst result: abomination. inaccessible and unusable.

picks on Bobby because it’s the name everybody knows; lots of others with their own idiosyncracies. if you give someone a hammer, that doesn’t make them a carpenter.

get away from “gaming the system” to get the “tangible artifact that represents accessibility.”

was at sxsw 04, advanced accessibility presentation. austin accessibility rallies. veen: “I don’t care about accessibility.” “what we in the industry call an ‘oh sh!t’ moment.” he said this, posted it to blog (I think I remember that): because he believed in standards-based design, and thus accessibility problems dropped by the wayside (automagically). (in theory; then again, “communism works, in theory.” my simpsons reference!)

#1 indicator of accessibility: conforming to standards.

but what people heard was the kick off line, and then the hype that standards is all that needs to be done.

three things to do….

designing for people. (hm, tools have trade secrets; can’t trust them.) NIH study in 2004, with ranges of disabilities. “can’t grasp or handle small objects” — 2.2 million (18-64 yrs old). huh. total is 3.8 million. hockey-sticks at 65…which is the fastest-growing group of users (and just people in general). vision trouble: 13.3 million (18-64) — not blind but poor vision. and the usual problems with IE and text-resizing.

getting into hearing loss issues, costs of captions. 11.9 percent of 18-64, and then 1 in 6 over 65. “numbers don’t lie”

you don’t hear from people, because they learn to work around.

I missed the ADHD number. And then attention deficit trait (13% of people in IT industry?!). attention span issues.

designing for standards. beneficial specifically because contract between producer and user agent. when you understand how things are rendered in alt devices, etc., you are providing better information. (this is a list with 35 links, vs. line-breaks, etc.)

people using CMSs? about half. (we are a mishmash.) can be a good thing. when you have a system that goes off the rails, fixing accessibility problems can be a bigger problem. fixing templates: easy to improve lots all at once! but if the stuff inside is junk, job could be a lot harder. “just type your stuff into here” but no knowledge among those people about what to do with images, etc.

question about finding people. I missed some of the answer, because I wasn’t paying much attention, because I know where I can find them, I just need to DO it.

problems with captioning and small sizes of video? script? recreate captions? audio description? help? cross-player problems with UI. ability to display SMIL/SAMI (?) captions outside of the bounding box. best advice: …question about what’s being used… Magpie is captioning software. WGBH. sounds like she’s getting already-captioned video from video group.

Oregon institute on disabilities, etc. really fscking long name. local guy, offering his assistance to locals. more similar centers nationwide (ADA). good for finding testers.

the basics on images, link images. text equivalents for rich media, captions for video.

layout: hope-building exercise. layout tables are not necessarily inaccessible. but often symptomatic of an older system of design. (I don’t do tables. Haven’t for years.) fixed vs. liquid. good news: browsers that scale up the whole page. IE7, apparently. Opera does the same thing. also familiar, the trouble with px in font sizing.

contrast. outside of black on white, the math gets complicated. grey on grey = bad! soothing, but unreadable. who has best refs? I thought he mentioned joe clark (I was distracted again).

tables, data style. communicate using the semantics of the table. there was a conference in DC just on data tables; he’ll cover in 90 seconds. use th! *scope* for disambiguation. okay, that does work.

forms. label is really cool! (I explained it to Tom yesterday.) find alternatives to form.submit and other JS-only stuff. bigger problem than

navigation. ugh, it’s 2006. “click here” is evil. use structure. device-independent menus (dropdown)

honestly, nothing is especially new to me here in terms of techniques, although I greatly enjoy his presentation style.

CAPTCHA: his particular hobbyhorse. pandora’s box.

q on captcha: on blog, had 2700 spam comments, and captcha stopped it. what else to do? Matt uses spam karma 2 (for WP), and has great success. a high value target will attract mechanized defeat of captcha. “the club lock of web security.”