These are my complete, unedited notes from the “Should Accessibility Be Perfect” session at CSUN12. Any errors are my own – if you would like to correct anything (say, spelling of your name, add your name if I failed to get it, or correct a gross misunderstanding I had in attempting to rapidly summarize your statements) please do let me know!

Comments in square brackets are my own comments while writing. I missed some names or don’t know enough information to identify people in some cases – if you can provide information, that would be helpful.

If you do not want your information personally identified in this document, let me know.

Panel: Kath Moonan, Lisa Herrod, Leonie Watson, Sarah Lewthwaite, Henny Swan

See also Henny’s post on the questions raised for discussion.


Leonie: Development of the Equality Act, specifically mentions accessibility of disability services. Section 508 in process; Australia has targeted WCAG (Web Content Accessibility Guidelines) AA as achievement guideline. Cases: Budget Airline, Target, Sydney Olympics


Lisa: Primary audience; can you cherry pick guidelines for your primary audience. Better to get something done than everything?


Sarah: Discrete skill sets with different pedagogies: hard code and symbolic code (perfectable); about people, and messy realities. Abstract vs. concrete. To what extent must a11y be a research-rich discipline involved with people with disabilities. Is a perfect education possible?


Kath: Lot of accessibility activity focuses on technical compliance, though starting in 2005 the DRC in UK stated that most issues were outside of WCAG 1 (at that time). Can one claim perfect accessibility without user testing? Should we classify people by impairment: medical model. Usability suggests 6 participants per type of user – how many users does this mean for user testing?


Henny: Perfect accessibility may not be possible. May not have been involved in scoping or user interface; given wireframes and have to code. If you have a global reputation, what’s your responsibility for your production, given it’s usage as an example? May have to sacrifice one small slice because of other needs. As consultant: when do you allow an exception. Where’s the cutoff? Working with imperfect assistive technology, browsers, technology — can we reach it anyway?


Sarah: Can we reach a definition of perfect accessibility? Levels and types of disability vary widely. Medical model of disability vs. social understanding of technological mitigation. Awareness of stigma and cultural bias. Post-structural: other power relations besides economic — family, duties, religion, etc.

John Foliot: Question: We as content producers and developers have a social contract with the consumers of our products. How does that factor into the process. Do they also have a social contract to maintain their tools to access what we create?

Henny: Formula – users should be able to use their assistive technology. If it’s a small edge case, it might be let go – if it has a huge impact, then we need to take responsibility even though it may not be our problem. Creates risk that we perpetuate the problem by supporting low-quality solutions.

Leonie: From a user end, the assistive technology vendors, browser manufacturers, ourselves should all be experts — audience should not have to be. We all have a responsibility, but we may carry more (all three groups) due to the necessary expertise burden.

Jessi: Similar situation with web browsers in general; up to the consumers to stay up to date. If consumers are using too old technology, the user should take some responsibility to update.

Jimmy Chandler: tries to talk to managers/clients by describing usability as a continuum — usable by one person towards usable by all. Want to encourage participants to move as far as possible, and keep moving further. Government Section 508 very black/white — if they can’t achieve standards, sometimes won’t even try. [Note: person next to me nodding; she works with United States government organizations on 508 compliance]

Lisa? Can’t see: Cherry picking guidelines — how do we handle that? Talking about continuum and what impact we have on a project, we can narrow down to specific guidelines for specific stakeholders (content developer, designer, developer, etc.) Large impact on continuum by distributing burden.

Jimmy: Encourages picking small tasks to achieve so progress keeps moving.

Wendy Lawton? electronic document accessibility services: Raises the concept of perfection — no companies pursue it; most pursue “Best practice” Organizations must define best practice and deliver it. Service providers must define their best practices. Best practice is just as cost effective to achieve in most cases as “good enough” If SP won’t provide “Best in Class” they should disclose

Ryan Benson – Problem with statement is that content providers don’t know that (don’t know what best in class is), and user can’t teach a contractor how to do their job. Contractor doesn’t like to be told everything was wrong.

Wendy: Hears this all the time; procurement people (hiring professionals) need to know what to look for when hiring a contractor in order to achieve accessibility. They need to know what quality requires.

Kath: summarize – “Ryan – In real world, contractors don’t have any accessible practices let alone best, why should consultant have to train the contractors under a tight deadline?” Frequently accessibility is a line in the sand and where do we draw that. Is User testing part of best practice?

Steve Faulkner: Implicit social contract between developer, implementors, and implementation of accessibility support within sites. Developers and browsers are on board, but IT vendors are NOT. Need to have purchasers involved in the contract. AT vendors suck millions of dollars out of users, but can’t take time to involve selves in standards definition – NVDA devs are heavily involved, but are the exception

Sally Cain, Royal National Institute of Blind People — Lot of AT being developed for people who don’t want to learn Windows environment, but they aren’t being supported outside. People can’t do their shopping, because the AT vendor of the product hasn’t made changes to provide outside support. AT gives simpler experience, but doesn’t have sufficient support for real-world tasks.

Jim Henna? – Applauds Steve, but dislikes blame games. Came across inaccessible service, AT blamed browser manufacturer; browser blamed AT. It’s not clear enough – (could be because AT not involved in standards? JCD)

Kath: Subject Up to user to update; references Antonia Hyde? — strong points is that if onus goes on user, it can really exclude people with learning disabilities.

Derek – UK association for blind? Association arranged meeting of companies to sit down — nobody actually said anything.

Lainey Feingold: “Cherry Picking” should be called “Priorities” — from a legal perspective, Priorities indicates intent to continue – “picking” doesn’t come across well; more like a final decision. Legal doesn’t look for perfection, though they want it, they look for priorities in line with needs of user base.

Lisa: Is prioritization “perfect” accessibility — doesn’t believe there is such a thing.

Leonie: What happens if you don’t “nail” all of those issues? Is that a risk due to proscriptive legislation with exact definitions of accessibility.

Henny: At what point do you determine that you’ll launch despite some being unable to use it — but if you don’t launch, NOBODY can use it. The line is constantly shifting.

Glenda Sims: Has to deal with this a lot, always describes it as would be willing to stand with them in a court of law to defend their decisions — avoid that punch in the gut. If you’re willing to defend the decision, you can probably go ahead.

Ian Pouncey: Usability testing is a great way to tell whether it’s accessible or not — user testing a measure of access. From an engineering perspective, not very scalable! Not enough testers, too many things to test. Very few tests get tested in any way shape or form.

Henny: One strategy: BBC has had a brain drain; they are focusing on high profile, key products, and putting effort and training into testing those products — documenting everything, and pushing that knowledge back into the organization. Corporate needs to retain these skills and knowledges, eve, if people leave.

Kath: Vodafone — not currently best in class, but trying very hard. Some products will take a year to get out to market. Lots of user research and testing — strategy – trying to bake working with disabled users into their standard usability tests, from all stages — from product creation through end testing.

John Foliot: Disclaimer personal opinion — tag on to Glenda. When he’s doing evaluations, primarily thinks about everything as being task oriented. Perspective is a) can I do the task and b) how painful is it. At what point is it too painful, regardless of being compliant. Petty issues like heading order doesn’t really make a practical difference. Differentiate failure from annoyance.

Lisa: User testing – works as an independent consultant, many different orgs. Small web agencies, non profits, etc. Has been running compliance/technical audits up front, then looking at those results to conclude how to run user testing and evaluations. Every project includes user testing working with people with disabilities among others — seniors, etc. Audit gives developers some work, then user testing refines focus and can greatly improve usability of sites.

Henny: Notes #PerfectAlly tweet streams; couldn’t keep up. Notes should take the steps to file bugs, talk heavily with major vendors — they frequently are open to considered and constructive feedback.

Karl Groves: re Ian Pouncey — disagree that large scale doesn’t scale, requires good Project Management and great planning, but is possible.

Henny: Need to learn from reports and test results, same code across different products should be reusable.

{Unknown}: Consultant — teams have picked base platform which is fundamentally inaccessible — address this issue early rather than piling band-aids. Can be a major problem. What do you do if you’ve inherited fundamental problems?

Kevin Chao: In regards to accessibility in general; various products, services, standards, countries — difference between different national standards makes the issue complicated.

Sarah: Question of international standards is interesting — vast majority of disabled people in very low-income non-Western countries. Are we really talking about a technological elite among people with disabilities? What about mobile-first countries where access to people with disabilities may be very low? Are we exporting expectations about disability which may not be as useful to individuals in these different locales?

Shadi: Specific area is developing countries – this question is big. Lots of research gaps and open questions. Doesn’t think there’s a real export of expectations, but the issues are very diverse and different. Don’t get to talk about accessibility of systems; lucky to have systems. Distance education programs can have very beneficial use [where technology is available]

Leonie: Legacy technology is huge issue. Go back to the supplier of technology to get better accessibility – don’t underestimate the power as purchasers and procurers. Big strides have been made forward in last five years. Amount you can tackle in house can be intimidating but impressive. In house can give you fabulous front and back end products which you may not be able to get from a 3rd party product if that product doesn’t natively meet your needs.

Denis Boudreau: Believes that the very fact this is happening means that panel/environment believes that imperfect accessibility is acceptable. Probably a good thing, we’re moving out of accessibility [elitism]/perfection, which can slow down take up for accessibility. Getting all web sites a little accessible much better than 10 web sites which are perfect —

Kath – Trying to be perfect can make us pedantic

{unknown} – Lot of research in communication technologies in development/mobile telephony – information science. Doesn’t entirely address accessibility issues, but addresses barriers with simple technology as area of literature to look into.

Linda Dardarian — Law doesn’t require perfection if it creates undue burden.