WC: The working group has been working for day and a half on making WCAG better - what can we be providing, and how can we do better with WCAG 2. We would like to hear from the people who have joined for IG on your experiences / problems / etc.
EH major hassle is an otside contractor who refuses to follow the guidelines, and insists on non-scalable flash for the front page with no accessibility support. We have just had sign-off that contractors who don't follow the standards don' get paid.
RP Problem of getting contractors to do accessibility is a big one. Specifying compliance in contracts, and understanding the capacity of vendors to do accessibility are big issues in getting implementation.
AA During workshops, common problem is what goes into a contract. Lookng forward, getting something in the contract is good, but asking for "understanding of" is not enough. So requiring demonstrations of expertise.
LN Go further than Andrew. My experience is it needs to be in the contract, need to know the contractor is capable, and it is important to specify compliance as functional requirements, that is included in master test plan. Best to specify role of designers, developers, etc, and constrain the roles so it is possible to manage when mistakes or problems arse. Question: How do you stop people from lying? Developers will claim accessibility that just isn' true.
DF Reiterate - hardest thing is educating the people who are writing the tenders. Government sector has a requirement for W3C guidelines here, but not done anything else to support that. Trying to push IIA to force the standards to include some kind of enforcement mechanism - people say "I only provide information, I don't need to be accessible". In contractor issue we should stress the endorsed supplier - the IT industry has to have a list of people who cn do this.
RC I am often explaining to designers about accessibility. I've got a perception that it is setting limitations on them, that isn't well received - if you explain to them what you are trying to achieve it is better than a list of don'ts.
CMN shoulld we be listing people who are working in teh area? We have a lot of constraints, but... what can we do to support? certification (resources...) etc
GSW There aren' specific tests to apply, so this is a problem, which is bigger in Australia than in most places - the government hasn' accepted the standards - they have said they are the best out there but HREOC hasn't said they are sufficient. HREOC needs to make somehting like 508 so we can set up a testing body. Several departments have tried this but it becomes unfeasible. having a group like Vision Australia that is already set up might be the way forward.
LN There is a problem in looking at WCAG in isolatoin. Most organisations are producing content plus ways of using it, and specifying how people will look at it - i.e. requires Netscape. Response to Charles - some questions to ask developers would be helpful
LS Good idea. Making demo tests would be good. There are a lot of quicktests for WCAG - these could be built into a contract. Test criteria for QA would be good - a big place can come up with it and then share it. e.g. Define goals for the site, confirm that without a monitor this is achievable, or without a mouse, ...
DF Looking at developers, problem between companies and the individuals who represent the capacity. Maybe some form of competency similar to the MCSE program and a certification.
RP WAI are not doing certification for good and understandable reasons. Is it a good idea to specify a body of knowledge needed to implement accessibility?
RC Things that would help: on one hand, checklists and checkpoints, which need to be clearer and more obviously associated with functionalities and objects that people are thinking. WCAG 2.0 is more focussing on ways to be able to use a site - that can sometimes be easier to use because it is easier to think through and test. Certifying web developers is difficult - it is the developer, managers, content developers, designers, all working on stuff together.
AA There could be different people maintaining.
JW To consider certification, rather than certify people or organisations, certify processes. This is what has to be done in design phase, testing, updates, etc, so that if the process meets the requirements then there is a better likelihood that the outcome will be accessible.
CMN Law and HREOC means that recognising WCAG as be-all end-all is not likely, but they are recognised. demo sites and test suites. Authoring tools are changing the state of the art. Certifying a process against WCAG has been discussed - maybe it would be useful to dredge that up. finally, contributing to WAI can led to wide distribution and helpful stuff
SC Some issues are technical - there is a raft of information to get out, and people are using different ways to get it. Where do you draw lines to exclude something, how do you deal with different browsers and formats. That's an interesting curve - we have thousands of pages of content, and keeping control of it and making it accessible is tricky.
WC So how do you make those kind of decisions
SC It's a group effort. Is stuff going up in PDF, is it static HTML or from a database, does it have scripting? There is no one answer
WC Do you have the same audience all the time
SC No. We have inherited a lot of legacy content. We have used a bunch of tools to try and make things accesible, and there are new tools coming through. We are doing a bit at a time.
WC The WCAG group has been thinking about conformance and where to start - how do you decide what is the starting point, how do you set priorities - can we give advice on what is the most important thing to start with.
AA Someone said "what questions do you ask a developer". Maybe that is a task that the EO group could take on. NOIE are putting example sites into the material they are publishing on the Web. The EO group have two documents in development "an Implementation Plan for Web Accessibility"and "evaluatiing web sites for accessibility". The group welcomes feedback from anyone.
WC Certification of developers - there is a group called brainbench who do certification. I reviewed a version of their certification on accessibility.
AA Kynn Bartlett was doing that?
WC We have heard something about Australian law, but what' the story in NZ
RC There are some drafts out about government and government funded sites, looking for WCAG levelA, resolution independence. It's easy to find. They put out a draft, took feedback, put something out as a new draft that might be alegal requirement.
Action RC send pointers to IG list
GO There is no legislation directly - there is the Human rights act that has an untested clause - when it was introduced the then government tried to get a permanent exemption for the government. Fortunately that didn't happen and they got a temporary exemption until the end of this year. At that point somebody may take a government website as a test case. There is also the NZ disability strategy - not legislation, but a document produced over 3 years with a big consultation process and set out guidelines the government committed to for improving the life of people with disabilities. that included a section on online information, and about a dozen departments have signed up to say they will comply with that. It is early days.
WC Australasia is great - everyone here is so well informed!
RP HREOC do have a disability standards that is looking at development of government best standards, but does not include IT. They do point to WCAG as the best practise available. Working with ISO on human factors and usability, there has been work in this area for many years and there is a substantial body of knowledge available.
CS We need people to provide techniques - things that have been tried and work (or don't) that we can collect and add to the guidelines. The more data we have the better.
LN We ought to note the difference between legislation and a specification. Looking at compliance issues there may be a need to assist the common law in understanding.
LS There is a body of information. The problem isn't that people have read it, it is that they haven't understood it.
WC We have talked for a day and a half about removing priorities. How would that affect your life?
RC It is good to be able to give people step by step goals. Stuff has to be broken up into chunks.
GSW One of the motivations is that people stop at A - is that your experience
RC People are taking resources to do this - if there is too much they will stop. Sometimes it is frustrating to see if what is done works - there isn't feedback from people who are using it.
DF When you take away the priorites there is more comprehension - people see the priorities and get scared - there are too many factors. 2.0 seems friendly.
SC we are trying to rebuild a major legacy site, so we are working on all the things we can at once, but there are bits here and there and it is hard to bring it all together at this stage.
RC It is hard to tell from WCAG 2.0 becuase it doesn' have the technical content. If people were given all of 2.0 as a lump I don't know if it would be done. Some of the sites in NZ are really small - a town with 1hr per week of a webmaster, and it could be quite hard.
GSW Coming to a company that had no idea about accessibility (except that it is different from usability!?!?!?) I spent ages trying to explain the difference between A, double-A, triple-A. I am sitting on the fence, but there are people getting confused by the priorities instead of trying to make stuff accessible. I have to justify testing against what is in the checklist. If there weren't priorities it would be easy to leave out the things that are "too hard". If there are priority 1 issues, it is easier to hold people to them as really important. There should be workarounds for things where there is a conflicting requirement so people try to fulfill the requirement as far as possible. We are going to have to do a lot of PR work if it is too hard.
/* Brian Hardy arrives, Vision Australia
/* Mat Mirabella arrives, Telstra Research Labs
GO for places with little resources it is good to get some ready-made guidance from WAI. If we don' have priority guidance people could take the list to produce single-disability conformance guidelines
RP Priority is a fairly well accepted notion for selling the idea - it gave a sense to people of how to make a pathway rather than taking on everything at once. It is essential that there is some pathway to progress through it.
PH Don' have a lot to respond to - you're looking at it differently. To make an analogy, I ahve 3 kids and they use the net a lot. Sometimes I get stuck for information and I don't know where to go, and they do. So if they get stuck with access they have to find somewhere else to get the information. So it is like what we are doing here too. A lot of deaf people who don't know how to use the net they don't know where to get the information from taht will enable them to get the information they are after. I do have some other comments to make. From my perspective, developing what is happening with IVR being popular for games and it is a huge access issue because often there is no text. That is a big problem - we are racing to try and make sure that people add text as well - also applies to the phone system. Other thing about visual access - the guidelines are great and talk about being able to see what is happening, but also a guide to the language the information is in - I see more and more emphasis on visual information but often there is too much focus on text - I know it is horses for courses but at the same time there needs to be more consideration for visual information on the internet as well. For example there are typical icons like search. What I would like to see, it would be good if designers who use those icons, there are different icons for the same thing. If they could be explained, and there were universal icons that were consistent people could immediately know what icons meant. One more thing, for deaf blind, the giudelines do mention people but there are people who are profoundly deaf and completely blind, and they really get stuck again and again trying to access the internet and the web. Most have some access, but mostly people who have a bit of vision or a bit of hearing, but the area for people who are completely deaf and completely blind is another area again.
GO Is that something that the WAI can take on?
CMN Yep. Advanced development stuff. WC minutes me.
/* 15 minute break
LN There should be a distinction between prioties and levels that they are not linear - level A isn't one third of triple-A. Level A is trying - you haven't got there, but you are trying. Double-A is where you are starting to have a real impact. Second is that we ought not to be priveliging one or other group. In Lego, there are only a handful of pieces, and to make a new piece it has to fit into 15 different models, and has to sell 15 million in a year. I began tothink about chunking - we make some groups of things that provide help for different groups, so you can get credit for a chunk of different things and there are a number of different chunks. This gets presented as there are different chunks, that aren' better than each other, but are a group of things that provide credit.
BH There are a few things we have experienced in 80 projects over the last two years. Having priorities and determining what is more urgent is really important - managers like to be able to tick something off on a progress chart. With priorities there is no sense about how important the functionality is - you can fail level A becuase there is a spacer gif missing an alt tag even though everything else is really good. There needs to be more importance attached to the effect on the user. There is a fair bit of sense in Jakob Nielsen's implementation stuff, and he says it so people will listen. The conformance levels are a very blunt instrument. We worked on a project where they committed to doing it well, but they did it too late, and were trapped by early design issues so couldn't get past level-A. There was a big backlash because they thought they were not getting much credit. That is less of an issue now becuase people are understanding the subtlties better. Something richer would be better, but to throw the priorities out altogether would be a problem because people will cherry-pick the things that are easy.
AA What Brian said. People want to know where to start. Even is there are no priorities, the idea of having various implementation plans available would be good to give people an idea of where to start. That could be worth exploring further
MM Having priorities and things for managers to tick off and make progress plans that take account of bottom line, cost benefit, helping users, is important. In a corporation that gets into management plans by having a conformance scheme that gives a progress plan. At best a policy will be adopted that helps some people and not others, at worst they say it is too hard and don't do it at all. I have also been building a site for young blind people, and the priority scheme hasn't been definitively useful - getting into the higher levels there are other factors that alter the implementation priorities - user feedback, ease of implementation, etc. I'd like to see a conformance scheme that could be more simplistic, but gives people the opportunity to say what it is they have achieved. It is important to address the issue of people giving up because they don't feel they are getting credit. The big issue is what would be in a core set of requirements
BH Is that to do with functionality? In some sites there are important axes of functionality that would change the way things are prioritised.
WC It is my turn!
CS When we tried to implement WCAG 1 on the msn home page, we were able to do a great many checkpoints from all levels, but there were a few from level-A we couldn't do at all. It would have been nice to be able to make a claim about what we had done, and build a policy for sites in our environment that other sites could follow.
WC Two main points. One, people want to be able to claim for what they have done. The other is that people want to know where to start. I think we have had some good discussion about how to do those, and the big open issue is working out where to start - people are in different situations and need to start in different places. And finally, that we need to make everything we produce easy to use.
/* Q closed - LS, CS, LN
LS Talking about getting rid of priority it isn't because the concern of implementation plan, and so on, are real. We now select priorities that legitimise making things work for certain groups over others, because almost every checkpoint is essential to some group of people. The problem is that unless the core requirements is about 90% of the checkpoints, we are leaving out some groups. In doing things progressively, it might be that there are different priorities because some things are easy to fix later, and others are harder to change. If someone is just going to take 5 ceckpoints, maybe it should just be random. For us to legitimise a process of discrimination is bad.
LN Can we have real world jazzy success models because big organisations lack confidence that they can have a cool and accessible site.
CS Rationale behind removing priorities. We talked a lot about adding metadata to sites describing which things had been done, so it is ppossible to find out what things are done, and enable people to select content that meets their needs, regardless of what else has been done or not. Removing the priorities that this group set allows other organisations to set policies based on their environmental rquirements. With 508 they took the WCAG checkpoints and reshaped it, and had to rewrite it to fit. If they could take a checkpoint-based policy it would have been easier to show how it related to our work. The knowledge that the group has is not "this is the priority for everything" because that varies, but being able to say what things can be done. Every checkpoint is better than not having one implemented.
JW The arguments that have been provided have not achieved consensus in the working group - this is an ongoing issue, and the input from today will be taken into account as we continue to deliberate.
WC The goal of W3C is to help evolve the Web. It makes specifications for the Web. There are several different domains - Architecture, User Interface, Document Formats, Technology & society, and WAI. WAI is the group who works on accessibility, and has several working groups
AA EO group is about educating the community on the need for accessibility. e.g. Quick Tips card - things to get the message out in easily digestible formats. have been working on an update on "Getting started" - where to begin? Been working on "the business case" - why it is important and how to implement it in an organisation. Drafts in development include implementation plans for different types of organisation, how to do evaluations. Recent work on "Auxiliary benefits" - side benefits of making a site accessible. There are documents that are currently in the form of 6 bullet points - if people want to join up that would be helpful.
GO What is the philosophy as regards outreach in different countries?
AA we are well represented in north america and europe, and australia. The rest of the world isn't well represented.
LS In at least 20 countries there are big problems because there are no known techniques for solving some problems in bidirectional languages - hebrew, arabic, some asian languages. There are other problems based on features of languages.
CMN I work in Spanish group across Sth America / Spain. It is difficult to work in other languages without expertise that crosses over, and that requires participation from people who have expertise - W3C has lots of french speakers and japanese speakers, one finnish speaker, two arabic speakers, ..
CMN AU report.
BH Feedback - we had a bash at doing a review of a content management system using ATAG. It was really useful for the organisation to understand the limitations of its tools, which allowed us to identify where we should create workaround strategies.
CMN PF report.
JW WCAG group is working on version 2.0 - just had a meeting (as you are all aware by now). Latest draft is 26 october, available. It is different in structure from version 1 - abstracted the principles and need to then work on specifics for technologies - e.g. how to apply them to HTML pages, or to mutlimedia presentations.
BH Rough timeline?
JW Officially, no idea.
WC We are aiming for the end of next year.
WC ERT. The group is going to fold itself into other groups - Authoring Tools, User Agent. The group was started to solve important problems that are now being taken up by Authoring Tools. Before we close down we are going to finish some work on EARL - a language to describe conformance that can enable claims to be made at the level of conforming to the HTML specification, or conforming to particular checkpoints or sub-checkpoints.
CMN tools looking at implementing include TAW, dreamweaver accessibiltiy extension, validators, AccRepair, ...
GO Where can we see it happening?
WC/CMN At the moment it is still in development labs, but moving fast. The result won' be a language that you see, it will be the fact the tools are more useful - they don't bug you about the same thing 5 times.
WC User agent group is proving they have implementation of the guidelines, developing test suites, and working with developers.
CMN Interest Group...
GO It is slightly scary. It takes a lot of courage to start posting. There isn't much guidance, and searching the archives is not very helpful. It is difficult to build a level of knowledge that provides confidence to participate. There is no FAQ. That is an area that could be addressed to bring people into participation - make the group more user friendly.
WC Resource materials?
GO FAQ would be really good. Improving the archive search. Some effort to codify boundaries about acceptable behaviour. It would be better if there was more collective responsibility for how to behave there.
CMN I have spoken to two people who pulled out of this meeting because they felt the IG was too unpleasant - is that a general experience?
/*nods
AA It is also a very helpful resource
GO Yes, but you have to get over a high scary barrier
RC There are a lot of messages about sites that are bad. Maybe that could be a seperate list. It fills up the list with flames and makes it scarier.
AA Does it need to be more strongly themed? The breadth of topics, and the tone, can be overwhelming.
Action GO draft FAQ for WAI IG.
CS should I look into the MCSE project as a model for certification of accessibility?
Thanks everyone - Good bye.