W3C logoWeb Accessibility initiative

WAI: Strategies, guidelines, resources to make the Web accessible to people with disabilities

W3C Workshop Synopsis:
Referencing and Applying WCAG 2.0 in Different Contexts
W3C Workshop, 23 May 2013, Brussels, Belgium

Workshop Sponsors

This Workshop is organized through the EC-funded WAI-ACT Project
EU FlagSeventh Framework Programme logo

Additional sponsors include:


Shadi Abou-Zahra from W3C opened the workshop, welcoming all participants and explaining that the aim of the workshop was to exchange ideas and perspectives, in exploring ways to make web accessibility a reality.

He explained that the workshop was organized through the EC funded project WAI-ACT, which has a number of different aspects, including expanded cooperation in Europe, which is a key point because we all need to work together to implement web accessibility. Other aspects of the project include developing authoritative guidance to support implementation of the WAI guidelines, developing evaluation methodologies, and supporting the coordination of research and development activities. More information on the WAI-ACT Project can be found on http://www.w3.org/WAI/ACT.

He invited participants to discuss the different approaches that they or their organizations have taken in referencing and adopting WCAG2, discuss policies referring to it, and share resources. One aim is to identify priorities for further development of resources, to further accelerate the implementation and the adoption of WCAG 2.

He went on to give an outline for the day, describing the three session topics, introducing the panelists and presenting the format as an interactive exchange to gather perspectives from everyone, culminating in tangible results that can be carried forward into future work.

Finally, he gave the intended outputs for this workshop, which include a W3C Workshop Report summarizing the discussion. This may provide further input into other W3C work to help identify how to better address priorities and to stimulate complementary collaborative work to accelerate the deployment of accessibility. It should also contribute knowledge and common understanding of the way forward and what is needed to move ahead.

Opening statement from Judy Brewer, Director of the Web Accessibility Initiative

Judy Brewer welcomed participants from many different countries and thanked them for attending.

She hoped that this workshop would provide an opportunity to increase standards harmonization in the area of web accessibility, as this is a key to help accelerate the pace of accessibility on the Web. This would be achieved by collaboratively identifying barriers in the take-up of web accessibility standards, in implementation scenarios, and in policies which can be a key driver of increasing awareness and implementation of web accessibility and exploring together some ways to reduce those barriers.

She asked participants to share implementation challenges and best practices from their own perspectives. The solution to a challenge for one country or one particular group may already be solved by another group. So the hope was that some of the thoughts and challenges could be addressed by others and that some of those strategies could be shared with the Web Accessibility Initiative (WAI), so that they can be built into supporting resources that people in other countries can learn from.

The Web is the access to the world. For people with or without disabilities the Web has become far more important than when the Web Accessibility Initiative was started in 1997. At that time it looked like a resource that would be critical for accessing many different kinds of information and interacting with many different kinds of services.

It has now become "the" premier way of accessing news, information, online learning, employment opportunities, engaging in the activities of employment, civic participation, healthcare, telemedicine, social networking, and entertainment. It is becoming a channel for accessing television. It is already heavily used for video. Location based services such as on mobile devices; also digital publishing, speech telephony, 3D simulations, immersive virtual environments/reality; all of this can be accessed on the Web now, including the Web on mobile devices. In some cases the Web is the only way to access them and to participate fully in society.

It is critical that people with disabilities have access to all aspects of the Web. Any time there are barriers, it may mean critical delays in getting educational materials, getting employment opportunities etc.

This is one reason that the World Wide Web Consortium, W3C, has been working on web accessibility since fairly early on in the development of the web. W3C started in 1994. The Web Accessibility Initiative (WAI) started in 1997. So we have been working more than 15 years on web accessibility, and there is still a lot of work to do.

The World Wide Web Consortium encompasses an expanding range of technologies. Typically there are over 60 different Working Groups operating at once, developing or updating existing web standards. There are over 100 community groups exploring potential new technologies or exploring ways to bring existing technologies to converge onto the Web platform.

Web Accessibility Initiative, WAI, has had a multi-stakeholder approach since it started, including industry, disability, government and research perspectives. It is consensus based in that it tries to address solutions that will work across all stake holder groups. It is community responsive. In some cases, as in WCAG2, it responded to hundreds and potentially into the thousands of comments. It uses an open process. Almost all of the work happens in working groups that are publicly visible.

The Web Accessibility Initiative includes six major areas of work, ensuring accessibility support in web technologies. It looks horizontally across all W3C's work. It provides guidelines and standards for web accessibility. Some of the workshop discussion will be related to the guidelines work, particularly in web content.

There is a very robust education and outreach effort within WAI. That is important because there are still a lot of people who may not understand the importance of web accessibility. The organizations that participants need to work with may need more and more effective, educational materials.

Standards harmonization is something that WAI has spent a lot of time on because it is such an important influence for the adoption rate for web accessibility. Why does standards harmonization matter so much? One of the reasons is that it can impact web accessibility. There is a resource called "Why Standards Harmonization is Essential to Web Accessibility", it has a draft that was created last year and needs updating.

When the uptake of web accessibility standards is fragmented, it incurs additional cost on developing alternate versions for particular local areas, organizations, or for government, as there may not be support for the alternate version in authoring tools and evaluation tools.

The support in authoring tools for web accessibility can be very productive as where one is implementing to a unified market of web content accessibility standards, and when there is support for production of accessible content in authoring tools, then the web developer can have more automated support for production of accessible content. This also applies to evaluation tools. If one is validating conformance to a unified standard, there are economies of scale. When there are separate versions, one loses one of those benefits.

When different varieties of standard are taken up, there can be reduced interoperability with mainstream technologies and with assistive technology. It can also be difficult for organizations with international audiences to track the conformance requirements across multiple audiences.

So WAI encourages people to train and implement locally but to standardize globally.

Web technologies evolve very rapidly. Working groups at W3C are updating and developing new standards and coming up with all kinds of ways to make the Web work in different environments, handle different applications etc.

The job for web accessibility is to run ahead of all of that and try to make sure that it's all accessible.

In the Open Web Platform there is document object model level 3, Math ML and probably a family of 20 or more different technical standards that provide a fully interoperable platform, that's essentially a computer moved onto the Web. All of that needs to be made accessible. There are also accessibility specific standards that are part of that Open Web Platform, including WAI-ARIA, the Accessible Rich Internet Applications and Indie UI. HTML5 is currently in Candidate Recommendation, in the implementation testing stage, but it's already being used heavily. And it is scheduled to be completed in 2014. It is being used particularly for the mobile environment.

There are many improvements in HTML5 that can help accessibility. If this is done in a harmonized way, it's a full programming environment for cross platform applications and access to device capabilities, for any kind of mobile device. It allows for integrated video, animation, graphics; it allows for style, typography and other tools that can be used for digital publishing; supports plug in free rich media games, so if harmonization is done right, there will be accessible gaming online; in terms of accessibility, it means a broader range of functionality to support in a smaller but higher performance footprint.

How can all of that be made accessible? Part of it is by having an ecosystem for web accessibility. This includes a technology foundation with universally designed technologies and interoperable APIs, the application programming interfaces. Those universally designed technologies are ones that WAI is internally cross reviewing at certain stages of their development.

It means having a clear motivational context for policies so that there is uptake of accessibility guidelines in information technology policies. It means clear requirements for accessibility, a well understood business case for it. It means implementation support such as the educational materials, training, demos, sample code etc. It means validation and conformance so that people know what to check for in the specifications so that the interoperability actually works.

The guidelines for accessibility are a key part of that. If one thinks of the whole web development scenario, there are developers who are using authoring tools and evaluation tools to create content that is accessible. There are users who are accessing that content through browsers. Sometimes using assistive technologies or existing capabilities of the browsers. The Web Content Accessibility Guidelines (WCAG) defines a unified standard for how that content can be as accessible as possible across a broad range of disabilities.

There are guidelines for the authoring side, the Authoring Tool Accessibility Guidelines (ATAG). There are guidelines for the browser side, that includes web browsers, mobile devices, media players etc; the User Agent Accessibility Guidelines (UAAG).

Then there is the technical foundation, where we are doing the cross review.

Those complementary guidelines are part of the ecosystem that we have developed. When this is applied to different specific environments, such as mobile, using the same set of standards, the whole thing can work together well. There is accessibility support built into HTML5 and other portions of the Open Web Platform. There are many different WAI resources that inform different aspects of mobile accessibility. We have a resource page on mobile accessibility. This explains some of the existing coverage. WCAG 2 does apply to mobile and WAI-ARIA and IndieUI actually help.

There are also media capabilities in the Open Web Platform that are new. There is a harmonization issue that was not completely resolved that has to be made up for now on the accessibility side with competing web standards, Timed Text Markup Language (TTML), SMPTE-TT and WebVTT. So we have to ensure accessibility for each one of those. The Media User Accessibility Requirements goes into the detail of what's needed on accessibility. There are captioning capabilities that W3C has been building in, partly driven by some of the regulatory discussions in one particular country, in the U.S. recently. The accessible digital publishing work, is an example where you can see the Open Web Platform being used with six or seven different standards, Daisy accessibility, HTML5, MathML, and JavaScript to create accessibility, all relying on the same Web Content Accessibility Standards.

In the Open Web Platform environment, the unified adoption of accessibility is essential. There are several different resources to support harmonization that have been developed over time and that are being updated. WCAG 2, the Web Content Accessibility Guidelines 2.0, became an ISO standard in 2012. For some countries this makes it more directly adoptable. Referencing WAI standards can resolve some of the questions on how to reference WCAG, if that is an issue. Guided use of WCAG techniques is being developed now because there have been questions on that.

In this workshop there are three panels. The main questions to think about are what are the implementation challenges for WCAG 2? What is the current state of international harmonization? How can we improve that and move that forward to gain more of the benefits of a standardized, harmonized adoption of WCAG 2? What specifics arise when applying WCAG 2?

She asked participants to consider: Are there challenges they can share solutions when they hear their colleagues mention challenges? Are there specific ways that W3C WAI could expand or shift some of the support resources for WCAG 2?

Session 1: Implementation Challenges

Elin Emsheimer, EC DG Connect, on the proposed EC Directive

Elin Emsheimer from the European Commission, DG Connect, gave a brief overview of the proposed Directive on the Accessibility of Public Sector Bodies' Websites announced in December 2012.

She explained that the background of the proposal was that a non-functioning of the internal market was identified. So the proposal legal basis is the internal market with the fragmentation that Judy touched upon, as well as with uncertainty of procurers and web developers.

The proposal itself consists of five main elements. Firstly, it establishes harmonized accessibility requirements for a set of public sector bodies' websites. These requirements are in line with WCAG 2.0.

Secondly, the scope: The proposal covers a set of websites -owned by public sector bodies- offering 12 essential types of services to citizens.

Thirdly, the use of standards: The proposal includes a presumption of conformity with future harmonized standards that point to WCAG 2.0.

There are also compulsory monitoring and reporting obligations for the Member States to the Commission as to accessibility and additional measures encouraging Member States to extend the scope and its effects to further increase accessibility.

The proposal is currently being discussed at the Council and at the European Parliament.

Chair: Dominique Burger, Association BrailleNet

Dominique Burger, introduced BrailleNet, the not for profit association the he Chairs. It was born in 1998, at a time when the web accessibility question was raised by the W3C. It has seen the birth of WAI, the first publication of the WCAG in '99 and has been taking part in activities of the W3C, mainly with Sylvie Duchateau attending.

The Association takes part in consultation activities and promotes the WCAG in France through training and documentation. It also has a program for certifying the conformity of websites to WCAG 2. After more than 10 years of work on web accessibility, he asked: Why is the progress so slow?

He related an instance when the project Directive on Web accessibility proposed by the European Commission came to France and was examined by the Senate, which then published a recommendation that France's government should not support the directive. French associations are campaigning for this directive, as a step forward if it is voted and implemented. He questioned why politicians can publish something against a directive that is obviously necessary.

The understanding of accessibility, and its impact, he felt, was obviously very poor among politicians and decision makers, and their misunderstanding is a big obstacle to the progress of accessibility. Currently, if web accessibility is discussed with those people, they have no real idea of what it represents in terms of the expectations of users. They consider it as a supplementary burden, not really something that may help people.

He suggested that discussion points could include the expectations of users and how they can be formulated in terms of technical requirements. For instance, if a person with a disability has to consult a train time-table on a website, what does that mean in terms of technical requirements? The connection between standards, which are obviously necessary, and the understanding of how it helps people should be improved.

Other points to consider include concerns about skills, training, and competence certification. He said that almost no university provides any training about web accessibility in computer science curricula. Another question is whether people have the tools they need to develop accessible websites and web services efficiently.

Monitoring is another challenge. How people, especially decision makers, can get a correct image of the general level of accessibility.

He posed the question on the relationship between expectations and standards, particularly WCAG 2, and how to formulate these expectations in terms of functional requirements.

Panelist: Alejandro Moledo Del Rio, New Technologies and Innovation Officer, European Disability Forum (EDF)

Alejandro Moledo introduced his organization, EDF, which represents 80 million Europeans with disabilities. Its mission is to ensure that persons with disabilities have full access to human rights through involvement in policy development and implementation in Europe. One of their top campaigns focuses on freedom of movement; including an objective to promote legally binding legislation regarding accessibility of websites.

Discussing political commitments on web accessibility, he cited the clarity of guidance in the United Nations Convention on the Rights of Persons with Disabilities. However, he said, in the past 10 years, non-binding law has not delivered its promises, exampling the 2006 Ministerial Declaration of Riga and the 2010 European Union commitment to deliver website accessibility by 2015 under the Digital Agenda for Europe. In spite of which, at present, only one third of public authorities' websites are made accessible.

Now there is a new European Commission Proposal for a Directive on Accessibility of Public Sector Bodies' Websites which is more ambitious and widens the scope of the proposal. One of EDF's main concerns is the implementation of this legislation.

EDF has proposed that each Member State appoints an enforcement body to be responsible for the implementation and monitoring of web accessibility of the websites concerned. This enforcement body should involve persons with disabilities and their representative organizations. This would cover the three key factors needed for successful web accessibility implementation: political commitment followed by binding legislation; political and social will; and, involvement of civil society, in this case persons with disabilities, their representative organizations, web developers, researchers, etc.

Distinguishing between political commitment and political will, he said that experience shows that a strong political will is more effective than political commitment. This was exampled by the law against discrimination adopted in Spain in 2007, which included web accessibility and established sanctions of up to one million Euros. Although public authorities and large enterprises websites were repeatedly denounced for the lack of accessibility, these sanctions were never used. So the political commitment of legislation exists, but no political will to implement it. However, he argued that a will to implement web accessibility, without a political framework nor binding legislation would result in confusion.

He stressed the importance of involving citizens in the decision making process, as disabled user feedback and perspectives can guide industry and the policymakers to the best solutions.

The current task is to keep on raising social awareness of the importance of web accessibility, as well as pushing for better political commitments and legislation and a strong will to implement them. Finally, of course, stakeholders need to be ready to participate and take action whenever it is needed.

Panelist: Susanna Laurin, Funka Nu AB

Susanna Laurin from the Swedish consultancy Funka introduced her organization, which was founded by the disability organizations in Sweden in the mid '90s but became a privately owned commercial company in 2000. They specialize in accessibility and user experience.

Funka is in the final stage of delivering a study for the Commission, monitoring e-accessibility across Europe. Previous reports were called MEAC 1 and 2, this is MEAC 3. Together with Empirica in Germany, WRC in Ireland and Spanish Technosite, Funka made a study on web, TV and telecom throughout the 27 European Member States, plus USA, Canada, Norway and Australia. The Web work was performed by Funka.

Twelve websites in each country were studied, nine public websites and three commercial websites. Ten indicators from WCAG 2 level AA where chosen as test criteria. The full report is 200 pages, and is due for publication in spring 2013.

The study shows that countries with law enforcement generally have a better performance and that monitoring is very important, and the combination of the two makes for greater success.

There is also a difference in how the laws are written. If they are too technically detailed, there are two problems. One is that website owners tend to see them as the goal. As such, they aim for technical specifications only and they believe that this law enforcement is enough to make websites accessible. Also, laws with a very specific technical detail tend to be outdated very fast because of technical development.

Law is not the only way to be successful in implementing web accessibility. There are countries that score very well in this study without law and only policy and awareness, in combination with monitoring.

The top scoring country, the UK, has a mix of law enforcement, good policy work, and lots of awareness, strong organizations, and a market with both nonprofit and commercial consultants working with implementation. The combination of these elements really is a success factor.

UK still has lots of problems though, in our very small study, it scores less than 75% of the extremely few indicators that we have measured. So we all have a long way to go. Even with 100%, if any website had reached it in this study, it probably would not have been accessible.

The overall result is surprisingly bad. We really thought it would be better. It is very saddening and tiring to read the whole report because some of it is so bad.

Some of the problems the study found are issues that have been known for ten or fifteen years, but still many countries score very badly. It is a little more understandable that new indicators from WCAG 2 score even lower, of course. It takes time for things to trickle down. But even the very old issues that everyone should know of, like headings, still show low scores.

Other results from the study: public is generally better than private; but the implementation between countries varies.

It is also clear that focus is affected by the strength of influential organizations, disability groups, consultants or public enforcement which varies from country to country. The focus could be on technical issues or on content issues. This means that even if most countries have WCAG 2.0 level AA as their goal for implementation of accessibility, the way the different countries and the Member States implement it still vary a lot.

One surprising trend: In poor scoring countries, the private sector websites tend to score better than the public sector websites.

The current MEAC has a broader perspective, to do tests that affect more people. It found that blind users are the ones that are best cared for generally, though even for these, it is not perfect.

The study also shows there is far too little multimedia, illustrations, pictures, sound, mainly in public sector websites, which are predominantly in text. So many people have problems with reading text; the public sector needs to change this.

With smartphones, tablets and touch screens accessing the Web, even more training and guidance is needed because developers and web owners know even less about accessibility in these newer techniques.

Websites using new or modern techniques, like lightboxes or AJAX, have huge possibilities for accessibility, but have problems due to lack of knowledge. Website owners and developers still need to be trained in how to use these techniques in an accessible way. The implementations of the new techniques found during this study are really not accessible, even where it is possible. The huge gap of competence when the newer or modern techniques are used needs to be addressed.

If we are going to have an Internet for everyone we should raise our eyes and look ahead because currently we are very focused on the public sector, paying taxes and other boring things. The Web needs to be used for many other things that are also extremely important.

Panelist: Andrew Arch, Australian Government Information Management Office (AGIMO)

Andrew Arch, participating remotely from Australia, introduced himself and his organization AGIMO. Australia's Disability Discrimination Act was introduced in 1992, so it has no mention of online service delivery, only advisory notes for the World Wide Web. However, their Human Rights Commission has a strong interest in the accessibility of online delivery.

In 2000, the Australian government adopted WCAG 1.0. It was agreed at Ministerial level that this would be the basis for accessibility of online information. In spite of quite a bit of action early on, in the long term there was limited adoption. There was no plan to insure that WCAG 1.0 would be widely adopted.

So when WCAG 2.0 came out and it was put to the government that this was the new standard to adopt; the Australian Government and all the States and Territories agreed. That was followed by endorsement from the heads of the Australian Government departments. The Secretaries' ICT Governance Board endorsed the national Ministers' adoption of WCAG 2.0, but set this at Level AA, rather than Level A.

The Australian Government Information Management Office was then charged to develop a strategy; the good intentions of adopting WCAG 1.0 had been let down by not having a formal strategy or a plan to measure the success.

The Web Accessibility National Transition Strategy was prepared and commenced in mid 2010. The strategy runs to the end of 2014, and includes a detailed plan which has three phases:

There is a lot of work to be done to position the government agencies to take on WCAG 2.0.

The Strategy also had two milestones embedded in it: The incorporation and conformance with WCAG 2.0 level A by the end of 2012 and conformance at level AA by the end of 2014. A survey at the midpoint had initial results that looked promising. Not complete conformance perhaps, but very high levels of activity.

Agencies were asked to report on each website that they were responsible for as well as giving an appraisal at the agency level, at the department level.

The Strategy does not only cover public-facing websites; it also covers the intranets that the agencies use, so people with disabilities who might be employed in a government agency, will be accommodated.

It also includes web applications, which means the government's service delivery, all services have to be incorporated in the agencies' activities for the adoption of WCAG 2.0. That is one of the biggest challenges, because very complex service delivery has a much longer development life cycle than the National Transition Strategy allows.

So, while there's optimism about the delivery of information and small services, some of the much larger and more complex service delivery may not meet the original time frame.

The strategy also requires that any new development, website or application, must be released as WCAG 2.0 conforming.

The initial challenges were that the level of awareness was relatively low across government, but the last two years of activity have seen a very high level of awareness develop. The other is the availability of skilled people to help implementation, this is increasing in Australia, but it is still one of our biggest challenges, particularly in the application development area.

Panelist: Piotr Witek, Fundacja Instytut Rozwoju Regionalnego

Piotr Witek works as an accessibility and assistive technology specialist in Poland. His organization prepared a Polish translation of WCAG 2.0, and created the Accessible Cyberspace Forum, for organizations representing people with disabilities.

Poland had no clear rules connected with accessibility until 2010. In 2009, the government announced that it would prepare an amendment to the Act on Computerization of Entities Performing Public Tasks. This was identified as being an opportunity to include accessibility into Polish legislation. Piotr's organization took part in the initial working group and in 2010, got the amendment that all public websites have to be accessible for disabled people. This was only the beginning as they wanted to include WCAG 2.0 to the terms.

They joined the next working group, explained what accessibility means and its importance, organizing courses for the public administration. In 2012, the regulation with operational details showed that, not only web pages but all public data on computer systems have to be accessible and compliant with WCAG 2.0.

This wide regulation covers all institutions that take public money; all have to be compliant with WCAG 2.0. The regulation set the level of accessibility at level AA with two exceptions, connected with live transmission. It started in May 2012, and requires compliance by 30 May 2015.

This is only legislation though. They know that reality is different. At the end of 2012, they checked 3100 local administration websites, and couldn't find any which were in full compliance with WCAG 2.0. Armed with the regulations, they still have to prepare training courses for the web developers and explain how to understand WCAG 2.0.

He closed; hoping that by 2015 Poland will have much more accessible websites and ICT systems.

Open discussion

Dominique Burger opened the discussion to all participants with a question on user expectation. Is compliance with WCAG 2, at Level AA, enough in terms of functional requirements? Does it match the expectations of users?

Differently expressed, when formulating an expectation in terms of functionality, how do you translate it into technical requirements? Is it a problem you encounter on websites or when discussing with web owners or politicians or decision makers? And how do you tackle this question?

Raph De Rooij from Logius, part of the Netherlands government has had experience since 2004 in measuring accessibility. The outcome of measurements is: you comply, or you don't. When there is no full conformance, the question why non-conformances occur cannot be answered. He asked if it wouldn't be a good idea to focus on the reason for non-conformance.

If the reason is known, it is possible to find solutions. There isn't much information on this right now that could let us effectively steer for improvement in web accessibility and achieve more sustainable results.

Andrew Arch said that the same issue has been experienced in Australia. There is an expectation that it is all about conformance. They are trying to move toward accessibility embedded into 'business as usual' so it just becomes part of everybody's normal work, rather than something treated separately.

Their recent survey asked about what progress is being made across a variety of what are perceived to be issues such as multimedia, forms, etc. and identify the biggest problems. The aim is to target some education events and materials to help agencies address the issues that they've reported.

The aim is to get a lot more information to be able to progress the education and adoption of WCAG 2.0 as a normal process of development.

Detlev Fischer from Germany, Dias GmbH identified a difficulty in getting people to adopt WCAG is the approach of seeing conformance as a yes/no issue. In practice it's Often very difficult to tell whether websites conform or not. The guidelines are fairly general, even the techniques have gaps and room for interpretation. Two people evaluating a website can reach very different results. This isn't made easier by the new JavaScript-based techniques which can be difficult to assess.

In the German implementation of WCAG, called BITV, the test procedure has a graded assessment scheme, where each criterion is assessed across a number of pages and given a score. When applied to websites that have put a lot of effort into being accessible, the result is often a score of 90 points out of 100, which is considered fully conforming. The German test procedure acts as an incentive for websites that might otherwise be discouraged from making efforts, as 100% conformance is so rare and difficult to achieve.

Detlev also mentioned WCAG Evaluation Methodology (WCAG-EM) which is in draft and may be used as the reference method for Europe, checking web accessibility and determining conformance. He noted that while it has rules for setting the scope of an evaluation and setting and defining the sampling method, it does not mention the actual assessment, which reflects the problems in assessing real content. Opinions on ways to assess criteria differ so widely, the methodology does not touch it at all, leaving a real risk that the methodology itself may be imprecise. It is a question of how the criteria are operationalized.

Susanna Laurin agreed that monitoring and conformance are important, but they are not interesting to website owners, the customers. To succeed in interesting them in a good quality user experience and accessibility on their website, something more than standards or laws have to come into focus because it is not all about checkpoints. It is about quality. When the customers understand that they are working for a more efficient website, they can earn money or save money, and not just because they have to, it makes it positive to say the solution is going to be better for all of their customers. So while monitoring is very important, to be successful there needs to be another scope or another focus.

Dominique Burger broached two questions often expressed by website owners or developers: If our website is fully compliant, if we have succeeded, will we be sure that users will be happy enough with that? The other question is if we do not comply, can the user nevertheless do something or what can they do?

As these questions are currently difficult to answer, he suggested that some mapping of correspondence between functionality and compliance is needed.

Reading the statement by the commission, he was very happy to read the term "functional requirements" rather than compliance.

He said that what is wanted is usable websites, websites that can be used, and asked what does this mean in practice?

Kai Morten from the Ministry of Labor and social affairs in Germany introduced his organization and its responsibilities. He said that some practical challenges have been noted in meetings with online editors from federal ministries. In spite of guidelines and implementation guides for the BITV 2.0 ordinance, the size of online teams means that it is difficult to find the capacity to undertake accessibility measures as part of daily work. Multimedia is being used much more frequently, but it isn't always possible to provide subtitles or audio description in a timely manner.

Another problem is how to deal with large volumes of legacy PDF documents. It takes a great deal of time and capacity to transform these into accessible files. As a result many PDF documents are not as accessible as required.

Jorge Fernandes introduced himself and his organization, the Access Unit of Portugal's Ministry of Education and Science, which has been working in web accessibility on implementation in public administration since 2000. He said that in 2011, Portugal's 350 websites in central public administration bodies were checked and 75 percent were conformant at level A on the first page. He contrasted this with the same samples checked in 2006, when there was 25 percent of conformity level A in the first page. He attributed this great improvement to having a champion in central public administration.

After the 2006 results the champion concentrated efforts on conformity, in international benchmarking, and analyzing how automated web accessibility was measured. The delivery was then to try to respond to the metrics of the automatic evaluation tools.

However, he said that complaints from users persist. Users still fail to complete forms, such as the tax form, though the portal of finance claims AA compliance. However, as this is behind a login, it is not testable by automatic tools.

He welcomed the Directive's stress on twelve types of service; which could improve the services that users find difficult to manage, but believes that usability tests with people with disabilities are needed, and emphasized that these should not be confused with the opinions of two or three users.

Raph De Rooij commented that there have been some conflicting views expressed so far. First is the rules based approach where WCAG is often implemented. The outcome is either a pass or a fail. An alternative is a principle based approach, advocated by other people. When it comes to accountability, a principle based approach is much more difficult to make legally sound.

Another difference is conformance and compliance, both used in the same sense, though they are not the same. Conformance is fully implementing a norm and compliance is alignment with a policy. When the norm and the policy are identical, then compliance and conformance are the same. But it is not necessarily so. When there is room for explanation, then compliance and conformance are not identical.

The third difference also has to do with the rules based and the principle based approach. Rules based is favored by positivists and principle based by social constructivists. 'Truth is absolute and needs to be found' versus 'truth is relative, it is what we agree upon'. These two views are mixed and mingled in the discussion; it can be valuable to make a clear distinction between the two.

He felt that there should be a more common understanding of some very basic concepts.

Andrew Arch responded that the Australian strategy concentrates on the public sector, but they find that the private sector follows government. Banks probably lead within the private sector. The Human Rights Commission has a Web Advisory Note that recommends to all sectors, government, private, education and so on, that WCAG 2.0 is a good basis for not getting into trouble with the discrimination law.

He invited comments from the floor about WCAG 2.0 and adoption in the private sector. He also posed a question for Susanna, who earlier mentioned that in some of the low conforming EC countries, the private sector is doing better than the public sector and wondered if she had any reasons or observations why that might be so.

Andreas Cederbom answered on behalf of his colleague Susanna, as he leads the Funka Nu work on international studies of websites.

Many poor scoring countries are also countries with a struggling economy; where there may be a lack of both focus and money in the public sector. Commercial organizations such as newspapers need to be professional and will spend money to update and work on their website as well as their paper product. However, governments need to prioritize spending on the day-to-day work in the societies rather than spend on their websites. So we see private sector websites scoring better than public sector websites in poor countries. Private sector websites also work on search engine optimization, which in the process, improves accessibility without there being an accessibility focus.

He added that, to succeed with accessibility work, there needs to be an organizational shift in the view of accessibility. WCAG is a tool; it's the guidelines, not the goal. It is needed to achieve accessibility, but this shouldn't be the sole focus of the work. The focus of the accessibility work should be to get people in the organization to think how to write text that can be read by users outside the organization. How to create physical environments where people can get around even if they are in a wheelchair or have other types of disabilities? It should be a culture shift rather than a conformance requirement.

Dominique Burger commented on how technological evolution should impact the requirements concerning accessibility in coming months. The advent of HTML5 and WAI ARIA will change the way accessibility is approached and force consideration of the question of how people use our websites. The movement is towards including users in the process of evaluating websites.

Session 2: International harmonization

Chair: Donal Rice, Centre for Excellence in Universal Design, NDA

Donal Rice, who works with the Centre for Excellence in Universal Design at the National Disability Authority, introduced himself as a public sector worker for the Government agency in charge of disability affairs in universal design.

Summing up the approach to implementing WCAG in Ireland, he said that the 2005 Disability Act requires all public sector bodies to ensure their websites, their electronic communication more specifically, are accessible. Underneath that is a code of reference that refers to WCAG. The version of WCAG was left intentionally vague because a new version was planned.

He characterized WCAG implementation in Ireland by saying that the Government department that has the mandate is very unenthusiastic about it. And the Government department that is most enthusiastic about it has no mandate in this area. So there is a little bit of a mismatch there.

Panelists: Marsha Mazz and Bruce Bailey, (US Access Board)

Introducing the first panelists for this session, Donal Rice welcomed both Marsha Mazz and Bruce Bailey. Marsha is the director of the office of technical information services at the U.S. Access Board, but to give its full title it is the U.S. Architecture and Transportation Barriers and Compliance Board. Bruce Bailey is an accessibility IT specialist within the Access Board and has contributed to work that the W3C have conducted.

Marsha Mazz introduced the U.S. Access Board as a very small U.S. Government agency, with a very important role in establishing the accessibility standards for Section 508. The first standards were established in December 2000. A small effort at harmonization was made then, by repeating many of the WCAG requirements in the 508 Standards, which resulted in much overlap with WCAG. So there has been a relationship with WCAG since 2000 and the Access Board want to strengthen that relationship in the future.

An Advisory Committee was formed in 2008 to help to revise Section 508 Standards. Harmonization among the various standards for accessible ICT was very important to the Advisory Committee. Previous to that time, the Access Board had made a commitment to harmonization of all of our guidelines and standards.

The Advisory Committee moved a little closer towards harmonization. When the first advanced notice of proposed rule making was issued, full harmonization hadn't been achieved.

In 2011 a second notice was issued, making a direct reference to WCAG 2.0, rather than incorporating a unique version of WCAG requirements. By using WCAG 2.0 success criteria for content on websites and also applying it to documents and software it went beyond harmonization. The conditions were ripe for that move given the similarity in the desired outcomes.

The first official proposed rule will be issued, probably in Fall 2013. The comments to the 2011 advance proposed rule have been positive and has resulted in even more commitment to harmonization.

With all that attention, things move a little cautiously but with due consideration and lots more input.

Bruce Bailey said that he found it interesting how much having legislation made such huge differences in the accessibility sphere, especially in products from the larger companies. The first Section 508 came out almost under the radar with only the accessibility players paying attention. The refresh has had much wider attention and participation. But, with all that attention things have to move with due consideration and lots more input.

One of the results of the extraction is that it was decided to use the success criteria not just for web content but for documents and software. That was quite controversial, but also quite inspiring; it got all of these industry players to come together and talk about our standard and do a terrific amount of work reviewing the proposal for us. It would not have happened without the W3C facilitating the conversation, and having a neutral and open platform for those discussions.

He also mentioned the importance of harmonized standards. There is an international building code, but for ICT, the harmonization is even more important. Buildings are thought of as universal, but the software industry is more worldwide than the construction industry and has a faster turn-around of rapidly changing products.

Panelist: Inmaculada Placencia-Porrero, EC DG Justice

Donal Rice introduced the next panelist Inmaculada Placencia-Porrero, Ima. Currently Ima is deputy head of units for the Rights of Persons with Disabilities within the DG Justice. He asked Ima to give introductory comments on the importance of harmonization from the commission's perspective.

Inmaculada Placencia-Porrero shared a number of messages on why international standards and harmonization are important.

In parallel with Member States to have consensus and common solutions in Europe she worked on the initiative called eEurope, in which Member States voluntarily committed to apply WCAG 1.0. The challenge of how to materialize this soon became clear. The implementation varied widely across the Member States from voluntary to mandatory, from sticking strictly to technical solutions or coding the commitments in legislation. A big challenge is how to build in the necessary mandatory or compulsory character without losing the flexibility needed to change the standards so that the evolution of technology is considered. At the same time, how to make sure that there is no fragmentation being introduced in the process. Europe has Mandate 376 which can be an instrument that can be referred to in legislation by the Member States.

Around 2005 the need for a reference document at EU level was realized so a standardization process was started. Working together with the Access Board and W3C enabled learning from their experience and their practical knowledge, bearing in mind the need for entering in to a globalized market with globalized solutions. Some technical solutions still need to be found: Once we have a standard, how to evaluate compliance to it. In 2003 a Council Resolution referred to the need to develop a common methodology for evaluation of accessibility to achieve comparable data but it hasn't happened. There are many methodologies at a national level.

A request has been put in the Mandate 376 to develop the methodology, together with WCAG. Knowing that W3C is working on it and having seen the software applicability of WCAG she urged that a solution is found soon. Time is pressing.

Another challenge for harmonization is that each Member State and the U.S. have different policy and legal making processes that differ widely.

She noted the need to coordinate issues of timing, issues of the way in which the law is being developed, what is compulsory and what voluntary. Section 508 standards have a mandatory character which is different from Europe. European standards are in principle voluntary. But in some Member States the technical rules for accessibility are coded in the law. Further work on coordination is needed.

She then referred to the fact that the European Commission's proposal for the revision of the public procurement Directive makes accessibility compulsory, more in line with the U.S. Harmonization in this context is a very important element. The web accessibility directive puts obligations on the results but not on the process. Internally the commission is working to align accessibility across different directives. She also reminded participants that the EU is now a party to the UN Convention on the Rights of Persons with Disabilities and its obligations.

In April 2013, the Committee on the Rights of Persons with Disabilities set up jurisprudence for how the Convention will be interpreted. It established an important link between accessibility and discrimination. It explained how to interpret the provisions on the Convention referring to the need to have legally enforceable obligations. The Convention deals with obligations for the ICT including the Web both for the public and the private sector. So how this "view" of the Committee is going to set up a precedent for development of accessibility is very important. EU legislation on nondiscrimination in the area of employment has been interpreted by the European court with a reference to the UN Convention. The Convention is now part of the EU legal order and these rulings set a precedent from a legal perspective to accessibility. We will continue to work together to have a coherent implementation of the Convention in the EU and that also applies to the Web.

An initiative that aims to have this coherence on accessibility is the European Accessibility Act. It looks to accessibility from an internal market perspective, assessing the situation in terms of fragmentation in accessibility rules, laws, standards, guidelines and support. It aims to find out how to improve the situation providing clear rules that industry can benefit from in larger markets in Europe and by harmonizing, globally.

said that without European Commission policy work, Ireland would give low priority to implementing. Ireland is involved in the research by Funka, WRC and Empirica to look at questions around the Directive and what practically needs to happen for Member States to prepare for the transposition of the Directive. Also through interviewing policy makers and web managers in the public sector, what types of evaluation would work within the context of a transposed Directive within their own national context. A methodology to evaluate that would not just give sensible answers as to how accessible are their websites but also act as a force to motivate further action in this area.

Panelist: Mike Pluke, Castle Consulting Ltd.

Donal Rice introduced Mike Pluke, the founder of Castle Consulting and a colleague on Mandate 376 work. Mike is the Vice Chair of the ETSI human factors group that has developed the standards for ICT accessibility. He invited Mike to share his perspectives on harmonization of standards.

Mike Pluke, picking up from Ima, mentioned that developing the standard in response to EC Mandate 376 was an initiative of the Commission. Funding was essential to support the large team needed for the work. There were 12 in the team, covering a broad spectrum of stakeholders.

One of the focus points was global harmonization, which is in everyone's interest as many companies taking part in public procurement have a global market and don't want to operate to different standards across different marketplaces.

One of the aims was to insure that any European standard is harmonized with what is happening elsewhere - in particular with the US Section 508. As the standards were being developed close attention was paid to Section 508 as they had just published their first Advanced Notice of Proposed Rule Making (ANPRM). This was taken into account in developing the first draft and early drafts.

The Access Board then made the radical decision to suggest that WCAG 2.0 could be applied to non-web ICT, to documents and software. The first reaction was that this was not going to be easy or maybe not even possible. Work on checking its feasibility found that it did appear to be feasible. The initial investigation inspired a reassessment of WCAG 2.0, going back to first principles and thinking what does this really mean was a very refreshing exercise and found that the WCAG 2.0 principles and the success criteria seem to be very robust. A draft standard was developed to apply WCAG 2.0 to non-web ICT. Initially there was a dialogue with people who had been involved in WCAG and in Section 508. Then W3C took the initiative to set up theWCAG2ICT Task Force, with the sole aim of getting a common understanding of how to interpret WCAG 2.0 when applied to non-web ICT.

Mike was invited to be a co-convener of that Task Force. Many of the people involved helped the development of WCAG 2.0 in the first place. This work has caused them to go back and look again at the detail in WCAG 2.0 to see to what extent its applicability could be broadened. The WCAG2ICT Task Force came to the same conclusion as the ETSI team, that WCAG 2.0 in general, and its success criteria, is very well thought through in the sense that it can be applied in a wider context than the Web and still work very well. The draft standard supporting the Mandate 376 is now in a public enquiry process.

Donal Rice amused participants by asking Mike if February 2014 could be a nominal date for finalizing the standard.

Mike Pluke replied that it depends on what comments are received.

Judy Brewer made an important clarification: The WCAG2ICT work will not be a standard. It is explicitly informative. It is how WCAG can be interpreted to be applied beyond the Web. It is beyond W3C's purview to issue a standard on non-web technologies. The work of the WCAG2ICT Task Force is important in order to ensure that the interpretation of WCAG 2 does not change because there is so much shared space. The Task Force, through the WCAG Working Group, will be producing an informative Working Group Note that will be useful to many different organizations.

Donal Rice had been jokingly referring to the finalization of EN 301549, but recognized the importance of not mixing up terms like guidelines and standards and recommendations.

Panelist: Raph de Rooij, Logius

Donal Rice went on to introduce the next panelist, Raph de Rooij. Donal had visited Raph's team in 2009 and admired the rigor of the Dutch approach with many layers, not just technical specification and a set of guidelines, but many things built on top of the resources, methodologies, methods of conformance for the various actors in the web accessibility chain. He asked Raph to talk about the Dutch approach, but also posed a question: Are the Dutch guidelines a set of fragmented guidelines from WCAG? Would that be a fair point to make?

Raph De Rooij from Logius, part of the Ministry of the Interior and Kingdom Relations, the Ministry responsible for ICT in the Netherlands government, has been involved in web accessibility since 2001, including the awareness project for web accessibility, Drempels weg, (Barrier Free).

In 2004 he started a web guidelines project, to provide a solid basic layer for web accessibility, including guidance on best practices that together make a website great. Inspired by Tim Berners-Lee statement, "The power of the Web is in its universality" they include 'universality guidelines'.

These guidelines were not developed with accessibility for people with disabilities in mind, because that purpose is met by WCAG. They explicitly deal with topics like separation of content, style and behavior, layered construction (also known as progressive enhancement), use of open standards etcetera. All of which can aid web accessibility for people with disabilities, but they are not developed for that purpose.

The web guidelines became normative, though this is not how they were set up. In 2006 questions in Dutch parliament about web accessibility resulted in the web guidelines being made mandatory for new web projects by Government. They include WCAG. So WCAG also became part of the set of standards for Government.

A lot of time and effort was put into documenting how to evaluate websites and measure conformance. It was not a Government effort alone, but a cooperation with the foundation called Drempelvrij.nl, barrier free dot NL in English.

The documentation is not part of the standard; the standard remains the same when improvements are made to its documentation. This is comparable to WCAG 2. The Dutch web guidelines have also been updated to version 2, based on WCAG 2 without any alteration, though they still contain the universality guidelines. Switching to version 2 was very easy in the Netherlands because two parliamentarians submitted a motion that ordered the government "to align the guidelines, now and in the future, to the World Wide Web Consortium". The project finished in 2010 and in 2011 it became the new Government standard.

A recently produced evaluation methodology was used as input for the WAI-ACT project. When a stable document becomes available they will discard their own evaluation methodology.

Answering the question of why international standards harmonization is important to his organization, he said, is very simple: It saves us a lot of work. We see standards as solidified cooperation. They are acceptable to most people.

Open discussion

Donal Rice introduced the next topic, consensus, illustrated by two anecdotes. The first showed that you had to be very careful when you get consensus. It might be what you want but you will have to live with it a long time and it may not turn out to be quite what you thought. Some victories can be empirical. The second one was to do with human nature. We all come from a set of conditioning backgrounds, employment histories, etc. that can make it very difficult to meet people halfway in a consensus based environment. Based on these anecdotes, he posed a question to the panelists and the floor. What two challenges facing the development of standards and in particular the harmonization of standards in the field of ICT accessibility can they see at the moment?

Bruce Bailey spoke first, identifying the challenge of a lack of availability of some standards to the people who are bound by them. Standards organizations need to sell their standards to sustain their organizations, which is in conflict with the public interest having ready access to them. This goes beyond just citizens, sometimes employees responsible for applying them don't have ready access to the standards that their organizations have purchased.

He contrasted this with the way that the W3C developed the standard, which is not only readily available it is also highly accessible, unlike many other standards which are often published in inaccessible formats.

Marsha Mazz added that in the U.S. the Access Board is not the only agency with the authority to establish standards for web content. Other agencies have differing authorities under several civil rights laws. Consensus has to be achieved among federal agencies that are issuing regulations for web accessibility.

The Access Board has committed a lot of time and energy to working with sister agencies to try to keep them all on the same page. As a very small agency; sometimes it is difficult to exert the kind of leadership needed for consensus and harmonization. Every agency that issues regulations in the U.S. must do a cost benefit analysis of the regulation it proposes to issue. Using voluntary consensus standards makes it easier to demonstrate the cost/benefit.

In addition, the U.S. is an organization of 50 states and the District of Columbia. Every state has internal authorities, each with their own ideas about how they want to provide and ensure accessibility for their citizens. Although a federal rule, such as Section 508, may co-exist with a state's rules and regulations, it does not always overturn them. So efforts to achieve consensus with all the political jurisdictions within the U.S. need to be made.

Raph De Rooij returned to the point about other Government bodies with their own regulations. In his compliance management studies, he has found some interesting insights on being dependent on the willingness of others to carry out a policy from the field of environmental planning that are decades old. He felt that it could be interesting to study how the problems that occurred then were overcome.

Inmaculada Placencia-Porrero highlighted the challenge of involving all stakeholders. That objective is raised by everyone in European standardization policy making, formally with the revision of the new package on the standards, but doing it in practice is still a challenge. The other challenge that is very particular to Europe and different from the U.S. is the issue of standards serving the purpose for which they are required in the policy and legislation. Basically at the Commission we issue mandates for a particular purpose, the technical details are developed to support the policy and legislation.

Without participating in detailed technical setting, they still need to ensure that the outcome of the consensus process serves the purpose of the Mandate. If it would not serve the purpose or is not in line with the policy, it would be a problem. Ensuring that standards developed under Mandates serve their purpose is particularly an issue in the EU due to the way standards are used in the "New Approach" legislation.

Mike Pluke made the point that not all standards cost money. ETSI standards are available at no cost as are W3C's, but that is not always the case. On harmonized standards, he said there are so many to choose from, it can get complicated. Groups that develop standards become very attached to the standards that they have developed, which can cause opposition when one standard is chosen as the one to harmonize around. This is a difficult problem to avoid.

Raph De Rooij recalled that when they were implementing WCAG2 there were two different types of reaction. Some found what they were being too restrictive, others said it was not strict enough. They had a parliamentary motion stating that WCAG 2 had to be implemented 'as is', having such a simple but strict rule may help others tremendously.

Donal Rice moved on to what measures enable policies to roll forward to the latest technical standards. He prefaced this by saying that in the Irish context, the way in which web accessibility is referenced in legislation and policy is quite strange. The 2005 Disability Act covers a range of areas to do with disability. It has to do with the employment of Persons with Disabilities in the public sector. It has to do with assessment of needs of children going in to the education system. Within that Act there is one line related to web accessibility. His own observation is that sometimes disability in equality legislation is quite a strange bed fellow when web accessibility is introduced to it. It seems to work best with dedicated Acts, such as the 21st Century Communication and Video Accessibility Act that is a vertical piece of legislation dealing with a particular sector and a particular issue in that sector. But many countries have overarching disability equality legislative frameworks. Keeping web accessibility on the agenda, trying to achieve web accessibility through those mechanisms can be quite difficult. He invited comments on what measures enable policies to roll forward to the latest technical standards.

Marsha Mazz said that at the Access Board standards have to be referenced by a specific date. They cannot adopt a certain standard and all its future additions. The public must have an opportunity to comment on the standard and any changes in subsequent editions. However, once the first referenced standard is established, it is much easier to update. Currently there are some efforts in the U.S. to allow a simpler notice and comment process for updating to the next edition, where the changes are not controversial.

She then raised the issue of harmonizing the guidance surrounding a standard. WCAG has very robust guidance but, there will be interpretations from the Access Board and others. She raised the hope that enforcing authorities will develop consensus regarding the guidance so that enforcement is consistent and uniform among all the agencies and stakeholders. The Access Board is very much interested in making everyone aware of the guidance available.

Inmaculada Placencia-Porrero thought this went to the essence of the legislation or the policy framework in which the standards are used. Europe has a technique of doing legislation called a New Approach. It says that the essential functional requirements should be in legislation, separated from the standards and technical details that give presumption of conformity. It allows for an easier way to have a lighter procedure to change these standards quickly. Changing legislation is a much more heavy and lengthy procedure.

This presumes that legislation or the policy is really focusing on the functional requirements and that standards provide the technical solutions. To be able to enforce the legislation it is necessary to assess its implementation. The separation is very good provided both instruments are complementary and that this is maintained and properly addressed.

Responding to the point about vertical and horizontal solutions; she saw the advantages and disadvantages of both. Some of the issues that have been raised are that in the U.S. the standards for the Web being done by the Department of Transportation or in other legislation. Separating those processes risks having different technical solutions. There are different approaches for accessibility for example procurement, or antidiscrimination. A common technical solution in standards increases the chance of harmonization and coherence among those solutions. It is a very difficult balance to strike between vertical and horizontal. She felt that both approaches are needed, but with an architecture that permits a good linking of the two.

Judy Brewer commented on the question of roll forward of standards: One of the key stakeholder groups for W3C work is industry. In the context of WAI work the presence of industry representatives is a constant reminder of the rapid pace of evolution of the technology, and that contribution was instrumental in the design of WCAG 2, with the layers of principles and guidelines. The normative success criteria are technology neutral while technology specific techniques can move ahead as the technology advances. The same multi-stakeholder nature of some of the regulatory processes is probably one of the best measures to ensure that government regulations that help take up the standards from W3C are an important driver for reminding everyone that the standards need to move forward at a certain stage. As long as that multi-stakeholder presence is translated to the government adoption level it may help drive some of the rolling forward.

Jorge Fernandes remarked on two issues. One is the difficulty of achieving comparable results between Member States related to web accessibility. He exampled studies of Portugal's municipality websites, which showed 3% conformance to Level A and similar studies of Norway's municipality websites showing 20% conformance at level A. Comparison of each result was almost impossible.

His second point was on the transposition of standards to national legislation. Until recently, Portugal had two periods of importance concerning legislation about web accessibility: First, in 1999 the resolution mentioned web accessibility requirements but not WCAG explicitly, in 2007, WCAG was mentioned explicitly, level A for informative websites, and level AA for transactional websites. However in the 2012 legislation, this was changed to put WCAG 2 seated in mainstream legislation, not in web accessibility legislation. WCAG 2 is now made part of the National Regulation of Digital Interoperability that applies to all public administration. His unit is a small one, maybe with this legislation they could leave others to run the implementation of WCAG 2, while his unit supported them as consultants.

Detlev Fischer identified one challenge for harmonization, that individual national assessment schemes run different operational processes. The new regulation sets out to impose a unified method: there is a risk that this method would wipe away processes that have been well established.

While he agreed that it is good to harmonize he felt there may be a risk for existing schemes, as working processes may be broken if too great a or detailed a change is imposed. This could lead to insecurity for existing testing schemes as clients may be concerned about how long they would be valid. Existing scheme operators, he said, fear that European regulation could disrupt their ways of working.

Dominique Burger, responding to Detlev said that BrailleNet in France, AbilityNet in the UK and Technosite in Spain, have measured their three different certification schemes for websites. Their conclusion is that while each has a different way to evaluate a website against WCAG 2, as it is a common reference it is easy to trust the evaluations and find correspondences between the methodologies. Harmonization is on the basis of WCAG 2, regardless of the different methods used in each country.

Kiran Kaja from Adobe commented that one problem raised by software developers or web developers is that there is not enough documentation or guidance to explain how standards apply to them. WCAG does address this issue, having supporting documentation: The understanding document and Sufficient and Advisory Techniques. It is crucial to have supporting documentation with detail that for obvious reasons cannot be provided in the technical standard itself. There may be restrictions in what can be done with legislative requirements in different countries, but one approach might be to add WCAG or any of the standards to something like a national standards registry as Jorge mentioned and refer to this registry in the legislation rather than incorporating the standard directly into the legislation.

Inmaculada Placencia-Porrero agreed, commenting that having a toolkit where guidance information would be provided in the use of the standards was one of the things that, inspired by the U.S. and Canada, were put into Mandate 376. That guidance material could be from a policy perspective and highlighted to make it more visible.

Returning to the issue of having different methodologies for evaluating and providing results; the audience agreed that having compatibility of data is essential. To achieve this there need to be common methodologies, but no process of convergence leading to common methodology is yet available. Going through the formal standardization process in Europe, it would be a solution because that would take away competing national standards. It is embedded in the European standardization system.

Posing questions for those developing national methodologies, she asked was it done because they believe that it conformed better to some particular needs at home in their national environment or because of the difficulties encountered in the process of making this wider than just national. How do they see the future of this process? There is a clear reference in the Web Accessibility Directive to having a common methodology. Are there particular needs at a national level because they are different from all the other countries? What are the issues that would justify having totally different methodologies?

Detlev Fischer offered an example: It would be a mistake to believe that a standard like WCAG, even a methodology like WCAG EM can ensure that evaluations will be done in the same way by everyone. It is not possible to operationalize at the techniques level. He gave the numerous techniques that relate to Success Criterion 1.3.1 Information and Relationships as an example. This includes aspects like headings. There is no definitive guidance on how to treat headings. Some evaluators will fail for headings with illogical hierarchy. The only failure is for visible headings that are not marked up as headings. Following WCAG strictly, it is not possible to fail a website which had completely nonsensical headings hierarchy, or it could be failed, if it were interpreted differently. The problem is that if you go down to that level of detail you open a can of worms, resulting in endless arguments between experts.

He felt that it is probably wise that the new methodology will not go in to that level of detail and just refer to the techniques, but by the same token it also means that it cannot ensure that the results of different people using the methodology will be comparable.

Raph De Rooij said that the Netherlands has a normative Committee that discusses the cans of worms. In practice, the can of worms is very well contained. They already have a common understanding on how to interpret this kind of question. Having such a forum may also be suitable on the European level, since it makes interpretation issues manageable.

Joshue O'Connor, Senior Accessibility consultant with the NCBI Centre For Inclusive Technology said that the can of worms is an interesting example, regarding headings because this is a case where if there wasn't some format or presentation to give developers a notion that headings were important regardless of how they were ordered, it would have a severe impact on the practical accessibility of a wide range of websites and applications. He suggested that arguing for convergence of methodology is secondary to agreeing on a baseline of what are the practical things we need to achieve with accessibility.

Session 3: Application of WCAG 2.0

Joshue O'Connor from the National Centre for Inclusive Technology known as NCBI, in Ireland, introduced the topics for this session: The practical application of WCAG, and what that means; and the role of WCAG with the technical standards and its various supporting documents.

Panelist: Pirthipal Singh, Treasury Board of Canada

Joshue O'Connor then introduced the first panelist, Pirth Singh, team lead at the Treasury Board of Canada since 2008. Pirth and his team are responsible for developing web and mobile standards for over 100 departments in Canada. He works to address implementation challenges in the government of Canada.

Pirthipal Singh started by saying that one of the big questions that has been posed to his team by the Web Accessibility Initiative concerns their unique adoption of WCAG 2.0 and why that approach was taken. He explained that when they adopted WCAG 2.0, they mandated the use of the sufficient techniques and the avoidance of common failures.

He then provided some background to act as the rationale for this approach. There are three major roles of his office. The first is to develop the web standards, including the standard on web accessibility. The second is to lead interdepartmental working groups to address implementation challenges, and finally, providing compliance oversight.

The government of Canada has had mandatory web standards since 2000. This included the adoption of WCAG 1.0. Over 100 departments were required to implement these standards. There were successes in some areas but challenges in others. The web standards have recently been revamped or renewed. This provided the opportunity to address the challenges.

The renewed web standards include the standard on web accessibility that adopted WCAG 2.0 in 2011. They also have Web Interoperability standards that adopted HTML5. These are all mandatory requirements. They also have a new mobile standard that basically ensures that websites and applications are optimized for mobile devices. Although not mandatory, they are also recommending the use of WAI-ARIA.

While still using WCAG 1.0 as the web standard, they found that the departments had three main challenges: The first was lack of published techniques to implement WCAG 1; the second was the need for more WCAG training; the third challenge was inconsistent application and assessment of WCAG 1 by these 100 departments, which owned 1500 websites.

So when they started reviewing WCAG 2 in 2010, they thought it was an opportunity to address these challenges and put forward four initiatives to do that.

The number 1 initiative was adopting WCAG 2. So it is a mandatory standard, but with mandated use of sufficient techniques and the avoidance of common failures to meet WCAG 2. This was important because it allowed the consistent application of WCAG by those 100 plus departments and the hundreds of teams.

The second initiative was to develop an evaluation methodology that combined assessments of statistically significant sampling of web pages to arrive at percentage compliance scores for ,departmental websites. These two initiatives helped address the needs for consistent application and assessments.

The third initiative was to use sufficient techniques and common failures as documentation for renewed training courses, utilizing all the great knowledge that WCAG Working Group has put together.

The fourth initiative was to develop an open source web experience toolkit that includes a library of components to help organizations develop websites that are accessible, usable, interoperable, optimized for mobile devices, and multilingual. The project is collaboratively developed on GitHub by over 30 departments and private organizations and is available for free to all. The toolkit is WCAG 2.0 AA compliant and utilizes HTML5 and WAI-ARIA.

The initiatives also helped them comply with the rulings of the Canadian federal courts. Two Canadian federal courts have ruled that compliance with web accessibility is a requirement under the Canadian Charter of Rights and Freedoms. Specifically they ruled that Pirth's department, the Treasury Board Secretariat, has a constitutional obligation to bring the 106 departments into compliance with the Charter.

The new approach to WCAG adoption was finalized before the federal court decisions and it stemmed from the challenges that were identified. But the robustness and the objectiveness of the approach definitely provided increased confidence in its value in their organization, especially post federal court decision.

Joshue O'Connor asked for confirmation that the Canadian standard mandated the use of the sufficient techniques and the avoidance of common failures. This was confirmed by Pirthipal.

Joshue O'Connor then queried whether WCAG is a part of a web standard that is made up of other methodologies relating to usability.

Pirthipal Singh explained that they have a suite of web standards. One standard is specifically on web accessibility. There is a separate standard on web usability, Web Interoperability and a mobile standard. Web accessibility is a stand alone standard.

Joshue O'Connor then asked if they saw this as providing a cohesive approach to generally reducing barriers in the environment for people with disabilities.

Pirthipal Singh agreed.

Panelist: Loic Martínez-Normand, Universidad Politécnica de Madrid

Joshue O'Connor introduced Loic Martínez-Normand from Universidad Politéchnica de Madrid. Loic is expert in website accessibility evaluation; he participates in ICT ergonomics in ISO and also in Mandate 376. He is president of the Fundación Sidar, which is a non-profit organization promoting web accessibility in Spain.

Loic Martínez-Normand gave an overview of the situation in Spain when WCAG 2 was published. At that time, AENOR, which is the Spanish standards organization, had a standard for web accessibility (UNE 139803:2004) which was not exactly the same as WCAG 1 but was compatible with some small differences. That national standard is mandatory for public websites and for some private websites according to 2007 Spanish legislation.

The public administration then asked AENOR to update the standard to be in line with WCAG 2. This time it was agreed not to reinvent or modify anything in WCAG 2, so standard (UNE 139803:2012) is just a reference to WCAG 2 and points out to the elements in WCAG 2 that belong to the Spanish standard. So the standard has a description of what are the requirements for A, AA, AAA. It just says "you shall satisfy the success criteria of level A, AA or AAA". It also says that if someone wants to claim conformance then it shall satisfy the conformance requirements in WCAG 2.

For the techniques and failures, there is a recommendation to use the techniques in the standard, which is a "should". However, the avoidance of failures is a requirement as part of the standard.

The standard also contains an annex with an overview of WCAG 2, rather than rewrite WCAG 2, it just points to the outside reference. The annex says that there are principles, guidelines. There is also a second annex with a transition guide from the old Spanish standard to the new one so people will know how to go from one state of their work to another one.

The benefit of that approach is that we have reached harmonization, because we have not modified WCAG 2 in Spain, but one of the challenges we have is that the Spanish legislation is not updated. So the legal obligation in Spain is still UNE 139803:2004, the Spanish version of WCAG 1. The law could have been written in 2007 to be open to new versions of the standard, but it was not. So there is a need for the administration to update the legal requirements to refer to the new version of the Spanish standard, which has not happened.

The second challenge is that Spain is also organized into regions, called the autonomous communities. Some legislation, like directives from the commission in Europe, has to be transferred to the autonomous communities in Spain. For instance, we have national legislation on sanctions and fines for websites that do not conform to the accessibility requirements, but then that legislation has to be adopted by the different communities in Spain, and it has not been adopted, except by one community. So any service that has local scope cannot be sued if the autonomous community has not adopted the law.

The problems of harmonization, evaluation and conformance claims cause trouble. In Spain there has been legislation about web accessibility since 2002. That has helped to raise awareness on web accessibility, but it has not really raised the level of accessibility of the websites. One part of the trouble is that there are so many different ways of claiming or partially claiming conformance and providing performance or lack of it, not even a way of comparing what the websites declare themselves. He thinks there is a need for a common reporting format, something like in the ISO standards, similar to common industry format for usability testing reports, something that everyone could use the same way and claim in a similar way.

He also raised the point about the slowness of the process. Spain has had legislation since 2002; websites in the public administration had to be accessible in 2005. Then a new law was introduced in 2007 that says that public websites had to be accessible in 2008. The laws have to be applied so far in the future results never really arrive.

Joshue O'Connor responded that he found it interesting hearing that in Spain using WCAG techniques also seems to be mandated.

Loic Martínez-Normand replied that the Spanish voluntary standard has recommended using the techniques and has mandated to avoid the failures. But legislation is needed on top of the mandating to use that standard, which has not happened yet.

Joshue O'Connor thought this seemed to be something of a fragmentation considering the point made earlier about the convergence of these standards in different domains.

Panelist: Michael Cooper, W3C Web Accessibility Initiative (WAI)

Joshue O'Connor then introduced the next panelist, Michael Cooper, a member of W3C's Web Accessibility Initiative. Michael has long standing in accessibility, working originally as a part of the Bobby team. He is now staff contact with WCAG, but he has his fingers in many pies, including Protocols and Formats Working Group, IndieUI Working Group, and is the editor of various specifications, including WAI-ARIA and WCAG itself.

Michael Cooper talked about the resources that relate to the Web Accessibility Initiative and gave a refresher on how all the different resources relate to each other.

Starting with the Web Content Accessibility Guidelines (WCAG), these are organized by principles and success criteria. The success criteria are uniforms of conformance, merely organizational. We provide conformance criteria which define how to achieve and measure a conformance to the success criteria. The guidelines have had to be testable and have the conformance requirements as part of being a standard and as part of the anticipated regulatory uptake.

The guidelines are a relatively short document. They are supported in large part by a document we call understanding WCAG 2.0. The other parts of the guidelines have a page that explains the rationale behind the success criteria and describes at a very high level how you would meet the success criterion, how you would satisfy it. They also provide references to external resources, when appropriate. The understanding document is updated roughly once a year. There are also references to documented techniques in the understanding documents. The references to techniques may be references to a single technique or it may be reference to a combination of techniques that together would satisfy the success criterion.

The techniques for WCAG are a separate document that describes a number of techniques, both general techniques which can be applied to any website as more of a design pattern and technology specific techniques that go into the specific technology of a specific implementation of a technology, such as HTML or script or style sheet, et cetera.

Techniques are categorized as sufficient, meaning the technique is by itself or in combination with other techniques sufficient to satisfy the success criterion. There are failure techniques which are documentation of known ways to break a conformance to the success criteria.

Techniques all have test procedures. Mostly they are written as a human testable algorithm. They define how you would ensure the technique has been tested, has been successfully met. Code examples and running examples are often provided to show what the technique looks like in action. These are also updated roughly once per year.

For a technique to satisfy the success criterion, it just has to be judged sufficient. It does not have to be blessed by the W3C. And that is an important aspect of our structure. We have different ways techniques can be produced; it was always the intention that the techniques documented by the W3C would not be the only technique that would be available in the world.

A form for people to submit techniques is provided. Better ways to submit techniques is being investigated, including a current experiment of developing techniques on GitHub.

Techniques, when they come to the WCAG Working Group undergo very rigorous review. There are not very many people active in the Working Group, so some participation from other organizations would help speed up the review process.

Other organizations can develop techniques without submitting them to the W3C. That is encouraged, but it does not happen a lot, though information is provided about how best to create such techniques.

There is also a document called How to Meet WCAG 2.0. This is basically a distillation of the understanding and the techniques documents. You can filter by technology and conformance level and it provides references to the appropriate parts of the understanding and guidelines. So it functions as a checklist as you are going through. However, even though it is a distillation, it is unwieldy. We are looking at ways to improve that document.

W3C is also collaborating with a number of the organizations represented at this workshop. A number of additional support resources are being developed, such as application notes that describe how to apply WCAG for specific use cases. An evaluation methodology is being created to provide a harmonized means of evaluating the application of WCAG. There is a document that will be updated called techniques for automated and semi automated evaluation tools to help guide tool developers in evaluating to WCAG. There is work being done on an accessibility support database to help identify which techniques are accessibility supported in which user agents, so that authors can make informed decisions about whether a technique will work in their particular situation.

New techniques are continuously developed. The focus is on HTML5 and WAI-ARIA techniques. There are new technologies coming down the pipe so developing new techniques is ongoing. New techniques developed by people who know those technologies would be very welcome. W3C is also involved in a large testing effort for the open web, developing tests. It is focusing on automatable tests. There is also a series of test web forward events that the accessibility community could participate in. So that in collaboration with the database is another way we could help improve the knowledge of accessibility features in the world.

Panelist: Kiran Kaja, Adobe

Joshue O'Connor introduced Kiran Kaja, the Programme Manager from Adobe based in UK. He represents Adobe in accessibility standards activities in Europe such as Mandate 376. Kiran is also responsible for developing and maintaining a certification training program for Adobe developers and quality engineers and also works with the product teams to help them understand accessibility requirements.

Kiran Kaja reinforced a few points that had already been made, but from the perspective of an assistive technology user.

WCAG 2.0 has gained wide acceptance around the world. To someone from the industry as well as an end user, this is a great advantage because using WCAG across different geographic settings has benefits for everybody. It makes it easier for vendors to comply with or conform to accessibility requirements. Harmonized standards benefit both industry and end user groups and other stakeholders.

Another great advantage to WCAG 2.0 is that the guidelines and success criteria are technology independent. This was one of the major issues with WCAG 1.0 which only applied to W3C technologies.

Product life cycles are really short. It can't take three or four years to develop a product or service. So it's important that any standard or guidelines be adaptable to new technologies. WCAG can apply to new technologies. One of the good examples of that is HTML5. This has a lot more semantic elements that are good for accessibility. For example, some of the things that could not be done with HTML4 like having clearly marked landmark regions. WAI-ARIA techniques had to be used to achieve that in HTML4. But it is now possible by default in HTML5.

More complex features, such as date and time input control, will be much easier to develop. In HTML4, web developers had to use or develop custom JavaScript based controls or use ones that are available on the Internet as part of a java script tool kit. But HTML5 already has date and time input control, which is being used more on mobile and things like iOS and Android. Once web browsers make them accessible, web developers won't have to worry about it

He was delighted to be part of the WCAG2ICT Task Force that aims to provide guidance on how to apply WCAG to non web contexts, such as software, mobile and documents. This initiative makes sense because the difference between web and non web is gradually diminishing. There are web apps and apps that basically rely on web content for most of their user interfaces. This hybrid of web content and applications can fall into the purview of WCAG. He strongly suggested that people take a look at WCAG2ICT.

He also commented on the high level of supporting documentation that is available, including the understanding documents, the sufficient techniques and advisory techniques. He found that developers love the understanding documentation because it clearly explains the rationale for implementing sufficient techniques.

However, he urged caution on the idea of mandating sufficient techniques. There are techniques for several technologies, PDF, flash, HTML and CSS. But some newer technologies do not have published techniques. HTML5, for instance, is being widely used but still does not have published techniques. Mandating sufficient techniques could discourage the use of technologies that have built in accessibility features.

Finally, when looking at WCAG beyond the Web context, he suggested that the unit of conformance, which has traditionally been a web page, might be discussed and clarified. It is hard to apply the concept of a web page to software or web apps.

Open discussion

Joshue O'Connor raised the point that mobile is becoming the predominant platform. So what is WCAG doing to address these changes? What can be done to help developers continue to reduce barriers within the environment? One point already raised is that in WCAG currently there is a lack of techniques which can help.

He commented that we have gone from a semantic famine to a feast, from HTML4, primarily a document markup language to HTML5, which has a host of APIs which allows developers to do a massive range of things that were completely unimaginable 10 years ago as part of a native markup language. What are the current challenges that WCAG faces as far as existing techniques and accessibility support are concerned?

Loic Martínez-Normand identified one challenge as that some success criteria might need to be slightly modified to really work in new and future technologies.

This was discussed in the WCAG2ICT group when considering the success criteria outside of the traditional work domain. Some concepts such as the skip blocks success criterion, where WCAG 2.0 refers to content that is repeated on different pages, for instance in HTML5 navigation elements could be specified because there is an explicit navigation mechanism. This could change the concept of blocks that are repeated into explicit navigation elements as the ones that need to be skipped. Some of the success criteria, especially the ones related to sets of pages, probably need rethinking to work better in new technologies.

Bart Simons from AnySurfer expressed concerns over the accessibility support for new web technologies. As a project run by the blind association in Flanders, they are reluctant to promote the newest web technologies in case the assistive technologies in use are not up-to-date enough to work with them.

He wonders when there will be enough critical mass of up to date assistive technology to start using the newest techniques.

Kiran Kaja said that assistive technology support is definitely a major issue. Some products are better than others in keeping up with newer technology, but end users are more comfortable with the technology they have been using.

However, some assistive technologies are building solutions into their software that help developers achieve sufficient techniques. One of them is 1.4.2 success criteria for audio controls. This, in most cases, is done by assistive technology on iOS. So VoiceOver automatically releases the volume of other audio. It is interesting how assistive technology can be an issue when it does not support modern technology; while on the other end of the spectrum there is assistive technology that actually helps developers meet some of the success criteria automatically.

Joshue O'Connor agreed that the issue of accessibility support for new technologies is very important. It seems like an issue of confidence. Developers must have confidence that they can build something using HTML5 or ARIA and that it will work. When HTML5 reaches full Candidate Recommendation stage, perhaps assistive technology vendors will get totally on board.

Shadi Abou-Zahra said that an accessibility support database is being developed through the WAI-ACT project. A database that is actually crowd sourced. There is lots of knowledge out there. It is one of the essential aspects of WCAG 2 design.

There are two questions: What is the threshold? When is something sufficiently supported that it can be used in a particular context? This will need to be decided on a case by case basis, depending on the kind of website and context. The other question is to know how informed we are in making that decision. The information is very scattered. There are islands of information. This database will try to collect the existing knowledge that we have as a community.

He then tabled the question: How can we get the people really involved and willing to contribute, to add the knowledge that they have to the database? There is a lot of guidance needed. There is a lot in the community. How can we connect the two?

Pirthipal Singh replied that technological support can be relatively easily found out, but asked whether users know the new functions. Screen reader users before only heard "link", now it says submenu, tab, slider bar. Users need to listen very carefully and know what to do next. Do people who have the supported technology, also have the knowledge to work with it?

Joshue O'Connor commented that the issue of digital literacy is very important. Working for an NGO which provides services for blind and vision impaired people; he thinks that organizations need to step up to provide better training for people who use these assistive technologies. It is not a sufficient trick to throw vast amounts of semantics at a browser and hope that in the end the user will know how to do something with it.

Kiran Kaja suggested that it might be useful to produce short videos or audio clips to explain how new HTML elements work. Resources concentrate on helping web developers but there are none for end users, expecting that government agencies would provide training, which they do. But it doesn't mean that they are up to date.

Andreas Cederbom said that the problem with assistive technology is substantial, with issues ranging from the user's knowledge of using their technology to the variation of entitlement to assistance in different countries. Even in countries where users get assistive technology provided, it is not certain that an individual will get the latest version.

Users may also be reluctant to change systems that they know work together. However, they do not know what the future holds. Even in Sweden where users are quite lucky with assistive technology, many users have old technology.

However, he stressed that there are lots of features in the new web technologies that should be implemented. They should not be avoided just because there are problems with assistive technology. He gave an example of predictive text in search functions, which developers may avoid to prevent problems with assistive technology, even though it could benefit 95% of the population who would find it useful.

Raising another point about guidelines, he said that modern web pages need to work to several sets of guidelines to work on different platforms and be reactive to user needs. He suggested that developers would be helped if there were more correlations of related guidelines, such as the work done by WAI to show the mapping between WCAG and mobile guidance.

Joshue O'Connor suggested that there is a general tendency for increased complexity as the technological environment changes, so there are challenges WCAG has to meet in setting standards for others. How can WCAG and the other WAI Guidelines meet these challenges?

Pirthipal Singh said that accessibility supported ways of using technologies was discussed a lot in 2010 when reviewing WCAG 2.0. A major question was whether a problem was really an HTML or coding issue or was it an issue with the tool the user is using. It is a really complicated equation. That particular aspect was a big factor in going towards mandating the sufficient techniques.

Returning to mobile, the requirement in the Canadian standard is to adopt HTML5. The deadline for 100 federal departments to meet that is middle of 2015.

A lot of HTML5 can actually be used today. There are some elements that do not have sufficient support and obviously do not have sufficient techniques. But there is a lot of HTML5 that can be used right now. So the standards ensure that our websites and web applications need to be optimized for mobile devices. That needs to be our first consideration. One of the reasons why the Canadian government lost the federal court case is because we were not moving fast enough.

He said that we have to move forward. Just because the standards are not complete is not a reason for us to stop.

As a government, we are trying to discourage the device-based mobile applications, more commonly called native applications. Sometimes there may be a need to use device specific features. Then these particular mobile applications can compliment our website and web applications, in which case, the lack of sufficient techniques for the different mobile platforms is recognized.

For anything on the Web, the sufficient techniques are applied. There needs to be a separation between whether it is a mobile website or a native application. Those are two different aspects.

Joshue O'Connor found the idea of mandating techniques, especially when there's a dearth of them, interesting. There isn't a great deal of bandwidth available for developing new techniques, though WCAG is moving as fast as it can to do so. Does that create problems?

Detlev Fischer added that regarding the idea of testability, shortly before WCAG 2 was released, there was a belief that when the few missing techniques were provided, everything would be testable. This sounded as though when there were failures for all success criteria, assessment of failures would be certain. However, the Web is no longer a collection of static pages; it is very difficult to tell what a web page is. It is more difficult to define, evaluate and claim testability for dynamic content.

The problem for evaluators is that a failure of a technique does not automatically mean that there is actually a failure to conform to WCAG. Disclaimers under each technique explain that there could be another way of meeting the success criterion. There are many dynamic techniques being used, many of them more complex than the simple examples provided in the WCAG techniques.

Much of the dynamic content in use comes from scripting frameworks like Dojo or jQuery, which are really hard to assess for evaluators, the code is very obscure and it is getting very difficult to maintain the idea of testability. He felt that this was probably inevitable - as the Web moves away from static pages towards dynamic applications, claiming testability will get harder, and so will setting out exactly how to evaluate content.

Joshue O'Connor agreed, if Dojo or jQuery is used, if it said "this is an enhanced toolkit", the developer will know that what will come out will actually do the job. It is not fair to expect developers every time a new plan is comes out to learn to it the Nth degree in terms of user interaction for every type of assistive technology, for every kind of diverse mode of user interaction because that is just not sustainable.

Loic Martínez-Normand wondered if, given the definition of what it means to conform to WCAG 2 as an evaluation, less attention should be paid to sufficient techniques. The definition of conformance in WCAG 2 says a criterion is satisfied if there is no failure. There is no need to demonstrate that you conform. Only that you do not fail. That is an appropriate typical scientific approach.

The trouble is that the set of failure techniques in the WCAG 2 documentation is much smaller than the set of sufficient techniques. Some common failures are not documented.

A performance declaration should focus on failures, not on sufficient techniques.

One possibility could be to create a number of failures to try to move in that direction, to give more examples of things that really have to be avoided. That is the reason why for the Spanish standard, you must avoid, you shall avoid failures. If you really find that failure, you know you failed. There is no way out of that. The success criterion is not met.

Joshue O'Connor wondered if within Spain, failures are mandated.

Loic Martínez-Normand said that it is a voluntary standard but, from the standard viewpoint, avoiding the failures is mandatory. But using techniques is recommended.

Joshue O'Connor mentioned that failures could be an important part of insuring a greater degree of conformance to WCAG. He asked what other resources and tools are out there that can help people build more accessible web content and reduce barriers in the environment.

Loic Martínez-Normand brought up something that is needed across all fields of accessibility. Spain, for example, is one of the pioneering countries in legislating and mandating accessibility to the built environment. One of the good things about requirements of accessibility in the built environment is that they are clear, objective and measurable. The width of accessible doors is known, for instance and really needs to be taken into account in design.

Even so, they are not applied as much as they should be. Houses in Spain are built with stairs and no lifts. And architects hate ramps for whatever the reason. So there is something more essential in dealing with accessibility, greater awareness of the need to take into account the diversity of humans.

The social awareness of diversity and accessibility is something that is needed in the web domain as well as in others.

Joshue O'Connor related this to the notion of Universal Design and the notion of curb cut solutions that are of benefit to very diverse user groups to the crossover between accessibility and search engine optimization.

He then raised a proposal that usability could be part of a conformance claim, because he felt that it is a very important part of the chain. Having the contact or even the simple interaction that a developer may have with a person with disability, grounds accessibility in a very real way. It brings the point to all the guidelines and standards into a very sharp focus. It makes the whole thing much more organic and much more of a pleasurable process. It should be an exciting and engaging thing. It should not be what people have to do because of the big stick. We need the stick but we also have to engage people.

Bart Simons added a comment on usability awareness. As part of their web accessibility consulting, AnySurfer has started to demonstrate with assistive technologies, the accessibility issues encountered on client websites.

Previously, clients received only a technical report of the results of evaluating their website's accessibility. Demonstrating the issues has made the reports more digestible and raised awareness, though it is neither user-testing nor usability testing. It is popular with both clients and consultants.

Susanna Laurin suggested that something else that could help with implementing WCAG might be to have an association for accessibility professionals.

Summary of the day

Judy Brewer

Judy Brewer summed up her capture of the highlights from the day and asked participants: What stands out to you? What particular things do you think were important from the discussion? Or what did they think that was missing in her capture? These might be things that individuals felt were important to focus on for themselves or for their organizations. Or they might be things that they would want WAI, the Web Accessibility Initiative, to hear more clearly and to help ensure that there's a collaborative effort to move forward on.

Typically a workshop does not result in a fully formed plan for the next three years, but it does generate a lot of input. Workshops yield things to reflect on, both internally and then externally with a report. Those reflections can be plugged into and enhance our existing work.

She asked participants to think critically and see what they thought she was missing, what needs more emphasis and where we all might be able to do more to make web accessibility work better for people with disabilities.

The high level overview for the day: People have identified a number of specific barriers and suggested some different solutions. And many people noted in their comments that progress is still slow in many areas.

She was encouraged to hear from a few people that they are seeing some increases in web accessibility implementation. It seemed to be a few instances where web accessibility efforts have been ongoing for quite a long time and also where some very particular strategies had been used that were closely tuned with the environment in a particular country, and that enabled some encouraging progress. She challenged all to think: What are some of the best approaches for a given situation?

In the first session she noted that nonbinding requirements have not resulted in accessibility of the Web. Also existence of law makes a difference but not the whole difference. Particularly, very detailed requirements can become outdated more rapidly.

Several people had commented that initially they might not have had a strong implementation plan or may have only required the lower level, level A.

In terms of solutions suggested, there needs to be an obligation to be accessible, and a monitoring effort to ensure progress towards that objective, and enforcement. The inclusion of the involvement of civil society, nothing about us without us, is important.

Comments also showed that there needs to be not only more awareness of basics -- that people are still not successfully implementing in some cases -- but also a plan for more teaching and training on techniques for more advanced technologies. Also it seemed that implementation in the public sector and in the private sector are both important.

On the perception of accessibility, some people mentioned that accessibility needs to be seen as more than a check box requirement, and to have the people need to understand the broader benefits of it. And it would be helpful to embed the notion of accessibility in the performance of business as usual.

She also noted that a high level champion can sometimes break through a lack of progress on accessibility. She reflected from a discussion over lunch. She asked: how did it work having a high level champion? He did not say, "Well, it helped because we had somebody there." No, what really happened was this champion actually called departments and said, "You are not conforming, and so we are pulling your website down." He only did that four or five times and it got the message through fairly broadly. We need a cultural shift to get people to think about accessibility more.

There was a lot of discussion about conformance throughout the day. From one of the early comments on conformance it almost sounded as though in some cases where there are poor performance results, if that data is used effectively, it can help drive stronger implementations.

There was also a discussion of the need for scaled conformance options. It would be helpful if it were more than a yes or no question. So again we are wondering: Are these points relevant? Are there things missing in this summary?

In terms of policy orientation, there were some people saying they are using the legal requirements to help set the direction for public sector. And then there are many factors that affect how it is implemented. The public sector can help set the direction for the private sector but really we need time and resources to make things accessible.

The legacy technologies can be an issue.

It was helpful as an adjunct to the policy approach to ask people: What kind of technical issues they are having problems with? Is it forms? Is it media? Then to follow up with the training that supportively accompanies the policy.

In the second session there was a lot of discussion about different aspects of a commitment to harmonization. And what was interesting to Judy was the different way people approached this. Also noting that harmonization is even more important in ICT than in the built environment because of how much more internationalized it is.

There were a few people who said that there was a first version of web accessibility requirements in their country that was not as clearly harmonized, and consequently there was not as much progress as hoped. So with the second round of policy adoption, they had made a more deliberate effort for harmonization.

There was a comment from one group that harmonization saved a lot of work. Then in terms of harmonization, there were a few comments during the day where it sounded like a parliamentary statement at a certain point helped in committing to a standard, used the standard as is. That also came up in a "roll forward" discussion. With regard to legislation, the observation was made that the UN Convention on Rights of Persons with Disabilities requires legally enforceable requirements for accessibility of ICT.

In the U.S., there was the observation that having the combination of three things can be effective. Having legislation, increased public attention over the years on the need for ICT accessibility and the intersecting obligations from different departments, different agencies, helped.

In Europe, Ima noted that initially it was not apparent that legislation was needed, but when it did, then the interplay of timing between the standards and the legislation can be very tricky.

A few people noted that jurisdictional issues between a national level and the state or local level can be very challenging and therefore presumably are helpful to plan for if that situation exists.

In terms of standards again, these are just a few highlights. The free availability of standards from W3C is helpful. The transposition of standards can sometimes be straightforward if planned ahead for. And the suggestion was made to consider adding WCAG 2 to a registry of national standards. That might help facilitate uptake.

Conformance testing came up on this panel. Particularly with an emphasis on harmonization of conformance testing is not as far along but would be helpful for comparison of progress where that is possible. There were the questions from Ima, are we making enough progress? What is impeding it?

Operational challenges: Looking at how to translate operational commitment to practical steps while allowing room for flexibility and evolution without fragmentation. There was some sense that the structure of WCAG may facilitate that combination of challenges. Again, the need for a consistent approach to evaluating conformance came up.

The amount of supporting detail in WCAG techniques was cited as being helpful from at least one perspective.

Roll forward of standards. This applies in two different ways. Some organizations, some countries as well, had been initially requiring WCAG 1. This really did not have provision for precise testability. That is something we tried to get much closer to with WCAG 2. So with roll forward of standards, the question is: How do people do that best?

The participation in a collaborative international discussion seemed to be important for some. Also, industry has an important role in helping remind us of the evolution of the standards.

A parliamentary voice may have helped say we need to move forward. So moving into WCAG 2, potentially you could use the same strategies in the Web standards advance somewhere in the future. Of course WCAG 2 is stable and reliable for now.

In the third session we heard more about training, the need for consistent training and application across hundreds of different departments. In fact, that was a consideration in Canada's current approach of requiring techniques. It had also helped drive some of their development of a sampling approach for departmental websites. That is actually something that the W3C is looking at with its evaluation metrics report coming up from the research and development group is working eventually towards the sampling approach.

There was discussion of use of the sufficient techniques and documented failures as a focus for some of the training in an effort to get consistent training across the implementers. The structure of WCAG 2, Michael Cooper, reviewed the different layers in terms of principles, guidelines and normative success criteria. And then also some of the characters of the techniques, how they describe how you can implement the success criteria. And we have test criteria to accompany them. They are updated annually. But they are not vetted in the same depth as the success criteria in the WCAG 2 standard itself.

There was some discussion of technical independence and that effective use of techniques relies on developer expertise. These need to go hand in hand. As Susanna has noted, there is more interest in developing an association that might help increase expertise. W3C also remains involved in trying to help develop training materials.

Concerns were expressed about mandating sufficient techniques because it may restrict newer technologies that may have even more advanced built in accessibility features.

There were also some questions about accessibility supported technologies and a plea from Shadi for participation in crowd sourcing more resources on the accessibility supported technologies.

Showing concrete results to people when they've gone from many accessibility problems to improved performance, showing that in a workshop can be helpful.

She then invited further comments: Do people have any brief reflections on what particularly stood out during the day? Things that could be taken up or perhaps shifted, consider some new strategies? Perhaps collaborate with others, perhaps help develop more resources.

Open discussion

Susanna Laurin prompted that the target group "website owners" often get forgotten as they do not get mentioned at the strategic level nor the political level, and are not end users. All of which are very important stakeholders, but the website owners need help too.

Raph De Rooij said that one important role, especially when web accessibility becomes legally mandatory, is the supervising role. He felt that harmonization could be stimulated by providing guidance and instruments to people that have a supervising role.

Judy Brewer suggested that the resource being updated last year but not finished yet, Why Standards Harmonization is Essential to Web Accessibility, might be particularly helpful to convince supervisory groups.

Raph De Rooij pointed out that it is not only the standardization part. For instance Council members may have to supervise how a municipality performs regarding web accessibility. They lack information on the performance. And they often don't have the knowledge needed to act accordingly. How can council members be helped if they want to supervise web accessibility?

Andrew Arch said that in Australia, the challenge is not convincing the web developers and authors, but the senior management, to provide resources to enable accessibility to be implemented. Managers are another of the key players that need to be included in the mix of people considered when trying to get accessibility up and running.

Jorge Fernandes said that he is frequently asked for checklists, for example, for specific authoring tools, how to do it with WordPress, how to do it with Joomla, how to do it with SharePoint. This kind of checklist is not easy to find.

He also noted the need for guidelines or checklists specific to sectors like media (TV, radio, newspapers) and services like eBanking and eShopping, written from the perspective of the media and services, rather than the technology being used.

Judy Brewer had a few things to say in response as Jorge's comments intersected with some forthcoming work and planning. She wanted to let people know about these and also invite participation because they will only work as long as people get involved in them.

With regard to checklists, she had heard similar calls among a number of U.S. agencies, U.S. departments. One of the issues they have is that there are so many thousands of people in different organizations who are now responsible for doing accessibility that people have to work on this who have no background in it. Even if many new people are trained coming through universities, there will be people who basically want an instruction set to follow and the needs of somebody who can do a focused particular piece needs to be better addressed. To just do certain areas well, like WordPress, Joomla, SharePoint, for instance.

One question that comes along with that is: People are asking could W3C please help with those because we would have confidence. But we need more participation of accessibility experts in the WCAG Working Group to review those things.

She invited participation in the WCAG Working Group, to help with not only new techniques, but possibly checklists too. She also mentioned that there is a real need for techniques to be developed in areas such as mobile accessibility.

For some areas such as media, we have also developed in-depth user requirements. One of our questions to you is: How do we make these more findable? Because we have those. Maybe they would help increase understanding. But they may be buried on our website.

Other questions or comments go to vertical markets for all these different areas where you are going in digital publishing, where you are going to have the areas. So we are trying to work those areas but we need people at the table on that.

Shadi Abou-Zahra remarked that one of the other things that Jorge touched upon was authoring tools. This was not talked about very much in this workshop. WCAG being the goal or the means of getting to a goal was mentioned. But not how to really get there and the role of authoring tools.

Also scalability, by the time the guidance for Word 10 has been written, you know how many versions later we are? And is it the responsibility of the vendor, of the supplier themselves to develop those, how to actually meet WCAG using my tool, to actually scale the model a little bit more?

Detlev Fischer identified reporting as an important aspect of the European directive. There may be a requirement to report annually, which might still be debated. However, the whole issue of reporting on a large scale could be quite difficult.

He suggested that an information exchange on the European level is needed to establish how organizations involved in accessibility testing would envisage a reporting regime without it being a very costly or impossible task. If costly or impossible reporting is mandated, he said, it will not happen.

There needs to be a way of handling the reporting by an assertion which could be challenged or checked. If the Commission is going to implement such a scheme, how will it define national reporting duties? He encouraged everybody to think about it and discuss a realistic reporting regime if this directive comes into force.

Judy Brewer asked: Do you see some of the discussion today in terms of the importance of consistent conformance testing be valuable, perhaps combined with some of what is coming out with the Research and Development Working Group in terms of scalable evaluation metrics?

Detlev replied that the methodology being developed is geared towards expert evaluators. A web developer or person who has run a website for the last decade may not have the expertise to use the documentation and make a self assertion.

There may need to be a checklist-based approach, defining a level of checking which does not produce erroneous results and is reasonably valid. It will not be perfect, but better than nothing and more practical than expert evaluation on the full scale of the European Union down to every small city.

Judy Brewer noted that it would be helpful to have a conformance testing methodology that a non-expert could use.

She also mentioned that W3C is expanding its validation service to include some accessibility crawling.

She then invited participants to get involved in any of the W3C groups, including:

Shadi Abou-Zahra closed the workshop, thanking all participants, especially Judy Brewer, whose brain-child it was.