W3C published the Web Content Accessibility Guidelines 1.0 (WCAG 1.0) as a Recommendation in May 1999. This Working Draft for version 2.0 builds on WCAG 1.0. It has the same aim: explain how to make Web content accessible to people with disabilities. Incorporating feedback on WCAG 1.0, this Working Draft of version 2.0 focuses on checkpoints. It attempts to apply checkpoints to a wider range of technologies and to use wording that may be understood by a more varied audience.
This document is prepared by the Web Content Accessibility Guidelines Working Group (WCAG WG) to show how more generalized (less HTML-specific) WCAG checkpoints might read. This draft is not based on consensus of the WCAG Working Group nor has it gone through W3C process. This Working Draft in no way supersedes WCAG 1.0.
Please refer to "Issue Tracking for WCAG 2.0" for a list of open issues related to this Working Draft. The "History of Changes to WCAG 2.0 Working Drafts" is also available.
This is a draft document and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use W3C Working Drafts as reference material or to cite them as other than "work in progress". A list of current W3C Recommendations and other technical documents is available.
Please send comments on this document to email@example.com. The archives for this list are publicly available.
This document outlines design principles for creating accessible Web sites. When these principles are ignored, individuals with disabilities may not be able to access the content at all, or they may be able to do so only with great difficulty. When these principles are employed, they also make Web content accessible to a variety of Web-enabled devices, such as phones, handheld devices, kiosks, network appliances, etc. By making content accessible to a variety of devices, the content is now accessible to people in a variety of situations.
Not all devices are the same. Not all systems are the same. Not all people are the same. When following the guidelines, attempt to reach the maximum number of people in the maximum number of scenarios. This can be achieved through a single accessible rendering or multiple accessible renderings of the same content optimized for different situations.
The design principles in this document represent broad concepts that apply to all Web-based content. They are not specific to HTML, XML, or any other technology. This approach was taken so that the design principles could be applied to a variety of situations and technologies, including those that do not yet exist.
It takes a variety of people to make a Web site; it will take a variety of people to make it accessible. All of the information and guidance we offer to make Web sites accessible is gathered in a suite of documents. At the top layer are the most general concepts, and at the bottom layer are the most specific.
Issue: it has been proposed that we provide multiple, automatically generated "views" of the guidelines document for different purposes, using XSLT. What are the different "views" of the document that we should make available? Should we also have versions that include techniques or technology-specific success criteria (i.e., the technology-specific layer), along with the guidelines, checkpoints and generic success criteria?
The top layer, where you are now, is divided into an Overview of Design Principles and four Guidelines:
Each Guideline has several Checkpoints associated with it. There are a total of 21 checkpoints.
Each checkpoint is divided into normative and non-normative (informative) information. The normative text is what you must conform to. The informative text only helps to make the normative text easier to understand. A checkpoint consists of:
In separate Techniques Documents are code examples, screen shots, and other information specific to a technology.
These will become active links as the corresponding working drafts are published.
We expect that policy makers will use the Guideline and Checkpoint level, while developers will dive into Techniques and managers might only need to read the Overview of Design Principles.
Issue: it has been proposed to include a discussion of accessibility vs usability and how we define our scope.
Issue: provide a discussion of technology-specifics and core checkpoints.
This WCAG 2.0 Working Draft does not assign priorities to checkpoints, nor does it include links to technology-specific examples and techniques. Later versions of this document will assign priorities and will link to techniques. This Working Draft presents an initial reorganization and begins to incorporate other feedback received since the publication of WCAG 1.0 in May 1999.
In some cases, WCAG 1.0 checkpoints of various priorities are combined into a single checkpoint in the WCAG 2.0 Working Draft. In these instances, a priority cannot be assigned to the new checkpoint until the WCAG Working Group has extensively discussed the priority for that checkpoint. Priorities will be included in a future Working Draft.
The WCAG Working Group is proceeding carefully to minimize substantial differences between the WCAG 1.0 Recommendation and the WCAG 2.0 Working Draft. Refer to the Checkpoint Mapping Between WCAG 1.0 and WCAG 2.0 Working Draft for more detail on current correspondences.
WCAG 1.0 is accompanied by supporting techniques documents (non-normative) which include examples of how to implement WCAG 1.0 in HTML and CSS. The WCAG Working Group will continue to develop the HTML and CSS technique documents as well as create new documents for other languages such as SMIL and SVG. Links to these documents will be added in future versions of the WCAG 2.0 Working Draft.
Issue: How should conformance be defined? How many "levels" of conformance should be recognized, and on what basis should they differ? What information does an implementor need to provide in order to make a valid conformance claim, and in what forms can this information legitimately be supplied (RDF, web page, conformance icons)? We also need to discuss conformance of server-side solutions and the role of technology-specifics in conformance.
Issue: should this instead be called "executive summary"? Does it matter?
The four design principles in this document are:
All Web content has been created because some person or group of people wanted to accomplish something - whether to provide the weather report for a community or to create a family photo album or to sell cars. The purpose and design of a site will vary as much as the people who create it. Content that is usable, attractive and meets the needs of its audience is designed that way. In the same manner, accessible Web content design is a conscious effort.
Accessible Web content benefits a variety of people, not just people with disabilities. In the physical world, ramps are used by bicycles, people pushing strollers, and people in wheelchairs. Similarly, accessible Web content is usable by a variety of people with and without disabilities. For example, people who are temporarily operating under constrained conditions.
Accessible Web content is often accessible because it can be repurposed by a machine. For example, a browser can display text instead of an image if the text has been associated with the image. Likewise, a search engine can find a famous quote in a movie if the transcript is provided on the Web site.
Note: These principles apply only to Web content presented to a human reader. A structured database or metadata collection where the data is intended for use by another machine and thus requires no interface lies outside the scope of these guidelines.
Here are a few scenarios, by no means an exhaustive list of the variations and types of disabilities and needs:
If Web content employs the design principles described in this document, then users should be able to access the content using adaptive strategies and assistive technologies. A screen reader is an example of an assistive technology that reads the page aloud. There are many other tools people with disabilities employ to make use of Web content. For more in-depth scenarios of people with disabilities using accessible and inaccessible Web content, please read "How People with Disabilities Use the Web".
Think about the process used to build a new office building. The structure of the building - its skeleton - is built before the decorations are added.
Similarly, the place to begin in your design is with the structure. How many chapters does your document have? Which navigation links are the most important and where should they go? What is the best way to express this idea - do you need an animated simulation to help people understand?
Once the structure is finished, the walls are painted and pictures are hung. On your web site, this is like deciding the color of chapter titles and how text will flow around images.
The architect designed the building with both elevators and stairs. Some people prefer to take stairs, while some people find the stairs too challenging or impossible. On your Web site, some people will prefer images, animations, multimedia, fast-paced interactive games, while others will find them too challenging or impossible to use. As with elevators and stairs, provide a variety of ways for people to access and to navigate through your Web content.
We adjust objects in our environment to meet our needs and use tools to help us do things we cannot do on our own. Both in real life and on the Web, some people rely on these adjustments to work, to contribute to society, and to enjoy life. One person uses a stepladder to reach an object on a high shelf. Another uses a magnifying glass to read small print. A third uses captions to understand what a television news announcer is saying.
Here are some examples from the World Wide Web:
User needs and preferences are influenced by:
For more information about user capabilities, device capabilities, assistive technologies, and usage scenarios refer to the Working Draft "How People with Disabilities Use the Web."
You will have successfully provided a text equivalent for all non-text content if:
Issue: how should server-side solutions conform to this checkpoint? Note the consensus statement in the Requirements Document to the effect that equivalents must be available in all versions of the content in order for it to be considered accessible. Is this really what we mean?
A text equivalent
Non-text content includes images, text in raster images, image map regions, animations (e.g., animated GIFs), applets and programmatic objects, ascii art, scripts, images used as list bullets, spacers, graphical buttons, sounds (played with or without user interaction), stand-alone audio files, audio tracks of video, and video. Issue: Should "scripts", "applets" and "programmatic objects" be removed from this definition (cf checkpoint 4.4)? Should the definition be simplified by omitting "spacers", "graphical buttons", "images used as bullets", etc.?
Text equivalents provide access to non-text information for someone who cannot see at all, who cannot see well, or who needs to supplement visual information with auditory information.
You will have successfully provided synchronized media equivalents with time-dependent presentations if:
Issue: define "time-dependent presentation". Suggestion: a presentation which requires the user to respond interactively within a limited period time, or is composed of synchronized audio and visual tracks (e.g., a movie).
Media equivalents present essential audio information visually (captions) and essential video information auditorily (audio descriptions).
Issue: have we agreed that "audio descriptions" should be included here, and that the redundancy with checkpoint 1.1 is not problematic?
Captions provide auditory information for people who are deaf or who have hearing loss. Audio descriptions provide visual information for people who are blind or who have low vision. Captions also provide auditory information for people who have lowered the sound volume or are in a noisy environment. Audio descriptions also provide visual information for people who are temporarily not looking at the video presentation. For example, while following an instructional video they must look down at their hands and away from the screen.
Issue: Can this checkpoint be better expressed, e.g., "Provide content in a data format that allows repurposing"? If so, what is meant by "repurposing" and what would be the success criteria? Should this checkpoint be combined with checkpoint 1.5? If we retain the existing text of the checkpoint, would it be better to use the term "data structure" instead of "data model"?
You will have successfully used markup or a data model to provide the logical structure of content if:
The logical structure of content represents changes in context. For example,
Issue: does the definition distinguish adequately between presentation (e.g., italics) and the corresponding structure (e.g., emphasis), or is the present wording likely to confuse readers?
Issue: Provide an adequate definition of "data model" to cover such phenomena as PDF logical structure, XML information sets that are not represented in markup, etc.
When the logical structure is provided in markup or a data model,
Issue: Should the explicit association of labels with form controls/user interface components be added here as an example? If not, where does it belong? Note that, in terms of the definition, it is a good illustration of a "non-hierarchical relationship".
The working group decided to retain the expression "natural language", but to link the term, as it appears in the checkpoint text, to the glossary. This still needs to be implemented.
You will have successfully identified the primary natural language of text and text equivalents and all changes in natural language if:
Natural languages are those used by humans to communicate, including spoken, written, and signed languages.
Oftentimes, phrases from various languages are interspersed in writing. When these phrases are identified, a speech synthesizer can voice text with the appropriate accent and pronunciation. When they are not identified, the speech synthesizer will use the default accent and pronunciation dictionary which can make the phrase intelligible. Identifying changes in language will also allow a tool to ask for automatic translations of that content. When editing documents, authoring tools can switch between appropriate spelling dictionaries.
Issue: instead of saying "content and structure", would it be more accurate to say "text and structure"? If not, can we find a suitable substitute for "content" in this checkpoint?
You will have successfully separated content and structure from presentation if:
Issue: define "content".
Issue: define "presentation".
Content and presentation can be separated because the rules that control how content is displayed can be separated from the markup that denotes the structure of the content.
Typically, style rules are stored separately from the content to which they apply, in resources known as style sheets. To facilitate the presentation of Web content by a range of devices (high and low-resolution displays, printers, speech devices, etc.), it is advisable to associate a variety of style sheets with your Web content.
We interact with objects and events in our environment in different ways based on our abilities and often determined by the situation in which we find ourselves.
Issue: We are currently discussing the scope of this checkpoint and what is required. Does providing multiple site navigation mechanisms increase accessibility or are we trying to get at something else? Refer to the benefits for the issue we are trying to tackle. How do we set limits for when to apply this checkpoint? If a site consists of only 5 pages, a site map might look exactly like the home page.
You will have successfully provided multiple site navigation mechanisms if:
Issue: These success criteria are not easily testable. We have also discussed complex sites and content versus small sites. For example, WCAG 1.0 is one "page" but it needs several navigation mechanisms.
Issue: define "site navigation mechanism". Should hypertext links in the body of a document be excluded from this definition, so that only specifically "navigational" constructs will count for purposes of the definition?
Someone may give us directions to a new place, but based on our knowledge or landmarks along the way, we might take a different route. As people interact with your Web site, they might find different associations between content and take a different path than you anticipate. Some people will miss the associations you have made and not find the content on the paths you have created. Providing people with multiple navigation mechanisms provides a variety of paths to access content.
You will have successfully provided consistent and predictable responses to user actions if:
Providing responses to user actions is important feedback for the user. This lets them know that your site is working properly and encourages them to keep interacting. When the user receives an unexpected response, they might think something is wrong or broken. Some people might get so confused they will not be able to use your site. Common responses to user actions:
These actions should be predictable and sensible to the end user. Make interactions consistent, both throughout the site and with commonly used interaction metaphors used throughout the Web.
Some of these examples are very brief. Should they be expanded and clarified with further details?
Issue: the discussion and examples under this checkpoint are all oriented toward visual interfaces. Can we correct this by modifying the definitions, benefits and/or examples to take account of other application environments, e.g., "voice browsers"?
You will have successfully either given users control of mechanisms that cause extreme changes in context or warn them of pending changes if:
Mechanisms that cause extreme changes in context include:
Issue: do these have counterparts in non-visual interfaces?
If the user is unable to track visual cues that make extreme changes obvious, then they will not realize the context has changed. People who are blind, some people with low vision, some people with dyslexia and other people who have difficulty interpreting visual cues need guidance during extreme changes in context.
What is meant by "as much time as possible", and what factors can the implementor legitimately take into account in deciding how much time it is "possible" to allow?
You will have successfully either given users control over how long they can interact with content that requires a timed response or given them as much time as possible if:
Examples of content that requires a timed response:
Issue: is this really a definition, or just a series of examples?
People with reading disabilities, cognitive disabilities, and learning disabilities often need longer than most people to read and comprehend written text. People with physical disabilities might not be able to move quickly or accurately enough to interact with moving objects. Content that is updated often might not be processed and read in time or in the proper order by an assistive technology or voice browser.
Issue: Should this checkpoint be combined with checkpoint 4.3? In either case, doesn't it belong under guideline 4?
You will have successfully used device-independent event handlers if:
Issue: In criterion 2, one of the device-specific event handlers must be encoding (e.g., keyboard) although it has been noted that encoding device handlers do not appear to be usable in lieu of direct manipulation (e.g., mouse) in those cases where the action or the result can not be expressed in words. Our challenge is to express this simply.
Device-independent access allows the user to interact with the content through a preferred device. For example, a mouse, a keyboard, a microphone, or a head wand. There are a variety of input devices that exist today and who knows what might be possible in the future. In general, if keyboard interaction is supported, speech input or a command line interface will also be supported.
You will have successfully avoided causing the screen to flicker if:
People with photosensitive epilepsy can have seizures triggered by flickering or flashing in the 4 to 49 flashes per second (Hertz) range with a peak sensitivity at 20 flashes per second as well as quick changes from dark to light (like strobe lights). Such users configure their user agents and system devices to avoid screen flicker at rates that individually affect them.
Issue: This is a new checkpoint that is being explored. It does not have full support of the working group. We know that spelling mistakes are a serious issue for people with writing disabilities and dyslexia. We have generalized this checkpoint to include all input errors but highlight spelling mistakes. Are there other input errors we should highlight? Are there other success criteria we can define? If not, then we will not use the general case and make this specific to spelling errors.
You will have successfully handled input errors, such as misspellings if:
People with writing disabilities and people with dyslexia often have difficulty writing text in forms or other places that need text input. People with speech disabilities might not be recognized properly in voice input applications.
We each gain knowledge in different ways. Some people read something and understand, others have to experience it, while others need only to see something. To help people understand the information you are presenting, consider the various ways that people learn. Keep in mind the variety of backgrounds and experiences people will bring to your site. Use language, illustrations, and concepts that they are likely to know. Highlight the differences and similarities between concepts. A delicate balance is needed in design to make this work well.
Issue: Is this checkpoint redundant? Cf checkpoints 2.2 and 3.2. The explanation under this checkpoint suggests that it can clearly be distinguished from 3.2, so is there still an issue here?
You will have successfully used consistent presentation if:
Issue: can we provide better success criteria?
Presentation includes, but is not limited to:
Issue: this definition does not cover auditory and tactile presentation. Should further items be added to include these cases?
Issue: this definition of "presentation" should also be given under checkpoint 1.5.
Consistency helps users predict where to find information on each page of your site. It also helps users determine the relationships between items in the content. This understanding of the structure helps users navigate, orient themselves, and thus understand.
Note that differences in presentation also help users determine that they have succeeded in loading a new page. Differences also help users distinguish between content. Highlighting these differences is covered in the next checkpoint 3.2.
This checkpoint asks the author to make items with similar functions have a similar appearance or sound, while checkpoint 3.2 asks the author to highlight the differences through presentation.
You will have successfully emphasized structure through presentation, positioning, and labels if:
Presentation that emphasizes structure enables users to be
If the default presentation of the structured content does not meet the needs of your audience, use graphics, colors, sounds, and other aspects of presentation to emphasize the structure. However, ensure that the structural and semantic distinctions are provided in the markup. Refer to checkpoint 1.3.
Issue: how should an author decide what is "appropriate" for the content? What criteria are to be used in determining whether a text is written as clearly and simply as is appropriate?
What should be the role of the author's "intended audience" in determining appropriateness?
You will have successfully written as clearly and simply as is possible and appropriate for the site's content if:
Should we require that non-standard orthography be used (e.g., the inclusion of vowels in Hebrew text) for the purpose of aiding comprehension? If so, should this be included in the success criteria of this checkpoint? If implementors decided not to meet such a requirement (which seems onerous), they would then have to assert that their content did not comply with this checkpoint, even though it may do so in other respects. Is this a desirable outcome? Perhaps any such requirement should be non-normative.
Authors should strive for clear and simple writing to aid all users, especially those with cognitive, learning, and/or reading disabilities. This should not discourage you from expressing complex or technical ideas. Using clear and simple language also benefits people whose first language differs from your own, including those people who communicate primarily in sign language.
Under what circumstances should text be supplemented with non-text content? How should the adequacy of non-text supplements be judged with respect to their effectiveness in aiding comprehension? What should be the role and significance of the author's "intended audience" in determining whether non-text content should be provided, and if so, what it should contain?
Issue: provide success criteria for this checkpoint. To what extent, if at all, should the categories noted in the examples (see below), e.g., "concrete concepts", graphical depictions of data, depictions of processes and relationships, etc., be relied upon in the success criteria to define the kinds of material for which non-text supplements are needed?
Non-text content includes images, text in raster images, image map regions, animations (e.g., animated GIFs), applets and programmatic objects, ascii art, scripts, images used as list bullets, spacers, graphical buttons, sounds (played with or without user interaction), stand-alone audio files, audio tracks of video, and video. Is this definition adequate?
Sounds, graphics, videos and animations can help make concepts presented in a Web site easier to understand, especially for people with cognitive, reading, or learning disabilities or those who are unfamiliar with the language of the text of the site.
Note: "Designers need to be cautious in deciding when to use illustrations. Reading a picture is probably a learned activity that is easier for some than others. Some users skip the pictures; others read only the pictures. Designers must also recognize that visual conventions are not universal and that individuals develop their own mental schema and expectations in interpreting visual information.
For a detailed discussion of guidelines pertaining to illustrations, consult Tufte (1983) and MacDonald-Ross (1977)." Robert W. Bailey, Ph.D., Human Performance Engineering, 3rd edition.] pg 431.
You will have successfully annotated complex, abbreviated, or unfamiliar information with summaries and definitions if:
A summary may also describe how the table fits into the context of the current document.
Issue: the second success criterion is very specific to tables. Can it be generalized?
Issue: What other applications of the principle enunciated by this checkpoint should be included in the success criteria? How can the criteria be improved?
Content is considered complex if the relationships between pieces of information are not easy to figure out. If the presentation of the information is intended to highlight trends or relationships between concepts, these should be explicitly stated in the summary.
Examples of complex information:
Content might be unfamiliar if you are using terms specific to a particular community. For example, many of the terms used in this document are specific to the disability community.
Summarizing information that is difficult to understand helps people who do not read well. Providing a summary of the visual cues that show relationships between complex information helps people who do not use visual cues or who have difficulty using visual cues. For example, people who are blind do not use any visual cues, while people with dyslexia or people with low vision might have difficulty interpreting visual cues.
Defining key terms and specialized language will help people who are not familiar with the topic. Providing the expansion of abbreviations and acronyms not only helps people who are not familiar with the abbreviation or acronym but can clarify which meaning of an abbreviation or acronym is appropriate to use. For example, the acronym "ADA" stands for both the American with Disabilities Act as well as the American Dental Association.
The other three guidelines stress the needs of the user and how those needs might change as the situation changes. This guideline focuses on the technical considerations that are needed to support the changing needs of a particular user and the differing needs of users.
The user's needs and preferences are tied to the device that he or she chooses, as discussed in Guideline 1. These four checkpoints highlight the use of technology that will support the design of content that changes according to user's needs and preferences in presentation, interaction, and comprehension.
Issue: Should there be a requirement that authors
either (a) use widely supported content formats for which style sheets, assistive
technologies etc., are readily available, or (b) provide style
sheets for presentation on a variety of devices, or transformations
into more widely available content formats such as HTML, SVG etc.? The
underlying problem here is that an author can provide content, and a
style sheet to accompany it, which is in a non-standard format, or
uses "extensibility" features of markup languages (such as XHTML
extensions or the HTML
CLASS attribute) without providing
a mechanism by which to make the structure specified in the markup
available to the user agent.
You will have successfully chosen a technology that supports the use of these guidelines if the technology:
Issue: are these success criteria complete? If not, what should be added or changed? Should we provide a link to the XML guidelines?
Issue: should the checkpoint be reworked (or an additional checkpoint inserted here) to require that content be designed, as far as possible, so that it is amenable to automated accessibility testing?
Markup languages, multimedia formats, software interface standards, etc., vary in their support of accessibility. When choosing which technologies to use, consider how easy it is to apply these guidelines.
You will have successfully used technologies according to specification if:
Are protocols relevant to this checkpoint? If so, why, and should we require that they be used according to specification? Obviously there are interoperability advantages in doing so, but is this pertinent to accessibility?
When languages, API's, and protocols are used according to specification, tools that use the results will be able to do so as intended and expected.
You will have successfully designed user interfaces compatible with assistive technology if:
Asking someone to access your Web site without their assistive technology is like asking someone to access a building without their wheelchair. Assistive technologies are an essential part of the lives of many people with disabilities.
Issue: define "default" for purposes of this checkpoint. If "default" were taken to mean "a user agent's default rendering", then this would defeat the purpose of the checkpoint, because (for many user agents) the default is to apply style sheets, invoke scripts and programmatic objects, etc.
You will have successfully ensured that content remains usable when technologies that modify default user agent processing or behavior are turned off or not supported if:
In determining the extent to which older technologies should be supported, keep in mind that
Issue: Should we include the entire glossary or just the definitions of terms appearing in the guidelines? We also need to make sure that the glossary definitions are the same as the definitions which appear in the guidelines themselves. Some definitions are already included below.
Regular participants of the WCAG Working Group:
Kynn Bartlett, Paul Bohman, Jonathan Chetwynd, Wendy Chisholm, Katie Haritos-Shea, Charles McCathieNevile, Matt May, Jo Miller, Sean B. Palmer, Anne Pemberton, Adam Victor Reed, Loretta Guarino Reid, Emmanuelle Gutiérrez y Restrepo, Gregory J. Rosmaita, Joel Sanda, Lisa Seeman, Cynthia Shelly, Andi Snow-Weaver, Gregg Vanderheiden, Jason White
Dan Aunspach, Bruce Bailey, Harvey Bingham, Judy Brewer, Dan Brickley, Dick Brown, David Clark, Joe Clark, Michael Cooper, Nir Dagan, Daniel Dardailler, Alan J. Flavell, Geoff Freed, Greg Gay, Al Gilman, Jon Gunderson, Shawn Lawton Henry, Donovan Hipke, Chuck Hitchcock, Ian Jacobs, Marshall Jansen, Phill Jenkins, Leonard Kasday, Thanasis Kinias, Chris Kreussling, Chuck Letourneau, Steven Livingstone, William Loughborough, Greg Lowney, Scott Luebking, Lisa Kestenbaum, Marja-Riitta Koivunen, Marti McCuller, Tom Martin, Masafumi NAKANE, Robert Neff, Tim Noonan, Anuska Perkins, David Poehlman, Chris Ridpath, Greg Rosenberg, Heather Swayne, David Tanner, Jim Thatcher, Claus Thøgersen, Peter Verhoeven, Cynthia Waddell, Mike Williams
Since the release of WCAG 1.0 in May 1999, the WCAG Working Group has received feedback on priorities of checkpoints, the usability of the set of documents, and requests for clarifications on the meaning of specific checkpoints and what is needed to satisfy them. Thus, it is intended that WCAG 2.0:
For a checkpoint by checkpoint comparison, refer to the Checkpoint Mapping Between WCAG 1.0 and WCAG 2.0 Working Draft.
We hope that WCAG 2.0 will have several improvements over WCAG 1.0.
Note to reviewers: Links within the document will be turned into references and the links to those documents will be listed here as references. They are inline for the time being.