Mock-up of WCAG 2.0 with both Format and Authoring Requirements
by Ian Jacobs
Status of this document
Based on comments I made on WCAG 2.0 in
September 2003, I have created a mock-up to better explain the type of
re-organization of WCAG 2.0 I had in mind. This mock-up includes most of the
requirements of WCAG 2.0 (based on the 27 Oct 2003
WCAG 2.0 internal draft) and many of the requirements of XAG 1.0 (based
on the 3 Oct 2002 WD
and ). It is obviously not intended to be a complete set of guidelines, but
should be sufficient for the purposes of illustrating how this
re-organization can simplify the task of the WCAG WG in creating WCAG 2.0.
This document does not say much about conformance, which I would suggest be
addressed separately after discussion of this proposal.
This document does not represent the consensus of WAI or of the WCAG
Working Group. It has not been reviewed by anyone. It is only intended to
represent the opinions of the author.
Last modified: $Date: 2003/11/21 02:36:46 $
Copyright
© 2003 W3C® (MIT, ERCIM,
Keio), All Rights Reserved. W3C liability,
trademark,
document
use and software
licensing rules apply.
Table of contents
- Introduction
- Format Accessibility Guidelines
- Authoring Accessibility Guidelines
- Conformance
- Definitions
- 1.1 Purpose of document
- 1.2 Scope of document
- Target content. It may be worthwhile identifying main classes of
content, identifying requirements that are specific to those classes,
and requiring claimants to identify those classes in a claim. E.g.,
"hypertext documents" or "user interfaces".
- Known limitations to the document
- 1.3 Relation to other WAI Guidelines
- WCAG 1.0
- UAAG 1.0. In particular, the document assumes a conforming user
agent that provides the following functionalities:
- Allows user to control colors, text size, fonts
- Allows user to start, stop, slow, pause multimedia content
- Provides the user with a time-insensitive view of content
- Allows user to toggle background images
- Allows the user to view captions unobscured
- Allows user to toggle blinking and animated text.
- Allows user to choose from among alternatives
- Allows the user to override author-supplied keyboard
bindings
- Allows user to navigate to navigation stops with the
keyboard
- Allows users to activate event handlers with the keyboard
- ATAG 1.0 and ATAG 2.0
- 1.4 Additional benefits of conforming content
The expectation is that the format satisfies all of the applicable
requirements.
2.1 Requirements common to all formats
- The format must support text content in a manner that conforms to
Unicode.
- Where the script or language cannot be recognized by the agent, allow
the author to identify the script of any text content (including
equivalents) and the language of any non-text content that communicates
meaning through words and sentences.
- For every element:
- Allow an explicitly associated text title
- Allow an explicity associated text summary
- Specify element types whose meaning does not rely on rendering. This is
not required when the format itself is "final-form".
- Ensure that for any final-form content, the author can provide the user
with access to explicitly associated original-form content, whether on
the server or the client
- Allow the author to control rendering through style sheets.
- Describe an accessible default rendering for common output devices.
Include in this default rendering:
- Sufficient contrast for default colors
- Rendering of section titles and emphasized text distinct from
paragraph text.
- Allow the author to group content.
- Allow the author to break large pieces of content into meaningful
(smaller) blocks
- If your format resembles HTML, conform to a W3C XHTML Recommendation.
When important an XHTML module, import the accessibility features.
2.2 Requirements for audio, video, animation, and other non-text
content.
- For each piece of video or audio-visual content:
- Allow any number of captions
- For audio:
- Allow any number of collated text equivalent
- For video and animations:
- Allow any number of collated audio equivalents
- For other static non-text content (e.g., images) that conveys
information visually:
- Allow any number of text equivalents
2.3 Requirements for content requiring user interaction
- Allow the author to specify user interaction in a device-independent
manner, or provide an automatic mapping to device-independent bindings
for each device-dependent binding.
- If the content is time-sensitive, allow an explicitly associated static
equivalent.
- If the content is time-sensitive, allow the author to specify a choice
of time scales.
- If the content requires interaction, allow the author to create a
serialization of the interaction requirements.
2.4 Requirements for navigation, orientation (for all content)
- Allow the author to specify a table of contents
- Allow the author to specify navigation stops on any element
- Allow the author to specify an order for navigating sequentially among
the navigation stops
- Allow the author to specify direct navigation to navigation stops
(e.g., keyboard shortcuts)
- Allow the author to provide documentation of serial and direct
navigation aids
- Allow the author to specify navigation groups (e.g., navbars).
2.5 Requirements for rendering (for all content)
- Allow the author to specify the foreground and background color of
text
- Allow the author to specify at least one distinctive rendering that
does not rely on color alone for the following types of content:
emphasis, grouping, section titles.
2.6 Requirements for specialized elements
- If the format supports data tables, allow the author to specify column
and row headings, and to associate headings with each cell.
- If the format supports data tables, allow the user to encode any
relationship that would be available visually in a manner the user agent
can recognize.
3.1 Conformance to specifications
- Use a format that conforms to the requirements of the previous
section.
- Use format features according to specification.
- Use the relevant accessibility features of each specification.
- Do not convey semantic information through style sheets alone when
format features exist to accomplish a particular task.
- Avoid deprecated features.
3.2 Requirements for text
- The first time (in document order) that an abbreviation or acronym
appears, provide the expansion using the features provided by the
format.
- If the contracted form of a word is used in a manner that would make it
ambiguous, explain the contraction the first time it is used.
3.3 Requirements for audio
- For audio content that conveys information through words and sentences,
ensure that other sounds are at least 20db quieter than the words and
sentences.
3.4 Requirements for content requiring user interaction
- If the content is time-sensitive, specify an alternative time scale
that is at least 10 times the default.
- Provide a means for the user to confirm an unsafe action (where
"unsafe" is one that has as side effect per RFC2616).
3.5 Requirements for orientation
- Organize content so that it makes sense when rendered serially.
- If the content does not allow the automatic numbering of sections,
provide hierarchical section numbers.
- When the purpose of the content is inherently unpredictable (e.g., a
mystery or adventure game), provide a warning to the user.
3.6 Requirements for navigation
- Provide navigation blocks
- Provide navigation stops for important information
- Provide consistent navigation blocks on related content (e.g., a set of
pages)
- Provide navigation stops at and just after navigation blocks
3.7 Requirements for rendering
- Provide information in a manner that does not rely on color alone for
understanding.
- Specify foreground and background colors with sufficient contrast
- Don't cause the screen to flicker (in the 3-50Hz range) or warn the
user before causing the screen to flicker.
3.8 Requirements related to user agents
- For content other than HTML and plain text, document what is required
as part of the user agent for the usre to be able to access the
content.
3.9 Good practice for using language accessibly
The following good practice items involve the accessible use of language.
These are not requirements that can be easily verified since they are so
subjective, context-sensitive (e.g., poetry v. scientific publication v.
parody), and vary from language to language. Nonetheless, we encourage
authors to keep the following in mind as they design content:
- Use good grammar
- Use proper spelling
- Use simple language.
- Use simple graphics to illustrate important concepts.
- Break down content into reasonably-sized chunks (e.g., use sections and
paragraphs).
- Create titles that are sufficiently different
- Include images and diagrams when they help convey the meaning of the
content.
More work to be doner here...
- 4.1 Conformance to WCAG 2.0
- 4.2 Conformance profiles
- 4.3 Guidance for policy-makers.
- This document explains how to make content accessible. It does not
make any requirements for when to make content
accessible; this is left to policy-makers.
- Here are some example scenarios where it may not always be useful
to require conformance to all of WCAG 2.0:
- Specialized content such as poetry
- Time-sensitive scenarios such as games and testing
environments
- Real-time v. pre-recorded content
This section is normative. It is incomplete at this time.
- content
- The document object, composed of octets
- text
- A character sequence (i.e., a code point sequence)
- text content
- An octet sequence that, after processing, produces characters in
content that communicate meaning through words and sentences.
- non-text content
- An octet sequence that does not have the characteristics of text
content.
- element
- Used here both in the sense of "XML element" and the sense of
"logical format entity".
- explicitly associated
- A piece of content that is associated with another (e.g., through
markup) in a manner than can be recognized by a software agent.
Explicit associations should be bidirectional.
- audio-visual content
- Content that consists of at least one audio track synchronized with
at least one visual track.
- audio content
- Content that conveys information through sound, including speech.
- time-sensitive content
- Content that, when rendered, is time-sensitive.
- static content
- Content that, when rendered, is time-insensitive
- text equivalent
- Text that approximates an explicitly associated piece of non-text
content. In the general case, a text equivalent consists of alternating
sequences of:
- Text transcript: Text that represents words and
sentences.
- Text description: Text that summarizes
important information.
- collated text equivalent
- A static text equivalent for audio-visual content.
- collated audio equivalent
- Like a collated text equivalent, but consisting of pre-recorded human
voice or synthesized speech (either pre-recorded or generated on the
fly).
- captions
- A collated text transcript that is synchronized with the explicitly
associated audio-visual content.
- important information
- Content that the author deems important to understanding the meaning
of the content.
- recognize
- An agent can "recognize" something if it can process it according to
specification automatically, i.e., without user interaction.