TK

Overview

This document provides answers to commonly asked questions about the EPUB Accessibility specification and its associated techniques. It provides guidance and insight into how the document was created and the reasons for various choices in implementation.

The document is not intended as a how-to guide to creating accessible content. Please use the IDPF Accessibility forum to ask questions about how to implement the requirements.

General

What version of EPUB does this specification apply to?

Although the EPUB Accessibility specification was developed during the EPUB 3.1 revision, it is designed to be applicable to earlier versions of the format. The requirements and techniques do not depend on EPUB 3.1; they can be applied to make EPUB 3.0-conformant publications accessible, and also EPUB 2.0 publications.

Authors can begin applying the specification to their content immediately, in other words; there is no need to wait until support for EPUB 3.1 becomes commonplace.

Discovery Metadata

Why was schema.org metadata chosen?

Schema.org metadata is required in the package document because it provides a comprehensive set of descriptive metadata that fits the EPUB metadata model.

As EPUB continues to integrate with the Open Web Platform — where schema.org metadata is used by the major search engines when cataloguing information — it also represents a stable fit for future versions of the standard.

Why is ONIX not required for discovery metadata?

Although ONIX is critical metadata in ebook distribution channels, EPUB publications are not necessarily books nor are they all distributed through the major ebook channels. Requiring all authors of EPUB content to include an ONIX record would complicate production, potentially turning authors off from including accessibility metadata. It would also require authoring tools to be able to create and modify ONIX records, and Reading Systems to parse the records to retrieve the information. To minimize this impact on the existing ecosystem, and ensure that the metadata is always available in an easily retrievable fashion, schema.org metadata was chosen.

The choice of schema.org metadata in the package document does not diminish the need for rich accessibility metadata in ONIX records, and there is a requirement to include accessibility metadata in any distribution records that accompany an EPUB publication.

The accessibility group is looking to work with the various metadata authorities to harmonize their accessibility vocabularies as a future objective so that the same information can be encoded across formats.

Can I call my discoverable publication accessible?

If your publication only includes discoverability metadata (i.e., it does not meet the accessibility requirements and is not an optimization), it must not be promoted as accessible per this specification.

Meeting the discovery requirements allows Users to determine whether the content will be usable by them, but is not in itself an achievement that they need to be made aware of.

Accessibility Conformance

What if I cannot meet all these requirements?

Changing production practices is never easy, but there also needs to be a clear benchmark to meet to begin making progress. The EPUB working group understands that the ability to meet these requirements is going to vary depending on the nature of the content produced and how it is produced.

The working group has also structured the specification to soften the move as much as possible, without compromising the usability of the content. The specification only requires WCAG 2.0 Level A conformance, for example, even though many regulatory frameworks require Level AA conformance for web content (publishers already have to meet the higher level in such jurisdictions). Conformance is also not mandatory for an EPUB publication to be valid, so that accessibility does not become a barrier to distribution while publishers are improving their processes.

The working group could not lower the threshold below WCAG conformance, as trying to pick and choose rules to follow would lead to content that is only accessible to certain audiences, and certifying who those audiences would be a complex undertaking.

The issue with raising production quality is why the discovery metadata is so important to include. A publisher may not be able to state their content fully conforms to the specification, but if the work that has been done can be surfaced then users can make the decision for themselves about whether it will still meet their needs.

Is this specification an extension of WCAG?

No, it leverages WCAG and includes additional requirements, but is not an official extension and should not be referred to as one.

Why not include the EPUB requirements in WCAG?

The IDPF intends to work with the W3C and Web Accessibility Initiative in the future to harmonize accessibility for digital publications.

Due to different revision timelines and the need to establish conformance requirements for EPUB Publications, the EPUB working group decided to pursue a separate accessibility specification.

Why are there guidelines for only some HTML elements in the techniques?

The techniques document does not duplicate the requirements of WCAG 2.0. It only addresses situations where the guidance differs for publications, or where no guidance exists.

What if there is another way to make content accessible that is not listed?

The techniques documents for both the EPUB Accessibility specification and WCAG 2.0 are advisory in nature. New techniques are always being developed to handle new situations. If a new technique provides the same level of accessibility, it is acceptable to use it.

Is there a checklist of requirements?

The techniques do not include a checklist, as the requirement for conformance will vary from publication to publication (e.g., depending on the type of content included).

A validation process document is currently under development that will guide in checking the requirements for conformance. In addition, software is being developed to assist in the process. It is expected that the validation process and the software when used together will still require human intelligence for certain tests.

Why is the conformance metadata a URL?

The a11y:conformsTo property takes a URL so that the conformance statement is always the same and can be easily checked by machine (e.g., for validation, ingestion into a distribution system, etc.).

A human-readable statement about conformance should be included in the schema:accessibilitySummary property. The EPUB Accessibility specification does not require wording for such statements.

Optimized Publications

Why identify optimizations?

The identification of optimizations ties into the importance of discovery metadata. Content that meets accessibility requirements for specific user groups needs to be discoverable by those groups, even if the content is not considered broadly accessible (e.g., braille- and audio-optimized publications).

Identifying the third-party standard the content conforms to gives users this missing context.

How do I identify an optimization standard?

Conformance to a third-party standard is indicated by including a link tag in the metadata with a URL that points to where the standard is defined.

<link
   rel="a11y:conformsTo"
   href="http://www.daisy.org/guidelines/epub/navigable-audio-only-epub3-guidelines"/>

The preceding example indicates that the publication conforms to DAISY guidelines for navigable audio-only publications.

It is recommended to include a human-readable description of conformance in the schema:accessibilitySummary property, as well.

Can I call my optimized publication accessible?

The EPUB Accessibility specification does not endorse optimizations; it only address how to identify them in the metadata.

Any claims about the accessibility of an optimization must not make reference to the EPUB Accessibility specification or the IDPF. Such a claim should identify the standard the publication conforms to and the organization that created that standard.

What is the optimized standards registry?

The IDPF maintains a registry of known optimization standards to help facilitate lookup of their identifiers.

This list is provided solely for informational purposes. The IDPF does not develop optimization standards and neither endorses nor discourages their use.

Do I have to register my standard?

No, the registry is informative. The only requirement is that the standard the content conforms to be identified.

How do I register my standard?

Inclusion of new standards can be requested by opening an issue in the GitHub repository (requires a GitHub account). Advanced GitHub users can also update the registry page directly and make a pull request.

Requests should provide as much information as possible about the standard.