Accessibility Conformance Testing (ACT) Rules Implementations
Status: DRAFT 11 July 2019
This page tracks implementations of the Accessibility Conformance Testing (ACT) Rules Format 1.0 for the Candidate Recommendation (CR) phase.
As part of the W3C Accessibility Guidelines Working Group (AGWG), the Accessibility Conformance Testing (ACT) Task Force documents at least 11 test rules that meet the 'Exit Criteria' of the ACT Rules Format 1.0 CR specification:
This set of 11 test rules is a subset of a lerger set currently being developed by the W3C ACT Rules Community Group (CG). Thier development includes review, involvement, and other contribution from a broad representation of the community, including:
- Accessibility Foundation
- Agency for Public Management and eGovernment (Difi)
- British Broadcasting Corporation
- DAISY Consortium
- Deque Systems Inc.
- IBM Corporation Inc.
- Level Access
- Power Mapper
- Universidade de Lisboa
- US Access Board
In addition to completing the 'Review Process' of the ACT Rules Community Group, some of these test rules have documented use by automated testing tools and manual methodologies, including:
- Access Engine, automated testing tool by Level Access
- Alfa, automated testing tool by Siteimprove
- Axe-Core, automated testing tool by Deque Systems Inc.
- RGAA, manual methodology from the French government
- Trusted Tester, manual methodology from the US government
- QualWeb, automated testing tool by Universidade de Lisboa
This initial set of test rules developed through an open and consensus-based process as well as implementations of these test rules in automated testing tools and manual methodologies demonstrates that the ACT Rules Format 1.0 CR specification is sufficiently mature and supported by a variety of stakeholders, to meet the CR requirements. Other test rules are currently in development, and active involvement and uptake has been increasing since the CR publication. To date, there have been no objections to the specification.
The following sections list the 'Exit Criteria' that reflect the normative requirements defined by the ACT Rules Format 1.0 specification. For each exit criterion there is a list of test rules that meet this criterion.
(C1) All rules must conform to the ACT Rules Format
(C2) At least one atomic rule that has a requirement in WCAG 2
(C3) At least one composite rule that has a requirement in WCAG 2
(C4) At least one rule that is a satisfying test, and one rule that is not
(C5) At least one rule that has an accessibility requirement mapping that is not part of WCAG
(C6) At least one atomic rule has no accessibility requirement mapping, but is used in a composite rule that has
(C7) At least two different input aspects are used
(C8) At least two different web content technologies
Note: rules for CSS and for SVG are currently in development too.
(C9) At least one atomic rule has an applicability different from the composite rule it is used in
(C10) At least one rule with one expectation
(C11) At least one rule with more than one expectation
(C12) At least one rule without assumptions
(C13) At least one rule where exceptions are documented in an assumption
(C14) At least one rule where interpretation is documented in an assumption
(C15) At least one rule where no accessibility support information was necessary
(C16) At least one rule where accessibility support information was included
(C17) At least two rules that include a changelog
(C18) At least two rules with more than one definition
(C19) At least two rules that include an issues list
(C20) Every applicability must have at least one inapplicable test case
(C21) Every expectation must have at least one pass and one fail test case
(C22) At least two rules with a background section
(C23) At least two rules with an acknowledgements section
Implementations and data format
(C24) At least two fully automated independent implementations of the same ACT rule (All test cases that should pass are passing, all test cases that should fail are failing)
(C25) At least two semi-automated independent implementations of the same ACT rule (Some results running against test cases are inconclusive, none of the results contradict the test cases)
(C26) At least two independent implementations of the same ACT rule in manual testing methodologies