Open Jym77 opened 1 year ago
See also https://github.com/w3c/wcag/issues/2820
Also, Attribute is not duplicated is not a 4.1.1 failure (in the light of the WCAG discussion), and never was. We should remove the mapping altogether.
From the CG call:
Is it worth maintaining a mapping of how 4.1.1 failures map to other SC? I'm happy to do this if there's any interest.
There have been 2 conflicting interpretations in the WG of the "nested according to specification" requirements:
Someone on the WG put this together, but it's incomplete: https://docs.google.com/document/d/1MJ6FxO7ujQ4X9BQtAnDDoWyvpAKU44MR4h-bob9SG7M/edit
There are other issues like wrapping a <table>
with a <button>
is not allowed in the content model, and makes the table presentational by removing the table semantics (related to https://www.w3.org/WAI/standards-guidelines/act/rules/307n5z/proposed/)
Following this WCAG discussion, decision has been taken to remove 4.1.1 from WCAG 2.2, making it effectively not backward-compatible with WCAG 2.1.
This means that our rules that map to 4.1.1 (Attribute is not duplicated and Id attribute value is unique are going to map to a non-existing SC (although they might still be relevant for people using 2.0, e.g. Section 508…)
How can we handle that? Should we remove the requirement mapping from the rules? Deprecate the rules altogether? Find a way to say "this maps to a 2.0/2.1 SC but not to a 2.2 one"?
(note that we have a similar but much less strong problem with 2.4.7 changing level)