search

act-rules / act-rules.github.io

Accessibility conformance testing rules for HTML
https://act-rules.github.io/
Other
136 stars 68 forks source link

Reasons for Reject of "Document has an instrument to move focus to non-repeated content" #1588

Open kengdoj opened 3 years ago

kengdoj commented 3 years ago

Document has an instrument to move focus to non-repeated content

Applicability of "This rule applies to any HTML web page."

As stated in the Background, "this rule only checks [a skip link's] existence"

WilcoFiers commented 3 years ago

Additionally, it is not clear what it means for the focus to cycle to the browser UI. Do I just press buttons until it happens? How do I do that? Do I press those standard buttons at random? How many times do I do that?

This seems ambiguous to me.

Jym77 commented 3 years ago

Document has an instrument to move focus to non-repeated content

Applicability of "This rule applies to any HTML web page."

  • Not all web pages have repeated content so this applicability will result in false positives. Can the assumption that all pages have repeated content be included in the applicability somehow?

This bit is not a problem. The rule is an atomic rule part of a composite one (Bypass blocks). Documents with no repeated content may fail this atomic rule, but they will pass some other (namely Block collapsible) and thus pass the composite one.

I do agree it might mean the requirement on the techniques is not correct.

kengdoj commented 3 years ago

Documents with no repeated content may fail this atomic rule, but they will pass some other (namely Block collapsible) and thus pass the composite one.

Thanks for sharing the composite rule.

If there is no block of repeated content in Block collapsible, would it be Inapplicable or Pass?

Jym77 commented 3 years ago

It would pass. Due to the ACT rule format, all atomic rules from a composite rule must share the same Applicability (well, no exactly, but almost). And "block of repeated content" being a subjective definition cannot be used in the Applicability.

So, in that case, the composite rule, and each atomic, is Applicable to any HTML document.

"Block collapsible"'s Expectation state (roughly) "for each block of repeated content in the test target, something something". If there are no block of repeated content, this sentence is true (universal quantifier on an empty set).

We can probably add a background not and/or an example in "Block collapsible" to make that explicit.

(also, note that we did discuss during writing of these rules whether it make sense to have a rule "document has no block of repeated content" which would make it explicit that "Bypass Block" doesn't fail in this case; we ultimately decided against it since anything passing it would also pass "Block collapsible"… but this is clearly a decision we can revert if we feel adding it makes the full rule architecture easier to understand…)

kengdoj commented 3 years ago

Adding an explanation and a test case would be very helpful to support Pass when there's no repeated blocks. I better understood Pass (instead of inapplicable) with your explanation.

It would be nice if the same could be done with the other rules like Document has instrument, i.e., rule passes if there are no repeated blocks.