Conditional Requirements

[rachaelbradley]

One of the proposed improvements to WCAG 2 is the ability to create conditional requirements. In WCAG 2, requirements always apply unless an object (such as multimedia) is not present. This limits the ability to create requirements that apply to specific languages, writing systems, technologies or other situations.

We've begun a discussion of conditional requirements as a way to support internationalization. This discussion is to work out how conditional requirements will be documented and work. Questions for discussion include:

• What information needs to be included in the normative and informative text?
• Can conditional requirements be foundational?
• What happens when a conditional requirement exists but some conditions are not yet documented (ex: language based but not all languages are covered)?

Examples:

• Line spacing requirements will vary based on the alphabet and orientation of the text
• The type of tense that best supports plain language varies based on language

[rachaelbradley]

[chair hat off] Here are some examples of a possible approach:

Guideline: Users can access explanations of or alternatives to ambiguous words.
Requirement: Content includes all letters and diacritics to understand the meaning of the word.

Conditions where this applies: Hebrew, Arabic
Conditions where this does not apply: English
Process for determining whether this applies: One or more native speakers of a language should determine if words are used with missing letters or diacritics that can change the meaning of the word. Once this is documented, send the documentation to WAI for verification and inclusion in WCAG 3.

Generic Method
Include all accents, characters, and diacritics that are necessary to phonetically read and understand the words

HTML Method
When text is missing letters or diacritics, provide accessible hover text that shows the word with all accents, characters and diacritic necessary to phonetically read and understand the words.

AI Method
Train an AI to correctly interpret the text and make this available to users who need it.

[GreggVan]

That looks complicated but I think the conditions are not necessary. What you wrote at the top...

Guideline: Users can access explanations of or alternatives to ambiguous words.
Requirement: Content includes all letters and diacritics to understand the meaning of the word.

... is sufficient by itself. There is no need for conditions.

• if a langauge has none - then you have, by default, included "all" letters and diacritics (there are none). -- you pass
• if a language has UNambiguous diacritics or missing letters that ARE NOT NEEDED to make it non-ambiguous. --you pass
• if a language has ambiguous diacritics or missing letters that ARE needed to make it non-ambiguous. -- then they need to be provided. (if done automatically by the user agent - then the author needs to do nothing. Otherwise they do. What they need to do is dependent on the technology they are using (e.g. HTML or PDF or ...)

but I THINK the two lines at the top are all that is needed.

Do you see something else?

Where I see conditions being used are for things like

Where content includes more than three flashes in any one second period, the content needs to xxxxxxxx
If there are no "3 flashes within a second" -- they can ignore the rest of the provision.

also like these

Where content includes pre-recorded video sychronized with audio, xxxxxxxxx

Where content provides real-time bidirectional voice communication, xxxxxxx