Re: Plain Language

Hi Stein Erik,

> One problem with your suggested generic user focused SC is that it is only testable/possible to validate from the position of a user

That isn’t what I meant, let me try to explain it from another perspective.

The WCAG 2.x structure is:

  *   Principles (4)
  *   Guidelines. Not testable (13)
  *   SCs, testable (~78)
  *   Techniques, with test procedure for one method that fulfils the SC (~100s)

(With understand docs to the side of SCs.)

Currently the “guidelines” are used for categorisation, e.g. reflow is in Distinguishable with 12 other SCs.

I’m suggesting we drop guidelines as a categorisation level and have one ‘guideline’ per set of technology-specific SCs.

So that would lead to:

  *   Principles (4 now, but presumably this could be re-thought)
  *   Guidelines (~78 from WCAG 2.1, presumably more later)
  *   SCs (~78 multiplied by however many technologies we cover)
  *   Techniques (~100s)

So the plain-English ‘guideline’ is not what you are testing, but it is the home for each tech-specific and testable SC. From an interface point of view, if you selected to view only HTML-stack SCs, you’d have just those SCs showing.

If you are a non-technical user, you might not have any technology SCs showing!

The BBC sort of takes this approach, I’m suggesting a slightly more extreme version:
http://www.bbc.co.uk/guidelines/futuremedia/accessibility/mobile/design/content-resizing


It would lead to a flatter structure under principles, so we might want to create more granual principles for easier categorisation. However, but the bit I’m focusing on here is the plain-English guideline & SC relationship.

When I read through the Silver research, two of the problem statements stood out to me as contradictory:

  *   1. Too Difficult to Read
  *   3. Ambiguity in interpreting the success criteria

This is my suggestion for resolving those conflicting requirements.

Cheers,

-Alastair