WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Proof of compliance?


From: Duff Johnson
Date: Dec 13, 2017 9:36AM

> You've answered your own question: due to the non-static nature of web
> content today, it's kind of hard to have a static, machine-readable
> "conformance claim" that would always be accurate.

I didn't mean to imply that it would always be accurate. Clearly, any change to the site would require some sort of validation protocol to justify the continued use of the conformance claim.

> And while a 'dynamic'
> claim could certainly be crafted using things like mechanical testing tools
> and change-log parsing, the value of a claim such as that would likely also
> be diminished from a "legal-defensiveness" position, which is often the
> root of these types of claims and requirements, due to its malleability
> ​ and ever-changing reporting values​

I take this point entirely.

<shameless PDF plug>
One of PDF's useful qualities is that it IS entirely self-contained in this regard, which makes the conformance claim problem that much easier to solve.
</shameless PDF plug>

> That said, a machine-*referenceable* conformance claim could be addressed
> using metadata (perhaps something like Schema.org's "Claim Review" -
> http://schema.org/ClaimReview), but the conformance claim would still be
> based upon a "snapshot" and date of review. This is not unlike a CPA's
> stamp, where the Chartidanered Accountant claims that what was reviewed on a
> specific date met the claim's assertions, but beyond that date, all bets
> are off.
> ​Or were you thinking of something different?

I was thinking of precisely what you've outlined above. It's too bad that WCAG 2.0 (and 2.1, it seems) provide so little guidance on this.