Input Purposes for User Interface Components is an appendix, so should be marked as non-normative

[fstrr]

Discussed in the backlog task force meeting on 5th April, Input Purposes for User Interface Components is an appendix, so should be marked as non-normative.

The trick to this seems to be adding a class="informative" attribute to the containing section element, but building the site locally doesn't create the homepage, so can't check this.

[patrickhlauke]

you beat me there by 5 minutes...was just about to push up a PR 🏃

[fstrr]

Well, you took that merge conflict off my hands, so I had a little time to look at something else

[mbgower]

As noted in the PR, I have located the original official WG response showing that this was clearly intended to be an appendix.
Given the wording of 5.1, an acceptable solution is also simply to add the word "Appendix" to the heading.

Introductory material, appendices, sections marked as "non-normative", diagrams, examples, and notes are informative (non-normative).

That said, I think it would be good to see if there's anything in a style guide that shows how appendices are supposed to be structured, and make sure we follow that. Some may argue that the sections labelled with letters in our document are the indicators of appendices (again, with no label). @shawna-slh @iadawn are you aware of any guidance on this?

My old copy of Chicago states:

When more than one appendix appear in a book, they should be numbered like chapters (Appendix 1, Appendix 2, etc.) or designated by letters (Appendix A, Appendix B, etc.), and each should be given a title as well.

[mbgower]

It has been pointed out that WCAG 2.0 is significantly different than WCAG 2.1 in regard to structure, with each appendix specifically named and grouped under an Appendices section following the (unnumbered) conformance sections.

WCAG 2.0

WCAG 2.1

WCAG 2.1 introduced a right navigation for the outline. This may have contributed to changes in visible structure, but to point out some obvious differences, the glossary was previously (and problematicly) in an appendix BUT also marked normative. In 2.1, the glossary was moved forward (and its enumeration changed from a letter to a number), but while some Acknowledgements and References remained lettered B and C, all references to appendices vanished.

[alastc]

My suggestion from the meeting:

Add the appendix heading about change log, and mark those and appendix a/b/c. (Probably done with the respect based classes rather than in text.)

Leave section 7 as it is (removing transction-amount separately), and it essentially is normative by that marking, removing the ambiguity.

[dbjorge]

I 👎'd the change (and 👍's @alastc's comment above) because I maintain that regardless of the original intent, the fact that this section is incorporated by reference as a required part of specifying SC 1.3.5 means that the section must be normative (unless we were to update SC 1.3.5 to say something like “for example” when referencing it)

[GreggVan]

It looks like there might be a confusion between something "being" normative and something "containing" normative information.

"Sections" cannot be normative. Only the specific clauses that are required for conformance are normative.
(I know this was mislabelled in past WCAGs)

• the titles of provisions are not even normative (only the requirement itself)
• notes are never normative (they should only clarify what is normative )
• examples are never normative -- and they should never fall outside of or expand the requirement language

What is normative?

• the text of the provision itself - including exceptions if and only if they are included in the provision language itself
• definitions
• the conformance section/requirements

[bruce-usab]

Discussed on backlog call 9/6. Conceptually, adding <h2>Appendices</h2> after the Glossary and before Input Purposes, is the simple fix -- but it is not that easy of course. We will defer to @kfranqueiro to propose how to proceed. Errata that results in visual change needs to go through CFC process (of course).

[kfranqueiro]

Since this is under TR, probably better to consult @iadawn first rather than me.

[bruce-usab]

Discussed on TF call 9/13. We remain in agreement with the change and that it is editorial. Regardless, we closed the previously drafted PR #3778 since the fix will not be as easy as that.

[bruce-usab]

AG decision is memorized back in 21 March 2018. It was during a face-to-face meeting (maybe coincident with CSUN).

RESOLUTION: working group decides to move the list into an Appendix of WCAG 2.1 unless that change contravenes CR status.

[mbgower]

@alastc can we assume you and @iadawn have this in hand? I want to ensure that you have this in mind for the CFC.

[bruce-usab]

Discuss on TF call. I had meant to see I could find AG decision in minutes (and not just only the one GitHub issue).