SC 1.3.4 Another consideration?

In preparatory discussions on testing WCAG 2.1, a colleague and I got into 
a discussion regarding how 1.3.4 could handle other kinds of inputs, 
specifically selects and radiobutton groups in a fieldset.
I have the uncomfortable feeling that although these are all considered 
inputs, the HTML5.2 autofill spec would not support either this:

<label for="ctry">Country</label>
<select id="ctry" autocomplete="country">
<option>United States</option>
<option>Canada</option>
...
</select>

or this

<fieldset>
<legend autocomplete="sex">Gender</legend>
<input type="radio" id="m" name="sex" value="male">
<label for "m">Male</label>
...
</fieldset>

The text in the spec preceding the table of allowed values for 
autocomplete states the following:
"Some fields are only appropriate for certain form controls. An autofill 
field name is inappropriate for a control if the control does not belong 
to the group listed for that autofill field in the fifth column of the 
first row describing that autofill field in the table below. "

This fifth colum is labelled control group, and you'll see "sex" and 
"country" only have a value of "Text". I get why the authors may have 
thought this was reasonable where the expectation was the objective was to 
help a user fill something in, but we are trying to use those values for 
other purposes and so would want a select element about country to have 
the same meaning/purpose as a text input about country.

Is this as thorny an issue as I'm thinking?

Michael Gower
IBM Accessibility
Research

1803 Douglas Street, Victoria, BC  V8T 5C3
gowerm@ca.ibm.com
voice: (250) 220-1146 * cel: (250) 661-0098 *  fax: (250) 220-8034

Received on Wednesday, 28 February 2018 22:53:12 UTC