Related spec PR: whatwg/html#12424
Spec issue: whatwg/html#10293
The HTML spec is gaining a readonly content attribute on <fieldset>
that propagates a read-only state to descendant form controls that
support being readonly — mirroring how disabled already works.
HTML AAM currently maps the disabled state of a fieldset's descendants
to aria-disabled (or the platform equivalent). The readonly state on
those same descendants should presumably map to aria-readonly (or
platform equivalent), but I'd like to confirm:
- Should a descendant
<input>/<textarea> that is read-only due to
the fieldset expose aria-readonly=true (or the platform AX
readonly state)?
- Should the
<fieldset> itself expose any readonly state to AT, or
only its descendants? (The fieldset has no aria-readonly mapping
today.)
- Are there any platform AX APIs where a "readonly group" concept
exists and we should map it differently?
Filing this so HTML AAM can be updated alongside the HTML spec change
Related spec PR: whatwg/html#12424
Spec issue: whatwg/html#10293
The HTML spec is gaining a
readonlycontent attribute on<fieldset>that propagates a read-only state to descendant form controls that
support being readonly — mirroring how
disabledalready works.HTML AAM currently maps the disabled state of a fieldset's descendants
to
aria-disabled(or the platform equivalent). Thereadonlystate onthose same descendants should presumably map to
aria-readonly(orplatform equivalent), but I'd like to confirm:
<input>/<textarea>that is read-only due tothe fieldset expose
aria-readonly=true(or the platform AXreadonly state)?
<fieldset>itself expose any readonly state to AT, oronly its descendants? (The fieldset has no
aria-readonlymappingtoday.)
exists and we should map it differently?
Filing this so HTML AAM can be updated alongside the HTML spec change