Consider new abstract role "phrase"

[jnurthen]

Consider an abstract role for the ontology which would be used for roles equivalent to HTML phrasing elements. Section doesn't seem correct as a parent role for many of these and there doesn't seem to be any other place to put them... Example would be the new subcript/superscript roles

[pkra]

I'm wondering if this could also help with the following problem. If a paragraph contains an inline graphic which represents text (e.g., an icon font, a non-Unicode character/word), then screenreaders will (usually?) interrupt their voicing after the image description is done. I'm not aware of a method to get AT to consider a description to be part of a text flow. Could a role derived from phrase help?

[aleventhal]

Wondering if we can think of some alternative names to consider? A phrase sounds like part of a sentence to me. Although, I just learned that phrasing element is a thing in HTML. Not sure how many authors will know that. Just as an alternative idea, how about "span". Overall I'm fine if we go with "phrase", just wanted to throw it out there. Aaron

…

On Fri, Feb 22, 2019 at 4:07 AM Peter Krautzberger ***@***.***> wrote: I'm wondering if this could also help with the following problem. If a paragraph contains an inline graphic which represents text (e.g., an icon font, a non-Unicode character/word), then screenreaders will (usually?) interrupt their voicing after the image description is done. I'm not aware of a method to get AT to consider a description to be part of a text flow. Could a role derived from phrase help? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#908 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABUAZjI9DDEQJhyljKeyp1LkvT8v8Vxzks5vP7NfgaJpZM4bIUeF> .

[mcking65]

Also consider using as superclass of insertion and deletion added for issue #526.

[brennanyoung]

This seems very similar to the intentions behind role=text (which already has support from many ATs).

Disregarding the (un)suitability of the nomenclature, how is the mooted role=phrase different from the almost-already-implemented role=text?

[jnurthen]

take a look to see if this would help to decide whether to close or do in 1.3

[cookiecrook]

@brennanyoung role=text would be a concrete role, usable by authors. role="phrase_bikeshed" would be an abstract role, not usable by authors, but could be used with inheritance in the role ontology.

[cookiecrook]

@aleventhal I think "span" isn't a good choice for an abstract role, as authors will assume they can use it for generic elements

[jnurthen]

How about actually using the html term. Either "phrasing" or "phrasingcontent" - the more abstract it is the less likely folks are going to use it accidentally.

[scottaohara]

either is fine by me, but i'm personally leaning towards 'phrasingcontent'. that seems more fitting of an abstract term and may make developers pause to figure out what it is for. rather than 'phrase' which seems like a role developers may find reasonable to use when authoring (he says as he's looking at a web page which has role=section added to a section element...)

[jnurthen]

Working group agreed in https://www.w3.org/2022/10/06-aria-minutes#a02 that this has little benefit and closing as wontfix