Re: More thoughts on ISSUE-736 -- names for list items (Resending to fix links)

On 08/19/2015 02:19 PM, Joseph Scheuhammer wrote:
> On 2015-08-19 1:54 PM, Joanmarie Diggs wrote:
>> What are the odds that we can get most authors to supply the name? Could
>> we not instead modify the name-from-contents computation of list items
> 
> This implies that you agree that listitems require a name. 

Some of them; not all of them.

> Can you provide the rationale?

I think widgets (and their focusable/selectable children) should have names:

1. Widgets (and their focusable/selectable children) have names in
   desktop toolkits.
2. ATs expect to be able to present the name of the focused widget
   and/or its active descendant as the user interacts with a UI.
3. Automated UI testing tools depend upon interactive objects/elements
   having names for the purpose of identifying/locating them.

Thus *if* the list item is inside a list box or similar widget, then
that list item should have a name. Perhaps we overgeneralized -- or
failed to explicitly state -- the types of list items for which a name
is applicable?

On a related note, why do table rows in a data table have accessible
names created from their children? It's the same problem.

> paragraphs do not require a name.  Paragraphs are containers
> of structured text and can be short, medium, or long. What is the
> difference between a paragraph and a listitem such that the latter
> requires a name?

Indeed. I would argue that there is no difference between a paragraph
and a list item *which is not the child of a widget.* So were it left up
to me, textual list items in (un)ordered lists -- like those in the
Doxygen-generated docs I mentioned -- would NOT have a name.

It would be interesting to see if any AT is presenting the name and not
the content for a non-widget list item.

--joanie

Received on Wednesday, 19 August 2015 19:22:09 UTC