Re: Fw: drawSystemFocusRing, drawCustomFocusRing

On Mon, Nov 11, 2013 at 9:31 PM, Jatinder Mann <jmann@microsoft.com> wrote:

>  I agree with the feedback from Apple [1] that the current name is
> confusing due to the usage of "draw" and "ring". The user agent isn't
> drawing the focus, it's setting the focus.
>

The API is very confusing.
drawSystem/CustomFocusRing does not set the focus. It checks if the element
has focus and then draws the ring.


> The appearance of a focus outline is a side effect; per spec, some systems
> require enabling accessibility features before the focus outline is shown.
> Also, the path of the focus outline doesn't necessarily have to be a ring,
> it can be any shape or path drawn using the existing 2d context API. I
> think a more appropriate name for these APIs would be setSystemFocusMapping
> and setCustomFocusMapping. If we want web developers to understand and use
> this feature, we should make the name clear.
>
>
>
> On Mon, Nov 4, 2013 at 10:58 AM, Rik Cabanier wrote:
>
> > the a11y software should always be notified of the region where the
> focus ring would be located, but that is not in the spec.
>
> I think those details are currently missing from the spec. We should add
> that to the spec.
>

Another question is if the a11y software should be updated about the focus
region even if the element was not focused.
If it were to do so, the browser and the a11y software would know about all
the regions that get focus (and their ARIA roles) even if they hadn't been
focused yet.


>
>
> >  should the browser scroll to the region that receives focus.
>
> I don’t believe magnifier tools are able to scroll within the user agent
> without the user agent supporting this feature. I think the spec should
> require this behavior.
>
>
>
> > Why is there text for high contrast focus rings if no one implements
> them?
>
> From an accessibility point of view, is there value in having a custom
> focus ring method? If the user wants a particular focus ring, wouldn’t they
> just change their system settings? If the user has decided to use a
> particular system setting, shouldn’t we just respect the user’s setting?
>

Dominic had the same feedback.


>
>
> [1] https://lists.webkit.org/pipermail/webkit-dev/2013-October/025685.html
>
>
>