Re: Updates to Understanding 1.4.11 part 2

Also to add to the *wide *interpretation is the fact, as Jon Avila has
noted (thanks Jon for pointing that out - I had forgotten about that as I
rarely reference Failure Techniques), that we already have an existing
Failure Technique of not setting background when foreground is set and vice
versa, which states:

"



*Unless an author specifies both foreground and background colors, then
they (the author) can no longer guarantee that the user will get a contrast
that meets the contrast requirements...Failure Example 1: Specifying only
background color with CSSIn the example below the background color is
defined on the CSS stylesheet, however the foreground color is not defined.
Therefore, the example fails the Success Criterion.*"


...so I will additionally argue that the *wide *interpretation going
forward would be consistent with prior and existing guidance.


> Let’s try to bring this to a decision (for the group), I’ll summarise and
you (and others) can see if it’s a fair summary.


Thank you for the summaries Alastair. Is this another straw poll? (The straw
poll from Tuesday's call
<https://www.w3.org/2018/06/12-ag-minutes.html#item03> saw 6 in favor of
the wide interpretation, 2 who indicated "they could live with it", and 3
opposed - although the minutes indicate that David MacDonald voted both for
and against this, so I'm not sure where he actually stands today... his now
immortal "I didn't change my mind, I changed my understanding" comment :-)
).

If the Chairs anticipate that your summaries are going to materially change
everyone's thinking, then can we go straight to a WBS survey or CfC?

Otherwise, I continue to strongly support the *wide *interpretation.

JF



On Thu, Jun 14, 2018 at 11:35 AM, Alastair Campbell <acampbell@nomensa.com>
wrote:

> Hi John,
>
>
>
> Let’s try to bring this to a decision (for the group), I’ll summarise and
> you (and others) can see if it’s a fair summary.
>
> (NB: I *lean* towards excepting default focus styles, but see the
> arguments on both sides.)
>
>
>
> Either way we can and will encourage best practice in the Understanding
> doc, but for what we pass/fail under the SC, let’s make a choice.
>
>
>
> If we take the *wide* interpretation of the exception (i.e. if the
> developer touches anything related to the component, including the
> background of the page the exception does not apply), the positives are:
>
>
>
>    - Making authors set the focus style explicitly solves the problem
>    that for *some* users: the default focus style on a blue background
>    (and some other types of background) will make the focus practically
>    invisible.
>
>
>
>    - Related but less of an impact is that the visibility of the default
>    focus indicator can vary significantly depending on background colour and
>    browser used. (NB: Some browsers ensure visibility with multiple colours or
>    inverting the colour on dark backgrounds.)
>
>
>
>    - Testing is more straightforward if the focus color is explicitly
>    set, and given the commonality of setting the background colour, the wide
>    interpretation would in-practice enforce setting the focus style.
>
>
>
>
>
> If we take the *narrow* interpretation of the exception (i.e. if the
> developer doesn’t touch the outline they pass this SC), the positives are:
>
>
>
>    - It doesn’t add something for developers to do that (in theory at
>    least) should be dealt with by the user-agent.
>
>
>
>    - The person’s preference (especially for the better browsers) are
>    less likely to be overridden by the site.
>
>
>
>    - It would be consistent with 1.4.13, which has a similar exception
>    for the title-attribute pop-up.
>
>
>
>    - Using the default focus style is good for avoiding the
>    focus-on-click issue, where an ‘ugly’ focus style appears around components
>    on-click. (E.g. go to https://www.w3.org/WAI/ and click & hold on
>    “What’s new in WCAG 2.1”. You might need to hard-refresh to see the style
>    as it’s just been updated.)
>
>
>
>
>
> Anything else? (Just on this!)
>
>
>
> -Alastair
>
>
>
> PS. Re thickness of borders, I just meant that it was user-agent / user
> set, so if that is your preference it’s helpful if the site doesn’t
> override it.
>



-- 
John Foliot
Principal Accessibility Strategist
Deque Systems Inc.
john.foliot@deque.com

Advancing the mission of digital accessibility and inclusion