G183: Removing hover/focus steps from procedure

[alastc]

Closes #1272
Related to #201

[yatil]

So this PR means that it is now OK to have a hardly visible link that even does not change on hover/focus… I guess that is only consequential from PR #1500 and the discussion around using this, but it is still disappointing to me.

[alastc]

There are two necessary reasons for this:

1. Having an SC for non-text contrast does not work unless you are differentiated colours by contrast. I.e. find an icon that is not using colour (general sense) to differentiate the icon from the background.
2. There would be too many false positives, i.e. examples where people (including with low vision) can differentiate links by 'colour' (in the general sense).

So yes, it is possible to create very difficult examples, but the opposite is also true so the metric is not enough to make it binding.

[patrickhlauke]

Suggest tweaking the title of this issue to also indicate it touches on the understanding doc itself

x-ref also #1500 (ah, i see @yatil already joined the dots there, sorry for re-mentioning)

[patrickhlauke]

What was the state of this? Will it be merged, or does it need further discussion?

[patrickhlauke]

@alastc so now that you've added the step back in ... are we saying that the 3:1 ratio is not sufficient?

"For this technique, other visual highlights are also required"

so does it pass or not without the extra highlighting? this is now contradictory

[patrickhlauke]

"pointing (mouseover) to the link" ... why not hovering? don't think anybody does anything explicitly on mouseover from a JS perspective, it's usually :hover from CSS

[patrickhlauke]

if it's more a case of providing an additional cue as a technique that goes beyond the normative requirement, then i'd suggest not using the word "required" in the phrase there. instead, making it clear that although the 3:1 ratio is sufficient to pass the minimum requirement of use of color, this technique suggests adding an additional highlight/cue that goes beyond the normative requirement, for increased usability

[patrickhlauke]

👍

[yatil]

Super concerned about adding back hover/focus into what is supposed to be a sufficient technique. hover/focus are not needed for meeting 1.4.1 (as we have now – sadly – determined), so sufficient techniques should not include further advisory advice.

Consider removing hover/focus from this technique and instead creating an advisory technique that advices on using hover/focus in addition to the sufficient technique.

[patrickhlauke]

agree with @yatil this should be an advisory technique. not a sufficient technique. the sufficient part is the 3:1 contrast, the rest is window dressing.

may also be good to add a note explicitly clarifying that while this will be of benefit to mouse users, it won't help on touchscreen or stylus devices where there is no concept of hover (well, the vast majority of stylus devices ... some fancy ones do have hover stylus support, like Samsung Note ones)

[awkawk]

This should probably be in a paragraph before the list:

For each instance where color is used to convey information about text:

This will also affect the "checks #2 and #3 are true"

[mraccess77]

I think the hover part does go beyond WCAG so we need to keep that "beyond" aspect in.

[mbgower]

@alastc I've added in a modification which I think addresses my concern. I didn't put in a link to 2.4.7, as I wasn't sure on the syntax.

[alastc]

@yatil and @patrickhlauke, we discussed this in the meeting, and there's a key point that sufficient techniques can go beyond the minimum requirements.

The description now includes:

To meet success criterion 1.4.1: Use of Color a relative luminance (lightness) difference of 3:1 or greater with the text around can be used. This technique goes beyond the success criterion and...

We also discussed removing the technique entirely, but decided not to as it helps to have sufficient techniques, and this one in particular helped the SC overcome objections originally.

[patrickhlauke]

ah, shamefully i have to admit i was unaware of the exact definition of advisory techniques (that, in essence, they may not even address the actual ask of an SC). somehow got it in my head that advisory was "sufficient to pass the SC, plus something extra on top of that". so i can see how this is a sufficient technique, but not strictly the "minimum sufficient technique" as it incorporates extra stuff.

long way of saying yup, agree (now that i delved into the actual definitions of advisory vs sufficient)

[yatil]

Sometimes I just wished we would use words in a way that they are used commonly outside of our accessibility bubble where sufficient does not mean going beyond. This conflating of meaning is a constant struggle, especially for people learning to differentiate between best practices and absolute requirements. Having Pass/Fail requirements but muddying the water in examples does not really help learning.