Re: 1.4.11 if a button has a background less than 3:1 and no border does it pass

Jonathan wrote:
> An exception for buttons seems arbitrary.

It isn’t an exception for buttons as such, but as agreed 13 months ago, links & buttons generally don’t require borders/backgrounds.


> this is exactly why I was concerned with the results of that study being used to drive decisions.

It hasn’t driven any decisions, the buttons/links aspect was decided a year ago and expanded on in the understanding doc over 6 months ago.

What the Google research has shown is that there are a lot of factors that go into “Visual information required to identify user interface components” for things like links & buttons, and we need more research to help refine those requirements in future.


Jake wrote:
> Added a question in the Github about the programmatic determinable possibility we’ve talked about yesterday for another possible SC

That might be helpful in future, but there is a place for guidelines that don’t require AT or personalisation. There are some things that are difficult or impossible to achieve without AT and/or personalisation, but that is not the aim of 1.4.11.


Abi wrote:
> I also saw a research paper presented this week that highlighted again that users with cognitive disabilities (in this case dementia and aphasia) rely on  visual affordances to identify controls.

Is that paper publicly available? Also, did it define what constitutes an affordance for controls on a website? I feel like that has been missing from research I’ve seen, but I’m not in academia so I don’t see that much research…

For example, I suspect that if you treat a navigation bar as a component, the links within it would not need individual affordances/indicators. If that’s true, we could refine an SC around controls in groupings not needing their own indicators, but perhaps requiring stand-alone links to use some sort of affordance. However, I don’t know if there’s any basis for that!

The Google team’s research was a first step to the sort of detail we need to progress that, it listed out the relevant factors from an interface point of view.


> So I feel we should be clearer about what visual affordances are in scope to provide an alternative contrast rather than abandoning the principle of supporting users with cognitive or low vision who are not using assistive technology.

Providing an (extra) affordance for controls was not the aim of this SC from when I first learned about it with the LVTF, it was aimed at ensuring people with low vision could see inputs etc. An equivalence situation. I’m not even sure links (or buttons) were intended to be covered at all, the term ‘UI components’ brought those in scope, but I can’t remember if that was the intent.

I suspect that during the early SC drafting process the thinking was that it would be a good thing for buttons / links etc to have contrasting borders if there are borders. However, partly from examples provided from public comments we realised there could be negative un-intended consequences. (And yes, I was involved in that early drafting process, but it’s a bit hazy so I use words like “suspect”!)

As I’ve said elsewhere, an SC about affordances would be a good thing, but either we need more research/findings to help us define the requirement better, or we need a different method of conformance (i.e. something more usability based in Silver).

Cheers,

-Alastair