RE: FW: Pointer target size

The circle approach seems logical because of a finger or stylus, etc. but most UI controls aren't circles and I'd argue that most users don't touch or click targets just in the center.  I tend to touch where I can based on the location of my finger and what appears to be visually the target area.  So unless I'm missing something I'm not sure this actually solve the user issues of activating the wrong target.  I haven't read the research of where in controls people tend to touch - but perhaps this assume that the problematic controls are small in two dimensions rather than small in one direction such as long or thin.

Jonathan

-----Original Message-----
From: Patrick H. Lauke <redux@splintered.co.uk> 
Sent: Thursday, October 29, 2020 9:37 PM
To: public-mobile-a11y-tf@w3.org
Subject: Re: FW: Pointer target size

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.


On 30/10/2020 00:25, Alastair Campbell wrote:

> To keep it simple and allow for spacing I'd suggest:
> For each target, the horizontal and vertical non-overlapping  distance between the center of the target and the @@center of the other targets@@ is at least 24 CSS pixels, except when...

Yup, that seems the same idea as my clumsy "circles of 12px radius from
each target's centre that can't overlap", in essence. Except for the
"non-overlapping distance" part which ... doesn't make sense, and if
you're just going centre to centre, there's no need to talk about
"horizontal and vertical" (unless you intend to split the horizontal and
vertical part of the distance and treat them as separate?)

However, thinking this through, just looking at distances from centres
(either using circles, or just thinking about a line from the centres),
does drop one aspect of the more convoluted "from the centre to one edge
and then distance to the nearest opposite edge" formulation - it doesn't
preserve the actual clearance around the centres themselves. You could
have a 1x1px button that's right next to a 48x48px button and they'd be
fine (their centres are more than 24px apart) but obviously completely
unusable. So perhaps a more complete formulation (that avoids the
not-very-understandable "non-overlapping distance" wording) might be:

"For each target:

- the distance between the center of the target and the center of any
adjacent targets is at least 24 CSS pixels, and
- no other adjacent target's area is closer than 12 CSS pixels to the
target's center

except when..."

Thinking this through even further though, this idea of
circles/distances from centres/even the "non-overlapping distance" idea
now guarantee spacing...but none of them express any kind of minimum
size. No matter how the idea itself is formulated, what happens if i
have two 1x1px buttons that are, say, 50px apart from each other.
They're very wide apart, so each has a very comfortable
spacing/clearance around it that doesn't "overlap" with the clearance to
the other button. Yet still, they have ludicrous target size. I don't
think any of the formulations here, nor in the original document,
account for that - trying to define size by not defining size can be
gamed the opposite way, resulting in super-small controls but with ample
clearance?

P
--
Patrick H. Lauke

https://www.splintered.co.uk/ | https://github.com/patrickhlauke

https://flickr.com/photos/redux/ | https://www.deviantart.com/redux

twitter: @patrick_h_lauke | skype: patrick_h_lauke

Received on Friday, 30 October 2020 21:47:52 UTC