Re: Comments on proposed new SC 2.5.3

One slight problem of asynchronous collaboration is that a few hours after
Detlev's comments we had the weekly call and worked for a hour on it ... It
is no longer tied to touch, and addresses, I believe in a fairly elegant
way, all the concerns to date...

2.5.3 Up-Event Activation: Single touch and/or pointer activation triggers
on the up-event, or has at least one of the following characteristics
(Level A):
- provides confirmation,
- is reversible,
- a mechanism is available to trigger on the up-event.

Note: This is when platform assistive technology that remaps touch gestures
is not turned on.

Also have revised the understanding document and provided some alternative
language for the SC.
https://www.w3.org/WAI/GL/mobile-a11y-tf/wiki/Proposed_revision_of_2.5.3


On Sat, Apr 16, 2016 at 5:29 PM, Patrick H. Lauke <redux@splintered.co.uk>
wrote:

>
>
> On 14/04/2016 15:53, Detlev Fischer wrote:
>
>> Just taking a minute to think about 2.5.3
>>
>> Echoing Patrick's advice that we should not focus on touch if the
>> issue is more general, it seems fairly obvious that  "2.5.3 Touch Up
>> Activation" or "2.5.3 Single Taps and Long Presses Revocable"
>> describes an issue that is equally valid for mouse pointer
>> activation.
>>
>> Which suggests we might draw the boundary wider and rename it to
>> something like SC 2.5.3 "Support undo"
>>
>> Which contradicts the renaming I have suggested in the last telco.
>> "Touch Up Activation" sounds easier (which is a benefit), but
>> narrowing the issue to touch seems inappropriate for a SC - it would
>> be OK on the level of Technique.
>>
>> So itf we try to tackle the general issue of supporting undo by not
>> triggering things on touchstart / mouseDown, the question remains
>> wehther it is really inside scope for WCAG.
>>
>
> Going back even further, rather than "undo" was the original issue,
> fundamentally, about "Avoid that users accidentally activate controls
> and/or have a way to 'bail out'"? (which won't win any terseness awards,
> but thought I'd throw the lot in there).
>
> So the normative part can, in a tech agnostic way, hopefully convey this
> idea (which is just as applicable to keyboard, switch, mouse, touch, voice
> activation, etc users) that an app/site should be built in a way that a
> user doesn't accidentally click on things they didn't intend to, and that
> if they already started a click activation (e.g.  touch down, mouse button
> already pressed down, etc) they either have a way of cancelling this
> activation (by moving their finger or mouse while still pressed
> outside/sufficiently away from the control before lifting their
> finger/releasing the mouse button), OR by providing some way of
> undo-ing/reverting the action - IF the action is "of consequence" (e.g. if
> it was a touchscreen piano, or the fire button of a real-time [rather than
> turn based/tactical] space shooter, it's no big deal if it activated by
> accident, and an undo would not be practical/possible).
>
> Then, in techniques, it can go further into tech specific "bind event
> listeners to both touchend / the "up" AND the generic "click" / activation;
> for mouse, don't listen to "mouseover" but "mouseup" AND "click"; etc.
>
> P
> --
> Patrick H. Lauke
>
> www.splintered.co.uk | https://github.com/patrickhlauke
> http://flickr.com/photos/redux/ | http://redux.deviantart.com
> twitter: @patrick_h_lauke | skype: patrick_h_lauke
>
>
>