W3C home > Mailing lists > Public > www-style@w3.org > April 2013

Re: [mediaqueries4] - Pointer size/accuracy

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Wed, 17 Apr 2013 10:54:11 -0700
Message-ID: <CAAWBYDBAeyTeRLjb5HJqGThSotY+7PuX-ECZHek+Lb28LcJicw@mail.gmail.com>
To: Robert Kk <umbertoko@gmail.com>
Cc: www-style list <www-style@w3.org>
On Tue, Apr 16, 2013 at 12:44 AM, Robert Kk <umbertoko@gmail.com> wrote:
> Querying for pointer can be very usable. But it is not exact enough for
> me...
>
> When I use an android mobile with a resolution higher the 400dpi I can draw
> many things on the view port. But it will not be very usable for the user.
> His pointer is in real live about 1cm2 and on the viewport over 150px. On a
> desktop|notebook I have a small resolution (like 1280px or 1024px) but a
> very exact pointer. I can click on a 5x5 pixel field....
>
> For accessibility reasons it can be also good to know, how exact a user can
> click. Some people can not hit small fields at all. Other must hard try to
> do it.
>
> For smartphones we can get a solution like zoom-in on ambiguous clicks (like
> in chorme). But we must click two times. Real good responsive design should
> be able to scale the size of action fields based on pointer size info.
>
> My submission is: to use (alternatively) numerical values to recognize how
> exact a user can click.

I don't see in your examples any good reasons to need more than two
levels of accuracy.  Can you explain in more detail a situation where
you'd need 3 or more accuracy levels?

While it's not explained in the draft, "fine" generally means your
pointing targets should be at least 10px wide, while "course" means at
least 40px wide.  I think Florian has noted in the past that it might
be reasonable to expose a "very-coarse" value, if someone can show
that they're developing for a device where targets that are 40px or so
wide are still too small to accurately hit.

(Exposing precise parameter values is sometimes an anti-pattern, as
authors then have to guess what the boundaries of exact categories
are.  Providing the semantic categories up-front can be an important
and useful simplification.)

~TJ
Received on Wednesday, 17 April 2013 17:54:58 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:10 UTC