Re: [mediaqueries4] pointer: coarse and pannable, zoomable viewports

On Wed, Dec 12, 2012 at 10:54 AM, Brad Kemper <brad.kemper@gmail.com> wrote:
> On Dec 12, 2012, at 10:26 AM, "Tab Atkins Jr." <jackalmage@gmail.com> wrote:
>>> Is such a UA expected to match 'pointer: fine' when zoomed in, and
>>> 'pointer: coarse' otherwise? I hope not; I'm not enthusiastic about
>>> triggering layout on zoom. If such a UA never matches 'pointer: fine'
>>> but only 'pointer: coarse', it's more likely that page authors will use
>>> this MQ to make content that behaves poorly when the user is zoomed in.
>>
>> It sounds like you're saying "damned if you do, damned if you don't".  ^_^
>
> It does sound like it, if triggering layout on zoom is such a bad thing.

I have no strong opinion on this, but I agree with Florian's current
conception of it being based on the default screen - what the user can
see without having to do anything special.

>> I don't really understand this objection, though.  Can you elaborate
>> with a more specific example?  In general, page authors should respond
>> to this MQ by making touch targets larger.  How does this make a
>> poorly-behaved page upon zooming, unless the page was badly-designed
>> to begin with?
>
> Someone might consider it badly designed, if the touch targets are enormous even when zoomed in to a single column, and someone might consider the same page well designed when viewed at the default zoom. I think it might be a good thing to have the touch targets be finger sized at any zoom.

This is simple.  If you expect the page to be zoomed in while being
used, then don't make things so large!  By basing the MQ on the
*default* screen, though, we let authors respond to what users will
see most commonly.  (I've heard data that a significant fraction of
people don't know you can tap-zoom pages on most phones, for example.)

> Regardless of whether you agree with that, I think having a unit that was equal to pointer discernment size (or estimated pointer discernment size, in the case of fingers) would be better than having a media query for that. Then I could do something like this for a link: 'padding: 0.3pointer-discernment-unit 0.5em; min-width: 1pointer-discernment-unit;' (unit name subject to change). For a mouse, the unit would be equal to 1px.

This doesn't seem useful, because the scale is too large.  While a
mouse *can* point to things with an accuracy of 1px, it's not good to
design things that require that precision.  Furthermore, touch
accuracy means the unit would be about 30px for touch (8mm, a common
size quoted for minimum touch target sizes), a 30x increase, so using
the unit like you proposed would be *terrible*.  While I think it
would be useful to have a "touch" unit that represents a good size for
a touch target, it shouldn't change much based on pointer type.

~TJ

Received on Wednesday, 12 December 2012 19:03:08 UTC