Re: [mediaqueries4]Differentiating touchscreen+mouse from touchscreen only scenarios

On Fri, Apr 25, 2014 at 7:50 PM, Oren Freiberg
<oren.freiberg@microsoft.com>wrote:

> >The fact that all the keyword MQs have mutually exclusive values so far
> isn't thus a necessary aspect, just an >incidental one.
> I always assumed this was intentional very interesting!! Although it is
> incidental I think it has become the way developers think about it and to
> me it doesn’t seem like this is the right place to break from that mental
> model.
>
> >We also thought of this but rejected it.  Using a special value for
> "both" means that anyone who designs a >stylesheet for "course" or "fine"
> will *not* match this value.
> I think that is the expectation that  "course" or "fine" will not match
> the combined value.
> I think this is desired more than a burden from my perspective.
>
> >Sure, it's possible for people to have styles for both "(pointer:
> >coarse)" and "(pointer: fine)" that fight, but I think most of the time
> the styles'll work fine, and letting people >naively apply their (pointer:
> coarse)" styles on a "hybrid" device will be fine.
> I expect "coarse" content and "fine" content to be very different so
> naively applying both won't be okay. I would guess that mobile devices
> (tablet and phone) to be the majority of  devices with only
> "(pointer:coarse)" so I expect their styles to be very different from
> devices with only "(pointer:fine)" typical desktop computers. As hybrid
> devices are not as prevalent it makes me feel that they will end up with
> bad experiences or more bugs due to accidental clashes.
>

I think there's always some risk that sites won't handle the hybrid case
well (but that risk goes down obviously as more and more laptops are
shipping with touchscreens).  I think the same arguments you're making
could be made for pointer:both - rather than have competing rules active,
pages may not get any rule and so be even more broken.

I think a less error-prone approach to separating them is to have separate
boolean rules, eg. 'fine-pointer' and 'coarse-pointer'.  This is also more
extensible (eg. what if we decide there's another option than just coarse
and fine, 'both' would be pretty confusing <grin>).

We also haven't talked about 'hover' yet.  The spec language there still
uses the "least capable primary pointing device" terminology.  Are you guys
OK with that?