Yes, I love the
computing device because all that it presents is a solid slab of glossy glass. If it tried to present any other affordance it would be all the lesser for it.
If the rant is supposed to be about the future, then IMO it spends too much time disparaging the present. I think it’s a stupid and pointless rant if the premise is “this isn’t the best it will get.” Duh? That’s like the ultimate straw-man.
But obviously the point of the rhetoric which I originally took issue with is stronger than simply “this is super awesome and amazing but we will eventually do better”.
Rhetoric like “permanent numbness” or claiming the experience has “no connection whatsoever with the task you were performing” is inaccurate aside from being hyperbolic.
In fact I’m comfortable stating that I’ve never felt more connected with or immersed in a map than on a sleek multi-touch display. Additional controls around the screen which I would have to move my hand to manipulate, or an attempt to shoehorn a haptic interface into that — honestly I believe it would only serve to mediate and therefore dampen the connection between user and map.
A globe is haptic. But a globe lacks the affordances of multi-touch zooming. A digital globe that allowed zooming and panning, maybe even with a dynamic physical structure to represent topography, that would be fucking cool, but hardly portable. And so single purpose as to be a novelty.
The beauty of picture behind glass is how adaptable it is to any arbitrary task and how easily it can be reshaped and reformed, and new interaction modalities invented on the glass surface which are joyful and useful and efficient.
Haptic interfaces lock you in, they dictate function based on their form, they provide fixed manipulation points which can not rearrange to best suit the task at hand. As such, haptic interfaces given our current constraints on material science are often inferior and frustrating. Particularly for tasks which require dynamically adaptive coarseness and dynamic range of inputs, such as navigation.
Haptics are particularly difficult to allow users to manipulate interfaces through varying levels of acceleration and jerk, whereas the glass surface excels at this. Faster swipes, slower swipes, flicks, jabs, back and forth movements, and that’s all just one finger.
A click button is a one dimensional interaction point. Rarely is there a “click” affordance which measures harder or softer, except on “glass” type touchpoints.
The scroll wheels in the TM3 steering wheel is a great example of haptics striving to achieve this. The scrollwheel is awesome for fine grain incrementing and decrementing. But it’s also awesome because it has a “flick” mode where if you flick it fast the behavior changes in an interesting way!
For the right hand wheel, single clicks up and down increase and decrease cruise speed in 1 mph or kph increments. But a fast flick up or down (which may actually rotate the wheel 3 or 4 or 5 click points — who knows?) will move the cruise speed up to the next closest 5mph/kph increment (not up by exactly 5mph unless you were already divisible by 5).
This is a very neat HCI trick to provide a more usable scroll wheel. First, the fact that the wheel has click-points at approximately every 15 degrees (and the exact number of clicks per rotation is crucial based on the particular task!) rather than a smooth flowing wheel, translates naturally to integer changes in state.
Second, that it does have some concept of coarse vs fine-grained adjustments through the altered behavior when you flick it (likely triggered by at least 3 rapid clicks). But this all had to be very carefully designed into the hardware and firmware to make it all work for this particular task.
Third, because the particular task at hand doesn’t have a large dynamic range in the scale of the adjustments that need to be made. Rarely am I cruising at 20mph and then quickly find I want to be cruising at 80mph, or else the wheel would be terribly frustrating!
This is why trying to use the wheel (or really any haptic control) to pan or zoom a map is an exercise in total frustration. I know because that’s how my wife’s Mercedes tries to do it. Because sometimes I want to zoom out from a scale of 1mi on screen to 100mi on screen for a trip overview, and sometimes I just want to zoom from 1mi to 1.3mi to see the next turn. There are no obvious integer set-points to allow a haptic interface to lock on to, and the GPS interfaces which impose them tend to suck. The fact that I can’t directly address any arbitrary XY coordinate on the screen instantly compounds the frustration.