Hacker News Reader

Comment by downerending | Hacker News Reader

Now that we're past the "gee-whiz" part of touchscreens, it's becoming more apparently that they're an awful solution for almost every application. Physical controls are simply better.

A couple of small examples: I frequently accidentally turn off my monitor, as it has a "touch" (or even "touchless") sensor that acts if I brush my hand by it. I'm not even sure exactly where it is, as it's not really marked.

ATMs now often have touch screens, but the angles often make it difficult to figure out exactly where to touch the screen to activate the "button".

The thought of something like this appearing in safety-critical applications is frightening.

View on HN

This is not a winner-take-all situation. Physical controls are better for some things, touch screens are better for others. Imagine a map without pinch to zoom, yuck. ATMs are an entirely different case. A lot of ATMs I use have _terrible_ touch screens.

Written on a touch screen.

I'd be interested to hear of cases where touch screens are an obvious win. Zooming a map is great, but you don't need touch for that. Physical controls.

I find zooming/panning/rotating maps without a touchscreen to be awkward on even the most modern car interfaces that I've tried with trackpad joysticks. IMO there's an obvious win there for touchscreens which can do all three in the same fluid movement.

The biggest advantage of touch screens is that they are infinitely reconfigurable. So things like phones where you can have dozens of different types of applications running in the span of a day they provide a clear win.

Another advantage is that you can share control space with display space but I think this is rarely a big advantage (other than the benefits described above in that you can reconfigure what is control-space and display space).

For most devices that aren't general purpose computers I don't think touchscreens are beneficial. For example in a car there isn't that much variability in what you do so there is little benefit of a touchscreen so it doesn't outweigh the downsides.

> The biggest advantage of touch screens is that they are infinitely reconfigurable

Not sure I buy this line of thought. A computer keyboard is an example of an infinitely configurable physical interface

Keyboards aren't really configurable at all. I mean sure, you can decide what each button does in a given context but at the end of a day you have a fixed number of buttons.

With a touchscreen you can change the size of the buttons, because if you have fewer options you can make them easier to hit. Or you can have (limited) other forms of control. You can have "analog" sliders and rotating dials.

Additionally you can configure between output and input which is valuable when you are space constrained. (which is probably not an issue in cars)

I have an old handheld gps unit without pinch-to-zoom (or a touch screen!) It has physical buttons for zoom in and out and honestly it works fine. In that I can easily use it one-handed and in the rain it may actually be superior to a touchscreen device.

I'd happily take a jogwheel over pinch-to-zoom tbh.

The biggest problem with tactile controls is that once they're in the field they can't really be changed. Touch screens allow updates to add or remove elements without changing the physical design, and allow the same physical design to be used for multiple product lines with more versatility than tactile.

Everything has tradeoffs.

Yup. But for cars, a lot of functions will always be needed and will never change. Also, there is such thing as physical function keys that change contextually and can be remapped in the future.

That's true, but as noted in another post, this seems to result in first shipping the crappiest interface anyone can think of, with the knowledge that "we can fix it later". And often, it never does get fixed.

Also, if it's my car, no, hell no, you're not allowed to reach out and change the interface. Get it right the first time, or go home.

The ability to change the product after shipping itself is a huge tradeoff. More and more, companies have taken advantage of this ability by pressing their customers into being unpaid testers. Nobody wants to commit to a final product anymore and quality has declined.

Reply on news.ycombinator.com