The New Interface Is There Is No interface | Hacker News Reader

The New Interface Is There Is No interface

tightwind.net

I am a music producer and audio engineer, which these days means I labor at the computer. I have come to the conclusion that the GUI as we know it is a really bad interface at which to labor.

One of the fundamental problems is that it breaks our normal mind/body connection. It is virtually impossible to develop any kind of muscle memory using an interface that requires different physical movements for the same act.

Using a application over time, we do learn to work in an unconscious, trance-like state. But it requires a rigid posture to maintain visual connection with the location of the mouse and the state of the interface. This leads quickly to fatigue and potential repetitive strain injury.

I do think the touch interfaces are an improvement in this context (they reconnect mind/body/application). But for something as labor intensive and complex as audio production (or any media production) I don't think they will be usable for sustained work (at least with existing interface metaphors).

I have developed a tool for use with Pro Tools audio software, the Hotkey Matrix. It is a keyboard with pre-configured single key keyboard shortcuts. It has a number of features which are distinct from the many other attempts at control surfaces for audio software, and somewhat against the tenor of the times.

The shortcuts are fixed. We have iterated the specific shortcuts and layout over many years of production use. They are color coded, there are no modes or pages, so the same function is always in the same place and it is easy to target. It basically replaces and adds to the default keyboard shortcuts in Pro Tools. Pro Tools has a nice set of default single key shortcuts, but since it has so many, increasingly elaborate key bindings and two hands are required to use them (some have four modifier keys).

I'm convinced this is a superior interface enhancement on a lot of levels. It brings back some tactile response, and makes both muscle and mental memorization a lot easier. Since production work is extremely repetitive, it saves a thousands of keystrokes.

I think there is a long way to go in improving interfaces for complex media software. The computer brings tremendous benefits and cost savings to audio production, so there is no going back. It is now trivial to build up a virtual studio environment in a laptop that would have literally cost millions of dollars back in the 90s (not to mention insane electric bills). But the awesome thing about old school analog production was that each parameter in a project got its own dedicated knob/button/fader.

I miss those all those knobs/buttons/faders!

I'm also convinced that most software developers who make tools for professionals never use their software as a professional does (they don't have the time). Hence they make design decisions that turn out to be non-optimal over long use. I've used a lot of audio applications, and IMO Pro Tools is the best. My guess is that since they have a large installed base of professionals and some kind of feedback mechanism from the user base, their updates tend in the direction of increased usability (for pro users). Usability for apps that are simple and used infrequently is very different, and it is here that the comments of the OP are most on point.

Anyhow, I've wanting to run this by the HN community to see what people think. http://www.youtube.com/watch?v=qTvhwtvw2sI

Physical objects have/are interfaces too. What do you think handles, levers, door knobs, gear shifters, file cabinets and flat surfaces are?

Between our senses and external reality, there will always be an interface. The beauty of computers is we are much less dependent on the physical properties of actual objects in designing those interfaces. Maybe that's what the author is getting at, but I think they are saying it wrong.

when I started doing user interface testing on websites in the mid/late 90s, all the literature I could find about user testing in the engineering library was related to industrial design / testing consumer products.

EDIT: it was probably more accurately the late 90s

In many ways, the industrial realm is still where the forefront of interface design is, and covers an area far wider than anything software hopes to do right now. In fact, I'd consider the industrial flavor of interface design to be far truer to "interface design" than what we see in mobile/web, which is often just "graphics design" in camouflage.

In industry, UI has consequences. A confusing readout may crash a plane and kill hundreds. A warning that doesn't trigger under expected circumstances can slice a man in two. A button that doesn't require enough force to push can accidentally decapitate. The list goes on. What you get from that side are incredibly usable, incredibly easy to understand, incredibly fault-tolerant interfaces... that are ugly as hell.

Whereas from the mobile/web side we often see pointlessly obtuse, absurdly minimalistic, confusingly designed interfaces that are works of art. We often confuse these with "good interface design" because they are different, new, and beautiful.

I really wish those of us in web/mobile can get off our high horses and learn a bit from un-sexy things like metal presses, diesel locomotives, and airplane cockpits.

In industry, UI has consequences.

Good point. High stakes make for deliberate interface design.

Donald Norman's The Design of Everyday Things [1] and Turn Signals Are the Facial Expressions of Automobiles [2] are must-read classics about human interfaces in both physical and "computer" realms.

[1] http://www.jnd.org/books.html#33 [2] http://www.jnd.org/books.html#34

Interface: A point where two systems, subjects, organizations, etc., meet and interact. -(First Google definition, and the one that I was taught)

I think the author is confused. He isn't talking about interfaces, he's talking about functionality. In his examples of mobile applications, interface is actually everything: the app is a point of overlap between the user's singular need and the tool that provides a targeted solution. (We can argue over the distinction of user interface vs. interface, but I posit the former follows the latter.)

In other words, the mobile app interface is not non-existent, but perfect: a single point of overlap between need and function.

From the article: That’s precisely what a tool is: something which requires very little explanation for how to use it, because it is designed so precisely for its purpose, that how to use it is obvious. If you’re trying to dig a hole with your hands, you don’t need much explanation for how to use a shovel. “This is the handle” is about the extent of it.

That's an example of a perfect interface, not no interface!

In some cases, mobile apps force you to put shit IN THE WAY of what the user is doing because of a lack of hardware interface for the user.

Game on an iPhone? Sorry, I'm going to need to stick buttons in the way of your visuals. So, you had a 3.5 inch display area, already small enough, and now you have to stick buttons and joysticks and touch menus on it. If not qwerty, at least throw me two or three buttons. Bah.

I think you are making the author's point for him. A game that requires a virtual joypad is not a game that is well designed for a phone. It's using a hammer to put in a screw. There are plenty of iPhone games that do not use the virtual buttons interface.

And the vast majority of those games are terrible. Partially because of a low budget, partially because of a low barrier to entry for the developers, and partially because touch screens are terrible interfaces for human beings to interact with where precision (positioning, timing, speed of repetition, among others) is concerned.

Some genres make sense for touch (RTS, Angry Birds), some don't. Accelerometers are flaky.

My point was that I believe phone designers are sacrificing substance for style. The iPhone has space for at least 4 more buttons without a slide-out.

Touch-only phones are a fad. Buttons will be back. Watch.

And? The whole point is, you can have totally new modes of control. They are crappy modes of control, but they are fun, and that's the point of games anyway.

It would be nice to have a joypad, so you can still play stuff that's following the old generation. Nintendo would have sold a joypad with it, because Nintendo makes everything backwards compatible with the last generation. You can play GameCube games on the Wii, and some of them are pretty good, having learned all the rules for the old interface.

I'm not sure buttons will be back. A touch-pad on the back, which displays your inputs on the front might replace them for most things. They'd need to be pressure sensitive, and have some kind of dynamic calibration (because pressure sensitive stuff is either oversensitive, or doesn't register anything), but it's not impossible (I think). It would be an interesting AI question - how to differentiate clicks, "swishes", and fat fingers; but that's the only real barrier I can think of.

Recently I bought a Sanza Fuze. Largely because I wanted an audio player, and not a locked-pocket-computer that happens to play music. What I learned from the experience is that touch sucks as a universal interface.

0) It is not always obvious what gestures to make to have the device perform a certain action. For something that I will pull out many times and interact with while doing something else, I should never have to think about how to do something.

1) In the cases where there are multiple gestures one can make at one screen, it is very often that you will make the wrong one or the machine will interpret your gesture incorrectly. This sucks. Having to memorize gestures brings us back to the era where most applications were navigated by keyboard shortcut. (In fact, I would prefer the keyboard shortcuts. At least those are non-ambiguous.)

2) Touch screens are quite possibly the best way invented by man to smudge display hardware. Even when a touchscreen has been forgone in favor of a touch pad below the display. Between greasy, sticky, food, dirty hands, and just plain wear, I basically have to constantly damage my display to interact with the device. Which is why most people end up getting a case. (Case-hell is its own subject really.)

3) Specifically for the case of video games, there is nothing more frustrating then the interface effecting your ability to play the game. Considering that the latency of wireless controllers is considered too much by the most hardcore of gamers, it stands to reason that touch interface will be universally looked down upon by anyone playing anything more complex than say; wii bowling.

Addressing your concept of a touch pad on the back to save the poor screen. Your forgetting that most people hold mobile devices in such a way that a touch pad on the back would be constantly triggered simply by holding the device with a firm grip. (And encouraging people to hold it with anything less is a bad idea considering how fragile they are.)

I expect buttons to be around for a long time to come.

The problem with interface-less mobile apps is that the interface provides critical cues to the user about functionality. Phrases like, "...a finicky piece of artifice that we have to strain to understand" is only indicative of a poorly designed interface, not all of them.

The beauty of technology is that its software transcends the physical realm. Software does things paper cannot. While specific physical metaphors often apply to software, especially with touch-based interfaces, to avoid taking advantage of the inherent non-physical nature of an application is a terrible mistake at the expense of true usability and functionality.

That’s precisely what a tool is: something which requires very little explanation for how to use it, because it is designed so precisely for its purpose, that how to use it is obvious.

Oh, that's what a tool is? :)

This is exactly why Siri is important. It is context aware and stays out of your way until you need it. I am very bullish on voice-driven apps of all kinds.

I think Siri is important in a variety of ways, but voice driven apps in the traditional model (a la Siri) aren't how I perceive the future of interfaces to be.

You wouldn't use Siri on a date, or in public would you? There are just some important issues that are raised by Siri that a lot of people sort of neglect. I don't think voice apps will be successful in a broad, paradigm shifting way. But, they will have their niche and there will be some fundamental princples that are learned.

Voice driven apps like Siri are one step away from being thought-driven apps, though. Full conscience interfacing is still a pipe dream, but word recognition from brain patterns already has proofs of concept.

I definitely agree, and I do think Siri and the like are very important steps to get there. I just don't think in the short term (relative short term) that these apps can gain enough traction in the wide scope that is championed by some of the proponents for voice apps.

You effectively just described what "good design" is -- so effortless, it's invisible.

This isn't a new concept, but with the recent focus on design in startups, I guess it's finally coming to light outside the design community.

When I write, the majority of my time is spent in diligent focus, refining what I call the "firehose of emotions" that was my first draft.

Thus -- a connecting thought from communication theory: "I didn't have time to write a short letter, so I've written a long one instead." - attributed to Blaise Pascal, Mark Twain.

Good communication is clear and simple: in conveying the message, it connects both the hearts and the brains of the author and the reader. This idea of simplicity isn't unique to hardware design, or software code, or written communication, or math ... but I don't believe that's the key takeaway. The point is that it's so difficult to understand and appreciate how much time and effort it takes, spent by talented and focused people, before the solution becomes simple and clean.

There is still an interface. Anything you interact with, by nature, has an interface. It's just become so intuitive you don't think about it, or need to learn it. It's just obvious.

Wall of text. Definitely No Interface >.<

No. Computer interfaces are increasingly virtual.

"The New Interface Is There Is No Interface" is an oxymoron.

Is it just me or did the author just write a whole article about the Metro UI without actually naming it?

http://windowsphone.interoperabilitybridges.com/media/42139/...

http://weblogs.asp.net/blogs/bsimser/MetroDesign_thumb_6D6BF...

It seems to be the only UI around that eschews window chrome and faux 3d and focuses on the content.

I posit that the old interface could be the new interface. For example, I'm willing to bet that most people spend most of their online time using complex interfaces to acquire tiny, yet important to them, bits of information. By "complex", I mean something like Facebook compared to the information retrieved, which could be something as simple as your best friend's status. So, imagine replacing Facebook in the browser with Windows Notepad plus an EXE made using AutoIt. The EXE would run in the background, regularly getting that important friend status, then "typing" it into the assigned Notepad window. Sure, you could argue that this does not replace Facebook in the browser, but it gives you a new way of utilizing an old interface, while also giving you features that the complex interface might not have, such as being able to easily save a history of your friend's status messages. I'm not saying that this should be an actual product, although it does have some potential. Rather, I'm pointing out that old interfaces can be repurposed.

Your interface has to be a metaphore for a real-world process. If it isn't then you are a) doing it wrong or b) there is no direct mapping to a real-world process.

Your interface has to be a metaphore for a real-world process.

Why?

Reply on news.ycombinator.com