Services, Design, and UI - Addressing the need for interactive metaphors

With WWDC now past and opinions of the iOS 7 redesign have been aired, there is an opportunity to think about user interfaces and the future of the use of metaphors in computing. Regardless of your opinion on iOS 7, its release into the wild has opened up serious questions about what constitutes good user interface design with hardware. John Gruber has argued that hardware and software should not be separated, meaning the entire device should be treated as an interactive platform. The difference between hardware and software does not occur at the level of interaction, but at the level of limit to the ability of each to do its respective task . Nilay Patel has argued that Apple has reached a certain limit in terms of what constitutes good design. Apple's signature innovation was to blend software and hardware in such a way that it created a service that saved the user time and effort in computing whatever (file x, word doc this, photo or video that) It appears, however, that Apple does not want or does not acknowledge that more can be done in this space in terms of software integration and the corresponding user interface design and Finally, in different ways, Vlad Savov and Joshua Topolsky have argued iOS 7 has not taken (full) advantage of the redesign opportunity in various ways, leading to a confusing and perhaps "childish" (or in my words in different forum posts "nintendo-y") looking UI and For Josh, the design looks like a mixed bag, while Vlad takes the mixed bag to a whole new level contending it is the end for Apple design.

For the moment I want to set aside critiques of iOS 7 and focus on the purpose of a UI. All four authors, and many designers, have different criteria for what constitutes a good UI. For John, it could be (admittedly reductively) summed up as simplicity. Let the user get to content and create content easily. For Nilay, it is about service. Allow the user freedom to find a service that best matches how they work, provide the service the best way possible via the computing medium. For Josh (and perhaps undeservingly I'm placing Vlad here too) I think it is about beauty that communicates a sense pragmatism or purpose.

I'm not for sure these are the best criteria to judge good design, albeit important and deserving criteria. The problem is that they forget the historical purpose of the UI as metaphor. The UI was invented as a means of covering over the 1s and 0s and code of computers so individuals did not need a computer science degree to use the machine as an archive, creative tool, or communications medium. Indeed, the computer is a nexus for all these services. As these services have changed, evolved, and added functionality, UI's have only petered along. Why? Again, UI's function as metaphors. Insofar as the metaphor works, people get work done. They don't care about the metaphor's quality until it impacts their work output. Maintaining consistency allows for a sense of long term security in terms of being able to put out an amount of work. It also provides fairly consistent development of features so no one has to (re)learn MS Word or some new version of a program.

And therein lies the problem. The metaphor can be stretched only so far (in terms of time and representational quality) to accommodate different features into the same interface. Enter skeumorphism as one possible solution to better acquint the user. Skeumorphism is a design language barrowed from the real world to transport the user into the virtual world by means of real world metaphors for functions and services of a device. But skeumorphism as a design philosophy runs out of time as computing becomes engrained in our lives. We expect software to not represent the world around us because we know software is part of that world now. One could say it has penetrated the barrier between being an invading "newness" to the world and being a part of the world. The computer is now definitely a part of the world. Skeumorphism has run its course, to a degree. That is, in the case of Apple's skeumorphism, services of the computer can be intergretated so tightly that real world metaphors fall apart because the lines in the real world that keep things distinct are no longer there. Apple skeumorphism fails because it is now a catachresis and not a metaphor.

So this raises the rhetorical question: What is the use of the metaphor in today's UIs? Microsoft has all but abandoned the traditional set of icon metaphors for live tiles. Google maintains the old style metaphors, to a degree, and adapts them to different devices in Android and Chrome OS. Google limits those metaphors to represent the medium to acquiring information or creating information from a source (i.e. the web with Chrome, or mobily with Android). This is decent skeumorphism (i.e. Google Now, iOS Google Maps, the Holo theme), but goes too far putting a computer desktop in the pocket. Apple doesn't know where to go. I think this is because of their customer satisfaction rating. If I were Tim Cook, why change a good thing? Then again, why listen to the Wall Street Journal and Shaw Wu when redesigning iOS?

I think Nilay (in the Vergecast link) hits a good reason to change a UI: services. And that must mean the metaphor and representation much change as well - that is a good reason for not liking iOS is that it did not change the button layout to increase or decrease access to content or its creation (i.e. iTunes app where its basically the same player just without the previous black design). Services, if I understand Nilay correctly, are the nexus of software and hardware to create, communicate, archive, and consume content. Services, should be represented simply, to barrow John's assessment. And to borrow Josh and Vlad's thoughts, simplicity should be beautiful if it displays content in such a way work can be maximized and doesn't trade off too much with the content being worked on itself. It stands to reason a good metaphor for a mobile device (phone, tablet) and computer (laptop, desktop) could be different because of screen size and function of the hardware to be able to process software. These devices deliver different services in terms of scope and capability. So a good metaphor will draw the lines of the limits of the hardware in the software and not needlessly extend itself as the software and hardware capabilities shift, integrate, add, and change over time. This will also be conditioned by market forces, or in other words, designers need to think about the limit of the consumer audience to adapt or completely shift from one metaphor paradigm to another (which Microsoft has not done well with Windows 8 by making such an abrupt shift). I do think Jony Ive knows this as Gruber rightly pointed out.

In short, UI design is about metaphors that provide a service clearly and simply. But they also have limits. And with gesture based interfaces, natural user interfaces, and integrated services where functionality is unified into a single app, the metaphors need to change to make the use simple and easy as possible. Despite people being able to imagine better functionality, no one has so far made better metaphors for the overall computing environment than Apple - see their customer satisfaction rating. That doesn't mean Apple, Google, Microsoft or anyone else shouldn't take some new risks to invent new metaphors or begin transitioning to new metaphor paradigms for UI design as services converge with new hard innovations like gestures. Apple is taking some of this risks now, and should be applauded, even if it is a half step forward compared to the full step we would like to see. Metaphors are needed, but they need to be better because we can now imagine doing more work, more easily, because we know the potential for computing in our everyday lives. So we need to strike a better balance between consumer user, market demand, and functional representation of services. The alternative is to discard the metaphor altogether, or so much, that it hurts core business or the consumer's ability to adapt and read the metaphor (i.e. Windows 8 transition).