Pixel Envy

Written by Nick Heer.

Rosy Glasses

There’s an article from Amy Hoy that has been making the rounds in a big way. The premise:

No argument here: Jony Ive has produced some of the best industrial design in the history of consumer products. He’s done it by cutting out all the extraneous parts. By eliminating edges, by smoothing and streamlining.

But what works beautifully for hardware does not work for software.

Hoy is unhappy with the state of Apple’s software design, particularly on iOS, listing her qualms so:

  • misusing metaphors (e.g. turning buttons into links)

  • eliminating the only affordances that software can have — visual affordances

  • using fake physical metaphors for interactions, such as using “wheels” for data entry

  • eliminating information hierarchy – homogenizing spacing and typography, for “visual tidiness”

  • giving all types of interface widgets the same visual appearance

  • reusing the same interaction design for click UIs (on 13″-27″ screens) and touch UIs (on 5″ screens)

  • tiny tap or click targets with invisible boundaries

  • software and icons that all look the same

Most of these are very reasonable complaints about specific aspects of iOS’ user interface design. I think the lack of visible tap or click targets around buttons can be very confusing, and the mis-placement of emphasis on labels instead of interactive elements makes for discordance. But Hoy, like so many others, makes the mistake of contrasting Apple’s current approach to UI design with their historical approach. I find that this greatly undermines her argument. Consider:

Apple used to publish its famous Human Interface Guidelines for designers inside Apple and out. Every guideline was user-focused. It covered everything from how to lay out a screen, to how to write an error message, to the appearance and placement of buttons.

The HIG was a powerful force for software quality in the Mac world. It was a major reason why the Mac’s shareware business was so strong, the quality so high.

The HIG wasn’t about aesthetics, it was about interaction.

It was based on research, not trends.

That Apple is gone, now.

Here’s the thing: the Apple that Hoy describes hasn’t existed for a very, very long time. Beginning around the introduction of OS X, Apple stopped following the HIG so strictly and started experimenting. Remember the brushed metal window texture? John Gruber in 2004:

The big problem, obviously, is that Apple has simply ignored the HIG. The HIG states, “Don’t use the brushed metal look indiscriminately”, but indiscriminate is precisely the word to describe Apple’s use of it.

How about the dock in Leopard? Craig Hockenberry:

The floor displayed on the Dock does not use the perspective of the desk in front of you, nor does it appear as a shelf [as the HIG states icons should]. Because there’s a difference between the floor angles and the traditional desktop icon angles, many icons look wrong. […]

Hundreds of designers have been producing icons for tens of thousands of applications by following the Human Interface Guidelines. Changes to the Dock should respect these guidelines since changing existing artwork is not an option on such a large scale.

iTunes 5, in one of my favourite Gruber pieces, “The iTunes 5 Announcement From the Perspective of an Anthropomorphized Brushed Metal User Interface Theme”:

BRUSHED METAL: I’m the bad-ass theme. I’m the one who flouts the Human Interface Guidelines.

MIKE: This guy trashes the HIG the way Johnny Depp trashes a hotel room. He even sports a custom radius on his window corners. No other window on the system has a shape like this. It’s wild. Just wait until the HIG zealots get a load of this guy.

I include these examples not to say that Hoy is wrong in her criticisms of recent iOS interface design, but to point out that the HIG was something that was frequently about aesthetics and trends. iTunes 5’s UI wasn’t a smooth grey because of research and interaction; it was because brushed metal stopped being trendy and started to look kind of, uh, bad. “Unified” windows in Tiger did not have a consistent use-case defined by the HIG, either.

But one of Hoy’s complaints really stood out to me: “software and icons that all look the same”. The irony here is that the HIG was created to unify the appearance of software, generally speaking. Michael Tsai in 2006, responding to John Gruber speaking at C4 proclaiming that “the HIG is dead”:

Violations in appearance are the easiest to see and get the most play. They bother me because I like looking at a consistent interface and because I think many of the custom controls are ugly. But it’s not that big a deal because, as the Web has shown, people know how to recognize weird-looking buttons.

Why was brushed metal bothersome upon its introduction? It wasn’t necessarily because people didn’t like the way it looked, but because it was inconsistent. Gruber, again, from that brushed metal article earlier:

A large part of the Mac’s historical usability advantage is that Mac applications all look and feel the same. Not exactly the same, of course, but certainly within the bounds of a single “theme”.

Even this isn’t necessarily correct; back in 2004, Apple certainly had a lot of “themes”.1

The problems with iOS’ UI aren’t because it doesn’t follow the HIG. It rewrote the HIG, establishing even greater consistency between apps. I do not think iOS’ interface design is perfect, but nor can its faults be ascribed to its newness. Even comparing iOS’ current design theory to that which it replaced isn’t a fault of Hoy’s — that’s the expectation.

But Hoy is looking at Apple’s interface design through the rose-tinted hues of history. There are so many trends that Apple either pioneered or followed, and plenty of them that are questionable. Many of the ones I’ve described are aesthetic decisions that fly in the face of the HIG, but Hoy says this isn’t an argument about the visuals:

It’s wrong based on 40+ years of computer-human interaction research. It’s wrong based on 30+ years of Apple HIG.

I agree: draw a border around buttons so we know they’re tappable. Make sure emphasis and hierarchy are respected. But, while the general principles described within older HIGs (PDF) are alive and well, the HIG has always been adapted through its history to meet the evolving needs and expectations of users. It’s not perfect yet, and it’s going to take some time to adapt, but the HIG isn’t dead: it has reborn, just a little differently.


  1. I’m referencing John Gruber a lot because he was one of the few people who wrote a lot about the HIG and violations/modifications thereof in the earlier days of OS X. ↩︎