(This may seem off topic compared to this blog’s usual focus, but it’s relevant as a discussion of user interfaces, as well as one of next year’s bigger trends that will affect the games industry.)
Metro, to hear the folks from Redmond tell it, is the future. It’s in phones. It’s in the Xbox 360. Pretty soon it will be in PCs and tablets too. It’s a new marketing story about a new paradigm for a new decade.
But (for me at least) something is wrong: Metro leaves me oddly cold. I dislike the metaphor of tiles, and the way that it seems overly concerned with guiding people to content. It looks pretty and yet fiddly. It’s over-reliant on user memory and arcane actions. Its implementation on the Xbox 360 is downright bad.
More than that, it’s telling a different platform story to the one that Microsoft thinks it is, a story of a dumb PC and a more restricted future. At a fundamental level, that bothers me.
The Delight Business
Every great user interface does four things well:
- Tell a marketing story
- Work with user instinct
- Prove powerful in the long term
- Be delightful
The mouse, for example, satisfied all four. When it first came into being it was a revolution in computing away from green screens and keyboards (#1). It worked with user intuition because it was essentially the point of an index finger on a screen (#2). And it proved very powerful, even without scroll wheels or context-sensitive interaction (#3). A mouse really could do anything from activating software to steering a first person shooter (#4), and yet felt entirely natural.
There are other examples such as the joypad, the keyboard, the steering wheel or the stylus-free touch screen. Yet there are also less successful examples. The WiiMote satisfies #1, #2 and #4 but proves lacking for #3. Touch interactions on the Mac are not setting the world on fire because they require the user to learn a variety of pinch, twist and multi-finger gestures that don’t come naturally. So it has a problem with #2.
Microsoft tends to do #3 especially well. The Office Ribbon, for example, is genuinely better than the previous menus and toolbars of Office. Windows 7 is both elegant and robust. It’s quiet, understated even, and yet extremely powerful. So much so that the company has sold 500m licenses.
But when it comes to delight, the company frequently stumbles. Microsoft, as Steve Jobs famously remarked, has no taste. It has no artistic sense of good or bad because the culture of the company is overwhelmingly left-brained (or play-brained even), and tends to misunderstand the importance of aesthetics. It equates experience with guided behaviours and has produced a long list of ‘delightful’ ideas that condescend rather than uplift.
Internet Explorer Channels, Microsoft Bob, Windows Media Center, Windows Vista and successive Xbox dashboards are all examples of what happens when Microsoft decides to be delightful. In using Kinect to access movies on the Xbox, for instance, the interface is so busy being impressive that issues of performance are ignored. It is fancy but slow, slick but easily confounded, and the gestural version quickly induces fatigue. It’s much faster to just use a boring old remote.
The Grand Strategy
So perhaps the company has poor historical form in certain areas, but that should not unduly prejudice all future work. In recent years it has been trying very hard to shift its image away from corporate software to entertainment products, and with some success. The Xbox 360 is the most prominent example and is a powerful presence in console gaming, especially networked gaming. There are also other interesting initiatives like Surface, the recently announced Kinect for Windows and other visions of the future which certainly are attractive.
However it doesn’t always work out. Despite ploughing billions of dollars into online services from Hotmail to Bing, Microsoft has never really managed to convince anyone that its ecosystem is worth using. It has also been actively trying to reclaim all the lost ground on Internet Explorer, but Chrome will eventually win that fight. And, perhaps most crucially, at a time when PC sales are pretty flat, Apple is managing to eat that space little by little with iPhones, iPads and iMacs.
A grand strategy was needed to reposition Microsoft at the heart of computing once more, a Windows 95 moment to define the new generation. Microsoft is one of the few companies that has pulled of such moves in the past (Internet Explorer, Xbox once again). When it decides that it’s really going to go for something, it tends to pull out all the stops and spend big in order to gain market traction. It plays to win, and its latest big play is best summed up in one word: unification.
The idea is that all devices can, and should, have the same interface. Touching a screen is not that different to pointing a mouse and clicking after all. But it needs to be more than that: The strategy needs to leap forward, to say it’s not just a unified set of icons and menus, but a departure. Something to get you, the consumer, excited.
That something is Metro.
Metro is based around personal dashboards, where rather than interacting with static icons, you interact with information-rich tiles. Web dashboards that do something similar have existed for years, where each panel acts as a window into a site or application. Metro uses this concept at the operating system level. Each tile on the dashboard is a widget, capable of displaying top-line information that updates in real time. Your email, calendar, photos from Facebook, music player and more can all be overviewed, interacted with and accessed at a glance.
Dashboards are also paginated. Rather than drill down with icons and folders and a hierarchical Start menu, you flip left or right (as in smartphone operating systems) to see new boards. You can zoom out from them, group them and assign names, and with other gestures such as holding or dragging you can activate settings, movement options and so forth.
The results are very swish. Metro is great in presentations, such as at the recent CES. On stage it glides, looks dynamic and smooth and is all about delight. The typography is beautiful, the background wallpapers are slick and the animations smooth. It’s very un-Microsoft (in the old sense) and many in the tech press love its vision and charm.
It looks very promising. So what’s the problem?
Doing Four Things Badly
My problems with Metro are:
1. It’s Too Busy
In any user interface there is a tension between pulled feedback and pushed feedback. Pulled feedback is that which the user goes in search of, such as opening a hyperlink or making a phone call. Pushed feedback, on other hand, is when the system alerts the user.
Users tend to prefer pulled feedback to pushed feedback because they want to concentrate on what they are doing. However this is not always the case. Fetching email is more annoying that being made aware of it. Activating a clock app to find out the time is considerably more cumbersome that simply having a system clock present in your field of view. Receiving a phone call is a vital feature without which phones simply wouldn’t work.
The problem with pushed feedback is that it can quickly overwhelm, creating an interface that looks busy but is actually hard to read. If, for example, you have too many apps on the iPhone enabled for full notifications then the phone becomes a blizzard of alerts and badges. If you have too many social plugins installed in Chrome, likewise. Notifications demand to be read, which leads to distraction or the need to curate the system rather than actually using it.
I think Metro’s tiles fall on the wrong side of this issue. While iOS and Android have notification systems of their own, they also use more conventional icons for general navigation. And notifications can be silenced. Whereas with Metro, the whole idea of tiles is that they are permanently updating.
For example, I have the new Xbox dashboard. Every time I launch the machine and log in, I am greeted with sets of bright, moving (frequently advertising-based) panels that seek to push content at me, and the net effect is to impart a very unsure feeling about its organisation. Everything is on the move with new photos, scrolling content and notifiers asking for attention.
It’s bothersome. I find that I have to piece through the system looking for ‘safe’ tiles to activate, and that the imposition of several tabs between the front page and games is just downright strange. There are some shortcut elements such as a list of your ten most recent used items, but the list does not work for everything. After the initial novelty wears off, that bothersome feeling becomes irritating. I feel the Xbox is permanently selling me something.
Windows 7 Phone is similar. There’s too much going on, too much distraction, too much change. PC demonstrations of Windows 8 likewise. On stage it looks awesome to have a dashboard of things updating, but in reality using a desktop that has too many widgets and notifiers is not a positive experience. It becomes an imposition, regardless of how personalised you make it, because the interruptions are just too frequent.
2. Task Switching
This may sound like the exact opposite of the previous point.
I love distraction-free writing applications. I use Scrivener for writing chapters of my book, for example, and OmmWriter for working on my Edge column. and I wish that some enterprising developer would compile a version of the WriteSpace plugin for Windows Live Writer. I also prefer full-screen games to windowed ones, again because it’s easier to get absorbed when nothing else is asking for attention.
However most software I use windowed. As I write this, I have a browser open in the background, a Skype conversation open, iTunes syncing my iPhone, the Sticky Notes application and so on on call. I can, and do, flit between all of them. Most of the time, this is why I would still consider a PC as a ‘work’ machine compared to a tablet as a ‘play’ machine.
For work purposes, an easy task switcher is essential. A user needs to be able to navigate among a number of applications without barriers, and maintain an ambient awareness of all of them. This has been the case for decades. When I used to work in technical support, for example, I would have Outlook, a browser, a CRM database and other applications open all of the time. For many this is what work computers need to be able to do.
Tablets and smartphones lack that sort of flexibility. iOS has a task switcher of sorts which relies on you double-tapping the Home button but it’s slow, and reliant on memory and attention span. You have to remember which applications you have open and how many steps or swipes away they are, and often you just don’t. On my iPhone I occasionally open up the task switcher to clean it out, and discover dozens of apps that were opened at some point in the past and which I had completely forgotten about.
Whereas on my current desktop I can tell at a glance what’s open and what’s not. I use Windows 7 for the most part, and the key innovation of combining Quick Launch with the Start Bar works very well for me. Since I use a widescreen monitor, I have 25 app icons all just one click away. I can see which are open, which are downloading, which need my attention and so forth.
Windows 8 is ditching that in favour of something more manual. Using gestures, you cycle through applications in order to find the one you want (a little like Windows-Tab) and change. You also see every application in full screen (except for backwards-compatible apps, which work within a legacy ‘Desktop’ mode) like a smartphone, and every application is held in an opened state. So you can backup to the dashboard, activate another and so on, but again as smartphones show this can get pretty tedious.
Microsoft has tried to solve this issue with the neat idea of side-by-side applications. You can have a browser open and pull another app in from the side to occupy about a third of the screen, in a column format. Apps need to be developed to take advantage of this, but the idea is still interesting. It only really works for two applications though, and involves a lot of swiping to find the correct app to dock alongside.
3. Arcane Actions
A mistake that Apple commonly makes is that of binding features to hidden actions. Multi-touch, double- or triple-taps of the Home button, voice, non-obvious pinches and squeezes and so on require the user to realise that they are there, and learn the habit of using them. In the context of games I have labelled these kinds of actions ‘arcane’ previously, and it applies here too.
Natural actions have a tangible quality, such as touching a icon to open it or clicking a button, but arcane actions require learning of artificial concepts. Depending on user expertise they may become natural in time (a good example of this is keyboard shortcuts) but equally many users simply never learn them. To this day, the number of users who don’t really get the right-mouse-button is high, and the number of automatic drivers who can’t grok manual gear boxes equally so.
Arcane actions are sometimes unavoidable. In the days of keyboard computing you really had to learn various shortcuts in order to be able to cut-and-paste. You had to learn how to type. However many arcane actions are a result of trying to push features around certain design principles rather than admitting that the design principles don’t really work.
Windows 8 in particular seems to be loaded with them. You have to learn gestures to side-by-side apps, navigate, browse the web and also invoke the Start Menu. You have to get the concept of ‘charms’, hidden preferences that require a special action to invoke. Even in moving applications around, there’s a kind of drag-then-pull action that seems tricky.
The system requires to user to remember more than they should have to, and this means they won’t. They won’t know how to use their PCs effectively, even less than today, and so they will feel even stupider than they already do.
4. Pagination vs Hierarchy
When I worked in interactive television games (pay to play casual games on your set top box), we used a paginated system the allow users to navigate games. Each page would show nine games in a 3x3 grid, with a video previewer for each on the right hand side.
A common problem that we encountered was just how many pages we could support. Although the system could handle up to eight pages, any more than three and users would get lost. They would forget where content had been and have to search back and forth to find it again.
I notice something similar with my iPhone. No matter how many ways I reorganise apps, I can only ever really remember the content of the first three pages. The fourth page is hazy and the fifth page I forget entirely. I forget that I have those apps installed, forget many of their names (so searching for them is also a pain) and so my use of the device is diminished. My Xbox 360 is the same, only worse because the content of each page changes.
While less fashionable, hierarchical systems are better in the long term if power use is the goal. Your brain is more easily able to work with categories it can see rather than pages it must remember, and establishes habitual pathways for use. Over time the number of things that you can use grows exponentially and so your computer turns into this magical device that has hundreds of uses, from documents to games.
The best example of the power of hierarchy is the World Wide Web. It is a hierarchy of search, domain, site section and page, and because each step is visible, each offers a large array of options. There are still issues (many people still don’t quite get URLs) but it definitively works to the extent that the internet is now central to global culture.
Habituation breeds confidence and the urge to explore further. Pagination plays on faulty memory and makes us feel limited, even foolish. It’s why Windows 95 through to 7 (with the possible exception of Vista) has been so successful. It has a system that can support thousands of options hierarchically (desktop, drive, folder, file and start, category, application), which combine with its task switching for near-infinite uses. It has some issues, but on the whole it works.
There is no reason to suspect that Metro pagination won’t have the same issues as other similar systems. It will make the computer feel small, limited and the user will feel forgetful. Just like my interactive TV game players, they will be easily lost because they won’t take the time to build a mental picture of the operating system in their heads (nor should they have to). So it will eventually lead to dumber computing.
Away from user interface issues, I have a significant concern over the role of Windows 8 in locking computers down. Signed operating systems, Metro compliance for applications and more are unsettling ideas in the Windows culture.
Windows is the biggest operating system in the world largely because of partnerships with vendors and its openness. It is the freest system for developers to make interesting software, and while that has some negative consequences (viruses, say) it also has some tremendous boons. It may appear old school in a world of swishy tablets and apps, but there are far more developers working in and with Windows than anywhere else. Hackers, IT departments, software providers and a multitude of game developers know and rely on this ecosystem.
In a recent talk, Cory Doctorow outlined the idea that various forces are trying to put an end to generalised computing and replace it with appliances. His pitch is that it is in the nature of these entities (such as content companies) to restrict the things that scare them, as a form of digital protectionism. And so the real frontier of computing is not over copyright or copyleft. It’s over smart or dumb. Powerful or protected. Emergent or experient. Network over broadcast.
It’s not the story that Microsoft wants to tell, yet it’s hard not to conclude that the story that Metro is actually telling is that the PC will get dumber. Sure, it will feel like Minority Report for a couple of days, but when you actually sit down to use Visual Studio and Photoshop and Chrome and Spotify and Skype all concurrently, then it will feel sticky. Like an app platform it will be more of a one-at-a-time consumer experience rather than a powerful engine.
The thing about grand solutions is that they often try to solve a problem that nobody has, and this is pretty much where Microsoft finds itself. Its problem is that it doesn’t really know what its platform story is supposed to be any more, so it’s trying to invent one. Metro is about solving that problem for Microsoft, about Microsoft feeling relevant once more. What does it solve for users? Nothing.
Sometimes being stable and ordinary is ok. The Windows 7 interface is really good because it’s stable and ordinary. The iPhone’s use of icons (and avoiding of Android widgets) is really good because it’s stable and ordinary.
Forcing delight on a user at expense of usability, especially in PC, is not a problem that anyone actually wants solved. What they need is computers that compute, consoles that make console gaming fun and mobile devices that work for mobile. And not to be made feel like sheep while doing so.
I think the market is noticing. Windows 7 Phone has gone down like a lead balloon. Various theories abound as to why, mostly to do with market timing, but the question not being asked is whether Metro itself is the problem. The Xbox 360 dashboard has updated to turn the machine into a social media powerhouse, and yet many players are ignoring that side of it because they don’t really care about how social their console experience really is. They want it to play games and videos, and for the video options to just work.
Does that mean that many Windows users will not want to jump to Windows 8? I think they will pause, and of those that do they will want to use the Desktop mode more than any other. Microsoft has convinced itself that it must be right because TechCrunch is impressed with Metro's swish, but I suspect users will see it very differently.