Is Windows 8 a bad OS for tablets? It's time to decide...

In a piece last week on the practical portability of the Surface Pro, I rather glibly mentioned that "Windows 8 is not a good tablet OS." Several people called me on this. Well, everyone called me on it in the comments, on Twitter, and over email. In hindsight, it was perhaps too broad a comment to make.

What I wanted to do with this article was have a look at how the tablet-optimised parts of the Windows 8 and Windows RT operating system--separate to any hardware--compared to the market leader, iOS. I wasn't concerned about the behaviour of Old Windows features on a tablet. What I focused on in my investigation was how native apps (Windows Store/Metro-style apps on Windows), the launcher (Start screen on Windows), and basic system configuration bits (setting up a Wi-Fi network, account configuration, etc) behaved across the different OSes.

The details

Usability

Let's look at usability first. I've broken this down into several bullets.

Basic touch

A key part of the Windows 8 reimaging was making touch a first-class input method. Old versions of Windows were touch enabled, but this only worked as an approximation of mouse input. By creating the new Metro-style UI, Windows Store apps, and the new Start screen, Windows 8/Windows RT implements basic touch operations in an optimum way, and it does this in easily as capable a way as iOS. (As mentioned, I'm ignoring touch when using Word and Old Windows desktop apps as there's no fair comparison and frankly this is always going to be a dirty hack.)

Equal scores here for Windows 8, Windows RT, and iOS.

Gestures

On paper, iOS' gestures are as complicated and non-intuitive as Windows 8/Windows RT's. The general problem with gestures is that you have to be shown them to know them. On iOS you generally don't need to know gestures to use the system, whereas on Windows you do. A user cannot find their way around without knowing about charms and app bars. For this, I'm going to have to ding Windows--although iOS loses on consistency (each iOS app can put the search function somewhere unique, but in Windows Store apps it has to be on the charms), Windows requires additional cognitive load and explicit training.

Winner: iOS.

Joining a network

On this point, both are equally difficult. iOS bafflingly buries this common system function in the settings app. Windows requires that you know that you need to use the charms bar in order to set up a Wi-Fi network.

Equal score here.

Launcher

People are starting to talk about how iOS' launcher is outmoded, but it is demonstrably very simple. The Start screen on Windows seems more complex, but I'd argue it's more usable than iOS as it allows the user to set up things how they want. It also allows developers to find new ways to push data front and centre to the user.

Winner: Windows.

Built-in apps

The two most important built-in apps on a tablet are the mail app and the browser. On Windows, the built-in Mail app is half-baked and essentially useless, especially in business. iOS' does a decent enough job. In terms of the browsers, I'm not a fan of IE, but for most tablet-based browsing activities both Safari and IE are just about good enough.

Winner: iOS, because Windows 8/Windows RT's mail app is so lacklustre.

Performance

iOS' raw performance is much, much better than Windows'. On the one hand, Apple has had longer to optimise it. On the other, Apple has always pushed developers to build apps with a user experience that is as good as possible and their toolset reflects this. (The toolset is hard to use, but gives good, smooth results.) Microsoft's philosophy has been to give developers tools that provide for rapid development.

Winner: iOS.

Security

Next, we'll look at security.

Exploitability

We know that both iOS and Windows RT can be jailbroken. In that regard, those two platforms score equally badly as it's the same problem. (Jailbreaking is about getting the OS to do something the original designers didn't want it to do. They rely on exploits, which could be used to do achieve something you as the device owner don't want to happen.) On Windows 8, this story remains a "PC story" and not a "post-PC story." In a post-PC world, the level of exploitability in a non-hardened OS like Windows 8 is not acceptable.

We have a split, with iOS and Windows RT doing equally well, but Windows 8 scoring about a billion minus points.

Hygiene

I'm using this term to mean how well separated apps actually are, both from the perspective of whether two apps installed together cause localised or system instability, and how well application data is protected on a per-app basis. Both systems are hardened, and native app APIs are protected to prevent intentional and accidental data loss and/or damage.

Again iOS and Windows RT score equally well, but from a post-PC perspective iOS and Windows RT score the same, Windows 8 minus a billion again, mainly because of stuff like this.

Updates

The ability for a platform owner to "force" patches down to a post-PC device is important, as this is the only practical way to smooth over problems in the security surface area. iOS has the edge here because its updates are simpler and atomic. Windows RT and Windows 8 have the battle-tested Windows Update system, which is fiddly but effective.

Both score equally pragmatically speaking--but I'd prefer to see Windows' update system more like iOS'.

Utility

Finally, "utility." This is a look at features that the platform vendor can badly implement so that any resulting device is less compelling.

Portability

My prejudgement on this one was that Windows 8/Windows RT would be less than usable without a physical keyboard and this would hurt portability. But, just like you can still use Windows without a mouse (Alt+Space, anyone?), you can use Windows 8 and Windows RT equally well without a keyboard. The virtual keyboard on Windows 8/Windows RT is just as good as the one on iOS.

Practically, portability scores the same. In reality, Microsoft's insistence on positioning Windows 8 and particularly Windows RT as needing a keyboard from a marketing perspective is unhelpful.

Battery

When talking about the battery, hardware is only half of the story. One of the major engineering tasks at Microsoft was limiting the OS' use of the battery, and limiting the ability of apps to abuse the battery. Windows 8--like the security points above--has no real protection against legacy Windows apps merrily draining the battery.

iOS and Windows RT have similar protections and score equally well.

Connectivity

This is relatively easy--iOS and Windows 8/Windows RT are equally good at connecting through to wireless networks and (when required) onto VPN. iOS and Windows RT score equally well and better than Windows 8, however, because of inconsistent implementation of "connected standby." (Surface RT has it, Surface Pro does not, for example.) This is an operating system mode that puts the device in a state where it's using very little battery, but is still network aware.

iOS and Windows RT again score equally, but Windows 8 is dinged because of that inconsistent implementation of connected standby.

Conclusion

I went into this article inclined to believe that Windows 8 and Windows RT--together or separate, I've frankly no idea how to split or combine the two of them--were bad operating systems for a post-PC tablet device. What I've discovered doing this exercise is that, objectively, they are not. Well, Windows RT isn't a bad OS for a tablet. Windows 8 has some...issues.

iOS and Windows RT score more or less the same throughout. Windows 8 gets badly wounded by the fact that under all of it it's still the same eminently malleable but endlessly abusable operating system that it always has been. Post-PC devices need to be closed and controlled, because only in that mode can they be trusted, and only in that mode are they useful to the majority of people who are not technologists. Which is, as we know, most of humanity.

I'm surprised that iOS and Windows RT are so close in their objective scores because--subjectively--my iPad is a joy to use, and I use it daily, but my Surface RT device I use only when I absolutely have to for work. Despite the objective measures, Windows RT is too unfinished, the reliance on Old Windows (desktop, Office, half of Control Panel) is too prominent, the Metro design aesthetic and UX metaphors too rudimentary, and the performance too lacklustre to be a truly great post-PC tablet device.

But I have to qualify that by adding "at the moment." Last week my ZDNet colleague Ed Bott and I debated whether "Windows RT was worthless, or the future of Windows." I was in the "worthless" camp. I guess now I can modify that to: "it is worthless, today." An objective assessment of iOS and Windows RT demonstrates that Windows RT could get there. It's just unfinished.

Of course, Intel's machinations to drive capability of x86 so that it negates the entire ARM proposition is a problem here. Windows RT is a good tablet OS, objectively. Windows 8 is not (hopefully objectively) measured on the fact that the security (whilst improving) is relatively hopeless on PCs compared to any post-PC tablet or smartphone you care to name. Intel managing to get its fingers in and keep Windows firmly in the "PC Plus" rather than "post-PC" camp where it could go with a more refined Windows-on-ARM implementation is disastrous to Windows actually being a good tablet OS.

But, let's say that Windows RT stays the course. Microsoft is famous on relying on time to help it conquer markets, building an organisational structure that collects instrumentation and feedback to help it make v2, and then v3 better than before and better than its competitors. (You don't think that happens by chance, right?)

The question remains: at the pace the post-PC market is developing, does Microsoft have time to make Windows RT great? Or will iPad, Android tablets, Kindle Fire and--dare I say it--the next-generation PlayBook (if there is such a thing), render it worthless after all?

What do you think? Post a comment, or talk to me on Twitter: @mbrit.