Android is a platform of huge potential and flexibility, but it needs an urgent makeover for its third revision.
Put Humpty back together again
(Image by Google/CBS Interactive)
With my crusty old Nokia E65 rapidly failing and the Android vs. iPhone choice looming in front of me, CNET associate editor Joseph Hanlon kindly lent me a Nexus One.
A month in, I've got a fairly good idea of Android's capability. It's a platform of huge potential and flexibility, but this raw capability also helps to highlight its painful shortcomings, even running Froyo.
Text selection is appalling. The media player is too basic. The keyboard makes us cry. Too many apps need to be forced close. Vendors have created software to help bridge these shortcomings, but this has led to further fragmentation of the platform.
The separate approaches of Google and Apple are interesting. Apple's ridiculous level of control, strange regulations and amazingly closed system have severely limited what its platform can do, but has resulted in a much more polished, complete and integrated operating system. Most apps will actually work when you download them. It's both its biggest strength and weakness.
Google's openness and flexibility is equally its biggest strength and weakness: it allows considerably more capability than the iPhone, but to the detriment of platform stability and a more polished experience. We've lost count of the amount of apps that simply don't work and need to be force closed.
Over the horizon though is the tasty Gingerbread, version 3.0 of the Android platform, rumoured to bring a huge user interface (UI) overhaul. After long and detailed chats with Mr Hanlon, here's what we think Google needs to do to continue its meteoric smartphone rise, and put Apple back in its place.
Customisability is clearly Android's strength, with HTC, Sony, Samsung, Motorola and LG all offering different experiences on the one platform. While this should be celebrated for the innovation it brings, it can also introduce performance, compatibility, stability and design consistency issues that make the Android experience not as pleasurable as it could be.
Not surprisingly, this fragmentation mirrors the Linux world. We'd liken HTC's Sense UI efforts to that of Ubuntu, in trying to bridge the gap between what the original developers have offered and what the users actually want (or in some cases, don't realise they need until it's enforced, there's a few million forum posts about hating it, then months later everyone realises it's actually an improvement). It's done an excellent job, but it's time Google reclaimed vast slabs of the Android experience, and limited what vendors and telcos can do with its operating system.
I confess I'm a bit of a puritan when it comes to user experience. I want everything to feel like it fits within the one ecosystem. It's here that Apple excels — Android, not so much. An enforced UI takeover by Google, however, is likely to rub vendors the wrong way — a distinct look is a unique selling point.
As such we suggest a halfway approach. Apart from hardware support, don't let vendors touch core systems. Instead, include a robust skinning option that both allows the vendor to customise the experience in a limited fashion, but also allows the user to turn it all off.
Anything else should be enabled by the use of custom apps which can be uninstalled. Anything put on by the telco should also be able to be uninstalled. Once all this is isolated away from the core system, it may even open up the possibility of Google being able to release updates for the operating system across all devices simultaneously, rather than waiting for a vendor's custom update, or having to root your phone and do it manually with a stock ROM not intended for your device. We have no idea if it's possible to decouple the system to this degree, but if so, it would surely speed up updates greatly.
Over the air (OTA) updates will obviously still need to be certified by individual telcos, but at least some of the pain could be eased.
Swype is a keyboard replacement that allows you to type by "swiping" your thumb or finger from key to key, with the software predicting what you intended to type. While there are alternatives like ShapeWriter and SlideIT, Swype buries them for accuracy. There's a video of Swype in action here, but you really need to use it to understand how much it changes the way you use your phone.
At the moment though, Swype's business model involves selling its app to particular vendors like Samsung and Motorola and customising it for a specific device, rather than making it available on the Market.
We think Swype is the most powerful killer feature Google could have in its arsenal to compete against the iPhone. At the moment as far as we can tell Swype is privately owned, which makes any buy-out deal a little more difficult. Assuming an appealing dollar sum could be agreed upon though, the software should absolutely be integrated into Android as the default keyboard.
Browsing apps on your desktop, for instance, is incredibly limited. Despite the advances in the mobile space, browsing on the desktop is still a much better experience. For a search company, we find it highly disturbing that Google has no search option on the desktop version of the Market.
You can, using annoying drop downs, select categories in either free or paid apps. But you're limited to top-rated entries only — what if there's a golden app out there that's never exposed because it can't get into those top-rated lists? You have to wait for Google to put it in the "featured" list.
Within the market infrastructure there's also no way to download apps to your desktop and transfer them to your handset. Apple does this by allowing you to browse apps in iTunes, download them, then sync your phone to install. You can obviously download APKS (Google's installer package) from third-party sites, install a file browser app and do it yourself, but for official market apps, this is a hurdle that shouldn't be required.
The handset version of the Market is equally frustrating. You can search, but there's no way to separate listings into paid, free or ratings, and the only categories are "Games" and "Apps". When you have as many apps as the Market does, this becomes unmanageable.
There's also the handling of the installation of apps. When an app fails to install, often it just tells you that it failed to install, not why. If you have a trial app that's expired, the app doesn't tell you when you try to install, it tells you after it's installed and you've tried to run it for the first time. It's messy and frustrating.
Then we come to the numerous force close issues. Many apps only work on specific handsets or Android versions, but in many cases there's no way to tell until you've downloaded the app and tried to run it. Why isn't there an option to filter out the apps that will break on your handset? While we don't believe Google needs to take the Apple route of sanctioning absolutely everything that passes through, everything that hits the market should be compatibility tested to guarantee it will work across the various flavours of Android.
Obviously this is a wide reaching change, and ties into Google taking greater control of its platform as we discussed before. Those who don't want this level of certification should still be free to publish their apps outside the market. Inside though, we just want stuff that works. Every broken Android app serves to undermine the platform's potential brilliance.
While we're talking about certifying apps, may we recommend the ability to move an app to the SD card be a mandatory requirement for it to be certified as Android 3.0-friendly?
There are cases where Android's default apps are either too simple, or simply don't exist.
The lack of note-taking app is particularly surprising. I'm a fan of Note Everything Pro — like Swype, I think Google should absolutely buy this app and make it standard. It supports not only text input, but allows painting, takes voice notes, allows you to attach notes to photos, create checklists and more. Give it a look more in line with the Android aesthetic, and you're on to a winner.
While messaging functionality is there in Android, it's nowhere near as flexible as Handcent. Rather than buying Handcent though, we think a little more attention to detail in the messaging app is all Google needs. Better contact integration, quick reply, contact groups and the ability to turn off the ugly Android smilies would be a start.
Finally, there are little features that should be included. Given the proliferation of QR codes, we believe the Barcode Scanner app should be included by default. Contact Owner's functionality should absolutely be incorporated into Google's lock screens, and we'd love to see some sort of phone-tracking ability similar to what MobileMe offers, without relying on third parties. We don't know about you, but we'd pay Google for a service that would allow us to locate our phone if lost, lock it or trace it even if the SIM changes.
The Music app is pleasing in its simple layout, but doesn't look impressive and is depressing in its lack of integration with things like automatic album art downloading. It also can't be limited to a single music folder in order to filter out things like system sounds that might also be stored on your SD card. Alternative players like 3 (cubed) and doubleTwist, while looking flashy and solving the cover art problem, also have this issue. Astro Player manages to overcome this, but has an interface that only a mother could love.
Then there's video — by default, the closest thing to a video player is the YouTube app, once again leaving users to run to the app store for something like Act 1 Video Player or mVideoPlayer. VLC may come along soon too, but we wonder if it will have the prettiness to "wow" Apple users away.
Compared to the iPhone, Android's media playback is an also-ran. Maybe if the oft-rumoured Google music service comes into existence this will change, but for now the default offering is definitely not impressive.
RSS integration is woeful. The Power Control widget doesn't give you much power. Calendar is fine once you enter the app, but doesn't provide much information in widget form. Once again, the gap has been filled by the likes of SwitchPro, FeedR and Agenda Widget, but are we asking too much to have this functionality as a default?
While Sony's included a zoom and select function in its Xperia X10, every other Android phone we've seen requires you to either use a trackball, optical pointing device, or to touch the screen to place the cursor — all slow, and in the case of screen tapping, inaccurate methods when you've got a touchscreen.
Selecting and copying text across the system is also woefully inconsistent, whether it be in Google's own Gmail or Messaging apps, or in browsers. A universal system definitely needs to be in place.
Is there anything you'd like to see Google fix for Android 3.0? Is there a neat customisation a handset vendor has included that you think should be standard? Let us know in the comments below.