Via Gizmodo UK, some data on Android fragmentation from Open Signal Maps. The post presents a lot more data on API version, screen resolution, and other interesting metrics. It was based on 681,900 devices that downloaded the client software — not an unreasonable survey size.
It looks pretty bad (and I’m sure that’s the angle most reporting of this story will take, particularly from Apple bloggers), but there’s an upside, OSM says:
Developers tend to bemoan Android fragmentation yet there’s much here to be celebrated.
We’ve collected signal data from 195 countries - the variety of Android devices and manufacturers has been crucial in allowing the OS to reach so many markets. For example the 5 countries where OSM gets most use are: US, Brazil, China, Russia, Mexico. From what we’re seeing the developing world is no longer developing but leading Europe.
While the number of different models running Android will continue to increase we’ve seen Samsung take the lion’s share of the Android market, most of that due to the Galaxy product line. Testing on the most popular Samsung & HTC devices will get you a long way.
In particular, I think this part is noteworthy:
With many devices under $100 unsubsidized, Android phones and tablets are able to reach a market that can’t afford netbooks. For the majority of the world’s population smartphones (and not computers) will be the must-have devices. We hope that OpenSignalMaps will be a must-have app.
I think that price point — and the advantages it gives OEMs in the developing world — is under-discussed. It’s impossible to get a contract-free iPhone in the UK for anything less than £319 ($510), and that’s the old-as-the-hills 3GS. You can buy three or four Android handsets of similar spec for the same money.
Go to the original post for more interesting data from this study.
MG Siegler: "Why I Hate Android"
MG Siegler hates Android because, he claims, Google kowtowed to the carriers. He writes that the original plan for the Nexus One back in 2010 was to sell it as an unlocked Android handset for $99, but the US carriers outright refused to support it. Google’s hand were tied so in the end it went on sale for $529 and no-one bought it. Later, Google started to support things like Verizon’s attack on net neutrality.
I see two problems with this theory.
Firstly, the Nexus One cost about $175 to make at the time, plus the significant R&D costs necessary to have designed the thing in the first place. Teardown estimates like this are far from infallible, but I think that’s enough evidence to indicate that Google would be losing a lot if it had sold Nexus Ones for $99. Even if development costs could be amortized down to $25 per sale, Google would be left trying to recoup $100 back per user just to break even. That’s a tall order for mobile ad display.
Secondly, Google’s strategy was only flawed in the US. Here in the UK, we enjoy a wide variety of decent pay-as-you-go options from our four cell networks and half-dozen or so MVNOs. This is the perfect environment for Google’s plan, as Siegler outlines it, but the Nexus One launched £330 ($510) here. Similarly attractive no-contract mobile plans are available all over Europe and in many other countries worldwide but Google didn’t even make the phone directly available anywhere else.
So if Siegler is right, and Google had grandiose plans to do an end-run around the carriers, why didn’t it simply go abroad? It could have sold the Nexus One cheap in any number of countries, then pointed to those deals to put pressure on the US carriers to open up a bit via some simple consumer activism. The fact it didn’t makes me suspect Siegler’s reasoning is incorrect.
Source: parislemon
Extremely impressive demo of Kal-El, which is nVidia’s next generation quad-core CPU and 12-core GPU aimed at tablets (the demo is running on prototype hardware and Android Honeycomb). The use of dynamic lighting really does show off how much grunt this chipset has.
I wouldn’t be surprised if some of the less perceptive Apple bloggers react to this with a haughty sniff of “Apple don’t need to engage in spec wars; specs don’t sell tablets, great software does!” These would be the same Apple bloggers who wrote breathless posts about how the iPad 2’s CPU and GPU were 2x and 9x faster than the first generation hardware, of course.
Personally, I think it’s fairer to say that higher specs don’t sell tablets until they enable compelling new applications, and I think it’s also fair to say that’s exactly what nVidia are demoing here: the possibility for new, more immersive games. Plus, why does everything have to be phrased in aggressive “Android vs iOS” arguments? Can’t we just applaud a cool demo, even if it does come from the other team?
Two final notes: firstly, as Andy Ihnatko points out, raw number crunching is only one dimension — battery life and heat generation are important too.) Secondly, Apple’s A5 and nVidia’s Tegra and Kal-El chips are all derived from the same ARM processor DNA. It’s not a stretch to imagine that Apple have something very similar to this up their sleeves for the iPad 3.
Epic’s Tim Sweeney on Android fragmentation
From an interview with Matt Buchanan at Gizmodo (which I saw via Daring Fireball):
Speaking of Android, you’re probably wondering why there’s no showstopper like Infinity Blade for the platform. Well, wonder no more. Says Sweeney, “When a consumer gets the phone and they wanna play a game that uses our technology, it’s got to be a consistent experience, and we can’t guarantee that [on Android]. That’s what held us off of Android.” The problem with Android is consistency. “If you took the underlying NGP hardware and shipped Android on it, you’d find far far less performance on Android. Let’s say you took an NGP phone and made four versions of it. Each one would give you a different amount of memory and performance based on the crap [the carriers] put on their phone.” Bottom line, for Epic to do the kinds of things they do on iOS, “Google needs to be a little more evil. They need to be far more controlling.” Even so, the main reason Epic has focused on iOS? “It’s really the best place to make money.”
This would be the same Epic Games which released seminal PC game Unreal in 1999 and — with Unreal Tournament, UT2003, and UT2004 — owned one of the PC’s largest ever first-person shooter franchises? The same Epic Games who ignored consoles until 2006’s Gears of War? Seems to me they did alright working with an extremely fragmented platform back then.
Each gaming PC has different amounts of memory. Each one has wildly different performance based on the specifications of the components the user has used. Each user has a different sized monitor running at a different resolution. Everyone has different mice tuned to different sensitivity settings. In general, the performance deltas across gaming PCs are far greater than you see across Android handsets and yet Epic made that work, as have countless other games companies. I’d say that’s clear proof that even a badly fragmented platform is not an insurmountable problem — it’s just one that’s more tricky to work with than fixed-hardware platforms like games consoles and iOS devices.
So, sure, today Android is a non-existent priority for Epic (and indeed it is mostly a console game company these days; since 2006, their only PC release has been Unreal Tournament 3, which had lackluster sales). If there are less Android users or they buy less apps, there’s little reason to take the trouble. But if profitability of Android gaming increases or there’s some other good reason like, perhaps, a successful gaming phone with physical controls, expect Epic to quickly stop disregarding it and follow the money.
Source: Gizmodo
Android Malware Scare Causes Google to Pull 21 Apps - Mashable
At least 50,000 Android users downloaded these apps, which are not only pirated versions of existing apps, but designed to steal data and leave a door wide open for downloading more malicious code.
Ouch. As Michael Gartenberg tweeted:
The problem with an open marketplace. Anyone can put anything on the shelf.
And m’learned friend Faruk Ateş also wrote:
“Open” is better. If you’re a malicious person.
I don’t think that’s a valid conclusion though. Whilst it’s true that Apple hasn’t had a malware problem anything like this one with the iOS App Store, I don’t think that’s particularly because the App Store approval process works against it. We’ve seen apps with hidden features Apple missed and we’ve seen knock-off copies appear too. It seems fair to suggest that the App Store approval process doesn’t:
- successfully examine every inch of an app for hidden functionality — static analysis can only take you so far.
- successfully compare it to everything else in the store to make sure it’s not a dupe.
Indeed, if it did, I’m sure App Store approval times would be very, very long indeed.
The reason I think we haven’t seen any serious malware in the App Store is more because if a black hat gets something on the device there’s not an awful lot he or she can do. The Android malware installed a rootkit, with background daemons and remote code injection; it could have gone on to install keyloggers and even turn the device into part of a botnet. In contrast, the more restricted multitasking present in iOS means these sorts of juicy (from a hacker’s perspective) goals are simply impossible.
Of course, this is a two-edged sword, because the ability on Android to install apps that can run in the background without restriction or can perform system-wide modifications like Swype is a key differentiator from iOS.
In any event, I’m not sure it’s fair or particularly meaningful to simply pin this on “openness”, however tempting it may be because of how fast and loose Google have played with the term in the past (as indeed have Apple, to be fair). I see this not so much as a unique failing of the Android marketplace so much as an unfortunate downside to Android’s permissive nature as an OS.
Source: chartier
