Oh, right, there’s an entirely new version of Android right around the corner. It could be days away, it could be weeks away. We’re not totally sure what Google has planned for what is easily the most ambitious and promising update to the platform since Android 2.1. It’s easy to forget that there’s a whole new world right around the corner, because Android is in this seemingly constant state of change now. We have core apps updating on a regular and consistent bases, manufacturers pushing their apps to the Play Store in order to update them in a timely manner, and the beating heart of the platform is on a six week release cycle. Of all the incredible things that we saw and heard about at Google I/O this year, Sundar Pichai’s announcement that Google Play Services would be updating and improving every six weeks is one of those things that didn’t get nearly as much attention as it probably should have.
It really is quite remarkable. In some ways, Android is starting to faintly look like a rolling release, with more and more core smartphone applications, as well as several core smartphone APIs, updated continuously through Google Play. The pace is quick, and I like it.
Still, the Android update situation has not been resolved. There’s a lot more work to do.
Installed CyanogenMod on my phone. Gets updated once a month now.
Unfortunately, with Cyanogenmod I’d lose Touchless Control and Active Notifications of my Moto X.
I do use a cheap Huwawei with Cyanogenmod for my Dutch number, though
.
Installed CyanogenMod on my tablets. Gets updated once a day now
It’s interesting to see more and more Google apps appearing on the Play Store now, so that people can update them independently of the OEM updates. Surprised a few “obvious” ones haven’t made the jump yet, such as Gallery (preferably with Chromecast support!).
Edited 2014-09-21 11:11 UTC
Insane. That’s the perfect word to describe Android: fragmentation, featurism, lack of coherence, lack of proper support by vendors. Android is CHAOS.
And chaos is not necessarily a bad thing, in fact, innovation is happening there and tons of people love its craziness… but let me be honest here: for conservative “just work” people like me Android is hell. Too much.
I think in a phone as a tool to solve problems, not a new problem to solve. But hey… I’m a completely old-fashioned unix guy. xD
So you should be more than used to fragmentation.
http://en.wikipedia.org/wiki/Unix_wars
Yeap, and It sucked in the same way that Android fragmentation sucks today.
It’s crazy… Android is the new sacred cow of the techie people like Linux desktops were 10 years ago. If you dare to criticize it… you are a troll!! wtf
When did it ever change?
When I look at UNIX ecosystem, it is as fragmented as ever.
I sometimes wonder whether the fanboy warriors of the internet think their own claims are even plausible.
My advice to people wanting an Android phone or tablet is to try to get a device that little or no “bloatware” slathered on top of it, which inevitably will slow down the likelihood of timely updates.
If you want to take this to the extreme, it means an overpriced “Google Play Edition” (yes, charge for the removal of bloatware), a Nexus device or – if you want to get techy and plunge into the world of rooting, bootloaders and custom ROMs – a model that can run CyanogenMod 11. All of these should mean reasonably quick updates (daily if you run CM11!) and give you a non-bloaty experience.
Or any recent Motorola phone. I got the latest KitKat update on my unlocked Moto X a week before it appeared on my Nexus 5.
Kitchen Sink
The APIs get updated, while one is required to use Java 6 if targeting all versions, with the official answer at Google IO for Java 8 support being “No comment”.
Or even better, the answer regarding better support for alternative JVM languages, “Why would you want to do that?”.
Kotlin is a more modern language than Java 8 and can be used to develop Android apps:
http://blog.jetbrains.com/kotlin/running-hello-kotlin-on-android/
http://blog.gouline.net/2014/08/31/kotlin-the-swift-of-android/
Going from Java to Kotlin is also trivial. So, who should really care?
Edited 2014-09-20 08:39 UTC
Because it is not an official approved language, lacks tooling support and there are still interop issues when targeting Android.
Additionally you increase development time by having to go both to Google and JetBrains to sort out issues.
Finally, most clients only accept consulting projects done in the official sanctioned languages.
Yes. I’m screaming inside for Java 8 support on Android. I make heavy use of RxJava, and the lack of lambdas on Java < 8 adds so much boiler plate code.
I wish I could just ditch Java in favour of Scala, but it’s simply too hard to get buy-in from the business. Now, if Google gave us a better language…
How tightly bound is Swift to their Cocoa framework and is it possible to to decouple it and provide Swift on Android or at least android-a-lise it. In the case of something swift like you get the benefit of being able to leverage the developer mindshare from iOS whilst pulling Android forward.
A language is nothing without its eco-system. Hence why Objective-C has never tookoff outside NeXTStep and nowadays Mac OS X.
Funny enough, Sun was considering adopting OpenStep for Solaris when Oak project was going on.
Java interfaces are loosely based on Objective-C protocols.
I don’t think that it would make too much sense to adopt Swift for Android development. Swift was designed to work well with Apple’s existing frameworks, which have been influenced a lot by Objective-C.
If you want a language that works on both platforms, you’d probably go with C# (or F#) via Xamarin instead. C# is much closer to Java than Swift is. I personally think that C# would already be an improvement over Java 6/7.
Edited 2014-09-21 16:07 UTC
C# and F# sound like good alternatives but the question is whether Microsoft is willing to give up the right to charge a royalty or make a patent claim if C# support was added as a first class language in Android. If Oracle keeps behaving the way they do and Microsoft have the smarts they would be trying to be as relaxed with their patents relating to their .NET framework and C#/VB.NET/etc as possible to encourage widespread adoption where possible. With Larry still in Oracle as CTO I doubt their culture of litigation will stop so maybe there needs to be a backup plan in place. How plausible is it to use something like LLVM to convert their Java code to an intermediate language then convert the code back from the intermediate language to C#?
Well, Google started it, by not wanting to pay for Java licenses and Oracle saw an opportunity.
However, having Oracle Mobile Application Framework as proposition for mobile OS, is a joke.
http://www.oracle.com/technetwork/developer-tools/maf/overview/inde…
Maybe they will announce somthing better at next week’s Java One.
Why use such a convoluted process when there is ikvm.net?
More interesting, is that the Android team is not even keen is using their employer languages (Dart, Go) in the platform.
Seems to me that Google is similar to Microsoft in that different groups are pretty independent and off doing their own things with their own technology stacks. Apple seems fairly unique in that their teams seem to be dogfooding everything from top to bottom.
They’d have no patent claim at all, if all Google used was the language. The language itself is an ECMA standard. What Microsoft has are patents and licensing for their runtimes and for .Net. As long as Google didn’t touch those, they’d be fine using C#.
Just watch the Google IO 2014, Android developers fireside.
https://www.youtube.com/watch?v=K3meJyiYWFw#t=1566
– Java is the official language
– No comment on Java 8.
– Only use Scala for business logic
I haven’t watched that talk yet, but my initial reaction/question is why should Scala be used only for business logic? Scala can easily be a Java replacement for Android development. As far as interop goes, Scala and Java are arguably even closer to each other than Swift and Objective-C are.
Most likely they only want the Android Studio and ADT tooling to work with Java code.
Okay, so I’ve now watched the part where Scala and Swift were discussed.
I’m a bit surprised in the way that the Android team dismissed Scala specifically. They imply that the framework would have to be rewritten to support the language. That’s not required.
The main thing that’s needed is better tooling in the IDE(s), and Google’s blessing. Apple is certainly not rewriting all their Cocoa and Cocoa Touch APIs in Swift. I can see the frameworks changing over time to better support Swift idioms, but this is going to happen over several years, and in a way that doesn’t break Objective-C compatibility.
I understand that the Android team is working hard to make their tooling better. That’s an area where they are lagging behind Apple and Microsoft. However, they are firmly entrenched in the Java-only world and don’t seem to understand the nature of Scala and Java interop.
So, based on the comment that Google deliberately designs things four years in advance because they know that they will want them to work a certain way in the future, and based on the jibe that Swift was needed because Objective-C is based on C, which is 40 years old, I think that we can look forward to at least four more years of Java. Shame. We could do so much better than Java.
Here’s to hoping that a Chris Lattner pops up at Android HQ and takes it upon himself to champion a new language.
Edited 2014-09-22 02:21 UTC
I like having an Android phone and all, but the iPad has been way in the lead for compact music creation. Show me anything like Loopy or ThumbJam or AniMoog on Android? You can’t. That seems set to change, though.
http://liliputing.com/2014/06/google-brings-low-latency-audio-proce…
http://superpowered.com/
If quality music apps appear, and someone really, really solves all the issues of syncing music and ratings and smart playlists to Android devices, I’d feel very comfortable with it. The fact that CyanogenMod and Replicant exist in the Android ecosystem caters to my preference for open source. The best of both worlds doesn’t seem too much to ask anymore.
And yet today there are still flagship Android devices launched with older versions of the OS pre-installed, and with no firm word on if or when they might be able to be updated to the latest version, or even any future version. Unacceptable. When you buy a new iPhone, you get a device that has the latest version of iOS installed ON it, and you have the assurance that your device will be capable of running EVERY future version of iOS for at least the coming 2 years and usually longer.
Yes, there are cheapo Android phones out there meant for “emerging markets” that have no use for the features in the latest OS version, but to launch a flagship “next great thing” phone with a 1-2 year old version of the OS is just ridiculous.
Apps do get updated on a regular basis, but that’s a whole other matter than if you buy a top of the line Android phone and it comes with a year-old version of Android OS and will never be able to be upgraded. Your article talked about the “insane pace of Android” but it’s really about the more frequent pace of app updates.
Edited 2014-09-20 09:42 UTC
Any version of Android since 4.04 is ‘good enough’. I have a 4.2.2 phone and a 4.4.2 tablet. The real world differences are negligible. [Even Gingerbread is perfectly usable.]
When you buy a new iPhone you are getting expensive and outdated hardware. The iPhone 6 is arguably no better than a 2012 Nexus 4.
Total BS. Flagships always have a very recent version of Android. Even the current sub $100 models have 4.2 (many have 4.4).
More BS. If you buy a CURRENT flagship you get the LATEST version of Android (eg KitKat). If you buy an OUTDATED 12 month old phone you (obviously) don’t get the latest version of Android. Nearly all flagships get at least one update.
Updating the apps upgrades virtually every bit of user software.
iPhone uses outdated hardware?
so your Android runs 64bit software then?
What about the TouchId that IMHO is far better than anything that Samsung can produce at the moment.
As for expensive H/W I saw a report that it costs Apple $338 to make an iPhone 6. It might cost a lot more than many Android devices including my old Samsung Galaxy Mini but they seem to be not cutting the corners that some other handset makers are doing in their race to the bottom.
Please tell us why you think that the hardware is outdated?
Apple uses 64bit as another marketing gimmick. It offers no real advantages at this stage. Real world benchmarks show that Apple phones are slower than other flagships.
Apple is a a design studio. ALL their hardware is produced by other companies. SAMSUNG makes the Apple 64 bit CPU.
There is NO way an iPhone 6 costs $338 to build. The hardware is only marginally better than the sub $200 Moto G. Apple simply sells mid range hardware at very high prices.
The iPhone 5C = Galaxy S2 (Feb 2011). iPhone 6 = Nexus 4 (Nov 2012). iPhone 6+ = galaxy Note 3 (Sep 2013).
A lot of misinformation in this post.
64-bit was not a marketing gimmick…. the only benefit was not address spacing. Please refer to Anantech with their thorough review of the 64-bit processor Apple used in the 5S… it had real world consequential benefits as a result. And it beat out the performance of ALL mobile phones at the time of launch.
Apple designs most of their own hardware… they are fabricated with their partners (including Samsung), but the designs are solely their own. Which is why they’re able to tweak their designs and have historically gotten more performance out of them than equivalent processors. Why would Samsung need to CHEAT on benchmark tests if they are always the best?
At the end of the day, it’s not about what looks best on paper, it’s about the entire unit as a whole, and how software and hardware coexist and optimized for each other.
Apple use modified ARM hardware manufactured by Samsung. Apple’s contribution to the designs would have been quite limited. The grunt work would have been done by others.
The whole 64 bit story is simply to get technically ignorant baristas (Apple’s core demographic) excited.
Lol. Do yourself a favor, read about ARM architecture license and Apple cores. You know absolutely nothing. Apple designed A6 cores with custom layout – totally unheard thing in ARM world. That thing alone placed Apple on top of CPU design houses.
A8 is almost as advanced as Intel latest cores. You need 2X frequency and 2X number of cores to match it for the rest of ARM crowd.
You clearly don’t understand that ARMv8 is, how performance advancements is achieved, and why single thread performance matters.
Your whole comment reeks of fanboyism, to be quite frank. http://www.anandtech.com/show/8514/analyzing-apples-a8-soc-gx6650-m… for example states that A8 seems to be about 25% faster than A7. That’s nowhere near the kinds of numbers you’re trying to claim.
I’ll grant you that the A8 is only 25% faster than the A7, but the A7 was in a different league from other ARM cores available at the time.
It is true that a the 1.4 GHz dual core Apple A8 can beat the quad core 2.56 GHz Qualcomm MSM8974AC (Snapdragon 801) in multicore performance. In single core performance, the 1.4 GHz A8 is almost twice as fast at the 2.56 GHz MSM8974. See for yourself: http://www.youtube.com/watch?v=v6d7Fes-eHA
The Apple A8 is not an ARM design. It uses the ARM AArch64 instruction set, but it is a completely new core whose MIPS/MHz rating is much greater than the ARM A15, A57, or Qualcomm Krait 400.
And here come the Mhz Wars all over again. Fun times.
Your whole comment reeks of fanboyism, to be quite frank.
Hehe. I’m ARM fanatic, that’s true.
And I think that OP comment, who trashed the work of hundreds of high-skilled engineers, plagued with fanboyism and illiteracy.
A8 seems to be about 25% faster than A7.
That’s nowhere near the kinds of numbers you’re trying to claim.
Ok, If you want geekbench results, let’s check
http://browser.primatelabs.com/geekbench3/581352
A8 @ 1.38Ghz | 1633 | 2920 (2 cores)
SD801 @ 2.46GHz | 1022 | 3011 (4 cores)
So we see ~3.5x IPC advantage over fresh QCOMM cpu.
Also Snapdragon will burn much more power at max load (up to 8X in theory)
Original poster is blind to the point he could not even understand that “limited contributions” to the core design could not be done years before this core is actually available.
People who can crank up to 4x performance with “limited contributions” worth their weight in gold.
I don’t even mention that A6 and A7 cores has nothing common with ARM cores.
One little diagram at iPhone6 announce was actually very important – A8 is the only mobile CPU capable to work without throttling for a long periods of time.
Perf gains will decrease every generation after radical perf leap. Intel was only able to bring <10% perf improvements per generation for several years.
Apple produced 2 different archs in a row. That means that most likely they have 2 teams, or worked really hard as usually CPU development takes several years.
BTW SGX6650 claim in this article is incorrect.
Despite the blurry photo, seems there are quad core SGX6450 GPU.
http://www.chipworks.com/components/com_wordpress/wp/wp-content/upl…
Edited 2014-09-22 03:55 UTC
BTW, for a starter you could read about the Apple A6 development instead of flaunting with the misinformation.
http://www.cnet.com/news/apples-a6-chip-development-detailed/
64Bit as a marketing gimmick?
I seem to remember much the same comments being made about the DEC Alpha in 1991/92. Not long afterwards the other Unix vendors started making their own 64Bit Servers.
Now where would we be without 64Bit CPU’s?
Apple design their SOC’s like many other ARM Licensees. Then the get a FAB to make it. Samsung and TSMC are two of the foundries making chips to Apple designs. That is Standard practice in the ARM world. To say that they do very little is IMHO a long way from the truth. If you are true then you could say the same thing about HTC Samsung etc.
So the pace of innovation on android is apparently insane, and yet there are negligible differences between the current version and that from over 2 years ago. Which is it?
One might argue you have no idea what you are talking about. Seriously the two aren’t remotetly the same aside from superficial stuff like screen size.
It’s telling that you think it’s normal for a slightly older model not to run the latest OS, or that you think getting at least one update is anything but embarrassing.
Edited 2014-09-20 23:21 UTC
A bit of both.
The truth is, bragging about a 6 week release cycle is mostly just hot air. It just the same as the 37 (!!) releases of Chrome and who knows how many of Firefox.
The development speed hasn’t increased one bit. What has changed with such a model is how long a patch rests in the master branch of the respective projects before it hits the users.
The catch of such a release model is that getting a known stable version gets far more difficult because there’s constantly landing new changes. This is the story they never tell.
Sorry, but I just won’t settle for “good enough”.
Examples?
My brother-in-law met a guy who knew a woman who’s uncle overheard somebody claim that an HTC One M8 used ‘Donut’. Either that or a guy with an M8 was eating a donut. I can’t remember clearly.
I mostly use my phone for calls and text messaging, it’s a communications tool for me, not an app platform or a device where I want a lot of features. The idea of regularly updating my phone definitely does not appeal in any way. I’d much rather Android have updates around once a year rather than a rolling release model.
There are many technical difficulties behind Android updates. It can take months to get a device ready for a new Android version. Too many things to take care of. display drivers/support for them in the kernel/google breaking things, etc…
Some vendors simply don’t bother for the sake of stability and because mostly because they would rather put their time into making new devices than porting older devices to new android versions.
A huge part of this is that most of the code the vendors ship is in their local tree and not the mainline Linux kernel.
Apple doesn’t have that problem because they make both the hardware and the operating system for their phones.
This isn’t similar to your regular computer where you can simply update to a new kernel, xorg-server and nvidia/intel/ati,etc… driver and simply expect the whole stack to work on your HP or toshiba or Dell or whatever.
So yes, it is basically a problem of fragmentation.
The difficulty lies in reducing fragmentation without alienating anyone or taking over.
Edited 2014-09-20 15:05 UTC
I think with Android One, Google has the opportunity to help standardize the hardware and therefore make OS updates easier. “Premium” Android hardware companies will be forced to keep up. Hopefully.
Software updates are financial problem rather than a technical problem. The carriers and manufacturers are only interested in selling new phones. [Apple is more cunning. It offers irreversible updates which slow older models to a crawl.]
In many cases a ‘new’ phone is simply an existing model with a slightly different case and a later version of Android. This is particularly prevalent in low end models.
If you wrote this article but changed Android to Unix would it still work? Unix is everywhere, it runs a huge chunk of the computer devices on the planet. It just keeps growing. But is it a single platform which can reap the benefits of it^aEURTMs massive installed base through the platform positive feedback effect? I don^aEURTMt think so.
Is Android a single platform which can reap the benefits of it^aEURTMs massive installed base through the platform positive feedback effect? I^aEURTMm not sure. Android seems to have evolved into a computing layer that is a bit like Unix in terms of it^aEURTMs role as a platfom, and if you count devices running that layer it^aEURTMs impressive but the platform feedback effect comes above that layer and there things get a bit more complicated.
It^aEURTMs absolutely clear that Android is not a particularly strong platform, in terms of generating platform effects and monitization. We can say that with certainty because we have a control case to compare it too, which is iOS, and which has a markedly stronger platform positive feedback effect even when cast into being a minority platform.
Why Android has a weak platform feedback effect is an interesting but complicated question and one I have not seen a definitive answer to. It may the result of lots of interlocking factors.
An even more interesting question is whether the platform feedback effect of Android will weaken or strengthen in the future as it evolves as a layer in many devices and businesses.
It doesn’t need to be. UNIX (mostly FreeBSD) is flexible enough that it can be monetized by multiple partners under different brand names. Data ONTAP, PlayStation 4, JunOS, OS X for example. Thus, name recognition goes to specific products, not the platform itself. This distances the core technology from the end-product allowing the core product to remain in use long after name brand recognition for a single product is tainted.
I imagine the same thing is happening in the Android world. Where we (the readers of a tech blog) see it all as a single platform, customers see it as either a Kindle or a Samsung or a Motorola or a Tesla. Thus, when a single company (say, HTC) goes under, the core product itself (Android) can continue to be used elsewhere, uninterrupted.
Protip: with most Tony Swash posts, you can skip the first half – because it’s invariably just filler, dancing around/leading up to his “…and this is why iOS is better” sales-pitch. Case in point:
Sweet merciful crap, you’re still beating this dead horse? Out of genuine curiosity, what’s your fallback plan/talking-point for when Android inevitably does overtake iOS in terms of “monitization” [sic]?
Yawn. You could go back in time in the mid-late 90s and use the exact same spin, with very few changes, to argue in favour of AOL vs the “open” Internet – and in a very shortsighted way, you would have been right. Throughout most of its heyday, and for some time afterwards, AOL had more subscribers than any other single ISP, and they probably controlled the lion’s share of online revenue thanks to userbase + pervasive advertising + they fact that they managed to attract many high-profile players from traditional media (many TV shows/networks, print magazines, etc, had an AOL presence before they had a website).
In other words, having a closed system gave AOL an INITIAL advantage. They were competing against a bunch of decentralized small, local (and a few larger national) ISPs – but the ISPs had the fundamental advantage long-term, because they all used basically the same underlying tech. AOL was even able to remain relevant for a few years after the legion of separate-but-compatible ISPs overtook them, coasting on the momentum from that original advantage – but eventually their profits followed their userbase. In other words, a “strong” single-vendor ecosystem lost out to an ecosystem of many smaller, “weak” individuals – despite their large initial advantage.
Apple is just repeating the same pattern.
There is no sign of that happening at a user per capita level and with the ‘minority’ platform users numbering in the hundreds of millions I cannot see how the platform feedback loop (where greater market share leads to greater investment in the platform that leads to a better offering than other platforms that leads to attracting more users that leads to greater market share) is ever going to happen.
Given the pretty well established patterns of platform activity on the two giant mobile platforms it seems to me that for Android to achieve a scale such that it would, through market share alone, leverage some sort of profound advantage over iOS would require such a preponderance of users that the user base of Android would need to exceed the earth’s population.
It’s interesting that a software layer like Android is so wide spread, and will probably migrate much further into all sorts of places and uses, but in terms of it affecting other significant dynamics (such as relative platform dynamics, OEM commercial performance, third party monitization levels) I can’t see it having a unified impact anymore than Unix has a unified impact. In fact if anything I think the impact of the size of Android uptake will get more dispersed, thinner and less impactful as times goes buy.
Suppose two billion people end up with devices of one sort or another that run on something that could be construed as sitting in a camp called ‘Android’, so what? What of significance flows from that? Who does it impact and how? Does it have any greater meaning than saying most of the world’s computer devices run on some form of Unix?
I’ve got a counter example for you: Microsoft, using a strong, single-vendor ecosystem, crushed everything before them. Even the mighty IBM fell, and Microsoft dominated the industry.
I really don’t think that you can draw the conclusions that you do. Firstly, selling access to dumb pipes is different from selling devices. Yes, Apple makes its money from selling devices, and not from selling iOS. Let’s not forget that. iOS doesn’t need to win an OS war against Android in order for Apple to do very well. Just like Mac OS didn’t have to win against Windows for Apple to be one of the most successful PC vendors. The fact that iOS competes with Android and that OS X competes with Windows – that’s incidental to Apple’s main business. Apple understands that it needs to build its own platform in order to be in control of its own destiny. That’s a strategic advantage that they have over say Samsung, who are at the mercy of Google. Just like Dell is at the mercy of Microsoft. But Apple’s money comes from selling devices. Not from selling access to their platforms.
As they say in Canada: “eh?” Since you didn’t bother to specify, I can only assume that you’re referring to the success of PCs running Microsoft OSes… except that that’s almost as far away from being a “single-vendor ecosystem” as is possible, once the clones started appearing (within a year or two of the first IBM PC going on sale).
If anything, that’s an even better example of my point than AOL was. Microsoft could just as easily have hitched their wagon to IBM and designed their software to prevent it from running on IBM clones – in which case they would have likely been relegated to the dustbin of computing history along with companies like Commodore and Atari (and Linux probably never would have come to into existence, since it would have had no hardware platform to run on – but I digress). Instead, Microsoft had the foresight to back an open system. At best, you might be able successfully argue that that’s an example of a closed single-vendor software ecosystem succeeding by backing & fostering an open hardware ecosystem – which is still world’s away from Apple’s “everyone else must fail in order for us to benefit” approach.
Now you’re merely splitting hairs – just because 2 different companies sell 2 different products/services, that doesn’t mean that they’re NOT subject to the same basic economic principles. In all 3 examples (AOL vs ISPs, Macs vs Windows PCs, and iOS vs Android), the part that is of actual importance is open and not limited to a single vendor.
It’s also disingenuous to portray Apple as selling devices and not services – the App Store & ITMS may be line items on the invoice when you buy an iProduct, but make no mistake: they’re selling those services just as much as they’re selling the physical devices. Not the least of which because those services are a way of extracting revenue from customers after/in addition to the device purchase (and are obviously intended to do so).
Non-sequitur, I never argued that Apple cannot continue to make money, even when Android inevitably overtakes it on all fronts. My contention is that iOS having higher “monitization” [sic] than Android is only a transitory hold-over from the days when iOS had a larger userbase. If Android’s userbase continues to leave iOS in the dust at the rate it has been (and there’s no reason to believe it won’t), then the relative profitability of the platforms will follow suit eventually – only the most naiive (or intellectually-dishonest) would expect that to happen overnight.
And, again, I never claimed otherwise. When Apple falls on hard times again, they’ll just do what they did in the mid-late 90s and early 00s: put the screws to their core demographic(s), and milk them for every penny they’re worth.
The success of MS was a freakish alignment of the constellations. MS was a tiny Seattle software services company that happened to be in the right place at the right time. PCs took of because Compaq reverse engineered the BIOS making clones feasible. IBM totally fscked up launching OS/2. The rest is history.
The rate of innovation in Android has been analyzed quite some time ago already.
Eric S. Raymond has observed in 2011 that the length of Android’s development cycle was approaching 90 days, which is way faster turnaround than any of the single company competitors can achieve.
http://esr.ibiblio.org/?p=2975
He then used military theory on OODA loops (orient, observe, decide, act) to predict the ultimate defeat of all competition in the marketplace. With Android now being close to 90% marketshare, it seems that he was right.
The Android update situation has been resolved. Nexus solved it for the few informed customers, and Android One solved it for the masses.
Really? Are you a fortune teller? Last I checked, Android 1 or L or whatever the heck it is hasn’t even been released to the masses yet. A bit hard to even know what effect it will have before it hits the market, wouldn’t you say?
Android One is a low cost hardware reference platform with multiple manufacturers. It will offer immediate software updates direct from Google. [Poor Man’s Nexus]. It will be released in December 2014.
Android L is Android 5.
You can buy Android One phones in India already. That’s 1 billion potential customers (“masses”) right there. Note that Android One and Android L are two different things.
Here is a review of a Reddit user who bought one:
http://www.reddit.com/r/Android/comments/2h1j0x/heres_android_one_b…
Android One is so successful that more manufacturers signed up, and in about 90 days (late December) we will see three more models in India alone.
http://economictimes.indiatimes.com/magazines/panache/round-2-in-an…
Of course it will come to other regions, too. Understandably, poor and developing countries will be served first.