A Murky Road Ahead for Android, Despite Market Dominance | Hacker News Reader

A Murky Road Ahead for Android, Despite Market Dominance

nytimes.com

60 pointsddlatham74 comments

What if a significant number of the people who adopted Android as their first smartphone move on to something else as they become power users? In Apple’s last two earnings calls, Timothy D. Cook, the chief executive, reported “a higher rate of switchers than we’ve experienced in previous iPhone cycles.”

Are they trying to insinuate that power users move to iOS? I'd consider myself a power user, but for that very reason I've stuck with Android. Higher-end Android phones provide much more for me than iOS can with its closed-ecosystem.

Also, are these "switchers" coming from Android, or a mix of other sources?

If I had to bet, it's probably Apple finally getting around to offering a phone product in the size that people want. For years, if you wanted a large phone your choice was only Android.

And now the situation is reversed, if you want a normal size (ie, less than 6 inches) phone, Apple is starting to be the only company in business that delivers powerful hardware in a non-phablet form-factor

I assume you're comparing only Nexus devices, because the breadth of Android devices in different sizes is by no means limited.

Yeah, I've got a Moto X and while it's not particularly small, it's basically the non-phablet version of the Nexus 6 with a few minor differences.

The trouble being that small devices are practically always very slow and forgotten the day they hit the market.

It depends on what they mean by power user. If you like customizing the OS, running another ROM, launcher, or theme, than Android is friendly to that.

If you like using more powerful software, than iOS / iPad seem the clear winner there. Look at the top 10 in both stores, they are very different.

In Apple Store, you will see lots of photography and music apps. In Android, its games and social media.

I've been developing for Android half a decade now, but still see the OS as a toy that can't grow up - mostly because of its design and APIs, but thats another story for another time.

I think both Microsoft and Apple are in better situation with convergence between their form factors and the capabilities of each platform.

For audio there's the known latency problem, but photography(the more popular category), android recently released new camera API's . won't they solve that issue?

Yeah, the imaging software can be really good on Android as well. Still, even as a fairly devoted Android user and not a big iOS fan, I'll point out that only the best Android devices have camera hardware comparable to the newer iPhones. The flipside to being able to install Android on any hardware means that the OS can only do so much if OEMs cheap out on the camera hardware.

Still, for the average user, the cameras on just about any higher end Android phone or iPhone will be perfectly fine for their intended purpose: being the camera you have at hand. For a photography pro or enthusiast, even the newest mobile devices (from any OEM) will only win on convenience versus a dedicated camera with larger optics. There's still only so much you can do with a lens that's a little chip of glass and a tiny sensor.

I consider myself as a power user aswell, but I've left Android after 5 years on the platform - reason being Lollipop. I'd consider Android when Google starts taking it seriously again and not a pet project.

They may be talking about power users as users who uses more of the phone, not necessarily technically inclined users.

Are they trying to insinuate that power users move to iOS?

Well, if Google tries to make Android more profitable by adding more restrictions and trying to make it more like an iPhone-- I'll be faced with the decision to either switch to iPhone (which is probably better at being an iPhone than Android) or one of the Android forks. As the article points out:

Even in the rest of Asia, where many low-cost phone manufacturers do include Google’s apps on their phones, there’s growing interest in finding some alternative to Google’s version of Android. About 30 percent of Android smartphones shipped in the last quarter of 2014 were actually modified, or forked, versions of the OS that may not be very hospitable to Google’s services, according to the firm ABI Research.

I don't understand why they don't use Tizen, FirefoxOS, Sailfish, Ubuntu, etc if they dislike google so much. The reality is they need google for all the services they can't or won't implement. They don't want to build their own mapping software, browser, etc.

Its all moot anyhow. Google has shoved all the newest APIs into its google play services and without a google license for those items, those apps won't run. This was a play to stop fragmentation like the one you're predicting.

I suspect the 30% of phones are Xiaomi phones, which solely exist because the CCP won't let Chinese citizens access google mobile services as they're blocked by their national firewall. If this block didn't exist, Xiaomi would be using all the Google services everyone else does.

> This was a play to stop fragmentation like the one you're predicting.

The parent seemed incredulous that "power users" would use iOS. My point was that power users might just give up being power users if Google does things like you describe: locking up the APIs in google play, etc. A lot of so-called "power users" will just go ahead and give up on Android and use iOS devices.

I believe Xiaomi phones use Google apps in non-Chinese markets, e.g. India.

I had to switch based on my company only supporting ios on a BYOD plan. But once I switched I sorta prefer ios. But I could be swayed back, especially if Android phones start decreasing in price as they commoditize themselves like PCs did.

This assumes "power users" want more engagement with their phone.

If I want to do something complex, I'll do it on my laptop or desktop. My phone is always with me and I just want something straight forward for everyday life.

I've always interpreted "power user" in this sense to be someone who occasionally wants to do something specific with their device that may require specific tools.

Like a "normal" user (of a phone) will make calls, texts, browse the web, check email, maybe watch a movie or play a game. A power user may just be someone who wants a custom keyboard for some reason or even just wants to be able to download a file via the browser, navigate to it in the file system, maybe unzip it, copy it elsewhere, open it with another app, etc.

Granted these are just my personal examples because they were things I found exceedingly difficult to do on my old iPad (although I guess the keyboard thing has been added). It wasn't something I had to do all the time but something as simple as browsing the folder structure and finding/moving the file I just downloaded was gonna require me to jailbreak the thing. On my Android phones and tablet, it's not something I often need to do but when I do, it's just a matter of using a file browser.

I think "power user" implies using more basic/powerful features outside of launching apps from a set of icons. Then again, some of the awesome projects I've seen controlled via an iPad prove that iOS users can also be power users as well.

It's an interesting time. Android 5.1 is easily the strongest version of Android that Google have ever produced; in terms of design, functionality and performance the bar has been raised several notches. The big problem that remains is getting it onto devices is a timely manner. This also reared its head on several devices that received 5.0 early, only to have fairly serious bugs that were not rectified for weeks or months until the next update finally rolled out.

Android Wear is the most complete smartwatch OS, backed up by a solid selection of hardware. Chrome OS continues to show strong growth, with Chromebooks and Chromecasts selling very well. Will we see more integration between Chrome and Android? Will Google try to directly interfere with third-party app stores and launchers?

> The big problem that remains is getting it onto devices is a timely manner.

This. I own devices from a manufacturer that supposedly uses close to vanilla Android to roll out quick updates. My Moto X 2013 is still rocking 4.4.4, despite the fact that it was just over a year old when Lollipop came out.

My Moto X 2014 (also pure) has Lollipop, but the bugfix releases (5.0.1 and 5.0.2) were never rolled out. The rumor is that 5.1 is now rolling out in Brazil (I am in the EU) and one US carrier. It's been 2.5 months since 5.1 was released.

No surprise that my next phone will either be a Nexus or an iPhone (I used an iPhone 3G and then 4 for years, and always had prompt updates).

The counterargument is always: the user does not care about updates. But I assume that people who drop 500 Euro on a smartphone care more than those who buy low-end.

> the user does not care about updates.

The Manufacturer does not care about updates. It nets them no cash flows. Delay the updates, force people to buy new phones for new (software) features.

Now if Apple did that it'd be called planned obsolescence and every tech blog would be breathlessly bleating about it.

Apple controls everything in their walled garden.

Android controls next to nothing - so you can pass the buck pretty well.

Supposedly the problem with the Moto X is the specialty chip they use for gestures and such. Which is interesting and troubling, because that kind of specialty hardware seems like the #1 place an OEM can add real value & differentiation to their product line, but adds obvious obstacles to rapid OS updates.

> Will Google try to directly interfere with third-party app stores and launchers?

They have already in a subtle way. Many apps depend on Google Play Services to function (including TextSecure). It's surprisingly difficult to find a list of good apps that don't depend on Google Play Store being installed.

As it stands, Android is almost unusable for the average user without the stock app store installed

This is kind of true. Though you're the first person I've heard say it without complaining that GPS isn't in AOSP.

Amazon has done a pretty decent job keeping up their version, as far as I know - and they try to maintain API compatibility.

This could be solved by some good third party architecturing a solution, but overall it's Google dominating by providing the better developer tools and ecosystem.

AOSP has GPS APIs. Maybe you're referring to the combined wifi/cell/gps location API in Play Services. It's built on top of the AOSP location API. Anyone can write a similar library.

I am, yes. And while anyone can always write a similar library, it increases the cost of switching out of the Google ecosystem.

A third party solution would have to provide some compelling benefit to get app developers to add support or switch. Otherwise it's not worth the time and effort to switch just to benefit the tiny amount of users who don't have GPS.

Provide a compelling alternative and you'll see developers catering to that audience. As it stands, the Play Store is the de facto target because it has the most users and the best developer experience. Amazon provides a very small fraction of users and revenue, other platforms even less.

But who wants a phone that is going to junk out on every second OS version update?

I've been seeing these gloom and doom articles on Android since its launch. Nowadays at least they have to have the reality of the situation included - "A Murky Road Ahead for Android, Despite Market Dominance". "Android is now not just the globe’s most popular smartphone operating system, but the most popular operating system of any kind. More than a billion Android devices were sold in 2014". It doesn't seem so gloomy to me.

At least nowadays these Android gloom-and-doom articles have stopped saying that Microsoft is going to come out and crush Google in mobile. Microsoft's threat has been reduced to being involved in cloning what Google did - "One software start-up, Cyanogen, has raised about $100 million from several investors, including Google’s arch-competitor Microsoft, to sell phone makers an alternative user interface that works on top of Google’s Android."

FWIW, there have been iPhone doom and gloom articles since its launch. Maybe it's just that doom and gloom articles get clicks?

I'm sure that's the reason. Doom and gloom headlines for any kind of platform or brand identity is perfect as clickbait because it sucks in both the opponents of that thing (due to schadenfreude purposes) and proponents of that thing (who want to see what is being said so they can refute it and be outraged by its wrongness).

"Apple Phones and Android Phones both pretty cool and here to stay for the foreseeable future." just ain't going to get clicked like any of these doom and gloom ones (in either direction) are.

I half expected the article to reference the Oracle Java API fiasco. That could definitely murky the waters for Android. This, at least, is pretty credible threat to Android.

I wish they'd replace Java stuff with something like Go , just get rid of Art/Dalvik . Java is very verbose and the xmls and build tools are all really hard work.

I am obviously being naive here. It probably can't be done

Earlier today there was a link to a talk about the state of Go. The link below shows that Go 1.5 is working on Android and experimental on iOS. What does "working" mean? No idea. I would bet that Google has been developing another development language for Android since a ruling in favor of Oracle would would be a huge financial burden for Google.

http://talks.golang.org/2015/state-of-go-may.slide#20

>What does "working" mean? That you can write JNI apps in Go. You can't access to almost any of the platform APIs that way, so it is really not intended at writing entire apps in Go. What it would allow you to do is to write something like a Rest library in Go and share it between the desktop, iOS (depending on the state of Go on that platform) and Android apps.

You should have a look at Kotlin. It removes almost all of Java's verbosity and the best part is that you can already write Android apps in Kotlin since the compiler emits java's bytecode.

Go is still garbage collected. Supporting Objective C, Lua or even Rust would be more meaningful in terms of improving performance.

As a free software proponent, I've spent most of the time promoting Android to my friends.

Android(Non-Nexus) is a toxic hellstew of surveillance and security vulnerabilities.

I'm currently recommending iPhones or Nexus only to my friends.

Google could do a lot to improve this but the carrier relationship with handset manufacturers allow them to MITM the CA systems through root certs and that is god damn unacceptable.

It's also why the rumor that Nexus devices would get 2 years of OS updates and one extra year of security updates (3 years) makes me want to stick with Nexus devices, despite some of their annoying shortcomings such as worse cameras (than high-end competition).

Two years of OS updates (2 major OS versions, since they do them yearly now) + another year of security updates seems quite ideal and I'd hope more OEMs adopt that. Unfortunately even if they do, they won't do it for all of their phones, only their flagships, which means lie 80% of the Android smartphone population won't benefit from it.

>Two years of OS updates (2 major OS versions, since they do them yearly now) + another year of security updates seems quite ideal

As someone who has recently had to support some Windows XP systems, I have to say Microsoft has set a high bar for support. XP was released in 2001, received major updates for 7 years until 2008 with SP3, and received security updates until 2014, a total of 13 years. Some customers are still receiving special security updates.

While Windows XP is not exactly a desirable OS in 2015 for many reasons, its sad that 3 years of security updates is seen as progress in the smartphone world, and I say that as an Android user with a perfectly functional phone facing EOL this year (a 2013 Nexus 4). I'm hoping google can wrestle enough control from OEMs and carriers to be able to support their OSs for 5+ years with at least security updates. Its not without precedent, they supported Chrome on XP until at least this year -- on a 14 year old OS.

TL:DR

A lot of poor people use android , poor people don't spend a lot and don't click many ads.Therefore Android is doomed.

maybe you didn't read the article it says that Google also benefits from iOS users(although it says it's a risk that they lose the higher end niche), and the "poor" people are the ones actually ditching the google ecosystem entirely. The article never talks about poor people, instead it refers to places where google services are unavailable/unpopular, e.g. China.

edit: For those down voting, could you please quote the part of the article that implies this: " A lot of poor people use android , poor people don't spend a lot and don't click many ads.Therefore Android is doomed."

it actually contradicts the parent comment: " For years, Android apps were a backwater, but sales have lately picked up."

"In 2014, Google Play sold about $10 billion in apps, of which Google kept about $3 billion (the rest was paid out to developers). "

"Google’s app revenue is becoming an increasingly meaningful piece of its overall business, and it is also growing rapidly."

This article commits a composition fallacy and ignores that Android is the cost-leader.

Yes, as the #1 low-cost solution you're going to have users you don't make a return from, but you have enough users that as a sum you make returns. Android's strategy isn't to maximize returns from each user.

It's a subjective article and doesn't really provide anything insightful beyond FUD of the future.

Google's strategy isn't to maximize returns from each user, but what about developers?

>Google's strategy

Android's strategy isn't to maximize returns from each user, not Google. It's all market driven. Google is trying to maximize returns for each AdWords user, for example.

App developers all have their own markets and their own strategies which are mostly irrelevant here.

I wonder if Google will need to, eventually, move away from Java or if it's simply impossible at this point. I've owned an Android device with every version of Android so far (all flagship devices) as well as iOS devices and I find it difficult that Android still has that annoying (but occasional) lag that everyone equates to Java and garbage collection (though with so many initiatives from Google to eliminate this I have to wonder if it's really garbage collection that's the problem). I see this on my wife's S6 Edge and on my Moto X 2014 both running Lollipop.

iOS is always buttery smooth and while I enjoy the power and flexibility of Android FAR more than iOS it's hard not to get distracted by the occasional lag / hiccup that I've seen on every single Android device I've ever owned or tested.

I also find it annoying that Google doesn't seem to be able to focus on delivering a full user experience. For instance Android Wear let's you respond to test messages but you can't if your phone uses the default Messages app on the phone (I had to download Google's or almost any third party SMS application to do it). Such a flagship product with features being advertised that you can't use by default on most of their phones is just...dumb.

> that everyone equates to Java and garbage collection It is a very complex problem and there I don't think it is possible to simply point at Java or GC as the culprits. It has more to do with how the platform architecture has been thought out and the lack of focus at that time on overall fluidity. The Sky experiment by the Dart team (so another GC language) shows a very interesting attempt at designing a runtime for 120 fps apps. The core idea of this project could very well be rewritten in Java.

It's interesting to see a wider acceptance of the reality of the mess of Android.

Really the shocking thing was how good the iPhone 6 was, and how ready to switch to iOS places like Korea were when the larger devices became available, when previously it had been assumed those markets were lost to Apple. The emerging social class type distinction emerging globally between Android and iOS users is a real and growing problem for Google, especially combined with Apple's probable search engine launch.

The noise from my network in Android land has been that Lollipop remains a disaster. Easily the worst version since Android became popular, and it will be very curious to see what, if anything, Google have proposed for the resulting mess at I/O.

It says quite a lot that the most exciting thing about Android for I/O is the rumoured stripped down headless version for Internet-of-Things devices, which may become an accidental foundation for a cleaned up future Android proper.

> network in Android land has been that Lollipop remains a disaster

That's merely anecdotal, and my own anecdotes run in the opposite direction.

I'm building https://recent.io/ for Android and iOS and have a Nexus 5, iPhone 5, and iPhone 6 Plus on my desk as test devices as I write this. I use both OSes, though I do use iOS a bit more.

It's true that early versions of L were less than stable, though Apple has had the same problem. Another problem is slow adoption; only in the last few weeks, I think, has the Galaxy A3 been L-upgradable.

The saving grace for Android L is Material Design, which is finally a strong unifying design language at least as good as what Apple has to offer and IMHO better with at spanning different device sizes. It's well thought out and will make Android apps easier to use (and Android generally easier to use) by increasing UI/UX standard interactions. That's anything but a "disaster."

I don't have the contacts to know if my opinions with Android are even vaguely representative. I'm interested in what your network think is wrong with Android Lollipop, and without any sarcasm or snideness can I please ask who your network represents just to get a sense of how representative it is?

From a user who's devices just all upgraded a short while ago, it has all the usability problems of the new google maps. Things are more animated/transition-y/slid-y, but hell if I can figure out what to push or where a desired setting is.

It looks great, but it's not great for usability.

Can you elaborate more? Since I'm finding Lollipop significantly more usable in pretty much all respects - UI is more thought out, runs faster, battery life has significantly improved, dead Wifi detection is more reliable, timed audio profiles are a godsend...

Can you be a bit more specific? Are you perhaps using a Samsung TouchWiz device, where Samung pretty much removed most of AOSP UI improvements?

Here's a quick example, on my shield tablet, when videos fullscreen, the button to make it go full-screen/embedded disappear for no reason after a few seconds when it's full screened, meaning to escape a full-screen video, I have to hit the home or back buttons. This is a new behavior and super annoying.

On my phone, under settings, there's no obvious button to push to change wi-fi settings. There's a skeumorphic switch to turn it off and on, but no obvious button. In fact there's no obvious button for all the settings, just the label and an icon, but apparently if you push the label for Wi-Fi it's actually a button, but there's literally no affordance that it is such a thing.

Strangely, when you dive deep and view the app info for an individual app, there's buttons and pushable things everywhere.

The new google apps are a mess as well. Here's the process to change users in the gmail app

1) hit the hamburger menu (what could be there?) A panel slides out, which for some reason doesn't go all the way across the screen, because I need that 20% of my inbox that shows me nothing at all to stay visible

2) I have 3 colored circles with faceless people icons, they switch between three of my accounts, where are the others? I have no idea which one is which, so I push one, get dragged in the inbox for that one, but it doesn't tell me which one I'm in until I hit the hamburger menu again.

3) I'm of course not in the right one, and the one I want to be in is not one of the three choices. Where to now?

4) I see a list of folders, labels, and other crap, does it scroll? Apparently it does (oh, and apparently settings is all the way at the bottom of that scroll list, that will be easy to find) but there's no indication that it's a scrollable thing.

5) Nope, no list of accounts to choose from. Where to now?

6) Oh...the name of the account that I'm currently in is a thing I can push (again no affordance) and if I push that for some reason the little down arrow triangle next to my account name turns to face up and a list of user accounts appear. I thought the arrow was just telling me to look down for stuff about my account. In other GUI metaphors, a down arrow means a selection list is unrolled, so it turning up means that something should have rolled back up. Why do I push my account name to find other account names?

Absolutely a mess.

I could go on, maps is something I use almost daily and it's a similarly painful experience. It's like google just said "fuck it, bury it all under hamburger menus". But didn't google get rid of the menu button on new android phones to force designers not to bury stuff under the equivalent? Remember when android devices had a built in menu button and a search button? Now that garbage clutters up every app screen and lazy designers just bury stuff under them. At least with the menu button I knew what to press to look for options. Now I can't even figure out what's a button, and the buttons they do show use faceless icons that mean nothing.

The new chrome also, with it's "let's have every web page be an activity on the app stack" is terrible. Now I have to thumb through a pile of long-since closed apps to find the web page I was reading yesterday. Thank goodness somebody had the good thinking to revert that nonsense.

It goes on and on, hiding needed interface options under 2 or 3 deep layers of hamburgers, making buttons invisible, weirdly out of place skeuomorphic sliders, removing affordances...hell, I have to hit "agree" every single time I turn on the location finding thing to consent to location sharing. EVERY SINGLE TIME.

They've completely lost the plot.

> In fact there's no obvious button for all the settings, just the label and an icon, but apparently if you push the label for Wi-Fi it's actually a button, but there's literally no affordance that it is such a thing.

This annoys me. I see what they were trying to do by giving you more options but the touch target for small for the label button that I fat-finger this a lot and end up doing something like turning off Wi-Fi.

> It says quite a lot that the most exciting thing about Android for I/O is the rumoured stripped down headless version for Internet-of-Things devices, which may become an accidental foundation for a cleaned up future Android proper.

Can you point me somewhere I can read more about this? i.e. not just the Google serach results, but any specific reading if you have suggestions. I'm an Android developer working with a company which is doing some very exotic things with custom hardware, and would kill to have something like this.

Why is Lollipop a disaster? It introduced many great things.

Like memory leaks. Greatly problematic.

Is this an example of Poe's law?

Android can run on $40 phones and therefore widely used .To me that's the most exciting thing about android ,that's almost as huge as what Microsoft did in the 90s ,you know making actual impact. There are more people talking to Google voice than Siri

There are many more people talking to Siri today than used Microsoft software in the 90s. Apple sells more iPhones in a quarter than the annual global sales of computers in 1995.

Mobile is big enough for at least iOS and several flavours of Android. Speaking of which, those $40 Android phones don't come with Google services do they, so is anyone talking to them?

Why are you comparing today's Siri with Microsoft from 1995? The comparison is between present day Google and Apple.

I am sure Android helped them collect data to improve their Maps, Google Now, retain people on Google search, Gmail, Drive.. and to some extent even Google+.

Google's biggest strength is their huge data and Android definitely helps them there.

> About one of every two computers sold today is running Android.

So, like, half of them?

This article takes a rather strained and artificial angle: It ignores revenue from Google getting to place Google Search, Google Maps, etc. on a billion devices each year. Then, it attributes Google Play revenue to Android. I can buy movies from Google Play on my PC. Profit from media sales are probably greater than profit from app sales.

It does not ignore Search or Play:

> ...Google collected about $11.8 billion on mobile search ads in 2014, with about 75 percent coming from ads on iPhones and iPads... 2014, Google Play sold about $10 billion in apps, of which Google kept about $3 billion

Yes, I can buy Google Play media on other devices, including my iPhone, but surely non-Android Google Play revenue is a rounding error.

The claim about "a billion devices per year" is false - around 30% of Android sales are forked AOSP, which Google does not allow Maps on. The concern is that these forked variants are most popular in growth markets.

My main problem with android is the play store. Its so heavily localised with crappy apps and many bad attempts at trying to find the right things for me. I miss out a lot.

Come over to the iPhone side of the fence where...

no, we're flooded with crap too. Tons and tons of it.

I like the walled garden, I like the idea of curation. I just wish Apple would really enforce them. It takes a single search in the iOS store to easily find stuff that is either illegal or just copied junk.

Google has announced a new search algorithm during I/O. http://www.androidpolice.com/2015/05/28/io-2015-google-is-wo... Clustering the results by topics sounds like a good improvement, at least in theory.

There really needs to be a way to filter apps that have "in-app payments".

While some legit apps have them for upgrade, etc., most are Skinner boxes that are only relevant for a very specific demographic.

I agree. How can google of all companies be bad at grabbing the results I want from my search?

I like the part about iOS representing the 75% of mobile ad revenue for Google without even linking to the source

Yes, some analyst at Goldman says something in a report and now is an stated fact like I'm saying in some other sites.

Reply on news.ycombinator.com