When Android Wear came out over the course of last year, Google promised that the young, new platform would receive updates “early and often“. While it wasn’t said with so many words, it’s easy to read between the lines: Google was going to make sure Android Wear users wouldn’t face the same headaches as Android users when it comes to updates. Wear would be a more tightly controlled platform, built in such a way that updates could go straight to users’ devices without meddling from carriers or roadblocks thrown up by crappy customisations.
Fast forward to June 2015, and Google has recently released Android Wear 5.1.1, which, despite its humble version number increase over 5.0.1, is a pretty significant update to the smartwatch platform. It enables WiFi on devices that support it, adds new ways to interact with your watch, and makes it easier to launch applications. All in all, it looks like a great update.
Sadly, I can only go by what others have told me, despite owning the poster Android Wear device – the Moto 360.
Let’s back up for a second and look at what happens when a new version of Android is released. Android development takes place behind closed doors at Google’s headquarters. Somewhere late Spring, during Google’s I/O developers’ conference, the company does the big unveil and shows the world what the next version of Android will look like and how it will make all our phones faster and better and smarter and other PR blabber. We all ooh and aah, we get excited, we want it.
Then, usually somewhere in the Fall, Google actually releases their new Android version. However, in the context of Android, ‘release’ doesn’t mean what it means, for, say, an iOS update. Instead of Google sending out binary packages to all eligible devices, the company just dumps a huge ball of code in Android’s git repository, after which each and every Android OEM has to dive into the code and ‘port’ it to their devices, test it, get it past any possible carriers, and then, and only then, will it eventually end up on your device.
There’s a lot that can go wrong in this many-step process, and not entirely unsurprisingly, it often does. We’ve had countless examples of enthusiastic OEMs announcing they’ll update their One Galaxy Compact Z6 Active or whatever to the new Android release, only to have them backtrack with their tail between their legs because it turns out they just can’t get the new Android version to run well on said device. Customers mad, salty quips on Apple blogs, hugs for my Nexus 5.
Presumably having learned from its mistakes, it was implied that the process would be different for Android Wear. Google would take more control over the update process, to prevent a repeat of the Android mess. Tighter restrictions on hardware and software, all to ensure that Android Wear users would get the latest and greatest version delivered within days after announcing it – not months or years later, or not at all.
And yet, here we are. I’m looking at my Moto 360, the poster device for the Android Wear platform, and it’s still running Android Wear 5.0.1. Not a single user of what is by all accounts the most popular Android Wear device is using the latest version of Android Wear. What’s gone wrong? Why are Google’s and Android’s most loyal users being left out in the rain?
Strip away the political correctness from Motorola’s official Twitter reply about this matter, and the answer is clear.
@davidcstevens_ Performance is not quite there yet. We’re working with Google to ensure the software works great on #Moto360. Thanks.
— Motorola Mobility (@Motorola) June 3, 2015
If you translate this from PR speak to plain English with a firm dose of Dutch Directness, you get something very depressing: We got Google’s code drop but uhm, we have no fucking clue how to make this work on the Moto 360.
Plus ça change, plus c’est la même chose.
To be fair, other Wear devices have been getting the update, so I’m assuming the tighter controls have had some positive effect in that the situation isn’t as bad as it is on Android, but that’s not saying much. With only a very small number of different Android Wear devices out there, you’d think Google would be able to ensure that all of them get updated simultaneously, but it’s clear they can’t.
The tweet from Motorola makes it clear that the update process for Android Wear isn’t all that different from Android. It’s still just a code drop that device makers have to port to their devices, while Google flutters on towards the next shiny project, leaving OEMs to their own devices. Even when starting anew, Google seems unable to address Android’s biggest weakness.
A recent in-depth article at Re/code about the sorry state of Android One gives a glimpse into the culture at Google’s Android division that gives some insight into why Google is having such a hard time with this.
Android One has some interesting parallels to Android Wear in that both of them represent a clean slate of sorts, as well as Google trying to retain more control over their platform through limiting hardware and software diversity. And, just like with Wear, Google is facing problems with updates, supposedly the initiative’s big strength.
And Google failed to foresee the internal stress of the feat. Central to the company’s mission with Android One is getting more devices on the latest version of its software. But the hardware manufacturers have little incentive to help this – they sell the phone and move on. For engineers working on Android, coordinating the updates across all the handsets, in seven very different markets, is taxing and undesirable. One person familiar with the matter described it as a “thankless job.”
I think that’s an understatement. Imagine being tasked with solving the Android update mess. First, you’ll have to find Google employees willing to work on this problem, pulling them away from new and shiny stuff for the next Android release, and coaxing – or ordering – them to work on something boring. After fighting internally, you’ll have to fight the OEMs. Timely updates are not in OEMs’ interests at all – they want to sell new devices, not work on old ones they’ve already sold. After fighting the OEMs, you’ll have to deal with the endless list of carriers in every single country, and get them to sign off on updates.
Even Mike Rowe would pass on this one.
And yet, despite all the roadblocks and hardship such a job would certainly entail, I remain convinced – as I have been for years now – that this is by far the biggest problem Google is facing with its smartphone platform. All Google’s progress on performance, design, and developer tools is for naught, because effectively nobody gets to profit from them. I can profess my love for Material Design until the cows with pink hats come home, but what’s the point when almost a year after its release, less than 10% of Android users are enjoying it?
There’s no denying Google is getting better at design faster than Apple is getting better at web services, but this has never been Android’s biggest problem. I would much rather have every Android user using a Holo Android 5.0, than a sliver of Android users using a Material Design Android 5.0. What am I going to buy when my Nexus 5 dies? The Nexus 6 is too big and too ugly, and I’m not spending €800 on a One Galaxy Z4 Whatever, only to have to worry if it’ll ever be updated to Android M or Android M+1.
This situation has gone on long enough now, and it’s time we, as users, put our money where our mouth is.
Meanwhile, every time Google talks about Android – such as during I/O – I’ll be waiting for that big ‘…one more thing’. That one moment where Sundar Pichai drops a silence, looks at the big crowd of eager Android developers, throws a short, charming smile at the camera. “Let’s talk updates”, he says. The crowd holds their breath. Millions of livestream viewers perch forward. Dozens of livebloggers stretch their fingers over their MacBook keyboards in anticipation.
And how do you suggest we do that if we, quite literally, don’t have an option to vote for? Should I buy a Nexus 6 even though I find it to be far too large and overpriced on the magnitude of an iPhone? Should I go back to the iPhone, which I can’t stand, or go to Windows 10 Mobile where almost none of the apps I use exist? Blackberry? don’t make me laugh. I’d love nothing more than to vote with my wallet but, save for buying a used Nexus 5 (which won’t make any sort of statement) I do not have anything to put my money into that works for me. In the end I had to compromise and go with a Moto G 2nd gen (and by the way I still don’t have the 5.1 update). The only way we can vote, I guess, is to never buy a phone from any carrier, but I do that already as a matter of course to avoid bloatware and expensive long-term contracts. Somehow I don’t think carrier or no carrier sends a message to Google, or the phone OEMs. And don’t give me any of that custom rom stuff. I know about it already. We’re talking about voting with our wallet here, and what rom you install sends zero statement to anyone. Now, if we start seeing a large string of Cyanogen phones hit the market and we get updates on those no problem, then I’ll cast my wallet vote there when it comes time for a new phone assuming they’re not gigantic.
Size, price, and overall fugliness aside, I’m not even sure if the Nexus 6 is getting updates as timely as the Nexus 5. I recall reading that Google went back to partnering with carriers for this device (mainly because of the steep price). I assume those that bought it outright directly from the Play Store will get timely updates, but what about those who got it from carriers?
Get a Jolla mobile! Android application support PLUS you get official system updates BEFORE they’re even released!
…Sorry, I know you asked me not to make you laugh but I just couldn’t help myself.
But seriously, I love my phone.
Even if that would work for me, I’m in the states. Damn hard to get them here, if they even support our frequency bands.
The Moto 360 might not have gotten the latest, but it had pretty weak hardware right out of the gate. The LG G-Watch which came out before is running 5.1 now, and so is the Samsung one that came out at the same time. Just because your 360 isn’t updated isn’t strictly Google’s fault, but Motorola failing in speccing it correctly (and hyping the damn thing as *the* watch to have). LG did a way better job in providing more future-proof hardware.
I never liked the 360 (flat tire, booo), and rocked my G-Watch for quite a while, and it works fine to this day. I gave it to my wife and caved in to buying the Urbane. Yeah, specwise it is the same as the G-Watch R from LG, but it is a thing of beauty.
Yeah, it’s hilarious that he keeps calling it the ‘poster Android Wear device’ when it was clearly underpowered to begin with. That’s what you get for going with looks above all else, Thom.
Edited 2015-06-12 17:21 UTC
This is why I moved to the Apple eco-system. While it has downsides, I’m at least “current”.
Please fix this Google, and I may come back…
I’m in the same boat. Lack of consistency and stability with Android, combined with Microsoft’s “we forgot how to release a flagship Windows Phone” has pushed me back onto the iPhone for the first time in several years. And to be honest, it has improved greatly as a platform. So far I’ve been pleasantly surprised. Still, I wish I had held onto my Nokia N900; for what I paid for this iPhone I could have put in a pre-order for a Neo900 board.
iOS was always superior to android and still is. iOS 9 will be out on a 4s. The 4s was released in September, 2011.
Out of the hundreds of devices, I doubt one single android phone released in 2011 is running Android 5. Prove me wrong.
Buy a new watch Thom to get your update. A fool and his money are soon parted.
That’s highly subjective, and not at all what I was saying. For me personally, right now iOS is the right tool for the job. Previously it was Windows Phone, and technically still would be if my phone hadn’t died. Microsoft/Nokia for some reason has refused to release a really nice WP device for two years now; I’m assuming they are holding out for Windows 10 Mobile but I couldn’t wait that long, and I wasn’t going to touch the HTC One M8 Windows with a 10 foot pole after briefly using the Android version.
I went back to Android late last year, but even Lollipop didn’t fix the long-standing issues that make using an Android device as an actual phone possible for me. The dialer still crashes, the interface is even more obtuse, and the only way to have a synced contact list is to use a gmail account. No thanks. After trying out my sister’s old iPhone 4s with iOS 8, I was sold on going back to iPhone for the foreseeable future.
with apple, what greatly matters is also when was the END of device sales. iPhones are pushed byApple on consumers obscenely long (I think I’ve seen ads for that 4s quite recently)
Very interesting (and depressing) post.
I’m not a developer, but I gather that Android development is very much a closed-development ecosystem, despite its guise of openness. As a matter of fact, it would appear that there are actually *two* layers of closed development in Android compared to Apple’s one layer. In the case of Android, Google develops a new Android release behind closed doors, throws a ball of code over the fence, and then the OEMs start a new round of development/modification/integration/uglification behind THEIR closed doors.
I honestly don’t understand why Android devices aren’t upgradeable in two big atomic chunks: an underlying kernel/hardware layer, and a frontend GUI that can be updated apart from the underlying OS. With desktop Linux, I can install/remove any number of desktop environments on any of my very different laptops, despite their diverse hardware components, without touching the kernel or the underlying system. This allows me to run cutting edge Linux on systems that are more than a decade old. So why can’t Google let the OEMs provide the underlying kernel tweaks and drivers specific to their hardware, while Google just releases a new GUI with new userland features?
Edited 2015-06-12 12:57 UTC
Actually, this is not true. There are packages that rely on more up-to-date libraries and you just can’t install them without upgrading also the libraries. It, actually, may trigger a cascade effect at which point your best option is to upgrade to a new version of your distro of choice. Been there, done that, don’t botter anymore. It the thing is too old to accept a new distro I just give it to someone and buy a new one. Didn’t have to do it in like 4 years, though.
I think that with distributions such as Gentoo, Arch or Slack you can run cutting edge stuff on older hardware. It depends on what parts you want to be cutting edge. You may not be able to have btrfs as a file system, run the Gnome 3.14 desktop environment or use the latest cutting edge graphics applications on your 10 year old laptop. But perhaps that old beast is perfectly capable to run the latest kernel, vim or python.
That is one reason why my hobby development on Android is done in C++.
My hobby development is mostly geared towards gaming, so although constrained, I can survive with the NDK.
Additionally, by using external C++ libraries I can enjoy the same APIs, regardless of the Android version and also port the code to iOS and WP devices.
There is a price to pay though, as each APK tends to be much bigger than a Java only approach, plus one needs to either replicate UI Widgets or use JNI.
There is more to an operating system than just a kernel, device drivers and a gui. I think this is where you’re misunderstanding lies.
With that division, I think everything you’d care about that you didn’t name would be in the “GUI”. And we’d still be stuck in the same place we are now. Can’t just update the “GUI” without consequences.
When FirefoxOS was shiny and new, they tried to do it right.
FirefoxOS had a bottom layer, the kernel and video drivers.
And a top layer, the runtime, renderer, GUI and system apps.
Mozilla tried to make it so that the Mozilla would do updates of the top layer and the botton layer would be handled by the manufacturers or providers.
But they just couldn’t get manufacturers and providers to go along with these ideas.
Now remember that the Android situation is much worse.
Maybe it’s just me, but if FirefoxOS with their great plan can’t succeed. How do you think Android is every going to succeed with getting this, right ?
Edited 2015-06-12 13:09 UTC
The reason we have the situation now is… why upgrade the OS and breathe new ‘life’ into older devices when they can get people to sign up for a new contract and get a new device?
That’s been my whole problem with the smart phone ‘revolution’ in general. It’s as bad as the computer industry not too long ago, where software would continuously be coming out but they’d require the latest hardware to run decently so you’d end up having to upgrade your computer every year or two. Now a 5 year old system can still be useful, which is why there are so many “Oh my god, PC sales are down, no one buys PCs anymore, they’re dying!’ uhm, no, they don’t sell as much because no one needs a new one constantly.
But as far as alternatives to the big two… Damn Jolla, why did you have to skip the USA… I’m still looking forward to my Jolla Tablet
Don’t get me started on that. I just hate it when smartphones can’t have their batteries replaced and so on.
Let’s just call it what it is: consumer economy and planned obsolescence.
These are just concepts that were created by people and didn’t exist a long time ago. The modern consumer economy was created by economists and modern advertising was created by psychologists. Don’t be fooled.
I don’t know which is worse. The fact that it was even invented or the fact that it works on most people.
Don’t get me wrong, consumerism can be good for the economy.
It’s really all very crazy when you think about it.
Let’s say the economy grows with 2% per year, every year. That means a doubling time of 35 years. It really means the economy is producing twice as many goods (denoted in money, GDP). And when it’s below 0%, it’s a recession, everybody panic !
But it’s all ‘controlled’, at least they try. For example the government ‘controls’ inflation. Their ideal is actually that same 2% I mentioned above.
So that means, the same dollar, pound, whatever over 35 years is worth exactly half.
So that mean that 2% and 2% is actually 4% ? So a doubling time of 17.67 years ?
Let’s say you are buying bread and in a little over 15 years the prices have doubled ? Could that be true ?
If not, I’ll have to watch Dr. Albert A. Bartlett lecture again:
https://www.youtube.com/watch?v=u5iFESMAU58
Instead, it would have been better if Motorola had pushed out a half-baked update that performed poorly or destroyed even more battery life? Just so you could get the update at the same time as everyone else?
Or perhaps ALL the other smartwatches updates should have been delayed because the 360 needed more optimisation?
Very funny and sarcastic article and I totally agree with everything you said.
Apart from security fixes, who really cares for updates ?
My Nexus tablet got the update from 4.4 to 5.0+, and I really preferred the KitKat version.
There is the strange expectation than it is acceptable to buy half baked products with the assumption that free updates will be available shortly. This is particularly true with video games.
Were updates expected on mechanical watches?
Google is perhaps the posterboy of “devops”, a constant churn of developer “shiny” over actual products.
It is a term originating in web development, supposed to symbolize a closer collaboration between ops (or sysadmins if you like) and developers.
But for my point of view at least it seems closer to devs gets a big club to hit ops with until they ok the use of some shiny new framework or language on production servers.
And on web you never “release” you just iterate. You write some code, push in front of people, and if something breaks you write and push more code. You don’t do stable, maintenance, or anything like that.
And the web attitude is spreading to other parts of the IT world.