Google’s iron grip on Android: Controlling open source by any means necessary | Hacker News Reader

Google’s iron grip on Android: Controlling open source by any means necessary

arstechnica.com

414 pointscoloneltcb225 comments

Most of the examples of Google closed apps that are not part of the AOSP release are in fact apps that are based off of Google data-center services. Would it really help Samsung if the source to the Gmail app was open? Since Google controls the server side, and the client-server protocol, it limits the amount of innovation they can do.

The web equivalent would be like claiming that Chrome OS isn't open because the source to Gmail isn't available.

Google is stuck behind a rock and a hard place. If they don't try to create incentives for a unified experience, they get bashed for encouraging fragmentation, if they do assert a level of control, they get bashed for not being completely open.

And, in fact, the same author made the exact same point about Google Play Services being a big tool Google is using to get past OS version fragmentation about a month-and-a-half ago:

http://arstechnica.com/gadgets/2013/09/balky-carriers-and-sl...

The author was all-praises on Google for its actions.

"Not having to package everything into a major OS update means Google can get features out to more users much faster and more frequently than before. ... This should all lead to a more unified, less fragmented, healthier Android ecosystem."

> Google is stuck behind a rock and a hard place. If they don't try to create incentives for a unified experience, they get bashed for encouraging fragmentation, if they do assert a level of control, they get bashed for not being completely open.

This is exactly the position Microsoft was in during the mid-1990s, even before they went after Netscape. The article is saying that Google is even applying some of the same strong-arm tactics to keep OEMs in line. Fortunately, Google can point to Apple and say they're not a monopoly the way Microsoft was in its market. Plus, Android still looks really open compared to iOS.

I don't think Apple will be a good defense if Android's market share is 80%.

That depends on how the market share of revenue looks as well, and to my knowledge the percentage of revenue Apple receives is not dropping nearly as fast (and Google's not the main competitor there, Samsung is, even if it's using Google software).

80% does not a monopoly make. Microsoft was in an entirely different position at the time.

Percentage of market share is not a clear indication of monopoly. From what I could find, 50% is typically the bare minimum to warrant consideration, and 70 - 75% will get more serious scrutiny. Also, it differs from country to country. In the UK, Tesco was investigated even though it had a market share of only 30%.

I also recall reading one of the EU's regulators (was it Joaquin Almunia?) say that (paraphrasing) "we start sniffing around when any one party gets a market share north of 60%." Unfortunately, I cannot find that article.

>Percentage of market share is not a clear indication of monopoly

True, and the reason is that a 30% UK-wide or a 50% Europe-wide market share can easily turn out to be a 100% market share in many local markets. Tesco defended itself by saying that 98% of consumers have access to five different super market chains within a drive time radius of 30 minutes. Make that a more realistic 10 minutes and the picture changes dramatically.

How can anything other that 99% market share even be considered a monopoly? If it's 80% or 90% then there is obviously choice and thus no monopoly.

Anti trust is such bunk...

Here's an example:

Let's imagine a hypothetical world where Company A owns 80% of the smartphone market and the rest is divided between the also-rans.

Now imagine that app developers can't profitably make a 1st rate app without getting the income from company A's app store (But can port the app to other OSes for extra income). If company A starts to use this situation to make it harder to port apps to other companies OSes via strongarm tactics, then company A is abusing its monopoly position of being the only profitable way to make an app.

Today's world is nothing like that, but it's an example of how near-total marketshare is not necessary for anti-trust to come into play. There is a certain point at which you have a large enough market share, that certain other players in the market must deal with you to get the volume needed to compete, particularly in high-volume, high-upfront cost industries (which many types of software qualify as).

Monopoly is not antitrust. Abusing control of a market is. A monopoly (which is usually considered to be dominance if a market place) is not itself unlawful.

Anti trust is the legislation policing monopolies, at least in the US, and it is such legislation which I consider bunk.

Antitrust laws exist to enable greater competition by stopping businesses behaving in an anticompetative manner. They don't necessarily exist for "policing monoploies" anywhere in the world.

>"and it is such legislation which I consider bunk"

Why? Don't you think that Microsoft should have been reprimanded over their abuse of the OS market to gain control of the browser market?

Or is it because it's being held over $favorite_faceless_corp like the proverbial sword of Damocles?

To be fair, Microsoft also created a browser that, for its time, is pretty widely accepted (even around here!) as being much better than the Netscape counterpart. So while they did throw their weight around, it's not clear that their browser dominance was purely because of their OS monopoly. There were many other ways that Microsoft was leveraging its monopoly which deserved antitrust scrutiny, but I think bundling a browser was the wrong thing to be prosecuted for.

I agree totally that IE 4 was a browser ahead of it's time, however Microsoft were shown to have needlesly integrated it into their OS and then told OEMs that their Windows licenses would be revoked if they sold PC with Navigator pre-installed. I'd say they were lucky not to be broken up!

The problem there was that Microsoft had already signed a consent decree with Janet Reno's DoJ which specifically allowed them to expand the OS but forbade them from tying.

However, Microsoft also rewrote IE into componentized form, so that components (eg the rendering engine) could be used by other applications, including third-party programs. This was considered a Good Thing at the time (it won the deal with AOL, for example) and made integration (as opposed to tying) a reasonable claim. This is, of course, independent of the way Microsoft handled the argument in court, which was very, very badly.

In sum, whatever Microsoft did was not independent of the actions of the US government, whose interference had extremely bad results for consumers. This includes the crapware explosion that resulted from the DoJ's removal of Microsoft's power over OEMs.

>Don't you think that Microsoft should have been reprimanded over their abuse of the OS market to gain control of the browser market?

Absolutely not!

Why?

It's their OS, they can do as they please.

As history showed, there wasn't even an MS monopoly in existence. OSX became viable in the early oughts, Linux was always around, and now people write of the death of the PC.

Regardless, abuse of market dominance (note I don't use the overused 'M' word) increases incentives for competitors to enter the marketplace.

It's got nothing to do with monopoly!!! Apple or Linux have nothing to do with it; the ruling being for behaviour in the period that preceded 2000. Microsoft were in a dominant position in one market and leveraged this dominace to gain dominance in another. This is anti-competative. for the final time; a monopoly is not illeagal in and of itself, neither is it antitrust.

Looks like someone didn't read my post :)

Looks like someone isn't reading anything!

There's actually no evidence that your claim about leverage is true. Microsoft simply built a better browser and marketed it better than Netscape.

It's also a fact that the DoJ lost the browser case: it was overturned 2-1 on appeal.

This sounds like an antitrust issue:

"This makes life extremely difficult for the only company brazen enough to sell an Android fork in the west: Amazon. Since the Kindle OS counts as an incompatible version of Android, no major OEM is allowed to produce the Kindle Fire for Amazon. So when Amazon goes shopping for a manufacturer for its next tablet, it has to immediately cross Acer, Asus, Dell, Foxconn, Fujitsu, HTC, Huawei, Kyocera, Lenovo, LG, Motorola, NEC, Samsung, Sharp, Sony, Toshiba, and ZTE off the list. Currently, Amazon contracts Kindle manufacturing out to Quanta Computer, a company primarily known for making laptops. Amazon probably doesn't have many other choices"

Yeah, that's a serious problem for me, essentially using their leverage with the Google platform to keep competitors out of the market. Not cool Google.

I once saw an economics paper finding that if the top four players in a given market have more than 60% marketshare between them, they'll naturally act as a cartel. Ergo preserving an effective free market for consumers requires preventing that circumstance.

Assuming the consequent...

Hasn't that always been the relationship between Microsoft and Apple in the desktop world?

I read the article and kept hearing echos of the Microsoft strategy of the 90s as well.

History may not repeat, but it rhymes.

I think that you hear whatever you want to hear, but the truth is very different: Google's apps and APIs are just cloud endpoints and weaving them into AOSP doesn't make sense and just delays updates, as the same author previously scribbled: http://arstechnica.com/gadgets/2013/09/balky-carriers-and-sl...

As for "strong-arm tactics", what is referred to is explained in this post: http://officialandroid.blogspot.com/2012/09/the-benefits-imp...

As Andy Rubin puts it, Google will not encourage non compatible forks, it is in their interest to have Android developed apps run on all Android devices. Anyone can still have a go but Google won't encourage it.

It's a very flawed and overwritten attack piece, seemingly coming out of nowhere.

I think the strongest argument the article makes is the claim about the new Location APIs. Location algorithms are not dependent on Google services and have always been part of the core APIs, but now Google is moving new location algorithms into their closed source app. It's an obvious power grab, they want AOSP to be less and less useful on its own.

Is it really that obvious? What power are they grabbing? By this reasoning, Google may not augment any 'traditional' AOSP service with improved functionality. That's obviously ridiculous, so what solution would you envisage?

The article was pointing out how Google is moving more and more functionality from open source to closed source. So, one way to counter that would be to augment the "traditional AOSP service with improved functionality", but license the code as open source. The service endpoints can certainly be closed, but the code to access those endpoints could be open. And the license to use those endpoints could be open source friendly.

Now, that may not be a good business decision, but it certainly counters the argument of the post.

An aside: I appreciate what google is doing to fight Android fragmentation, but also see some of the tactics as anti-competitive.

Actually location APIs do use core Google backend services (the wifi geolocation database). Even Keyboard does (spelling). In a way they are doing the opposite, building their services into everything, which is a natural googly thing to do to add features/improve them. At that point they close source them as they dont want other people using the services without signing up.

The way to compete is either to make standalone apps that dont require services (ie like the old unmaintained ones) or to build your own services, either open or closed.

I think the location APIs may also use system-level facilities that aren't open to normal applications, such as the ability to detect nearby access points even when WiFi is off and the device is in standby. So it's not clear that apps can implement it themselves without relying on Google. Also, the APIs for newer stuff like geofencing are Google-proprietary and don't support alternative providers, so mobile phone makers that don't want to rely on Google's services can't substitute a different backend without breaking existing apps.

It does not use system-level facilities that aren't open to normal applications. People love to hate on Google; if they ever embed features into the OS only they are allowed to use you will hear about it very quickly.

There is a class called WifiManager[1] which lets you ask the user to allow background scans. It also lets you see the results of the last scan.

It is unfortunate that the Geofencing APIs are not part of the android platform, but I suppose it's possible they use Google services.

1: http://developer.android.com/reference/android/net/wifi/Wifi...

If they rolled the improvements out as part of the platform, it would inherit the slow rollout weakness as any other platform API (it would take months/years to become viable and unavailable on existing versions). If they went the static library route, it would balloon the size of of the apps and they wouldn't be able to do the same sort power/efficiency call coalescing as they would by having an always updated client on the device.

Saying 'iOS is more closed' doesn't make Android meaningfully open.

What about the fact that competitors are using the OS without so much as asking Google?

The fact that I can download and build Android is what makes it meaningfully open.

You can't download and build what consumers think of as Android.

Can you sent me a link to the current HEAD?

Right. Being able to fork the OS and release devices like the Kindle Fire do make Android meangingfully open.

Saying 'iOS is more closed' is just obvious.

I'm interested in how you think it works well for Google if they do nothing and allow Alibaba's Aliyun OS to market themselves as android. When issues do arise and incompatibilities in apps occur.. is this not the fragmentation that Google already is looked down upon ?

Android is open enough to replace all Google apps with your own (Samsung) and open your own appstore (Amazon). Also companies are completely free to not pick Android (Tizen / Firefox OS / ?) and they are free not to license the Google apps and put their own or (other) open source ones. It would be hard to make Google out to be a monopolist via that route.

MS was different; there weren't much alternatives and the alternatives which were there were squashed by MS. People currently want the apps in the appstore; they want to play clash of clans; they don't care about the Google gmail app. This being the same bubble world as the chromebook thread on HN yesterday; tech people think non tech people actually notice what they are running; they generally don't. If they can play the games their friends are playing and if they can use 'the software' they are used to they are happy. What brand it is is not important.

As an example; when I sit with any of my family members (they are all non tech), they will say 'I will open up Word now' or ' I will open up Excel now' to me when we need to organize something or go over numbers of one of the companies. What pops up definitely never is Word or Excel but rather Libre Office or Google Docs or some free Android/iPad variety. No-one I know actually has or uses MS Office; they use the terms because they don't know 'spreadsheet' and 'word processor' is a mouthful. They don't miss Windows and would even mostly hate it if they had to work with it now (after tablets or chromebooks and even Macs, Windows for non-tech people Windows seems incredibly hard and tech to use).

All these alternatives and Android being deployed by many different companies in different forms would make it hard to call Google a monopolist on that grounds. Samsung could turn into one though.

Did you even read the article?

And if you want to stay on the technical side.. then what about contributors? I've helped port android to a couple devices. I had no idea google had a contract obligation with hardware makers that my work should have to be used in one way or another. I feel dirty.

edit: the Acer example goes exactly against what you mention. They tried to ship a fork, with some of the substitutions you mention. google released the layers.

I read the article; if you want out of their grip you can. You just need to provide alternatives for the Google apps and appstore. That is not trivial, but for a company like Samsung that wouldn't be that big of an issue either.

Acer tried to release it while still wanting to be in the 'Android family' (Open Handset Alliance); they didn't have their own substitutes and didn't want to make a clean break with Google (OHA). If they wanted that and would have provided an appstore, they could've.

> This being the same bubble world as the chromebook thread on HN yesterday; tech people think non tech people actually notice what they are running; they generally don't. If they can play the games their friends are playing and if they can use 'the software' they are used to they are happy. What brand it is is not important.

At the same time, those people are first to complain about a single button located somewhere different from what remember. Unless all they are doing is the bare minimum that any of interface elements doesn't matter, they would continue caring about what operating system they would be running on. I guess problems are people between complete novices and experts -- they know more than some basics (to recognize what they are running) but not quite up there to mitigate the difference themselves. (And somewhat, this could be said true so far as there haven't been so much of UI functional changes happened up to 7 from the Windows 95 era. People around me haven't really exposed to themselves to Windows 8.x would do to them when they finally hit them...)

So how long have you been working for Google?

You're partly right, but in regards to the Gmail app, what I don't like about Android is that it doesn't come with a good email client that's not GMail. Android's GMail client is very polished and there's not much in it that's specific to Gmail, it could be a generic email client that works with POP3/IMAP/SMTP and other standards as well. The experience of users that aren't into Android's own services do suffer.

But I do not agree with people that claim Android is not open. Building an operating system is a big challenge and stock Android is enough for anyone to fork and do whatever they want. I actually play around with an old Galaxy S on top of which I installed Cyanogenmod without any of Google's Apps. The only problem is that developers only publish their apps on Google Play, which is a shame, given that Android does allow you to install apps from third party sources and you could have a good experience just with stock Android. I ended up installing Amazon's Appstore on it, which is not as good as Google Play, but at least they've got special offers :-)

> What I don't like about Android is that it doesn't come with a good email client that's not GMail.

That would be like saying that the default camera app does not come with filters, shaders and other features that you like. What Google does is, gives a good enough app for generic use and leaves options open for developers to make better things and put them on the Play Store.

If Google starts incorporating all such features, it'll get much more difficult for developers to earn through the store.

Yeah, but the email app on iOS is pretty polished and nowadays people expect the stock browser and the stock email app to be pretty good, especially on mobile devices. I'm not saying that Android's out of the box functionality should encompass everything, just the basic necessities, like email and web browsing. And I don't want people without an Android to get the wrong idea - Android does come with a standard email client, it's just that it's much, much less polished than the GMail app, which is a shame.

> I actually play around with an old Galaxy S on top of which I installed Cyanogenmod without any of Google's Apps. The only problem is that developers only publish their apps on Google Play, which is a shame, given that Android does allow you to install apps from third party sources and you could have a good experience just with stock Android.

I went through the same process of installing Cyanogenmod without Google Apps and it made me wonder why developers of free apps don't distribute them outside Play e.g. on their sites. I understand paid apps and/or apps that have in app purchase but there are a lot of just free apps that just can't be downloaded. Any theories?

Well one theory is that simply setting up a site that you can download an APK from is a hassle. It either costs you bandwidth if it's your own server, or has terrible UX if it is hosted via a share site.

There's always https://f-droid.org - it has somewhat vetted free-as-in-freedom apps that must compile from source.

Oh, come on, putting a 20 MB binary somewhere is cheap and easy.

I think developers don't put APKs anywhere else because they think they won't have an audience for those anyway. But it's a pity anyway. For example Humble Bundle packs with Android games were quite successful and the distribution method was through download links.

I agree that it is a pity. When/if I get around to developing Android apps, I will be definitely distributing packages that way.

However, I would like to point out that the Humble Bundle packs, while available via download links, also is available for download using a Humble Bundle app that acts like a package manager. I don't know how much that would change things, success-wise, but I thought it was worth noting.

> Android's GMail client is very polished and there's not much in it that's specific to Gmail, it could be a generic email client that works with POP3/IMAP/SMTP and other standards as well. The experience of users that aren't into Android's own services do suffer.

GMail uses a proprietary protocol that is decidedly not imap. It has features like colorized labels and inbox prioritizing that just isn't present in imap. Furthermore it has completely different performance characteristics (e.g it is much faster) than what imap email can provide. No one seem to have any details on exactly what the protocol is or how it works. They definitely seem to not want any third party clients accessing gmail though. Otherwise they would have published a protocol specification a long time ago.

> GMail uses a proprietary protocol that is decidedly not imap

A thousand times, this.

Using a traditional IMAP client such as offlineimap makes this painfully obvious. Gmail synchronizes labels as "folders", not "tags", for some unknown reason, which means that mail in the inbox ends up getting duplicated on disk under "All Mail", and then again for every single label attached to that label.

Archiving hijacks the way that deleting works, and while clients can adjust to this, it breaks the way IMAP is supposed to work. If clients have to add special cases to interact with your service, that means you're going off-protocol!

Even the login is different. On IMAP, your username should be "foo", not "foo@example.com".

foo@example.com is a perfectly valid IMAP username. How else could a email service provider host multiple domains on one endpoint? For example Rackspace does this through imap.emailsrvr.com.

Is GMail subject to the same task killer that 3rd party apps are? I suspect it isn't.

If that's true then Google has given itself a good monopoly for email on its own OS, much like Microsoft did for Internet Explorer.

Are you referring to how Android will reclaim memory from inactive processes? Android handles memory differently than you expect, it tries to keep thinks in memory and only kills apps if it needs their memory back for a more demanding process.

Google consolidating most APIs into Play Services would ensure it's constantly active and never purged from memory, but the same isn't true of Gmail or other Google Apps. They can and are purged from memory as other apps demand more memory. That being said, I've encountered very few situations where any app gets purged from memory. About the only time this happens is if I play a game like Galaxy On Fire or decide to open 100 tabs in Chrome.

Yes I was, glad to hear GMail is subject to the same rules as other apps. Not sure what you mean by "Google consolidating most APIs into Play Services would ensure it's constantly active and never purged from memory"

> Not sure what you mean by "Google consolidating most APIs into Play Services would ensure it's constantly active and never purged from memory"

A lot of common APIs (Location API, Maps API, Play Games API, Push Notifications API, etc) have been consolidated under the Google Play Services umbrella process. Since these common APIs are used by lots of applications, this would ensure the Play Services process never gets booted from memory.

It also has the added benefit that applications using push notifications like Gmail don't need to maintain a background service because they can rely on Google Play Services to activate them as needed. Location based apps like Yelp can determine your location more quickly because Google Play Services is constantly monitoring it. All while reducing the over all memory foot print of applications because they're all deferring to Google Play Services.

Google can also change their APIs for services, upgrade to new protocols like SPDY, and developers don't need to update their applications because Google just needs to maintain the interface to Google Play Services. It's pretty clever.

Your suspicion is wrong and really, really, out there as well. I'm surprised to hear someone thinks this.

>Most of the examples of Google closed apps that are not part of the AOSP release are in fact apps that are based off of Google data-center services. Would it really help Samsung if the source to the Gmail app was open? Since Google controls the server side, and the client-server protocol, it limits the amount of innovation they can do.

I don't think Google's development of their own services is issue here. The issue, in my understanding, is that Google ties its services to Android under "Android compatibility" label.

Skyhook is a great example (the article mentions it) - mapping is important for Google, and Google literally pushed the company out of business by strong-arming manufacturers to stop using Skyhook services.

Can, say, Samsung make a deal with Yahoo to drop gmail, include only Yahoo mail on their phones and still pass "compatibility" test in Google, and have access to Google Play store and the rest of the services? Can you imagine HTC phone with Nokia maps? (speaking of maps and Nokia, anyone is quick to point that Nokia made a huge mistake not betting on Android as Symbian replacement. And couple of months ago we saw that mapping is so important to Nokia that they're ready to let negotiations with Microsoft fail just to keep mapping in their hands. Does anyone thinks that Nokia would be permitted to make Android phones with Nokia maps?)

The fact that Google services are (mostly) better than competition isn't relevant here - back in the day Internet Explorer was way better than Netscape Navigator, but that fact didn't made MS actions any better.

> Google literally pushed the company out of business by strong-arming manufacturers to stop using Skyhook services.

This is what Skyhook claimed, but the reality was much different. Skyhook were intentionally or unintentionally polluting the Google AP database. They refused to change and so failed the compatability test.

Look into this more.

I think there is some legitimacy to this line of criticism, but as Amazon has shown, large organizations can produce their own forks.

I can imagine too that selective replacement of chunks of the 'Google Experience' might make consumers get a negative impression of the Google brand if the replacement has issues. Like if you replace the location with Skyhook or Nokia, and the new Maps app is just called "Maps", and if there are serious issues, consumers might say "Man, this Google Maps on Android sucks!" without realizing it's not Google Maps, because Android is strongly brand associated with Google.

There's also a logical rational for Amazon-style forking, in the sense that if you're going for a complete reskinning, the end result will likely be a lot better if it is completely horizontally and vertically integrated by a single vendor rather than cobbled together -- 'bloatware' experience.

I'm already disappointed by my inability to remove the (rather poor) Samsung apps from my phone, as I'd like to only use the 'proper' Google apps (for me, gmail/google calendar app is 50+% of phone usage). It's not enough for companies to be able to produce their own forks - the forks still need to be competitive or better than Google versions, and that's not so easy to do.

Aren't we forgetting about the closed-source elephant in the room - Google Play Services - the "platform behind the platform"?

>> Play Services has system-level powers, but it's updatable. It's part of the Google apps package, so it's not open source. OEMs are not allowed to modify it, making it completely under Google's control. Play Services basically acts as a shim between the normal apps and the installed Android OS. Right now Play Services handles the Google Maps API, Google Account syncing, remote wipe, push messages, the Play Games back end, and many other duties. If you ever question the power of Google Play Services, try disabling it. Nearly every Google App on your device will break. [Source: http://arstechnica.com/gadgets/2013/09/balky-carriers-and-sl...]

Google Play Services are not (currently) required for developing and running apps, unless those apps integrate with Google's services, like Google Accounts and Google Maps. Of course nearly every Google App will break if you disable it, since every Google App is proprietary and Google's Play Services is their common framework.

The question is - do third-party apps break if you disable Google Play Services? And the answer is that it depends. If the app needs Google's services then it will stop working. However, making Play Services open-source, at least for the functionality provided right now, wouldn't do any good. Because those services rely on Google's server-side infrastructure, so you're still at their mercy.

The only thing that bothers me is the Wifi-based location tracking. Google collects info about nearby Wifi networks and so the phone is able to give a pretty good estimate of latitude/longitude coordinates, even without turning the GPS on. Turning Google Play Services off means that Wifi-based location tracking stops working. And this is a pain, given that many apps these days (Facebook, Twitter) want to get your current city and so Facebook for example is accessing my GPS every time I open the app, with no way to turn this functionality off and it's consuming my battery and the GPS status notifier is annoying. In case you're wondering I had to turn Wifi-based location tracking because of a bug in the latest Android 4.3, as it is preventing my Nexus 4 to go to sleep, thus draining my battery.

> And this is a pain, given that many apps these days (Facebook, Twitter) want to get your current city...

Them wanting, I can understand. But why should you give them that?

Because currently it isn't optional, which is a flaw in Android's permissions system that sucks.

Three points:

1) Newer custom ROMs can deny permissions on a per-app basis. I understand this isn't for everyone, but I enjoy having this option very much on my own device. (I also agree with you that it should be baked into the vanilla Android, but I can understand why they wouldn't want to confuse users with that option.)

2) Some apps will give you a choice in their own settings (if you trust that).

3) You always have the option of just not installing an app if you don't like the permissions. There's a fair bit of apps I refuse to install/update because their permissions asking for way more than I think their feature set requires.

I've been running CM since the Nexus One. I don't think it supports per-app system privileges. Unless it is hidden somewhere deep within (though I could swear I've recursively been through their entire settings tree).

May I ask which ROMs you are referring to?

I myself use AOKP, although I believe the permission management feature was cherry-picked from Cyanogen, and Google searches seem to verify the functionality exists.

On AOKP, you have to enable it by going into System Settings -> Permissions, which will then let you manage your apps.

Cyanogen uses a similar system[1], you'll need to navigate to CyanogenMod Settings, then Permissions, and finally Enable Management. After that, permissions can be enabled or disabled through the Settings -> Applications -> Manage Applications menu.

[1]Those instructions were pulled from a Google search, as I don't run Cyanogen myself currently. They may be old/incorrect/etc. Good luck.

Cyanogenmod used to provide this in the Gingerbread days, but they removed it somewhere around ICS/JB.

It was rumored that this was under pressure from Google, etc., but I'm not sure if that was ever confirmed.

CM 10.2 introduced something called Privacy Guard. https://plus.google.com/+CyanogenMod/posts/gk7X3HjNvnH

> I also agree with you that it should be baked into the vanilla Android

They appear to be working on that: http://www.androidpolice.com/2013/07/25/app-ops-android-4-3s...

I used it to block Facebook from accessing GPS. Worked well.

> Some apps will give you a choice in their own settings (if you trust that)

I noticed recently that Google gives you the option of disabling "Google Location Services" which allows apps to read location when they're not running. However, disabling this causes Google Maps to refuse to read your location when it is running.

...And that's about when I started looking seriously at OpenStreetMaps. Sadly haven't found a good replacement yet.

As a practice, I turn off all location services on my phone and only briefly turn it on when I need to use Maps. May seem paranoid, but don't find myself suffering too much by this change.

Ah. I don't use Facebook, and I haven't installed Twitter's Android app. So I never faced these issues.

This is either becoming or already is the core of Android, and it's closed source and no one wants to talk abut it. It's ridiculous.

iirc Google Play Services becoming updateable was to work around carriers and manufacturers not giving android os updates to their end-users.

Hypothetically, if you made all parts of android rely on "Google Data center services" would moves to close it off be justified?

This is what Play Services are moving towards.

>off of

Learn English, and then we'll talk.

"Off of" has been an English construction since the 1500s. It's entirely reasonable and appropriate in this context, much like contractions in informal written English.

But I'm fairly sure I'm replying to a throwaway account, so it's not like you were seriously interested.

When someone like Github does this (make some parts of their code open-source, but others closed-source), journalists don't write critical pieces about them, do they? I mean, Google leaves a bad taste in my mouth since they started shuttering services like it was Christmas at the Google Service Chopping Block, but I don't see them being actively evil here.

It's all according to the previously openly aired plan. Google keeps all of the existing code open source. Anyone who wants to build a fork can do so. Now if they want a hardware platform to run on, go find one outside the Open Handset Alliance ecosystem. It's fair game -- if a hardware partner thinks that one of Google's competitors can provide a better Android fork, they are free to leave the Alliance and go partner with that competitor. They will still get an enormous amount of code for free in AOSP. They just won't get all of the services that Google is building specifically for its own version of Android. How is any of this maintaining an "iron grip" in any way? Just contrast this with Apple where it is the sole owner of everything to do with the OS and app marketplace.

AOSP is so open that the previous AOSP leader at Google quitted over the lack of openeness.

That's a mischaracterization of his own comments on why he left. It had to do with open sourcing GPU drivers, which is probably an issue with other vendors, not Google.

See: https://plus.google.com/112218872649456413744/posts/9HHRURor...

well actually while I get where you're coming from, that's still correct. if AOSP was more open he wouldn't have left. Google doesn't give a damn about openness right now.

"If AOSP were more open he wouldn't have left." Can you substantiate that statement at all? AOSP's "openness" is determined by its licensing, which is a standard Apache 2.0 license. The issue was whether certain vendors would contribute to AOSP under that license, not whether AOSP was open enough.

It was Google's choice to put that hardware into the phone, and it was their choice not to pressure qualcomm into being more open.

Your argument itself is also somewhat of a mischaracterization, because at least Google could have wrestled redistribution rights for the binary drivers, thus making AOSP actually usable on the nexus, and they didn't even bother with that.

Vendors are a convenient whipping boy when don't care about openness but wanna look like you do.

There's a very, very simple problem with your argument:

https://developers.google.com/android/nexus/drivers#razor

and here is the problem with yours, from the EULA:

     Subject to the terms of this Agreement, Licensor hereby grants to
     You, free of charge, a non-exclusive, non-sublicensable,
     non-transferable, limited copyright license, during the term of
     this Agreement, to download, install and use the Software
     internally in machine-readable (i.e., object code) form and the
     Documentation for non-commercial use on an Authorized Android
     Enabled Device and non-commercial redistribution for academic
     purposes only of a reasonable number of copies of the Authorized
     Android Enabled Device Software (the "Limited Purpose"). You may
     grant your end users the right to use the Software for
     non-commercial purposes on an Authorized Android Enabled Device.
     The license to the Software granted to You hereunder is solely for
     the Limited Purpose set forth in this section, and the Software

This makes it entirely impossible for AOSP to distribute the drivers

Blame mobile chipset vendors for that one.

> Anyone who wants to build a fork can do so. Now if they want a hardware platform to run on, go find one outside the Open Handset Alliance ecosystem. It's fair game -- if a hardware partner thinks that one of Google's competitors can provide a better Android fork, they are free to leave the Alliance and go partner with that competitor.

A good example of this is Amazon. They are doing this successfully.

Another example is OUYA.

And Aliyun OS And Baidu Yi May be even the newly funded CyanogenMod company

Google used 'Android is open' as a dishonest slogan to convince developers that it wasn't just a power play.

That is why they are being criticized. It's not about who is more open. It's that Google misled everyone about their intentions.

Prove that Android is not open.

He meant free (as in freedom), not open. "Proof" of that is in the submission's article. It's not exactly a formal proof but it has convinced most people here.

Unfortunately, this highlights some weaknesses in the GPLv2, which is supposed to guarantee freedom.

Android largely isn't GPLv2.

Anyone who wants to build a fork can do so

Not really. You will have your mainline Abdroid implementation shutdown if you do this. Google is clearly doing this to make forking a very risky proposition for device manufacturers. It's the antithesis to open.

> Now if they want a hardware platform to run on, go find one outside the Open Handset Alliance ecosystem.

Imagine you want to sell a Linux laptop with your own Linux distribution. You're just a small shop in California, and you're shopping around to find a manufacturer for your laptops. And you find that anyone connected to Microsoft (Foxconn, Asus, Acer, Gigabyte, etc) can't do any job with you. Would you say "that's fair, Microsoft isn't evil by forcing hardware manufacturers to work exclusively with them" in such situation?

That's not what they're doing. HTC, Samsung, etc still sell non-Google OSs, like Windows Phone and Bada.

A good analogy would be Microsoft preventing you from selling your fork of Windows on those manufacturers, which happens to be exactly what every big OS maker does, by simply not distributing it under a Free license.

> When someone like Github does this (make some parts of their code open-source, but others closed-source), journalists don't write critical pieces about them, do they?

Github don't constantly go on about how 'open' Github is, though.

True, but Github does use the halo of "Open Source" a lot to promote themselves (or others promote them using it..e.g. http://www.wired.com/opinion/2013/03/github/) while overlooking the fact that most of their codebase is very close sourced indeed.

> Google was terrified that Apple would end up ruling the mobile space. So, to help in the fight against the iPhone at a time when Google had no mobile foothold whatsoever, it was decided that Google would buy Android. And Android would be open source.

This is blatantly false. Google bought Android in 2005, two years before the iPhone was announced.

Almost exactly a year after they bought Android, Schmidt joined the Apple board of directors.

The iPhone certainly was in development some time prior to its announcement.

Schmidt joined the Apple board August 29th 2006: http://www.apple.com/pr/library/2006/08/29Google-CEO-Dr-Eric...

Public iPhone announcement was January 9th 2007: http://www.apple.com/pr/library/2007/01/09Apple-Reinvents-th...

Even with a conspiracy angle, a three month head start isn't much.

Android clearly went in a different direction post-iPhone.

Agreed.

All those videos of Android demo or prototype units clearly showed Android was a Blackberry clone (just look at that tiny touchpad for navigating around Android OS) and changed its design direction after iPhone made its appearance.

nonesense. Android is a software platform that companies can build sofware and devices on. Android is not a device, and it's not an "operating system" in the sense that people use this word today. It is a fundament that you can build an operating system with, and a cross platform that allows applications run on different software, as long as this software is based on Android.

If you want to show off what Android can do, you put it on all kinds of devices that people are using at the moment. If everyone has a Blackberry, you want to show that you can make a software for Blackberrys based on Android. If everyone is hyped on digital cameras, you show a software and ui for such a camera.

Android used to have 3 prototypes in its early stages: a blackberry type, a tablet-type device, and the device that later became the Tmobile G1, the first android phone. The platform never changed course, and it clearly never was a Blackberry clone.

Note that the HTC G1 didn't have a software keyboard at first; it was a (non-multi)-touch device, but you were meant to use the physical keyboard for all text entry until Android 1.5

"If you want to show off what Android can do, you put it on all kinds of devices that people are using at the moment."

I'm not so sure about that. Blackberry didn't do that, and neither did Apple...

Not sure I understand your comment. Neither Blackberry nor Apple developed a cross platform technology meant to run different hardware/software configurations. So of course they did not demonstrate how a platform runs on such configurations.

> Google bought Android in 2005, two years before the iPhone was announced.

Not only, as has already been pointed out, did the UI change, but if you look at the Android architecture it's clearly evolved along an very different path to iOS.

I remember when the first Android SDK came out the programming manual made it very clear to developers that we weren't to expect devices to always have discrete GPUs or touch-screens.

In fact, Google was concerned about Microsoft more-strongly binding Windows Mobile (Windows Phone came later) to their Web ecosystem. Microsoft was working on Windows Live Search, which would later be replaced by Bing.

At the time, Windows Mobile was the epitome of a modern mobile OS. It had a real(ish) browser and a VM-based app runtime for a modern (C#) language (albeit optional and not nearly as tightly integrated with OS middleware as in Android). Google saw that an ecosystem such as Windows Live is much more critical to a mobile device than it is to a desktop PC. The nature of Android more closely reflects the threat from Microsoft than from iPhone.

The only real thing that seems "evil" is the requirement for OEMs to not manufacture _any_ devices compatible with non-Google forks. The rest of it seems pretty necessary in order to keep carriers and OEMs in line. A lesson Microsoft learned, and why Windows Phone started off by allowing the user to remove any pre-installed crap.

If Google didn't do any of this, and was totally altruistic, Samsung and others would already have completely screwed things up.

While it's certainly very much to Google's benefit, it also benefits most users because overall, Google has done a far better job than any OEM regarding user experience.

> the requirement for OEMs to not manufacture _any_ devices compatible with non-Google forks

Is that really where the line is? My understanding is that it is almost the opposite: an OEM cannot manufacture any devices that break compatibility with AOSP, not support compatibility with a 3rd party fork. And the constraint is a condition of membership of AOSP rather than something imposed directly by Google. They can make things as compatible with as many non-Google forks as they want. They can break all compatibility wherever they want if they give up AOSP membership.

In fact, the boundaries of the compatibility requirement are pretty fuzzy.

For instance, the Acer/Aliyun situation is more complicated than presented. The Aliyun app store was distributing pirated apps (including pirated Google apps): http://www.androidpolice.com/2012/09/15/aliyun-app-store-con...

It seems obvious that Acer would have wound up in hot water with Google over this somehow, though it is odd that Google decided to focus their complaint on compatibility (rather than on Aliyun's piracy).

On the other side of the coin, many Android OEMs have distributed (and still distribute, so far as I know) devices without Google apps for the Chinese market (including Samsung, ASUS, Huawei, and even Motorola [1]). So far as I can remember, these were always essentially the same as non-Chinese devices except that they came with a different ROM that didn't include Google Apps. Nevertheless, that means that there's some scope for making non-Google Android devices, we just don't know how far it goes. Maybe if the HTC/Amazon rumors turn out to be true, we'll find out more.

[1] Though I think Motorola's efforts were shut-down post-merger because of Google's China policies.

I think alot of people misunderstand what open source means. It's nothing more than allowing people to see the source code, and use it (including forking).

Open source doesn't require you to cooperate with anyone, it doesn't require you to give away access to APIs, it doesn't require you to do anything beyond whatever is explicitly stated in the license.

Google, Canonical, Oracle, IBM, Red Hat, SUSE, etc... aren't required to be good team players or corporate citizens. They're just required to abide by the terms of the licenses on code they use...

Well, there is a philosophy (or rather, various different philosophies) that go along with the legal structures, and different projects have different levels of adherence to those ideas.

It eventually comes down to perception of value. Part of the original attraction of Android was its openness, if Google is now closing off substantial functionality, to head off competition, then it's not unfair for people to re-evaluate their enthusiasm for the product.

> Part of the original attraction of Android was its openness, if Google is now closing off substantial functionality, to head off competition, then it's not unfair for people to re-evaluate their enthusiasm for the product.

Very well put! Its not a legal perception that theres something wrong about what they are doing(and nobody are saying that they cannot do it in the legal aspect).. its a betrayal of some part that make people defend android for what it is, a open source project.. its the ideal behind the project that is being broken

If its not like that anymore, the same people that support it for its openess should be aware of it.. and see that things are actually, gradually and silently changing..

Perhaps there should be a phrase to mean "open source in name only." I wonder if all of this is related to Richard Stallman's dismissal of projects that claim to be open source, saying that it isn't enough unless it's "free software."

RMS is satisfied with anything that has a GPL licence. He's always supported creators rights to manage projects however they choose, and sell software as they choose. Google would fall short simply because the BSD licence isn't copyleft, but not for much of what they're being accused of here...

You are perfectly right regarding stallmas posted opinion regarding the GPL license.

I doubt however that RMS would be happy with open/free code being replaced by closed one as exemplified in the article. That has of course nothing to do with permissive and copyleft, and all to do with lock-in, proprietary practices, and project management.

He'd likely just say it was never fully free to begin with, because of the license. Not to mention, all the Google apps in question connect to SaaS backends which are closed anyway.

Personally, I don't see the problem with what Google is doing, I would prefer 100% open software myself, but Google would never open-source their services' code, so the fact the apps are closed-source doesn't really make a difference.

Not true, rms has no problem with the BSD license. It's still Free Software, and rms has even supported licensing a particular piece of software (Ogg/Vorbis) under a permissive license instead of the GPL.

http://lwn.net/2001/0301/a/rms-ov-license.php3

> its a betrayal of some part that make people defend android for what it is, a open source project..

As an open-source project, Android was only ever really open in the way that, say, Oracle BDB is open; there are periodic code dumps, but it's very much not operated as an open-source project.

Android is open source in that sense, that Google can leave the AOSP app versions to rot, but they can't force them to be left alone - in fact, both Amazon and Samsung can work on the AOSP apps instead of their own, and keep them up-to-date and better than Google apps if they are able to.

As an Android developer, I love that Google is doing this!

Android has come a very long way in the last few years in terms of usability and design. A large part of this has been due to an increasingly uniform design language and feel. That, and the new distribution model for what are basically Android updates (Google Play Services) has made Android feel more polished and actually allowed it to stand on its own against iOS. It also means that developers like me don't have to spend nearly as much time worrying about fragmentation in the traditional sense. Each day the percentage of people using sub-ICS phones falls, and we all get one step closer to the day we can support ICS+ only.

However, companies like Amazon would force me to rewrite the maps integration, the sign-in portion, the wallet, etc... Amazon did a great job of replicating Google Maps API V1 but they have yet to mirror V2 and don't mirror the other components I mentioned.

Aside from fragmentation and developer sanity, the article mentions another key point here:

"[M]any of Google's solutions offer best-in-class usability, functionality, and ease-of-implementation."

Exactly! Google APIs are not perfect, and there's bugs (like when Google Maps broke map markers on high resolution phones like the HTC One). But generally speaking, I'm really happy with the quality of the APIs and services. In an ideal world, Amazon and Google would work together to provide great and uniform single-sign-in APIs, great maps, etc... As it currently stands though, I don't believe either party is interested in doing so. Prisoner's dilemma?

> While it might not be an official requirement, being granted a Google apps license will go a whole lot easier if you join the Open Handset Alliance. The OHA is a group of companies committed to Android—Google's Android—and members are contractually prohibited from building non-Google approved devices. That's right, joining the OHA requires a company to sign its life away and promise to not build a device that runs a competing Android fork.

...

> This makes life extremely difficult for the only company brazen enough to sell an Android fork in the west: Amazon. Since the Kindle OS counts as an incompatible version of Android, no major OEM is allowed to produce the Kindle Fire for Amazon. So when Amazon goes shopping for a manufacturer for its next tablet, it has to immediately cross Acer, Asus, Dell, Foxconn, Fujitsu, HTC, Huawei, Kyocera, Lenovo, LG, Motorola, NEC, Samsung, Sharp, Sony, Toshiba, and ZTE off the list. Currently, Amazon contracts Kindle manufacturing out to Quanta Computer, a company primarily known for making laptops. Amazon probably doesn't have many other choices.

That is fairly incredible, I'm surprised it is not an anti-trust/competition issue.

It's not anti-trust because they're not required to only build Android devices. HTC and Samsung also build Windows Phone devices...

Are you serious? What other commercial OS can I do a competing fork of? Don't you see that the same restrictions apply to all other consumer OSs (iOS, WP, etc) by simply being closed-source?

That's a fair point, but Android is open source (and indeed was heavily marketed on that basis). They don't get to use any mechanism they like to make the situation comparable to their closed competitors. Their situation is also not comparable to iOS, WP, etc, because of their position of market dominance, just the same as the situation with Windows and Mac OS back in the day.

Android is open source; their services and corresponding apps aren't, and that's what they use as leverage. Not allowing them the same rights over those as other companies have over their proprietary software is essentially punishing them for having released something as open source.

And yes, Android is open source, and the Kindle Fire proves it. The fact that some people expect Google Play et all to be bundled is not Google's fault.

Their situation is also not comparable to iOS, WP, etc, because of their position of market dominance, just the same as the situation with Windows and Mac OS back in the day.

But Windows was, and while Microsoft was criticized for a lot of anticompetitive plays, not allowing people to create forks of Windows was never one of them. Not to mention that Google does allow you to create forks of Android, they just don't give you access to their services on top of that.

> Google does everything in-house. The company gets Maps and all of its cloud services basically for free.

This statement is utterly false. In-house does not mean free.

All of its Maps data and many of its APIs already existed for Google Maps on the desktop. Within the mobile space, and relative to a mobile-only competitor, Google's access to that data is effectively free.

You mean it has a low marginal cost. However, it required a huge capital investment to create, from which they are now rightfully reaping the benefits.

I'm certainly not saying they are wrong to leverage that asset. I don't think the point is about right or wrong as regards Google's action, but simply the extra barrier this represents for a competitor.

Map data for mobile isn't free even for Google. Google buys licenses to the data from a few map data providers, and as there was effectively a duopoly in global map data, those licenses have very strict terms of use. To use that data for e.g. real-time navigation assistance in mobile requires a different, much more expensive licenses.

Good point, thanks for the correction.

Exactly. Google's mapping data is a legitimate competitive advantage, and there is no reason anyone else forking android should be entitled to it.

The article in no way says this is illegitimate, it is listed as an extra barrier that a competitor would have to cross.

Maybe not explicitly, but it was stated in the context of a critical article about Google's betrayal of openness, so we could at least be forgiven for interpreting it in that light. It's not a stretch to assume it was meant as a criticism.

Close enough to free.

While I can see the point of this article, it's being cast in a much more dramatic light than necessary. Phrases like "While Google is out to devalue the open source codebase as much as possible" seem hyperbolic to me.

I agree. While it does make some good, disturbing points, you can tell they had to stretch to make some of the other points.

What's really annoying is when the Google apps start becoming worse to make Google more money. I never open the app store without wanting an app, but it constantly tries to sell me books and movies and stuff like that. They even have separate Google apps for reading and movies, so shoving it in the app store is just a money grab making my usage more difficult to shove some ads in my face. Were these apps open source people could just fork, but we're stuck going along with Google until they mess up so bad it makes sense to switch over entirely to Amazon App store, Samsung Apps Seller, etc. and the equivalent for everything else.

This is an opportunity for Yahoo. The article mentions that OEMs can't leave because they're afraid of losing Google services, Yahoo can be that replacement. They've been making beautiful mobile apps lately and they have some very polished services. They are a pure play service provider who isn't trying to compete with the OEMs; they have no conflict of interest.

If you work at Yahoo, this is something you should run up the chain. A relationship with OEMs would solve their problems, and allow you to get a foothold in mobile that you haven't been able to before.

Yahoo internals were even discussing ideas of creating their own fork of Android, way back in 2010. Nothing happened.

Correct me if I'm wrong, but isn't Linus Torvalds famous for his iron grip of Linux? In a very different way, to be sure, but it's my understanding that just because you make something open source doesn't mean you have to (or even should) relinquish control.

I think it's also pretty standard to open-source the core and keep the baubles proprietary. GitHub, for example, made their git interaction library open-source but their git hosting service itself is closed, as far as I know.

Every open source project has someone (or some group) in control, and contributions can't land without their say so. But anyone who feels that the current leadership is doing it wrong can fork the code and try to persuade people to use their version. This is very important to the way open source software works.

Google have a level of control beyond that: they undermine anyone trying to fork Android by not letting them use Google's apps and services (even as they allow such use on competing systems like iOS).

I have a feeling that if an Android fork actually started to gain serious market share, Google would very quickly allow them to use GMail, Youtube and the other apps.

Amazon have a Kindle product that is an Android fork with serious market share. I don't know if Google has anything in the Kindle market place with GMail etc.

The Kindle Fire doesn't have the Google Apps, and might never have them, but the Barnes&Noble Nook HD does (I have one), starting a few months ago. It originally didn't (I suspect for the same reason as the Kindle Fire), but Google later allowed B&N to run the apps (B&N even replaced their custom web browser with Chrome, a move which not everyone liked).

Sure. That makes perfect sense to me - Google apps and services aren't open source in spirit or practice. If I wanted a pure open-source mobile OS of my own, I wouldn't want any of Google's apps or services tainting it; if I did, Google would evaluate that option on their own, as they should.

Part of my business is about creating Android-based embedded systems. So far, none of what Google has done impinges on using Android as a basis for an operating system for appliance-like devices. The main problem is that current development of Android is not done in the open. But, so far, the advantages of using Android's UI stack and other APIs in "appliance OS" applications outweigh the annoyance of sporadic updates to the AOSP code-base.

If you want to compete with Google, using Android poses a choice: If you make Google-branded Android devices that use Google's proprietary apps, you will have to give that up in order to use Android with other ecosystems.

Thirdly, if you want to use the Google ecosystem in a product, you have to use all of it. You can't substitute someone else's location services, for an example that was litigated.

Google could develop Android in the open and retain the same level of control over OEMs, and I think they should.

Google appears to be inconsistent in enforcing restrictions on OEMs. OPhone OEMs also make Android handsets, despite the fact that OPhone is an Android derived product. Maybe that arrangement pre-dates Google's current policies.

> Android has arguably won the smartphone wars, but "Android winning" and "Google winning" are not necessarily the same thing.

This is false. Google wins when more people use the Internet. Android is fulfilling its initial goal incredibly well: offer a free and open-source mobile OS to encourage mobile device proliferation.

Android is doing exactly what it was designed to do.

The article's description of what we're to happen if Apple did not have competition is quite accurate and write frightening. Even with the north American market share that Apple has they abuse it where possible.

Yes I agree.

Android is not about North America. Android is about India, Africa, and China.

It is banned from China (AFAIK, they use Alibaba fork). India and Africa are mostly untapped markets. The 80% market share come from America and Europe

That statement is entirely incorrect. http://www.tech-thoughts.net/2012/12/smartphone-market-share...

Android is not banned from (or, in) China. It is hugely popular. See, for example, a popular website's offerings for Android telephones: http://mobile.139shop.com/commend/170__001_1.htm

Interestingly, in China, and this is personal experience not a market survey, the more one moves away from big brand manufacturers of Android telephones, the more pure / bare-bones / no manufacturer or telecom added applications the Android install is (again, China experience only).

I am sitting in mainland China with an Android phone in my pocket. 518 Yuan can easily buy a bright 5 inch screen, 4 days battery life (my average usage) with WiFi on most of the time, only one manufacturer installed app (Baidu input; being in China, this useful), 1.2GHz processor, 1GB RAM, 2 SIM card slots, SD Card slot (free card included, only 2GB, but free is free), 5MP camera (2MP reverse), user serviceable battery.

Apple hasn't abused anything.

> This is false. Google wins when more people use the Internet. Android is fulfilling its initial goal incredibly well: offer a free and open-source mobile OS to encourage mobile device proliferation.

> Android is doing exactly what it was designed to do.

That's effectively what it says in the article: "It's easy to give something away when you're in last place with zero marketshare, precisely where Android started. When you're in first place though, it's a little harder to be so open and welcoming. Android has gone from being the thing that protects Google to being something worth protecting in its own right."

i.e. Android is doing what it was designed for - protecting Google's presence on the mobile web - but with such a dominant market position Google found that it ceased to be solely a means of protection, and actually an asset in its own right. It's in that extension of the original aim where the interests of Android as an open source project and Google as a for-profit company do not match up.

"In an ocean with great waves, whales fly into the air unnoticed, but in a calm pond, even the tiniest minnow makes a ripple." -confucius

When the iPhone debuted, no doubt Google sensed the impact, and Apple's ability to create an effective closed ecosystem had already been proven with iTunes. I believe that Google wanted to undermine the market long enough to understand it. True enough, "android winning" was not the same as "Google winning," but it did mean everyone else "losing." I believe that for Google, Android started as a strategy in search of a goal. It was a smokescreen to prevent Apple from taking a dominant position by default. As the data poured in, they began to understand how to leverage it, and the Nexus line became an expression of such understanding, working to establish more control, and hopefully emerge from the smokescreen they had created.

> whales fly into the air unnoticed

Is this the origin of the Twitter fail-whale?

> I believe that for Google, Android started as a strategy in search of a goal. It was a smokescreen to prevent Apple from taking a dominant position by default

Impossible. Google bought Android years before Apple sold a single phone.

I will only buy / recommend nexus phones. Knowing that the phone will be updated quickly and for a long time is worth a premium, rather than being a cash in strat for Google.

my impression of nexus devices is that you are tied to the package of google services. is this so, or can a nexus-user break free from google's domain and enjoy the benefits of a quality low-cost device without being beholden to the google ecosystem?

From what I know, for the 2012 Nexus 7 at least, there are lots of custom ROMs, some of them without Google Apps. There are even alternative operating systems like Ubuntu Touch. I don't know if other emerging mobile OS such as Firefox OS and Sailfish have been ported already or not, but if not, I'm sure they will, given the market share of the device and the amount of people interested. Ultimately, you can also run GNU/Linux with the usual light desktop environments.

Well, this is really arguable. The "Iron Grip" is the modules that happen to be dependent on authenticated API calls to the servers that Google owns and maintains.

I'd fully support their modules that connect to the cloud servers being open source / GPL / etc, but to expect them to open them up to unauthenticated requests is untenable and leaves them way open to abuse / lack of rate limiting / making the service a bad time for all involved.

> Well, this is really arguable. The "Iron Grip" is the modules that happen to be dependent on authenticated API calls to the servers that Google owns and maintains.

It also consists of witholding access to those APIs if a company uses a competitor service:

> Another point of control is that the Google apps are all licensed as a single bundle. So if you want Gmail and Maps, you also need to take Google Play Services, Google+, and whatever else Google feels like adding to the package. A company called Skyhook found this out the hard way when it tried to develop a competing location service for Android. Switching to Skyhook's service meant Google would not be able to collect location data from users. This was bad for Google, so Skyhook was declared "incompatible." OEMs that wanted the Google Apps were not allowed to use them. Skyhook sued, and the lawsuit is still pending.

What's interesting is how having a mobile OS is now only one part of the offering needed to be successful, and is arguably the easiest part.

To be successful on mobile you also need a fairly extensive layer of services. Some of those (web, mail and so on) are easy to bolt together but others such as maps and app stores are far harder and are about data and commercial deals as much as they are about software. While it would be wrong to say that these services can't be opened up, in many cases doing so isn't as straight forward as sharing source code.

It doesn't feel as if Google has changed so much as what it means be a mobile OS has.

Small correction: "Chrome is still open source" is incorrect. Chromium is open source. Chrome is closed source.

Google is creating a walled garden just like any other company does. The article points to how they are making their shift towards an operating system that is similar to ios (in terms of lock in). Android may be an open-source platform but, on the majority of devices that compete at the top level, it becomes far from open source.

It's understandable why Google would lock people out of seeing the back end of their closed apps. But you have to look at what the long-term implications of them slowly removing support for ASOP apps are. As Google continually pushes out fantastic products that tie in so well to the mobile experience, why would anyone/developer want to have/develop [for] anything else. As this power grows, Google can strong-arm phone manufactures to develop hardware/features/etc to work with what they are developing. They have to sign contractual agreements to get the top version of Android and are then locked in to keep up the good terms. Google is outsourcing the hardware manufacturing to other companies and ensuring that if a user wants a good phone, they will be using their services.

Many people here are claiming any company can leave Google's garden like Amazon did. While some companies may be able to do that, I'm struggling to think of a one with the technological background, money to invest, and callousness for risk who are willing to try. Amazon has a huge assortment of media that it can toss at its users who use their hardware. Other companies don't have a differentiating factor or the software development to be able to make a truly competitive product to drive people away from Google supported Android. Just look at how much Microsoft, a software giant, is struggling to gain any shred of market share.

No executive in any reasonable company is going to propose to invest billions in order to squeeze into the highly competitive mobile OS market. It’s a huge risk that only a startup could swallow, and yet few startups could even raise the money required to topple the Google supported android market.

What the future is starting to look like is the one Google was initially afraid of, that users were” faced a Draconian future, a future where…one company, one device, one carrier would be [the] only choice.” As Google gains more power, the open source part that Android users love is going to slowly disappear. This may or may not happen, there are many variables that could prevent it, but it is a future that would bring Google the highest return and that is the goal of all market traded companies.

google had a very good reason to move services outside AOSP, to update them without relying on carriers. they could release billing API v3 w/ 90% compatibility in day 1. this is how they could workaround fragmentation. as an android dev, i love being able to read framework source code for better design, performance and less bugs. that is all really matters imho.

for development yes, thats like providing a SDK. I can develop just fine on Windows phones and even iOS :) So basically, its not really all that opensource anymore, that's what most people mean.

it is not the same with providing an SDK. Well, if android was documented better, we would not need source code this much but most of the time, it requires checking the actual implementation to understand the whole picture. Probably it is the difference between being "just fine" to creating great things.

I don't understand the point of this article especially the bit about being an evil genius by ways of making excellent best in business cross OS api's. Really? being competitive is being an evil genius? If amazon is willing it can open there api's to none FireOS apps they have the infrastructure and money to support it, but they don't.

As a user I'm happy that Google is making sure that I can hop device manufactures without loosing my apps or functionality, if everybody would roll out there own app store and removed Google's you would be locked in with the OEM. Now you can safely change to a different phone, also they don't mind you downloading the Google apps when using an alternative ROM.

Android is open source but does that mean that you are not aloud to make money of it by providing closed source apps and service, many open source companies do that. The work that went in to Android if freely available for competitors. Lots of kernel enchantments went back in to Linux and now you have Ubuntu touch and Firefox OS both based on the Android kernel which in turn is based on Linux, how cool is that.

I once read an article in 2010 that criticized people for saying that Apple and iOS of the 2000s is like the Microsoft and Windows of the 90s. The article pointed out that Apple IS the Apple of the 90s and Google with its Android platform will become the Windows of the 90s. I think it's happening. In a few years Android is going to be as closed sourced as Windows, probably as ubiquitous, and most likely, just as prone to security issues.

It's already kind of like windows, no? It runs on hundreds if different devices. It's often bloated by OEM software that people hate. It's prone to security wholes. It's slow and clunky unless you run it on the latest hardware. It bends over backwards for compatibility sake. It's more and more closed sourced...

Android is Mobile windows of the 90s. I hope Ubuntu Mobile will be successful.

Sorry, but some of the things you wrote are about perception, not reality.

What does "prone to security issues" mean? And compared to what? When we compare iOS and Android, iOS had 304 vulnerabilities in 2007-2013 (294 in 2009-2013), while Android had 29 in 2009-2013[1]. That's order of magnitude difference, yet Android is blamed to be prone to security holes.

Also what does "clunky unless you run it on the latest hardware"? All systems are slow, when you run new system on old hardware. You cannot bend physics there. Ever used iOS4 on iPhone 3G? iOS7 on iPhone4? Or why there is no WP8 update for WP7 phones? These are exactly the same reasons.

Even the APIs in Play Services, as described in the article, are exactly marked as "google, not android". You project has to explicitly include Google APIs, it is not enough to target Android. Every Android developer knows that.

[1] http://www.cvedetails.com/product/15556/ and http://www.cvedetails.com/product/19997/

Vulnerability numbers alone do not accurately convey the mobile security landscape. To start with, even if those mean something, we must consider that Android devices tend to lag behind in terms of receiving OS updates. By contrast, iOS devices tend to be quickly updated.

It is widely known that iOS is far more secure in the practical sense for the average user. Report: http://www.f-secure.com/static/doc/labs_global/Research/Mobi...

Few centuries ago, it was widely known, that the Earth is flat. Just because something is widely believed, does not mean, that it is true. I would prefer to believe hard, cold numbers, or analysis like this: http://qz.com/131436/contrary-to-what-youve-heard-android-is...

That F-Secure report describes legitimate applications that could be potentialy used as a malware (e.g. just like Wireshark on desktops) or trojans. What happens with trojans is described in the link above.

Also note, that it is in F-Secure's interest to cause scare, it is good for their business. They are not interested in people rationaly thinking about presented issues.

This is exactly why most malware out there targets iOS, and why so many botnets are comprised of iOS divides. Right?

This was meant as a comment to vetinari.

Nothing wrong with what Google is doing. Google is essentially a consumer (and enterprise) cloud services company that is looking to commoditize (read open source/sell at cost) all other parts of the stack. That includes open-sourcing Chrome/Android, selling Chromecast/Google Fiber at cost.

Claiming that Google is "controlling open source" by working in-house on it's own Android applications is just really bizarre.

So, every Samsung phone effectively comes with three versions of the main apps - the AOSP version, the Google Play version, and Samsung's bloatware?

This seems like a terrible situation for users. Can someone with a Samsung smartphone confirm this?

If this is the case, how are the apps organized when you first buy the phone - are they all in one big apps list?

Samsung devices ship with two versions, remember that while Samsung can't opt out of the Google apps, they're free to remove any AOSP app they'd like.

In fact, I've had AOSP apps disappear from my Cyanogen Mod ROM when I installed gapps, although I don't know if that was due to being overwritten or just part of the install script.

Wow, I've heard that HN is often pro google, but this thread makes it blatantly clear.

I know this is potentially dangerous in the future (I worry more about NSA having direct access to all the phones in the world through Google), but in terms of user experience, I welcome this. In order to have an ecosystem that is "as unified and standardized as possible" you need to have one company controlling it, and the vision behind it. Too many companies pulling in too many directions is not that good.

Here's a different perspective:

http://techtainian.com/news/2013/10/20/editorial-how-kitkat-...

If AOSP is open source and Google updates let's say the location services, why can't anyone start a similar project for AOSP and have it funded (like by Apache or Mozilla)?

It seems that the main problem is the gatekeepers who manufacture phones.

Mozilla makes Firefox OS. They don't have the resources to also fix Android and what not, I think. Apache takes a few projects under its wing but they don't provide the developers, and thus many of their projects are stalling.

That, or someone to maintain a generic shim/interface so that the proprietary services are interchangeable

IMHO, the unfair part is the following:

"Since the Kindle OS counts as an incompatible version of Android, no major OEM is allowed to produce the Kindle Fire for Amazon. So when Amazon goes shopping for a manufacturer for its next tablet, it has to immediately cross Acer, Asus, Dell, Foxconn, Fujitsu, HTC, Huawei, Kyocera, Lenovo, LG, Motorola, NEC, Samsung, Sharp, Sony, Toshiba, and ZTE off the list."

Google Apps and APIs are fine and good, but I don't think any company should dictate to an OEM what products they can make for other companies.

I'm surprised that there's no mention of Tizen, Samsung and Intel's project and supposedly the former's Plan B to ditch Android altogether. With the same TouchWiz skin both Samsung's Android and Tizen, and with both OS's able to run Android apps, the plan would be to swap out the underlying OS without the users noticing.

Sometimes businesses take measures to increase profits at the expense of their users. However, preventing fragmentation of Android seems to be both in both Google's interests and the users. Only certain competitors could mind. Also, why should Google do all the work of creating an operating system and not get anything in return?

As much as the Android people hate the Apple people, Android is doing the same thing Windows and Apple have been doing for years - trying to shoe horn people into THEIR walled in garden.

This is the future of smartphones. You pick a phone and by doing so, you pick the walled garden you're going to most comfortable playing in, pure and simple.

I think Google's malicious intent is being over-exaggerated. It could simply be that they don't have enough resources to maintain old code. As to them creating closed-source apps, well, Google knows which side their bread is buttered on :) Anything that makes money for them is closed source.

I'm curious to know if there are any internal google project of porting all the Google play API on iOS (either on Objective-C , or using a cross-platform language such as mono develop, which would make more sense).

Why don't they (potential competitors) write an open source app based on openstreetmap? Their mapping data is usually on par, often even superior to that of Google. Plus, its free (in both senses of the word).

Nothing good can come out of Amazon or Samsung influencing or controlling Android. If those companies were in control, we'd still be in the tech ice ages where the phone companies control our devices.

So I guess the sign to look out for is Samsung licensing the Amazon APIs?

That would be interesting: an Amazon-Samsung software-hardware partnership.

This is an excellent piece of tech journalism and thank you for the journalist for examining each aspect of Google's strategy so thoroughly.

As a Google shareholder, this article warms the cockles of my heart. Why should Amazon be able to get all the Android improvements that Google creates?

I don't see how making closed source apps is "Controlling open source"

By any means necessary? So they'd murder to control the source?

Open always wins, until it conflicts with your business interests. "Open" used to be the oft-repeated advantage over iOS in the early days, I wonder if that will slip away from the narrative like "SD card slots", "removable batteries", and "real keyboard" did.

History doesn't really support this. Even Linux had to get some corporate backing (initially Red Hat, Suse, and the like), and later Canonical and others before it really took off.

Open source has proven to be a long term survivor, but not a winner.

Open source is most definitely winning in infrastructure, high performance computing, servers, etc...

Ask IBM or Oracle if Linux is losing...

You're conflating the reasons though...would Linux be winning in those areas without big corporate $$$?

I was going to say, "well, most people aren't actually using a corporate-sponsored distro", but I guess the real question is whether Linux would have become the impressive piece of technology it is today without corporate backing, if there had never been anyone working on it full-time. Maybe FreeBSD would have taken it's place if Linus didn't have enough time to work on it. Either way, it's really hard for me to imagine a non-open system in this position of dominance.

Does it matter? Corporations adopted Linux because of it's openness, and Linux is winning in those areas because of corporations who adopted it because of....

Open doesn't mean 'hobbyist', it merely refers to the freedoms enabled by the code's license...

They adapted it because it was cheaper, not because it was open. Now, open frequently tends to be cheaper, but not always, it isn't an inherent quality of openness.

IBM and Oracle both have in-house operating systems (AIX, z/OS, Solaris), and are both big enough that they have, for all practical purposes, unlimited resources.

Yet both use and push Linux, IBM uses SUSE Enterprise, and Oracle forked Red Hat's OS...

I really doubt either cared about whatever marginal cost they may or may not save. Linux's openness is kind of like natural selection - good features live on, bad ones die.

Open source does make better software (everything else being equal of course).

I don't see Google making any rules about OEMs not being allowed to manufacture devices with micro SD slots, removable batteries or physical keyboards. Just because Google doesn't want to spend its R&D money on those features doesn't mean they're stopping anyone else who uses Android.

Should read more like "Google has an Iron grip on Google Apps (Gapps)" - Not Android. Android is the OS, not the Google service based apps.

Isn't the whle point of the article that it's getting harder and harder to get the OS without being tied to Gapps? Which would justify the title?

The Android platform is open source under Apache/BSD/MIT licenses -- it's no harder to get that than it ever has been (not hard at all). Look at products like Kindle Fire for examples of "uses Android without Google Apps".

Yes you can easily just download it. But that's not exactly what the article is about. It's about why OEMs don't just download it and put it on there as such, but instead seem to get tied to Google Apps.

Reply on news.ycombinator.com