7 ms·

The exact same thing stuck out to me as well. Mostly because they had to start the article with this sentence: > With more than two billion active devices, And

The exact same thing stuck out to me as well. Mostly because they had to start the article with this sentence: > With more than two billion active devices, Android is the largest mobile platform in the world. So they'll get it onto 5% of active devices, or one in every twenty. Not great, perhaps not even good. I've not been a fan of android fragmentation for awhile now, and have been surprised that Google hasn't been able to do more to attack the issue. Even when Treble launched, I asked myself... is that all? Rock and hard place I guess. I actively avoid android devices because of the issue. My last android device was a tablet in 2011.

EddieRingle 9y ago
> Even when Treble launched, I asked myself... is that all? Rock and hard place I guess. Curious, what exactly were you expecting? There hasn't even been enough time for Treble to play out to see if it brings results.
i_cant_speel 9y ago
Personally I would prefer them to target the top devices with a great product instead of delivering a mediocre product that all devices can run.
- vlozko 9y ago
  That doesn't make it very appealing for app developers, though, if the size of their base is dramatically limited to just people with both the latest and the greatest devices.
  - kemitche 9y ago
    I'm not sure what's unappealing about targeting users that can afford to always have the latest and greatest. That's effectively exactly what you're doing if you choose to Target iOS over Android.
    - mcintyre1994 9y ago
      I guess the problem is that (going by Google's target of 100m users with this feature by winter) Apple have vastly more of those users.
  - makecheck 9y ago
    Today's "a few" is tomorrow's "everyone", this has been proven before (the original iPhone, car technology, etc.). You take the time to perfect what you build with a smaller audience and more powerful technology, and eventually there will be more people to sell it to.
    - MBCook 9y ago
      The question is if you're trying to earn your living doing that can you wait long enough for "a few" to become "everyone"? There are a lot of businesses that fail because they're too early.
  - orbitur 9y ago
    Considering Android apps make 15% of the money iOS apps make, perhaps it is for the best to only target devices that happen to be owned by people who spend money.
    - MBCook 9y ago
      If android apps make 15% of what iPhone apps do despite having a much larger market share, what would be the point of targeting a smaller fraction of that? Even if those users are all willing to pay you're probably not going to get higher than that 15%. The "high-end" at market on iOS is bad enough, would you want to try and fight that battle on android?
jmarinez 9y ago
Ditto. To add more wood to the fire consider that those 2 billion Android active devices are not “Google’s Android” devices. In China, where the largest bulk of those 2 billion devices are, Google isn’t allowed. Which means that SDKs like this one that are part of or require the Play Services framework are nowhere to be found. I would put Google’s Android active device numbers somewhere around 300 million. Translation - ARCore will show up in 1 out of every 100 users six months downs the line when Samsung, LG and other partners decide to release it. In the meantime, around Sept. 29 of this year Apple will have 100 million+ ARKit ready devices - based on the 80%+ install rate for new iOS versions - when iOS 11 is released. Just saying.
- foobarian 9y ago
  I'm still bitter how they abandoned the Nexus line in favor of Pixel. Hopefully things will be more stable with the in-house design.
- hooluupig 9y ago
  The 2 billion is the number of devices with the Play Store. Android devices in China do not have Google Play Store or Services on them. So devices in China do not factor into this number.
ehsankia 9y ago
As a user, how does fragmentation matter to you? It shouldn't affect you directly. The place where it matters is that you'll get less developers making software for you, but honestly, in the past 3 years, that hasn't really been an issue on Android. Except one or two big games like Mario, it's been fairly equal to iOS in terms of app support.
- khaledtaha 9y ago
  Not only that, but you can quickly solve that issue by going with Pixel/Nexus devices. So I'm not sure that the criticism is well-deserved.
  - freehunter 9y ago
    It certainly is an issue if the entire Android ecosystem of devices is reduced to just one niche $650 phone.
    - ehsankia 9y ago
      Sure, but you are also comparing it to a 700$ phone (the iPhone) and I don't think it's fair. Like, obviously, iOS will have a cleaner and less fragmented ecosystem, because they support far fewer devices (all of which they own).
- freehunter 9y ago
  Fragmentation impacts which device you might want to buy. The biggest argument for Android is the wide choice of devices and features on those devices, but if your phone never gets an update, you're left behind. It's possible to get a really nice and inexpensive Android phone, but what's the risk of never getting an update?
- acdha 9y ago
  The developers support isn't just a minor quibble, especially when it comes to quality of life factors like browser versions (e.g. LG, HTC, Samsung, etc. have all shipped Chrome builds with subtle differences). This also applies to optimizations for CPUs, graphics hardware, etc. — the app may be available on both but it almost certainly runs better on iOS. The other problem is that fragmentation increases the odds of a vendor walking away from a device since the per-model sales are divided more ways. Trying to pick a device which will get updates basically involves sales forecasting which wouldn't be necessary if the platform was open.
rsp1984 9y ago
The purpose of ARCore is for high-end Android devices to be competitive with the iPhone technologically, not to get on 2 billion devices, at least not now. Therefore it's fine that they get only 5% to start. I assume it's going to turn into 10%, then 20% quickly. For some perspective the iPhone only has about 15% market share globally.
- VectorLock 9y ago
  And obviously a very good driver to get people to upgrade to high-end Android phones. There aren't too many use-case drivers for that right now.
- s73ver_ 9y ago
  It'll only do that if ARCore itself takes off. With such small penetration, it's entirely likely that developers will just focus on Apple's ARKit.
  - rsp1984 9y ago
    likely that developers will just focus on Apple's ARKit Most likely someone will just build and offer an abstraction layer on top of ARKit / ARCore and solve the problem that way.
    - s73ver_ 9y ago
      Writing the code is probably the easiest part of all this. Testing and design are going to be huge.
    - k__ 9y ago
      AR.js seems to be doing it at this very moment
    - Benjamin_Dobell 9y ago
      Google have already provided this themselves, plus hooked it up to a 3D rendering engine: https://github.com/google-ar/three.ar.js https://github.com/google-ar/three.ar.js That's assuming you're okay with JS, but if you're going cross-platform then it makes sense if you need to have interactive scenes.
SkyMarshal 9y ago
You don't have to avoid Android devices, just buy Google ones. Most of the others are filled with crapware and bad customizations anyway, no idea why any savvy HN reader would want those.
- kuschku 9y ago
  Which is easy to say, if you’re on silicon valley wages. But a student trying to develop for Android can’t afford devices that start (!) at 908.35 USD. (Price of the Pixel 5" with 32GB storage, in Germany, right now). And Google just dropped support for my phone, the Nexus 5X (which I bought for 324 USD, on sale, just before they stopped selling it)
  - chii 9y ago
    How is it any worse for iOS? You even have to fork out $2k for a Mac as a dev machine!
    - kuschku 9y ago
      I never developed for iOS, because a platform that has in my home market (the EU) a marketshare similar to what Windows Phone 7 had in the US (below 15%) is useless. Especially when development starts upwards of $3000. I’m comparing between developing on Google devices, and non-Google devices.
      - lorenzhs 9y ago
        Your numbers are incorrect, iOS market share is around 22% in the EU5 (GB, DE, FR, IT, ES) according to https://www.kantarworldpanel.com/global/News/iOS-and-Android-push-towards-a-Two-OS-world https://www.kantarworldpanel.com/global/News/iOS-and-Android.... And according to https://www.kantarworldpanel.com/global/smartphone-os-market-share/ https://www.kantarworldpanel.com/global/smartphone-os-market... Windows Phone's market share (as opposed to device sales) in the US peaked at 5.6% in 2013. That's a factor of 4. You may not like iOS, and there's nothing wrong with not considering a platform for ideological reasons, but it's not irrelevant. Also, a used Mac Mini isn't $3000.

EddieRingle 9y ago

> Even when Treble launched, I asked myself... is that all? Rock and hard place I guess. Curious, what exactly were you expecting? There hasn't even been enough time for Treble to play out to see if it brings results.

i_cant_speel 9y ago

Personally I would prefer them to target the top devices with a great product instead of delivering a mediocre product that all devices can run.

vlozko 9y ago
That doesn't make it very appealing for app developers, though, if the size of their base is dramatically limited to just people with both the latest and the greatest devices.
- kemitche 9y ago
  I'm not sure what's unappealing about targeting users that can afford to always have the latest and greatest. That's effectively exactly what you're doing if you choose to Target iOS over Android.
  - mcintyre1994 9y ago
    I guess the problem is that (going by Google's target of 100m users with this feature by winter) Apple have vastly more of those users.
- makecheck 9y ago
  Today's "a few" is tomorrow's "everyone", this has been proven before (the original iPhone, car technology, etc.). You take the time to perfect what you build with a smaller audience and more powerful technology, and eventually there will be more people to sell it to.
  - MBCook 9y ago
    The question is if you're trying to earn your living doing that can you wait long enough for "a few" to become "everyone"? There are a lot of businesses that fail because they're too early.
- orbitur 9y ago
  Considering Android apps make 15% of the money iOS apps make, perhaps it is for the best to only target devices that happen to be owned by people who spend money.
  - MBCook 9y ago
    If android apps make 15% of what iPhone apps do despite having a much larger market share, what would be the point of targeting a smaller fraction of that? Even if those users are all willing to pay you're probably not going to get higher than that 15%. The "high-end" at market on iOS is bad enough, would you want to try and fight that battle on android?

jmarinez 9y ago

Ditto. To add more wood to the fire consider that those 2 billion Android active devices are not “Google’s Android” devices. In China, where the largest bulk of those 2 billion devices are, Google isn’t allowed. Which means that SDKs like this one that are part of or require the Play Services framework are nowhere to be found. I would put Google’s Android active device numbers somewhere around 300 million. Translation - ARCore will show up in 1 out of every 100 users six months downs the line when Samsung, LG and other partners decide to release it. In the meantime, around Sept. 29 of this year Apple will have 100 million+ ARKit ready devices - based on the 80%+ install rate for new iOS versions - when iOS 11 is released. Just saying.

foobarian 9y ago
I'm still bitter how they abandoned the Nexus line in favor of Pixel. Hopefully things will be more stable with the in-house design.
hooluupig 9y ago
The 2 billion is the number of devices with the Play Store. Android devices in China do not have Google Play Store or Services on them. So devices in China do not factor into this number.

ehsankia 9y ago

As a user, how does fragmentation matter to you? It shouldn't affect you directly. The place where it matters is that you'll get less developers making software for you, but honestly, in the past 3 years, that hasn't really been an issue on Android. Except one or two big games like Mario, it's been fairly equal to iOS in terms of app support.

khaledtaha 9y ago
Not only that, but you can quickly solve that issue by going with Pixel/Nexus devices. So I'm not sure that the criticism is well-deserved.
- freehunter 9y ago
  It certainly is an issue if the entire Android ecosystem of devices is reduced to just one niche $650 phone.
  - ehsankia 9y ago
    Sure, but you are also comparing it to a 700$ phone (the iPhone) and I don't think it's fair. Like, obviously, iOS will have a cleaner and less fragmented ecosystem, because they support far fewer devices (all of which they own).
freehunter 9y ago
Fragmentation impacts which device you might want to buy. The biggest argument for Android is the wide choice of devices and features on those devices, but if your phone never gets an update, you're left behind. It's possible to get a really nice and inexpensive Android phone, but what's the risk of never getting an update?
acdha 9y ago
The developers support isn't just a minor quibble, especially when it comes to quality of life factors like browser versions (e.g. LG, HTC, Samsung, etc. have all shipped Chrome builds with subtle differences). This also applies to optimizations for CPUs, graphics hardware, etc. — the app may be available on both but it almost certainly runs better on iOS. The other problem is that fragmentation increases the odds of a vendor walking away from a device since the per-model sales are divided more ways. Trying to pick a device which will get updates basically involves sales forecasting which wouldn't be necessary if the platform was open.

rsp1984 9y ago

The purpose of ARCore is for high-end Android devices to be competitive with the iPhone technologically, not to get on 2 billion devices, at least not now. Therefore it's fine that they get only 5% to start. I assume it's going to turn into 10%, then 20% quickly. For some perspective the iPhone only has about 15% market share globally.

VectorLock 9y ago
And obviously a very good driver to get people to upgrade to high-end Android phones. There aren't too many use-case drivers for that right now.
s73ver_ 9y ago
It'll only do that if ARCore itself takes off. With such small penetration, it's entirely likely that developers will just focus on Apple's ARKit.
- rsp1984 9y ago
  likely that developers will just focus on Apple's ARKit Most likely someone will just build and offer an abstraction layer on top of ARKit / ARCore and solve the problem that way.
  - s73ver_ 9y ago
    Writing the code is probably the easiest part of all this. Testing and design are going to be huge.
  - k__ 9y ago
    AR.js seems to be doing it at this very moment
  - Benjamin_Dobell 9y ago
    Google have already provided this themselves, plus hooked it up to a 3D rendering engine: https://github.com/google-ar/three.ar.js https://github.com/google-ar/three.ar.js That's assuming you're okay with JS, but if you're going cross-platform then it makes sense if you need to have interactive scenes.

SkyMarshal 9y ago

You don't have to avoid Android devices, just buy Google ones. Most of the others are filled with crapware and bad customizations anyway, no idea why any savvy HN reader would want those.

kuschku 9y ago
Which is easy to say, if you’re on silicon valley wages. But a student trying to develop for Android can’t afford devices that start (!) at 908.35 USD. (Price of the Pixel 5" with 32GB storage, in Germany, right now). And Google just dropped support for my phone, the Nexus 5X (which I bought for 324 USD, on sale, just before they stopped selling it)
- chii 9y ago
  How is it any worse for iOS? You even have to fork out $2k for a Mac as a dev machine!
  - kuschku 9y ago
    I never developed for iOS, because a platform that has in my home market (the EU) a marketshare similar to what Windows Phone 7 had in the US (below 15%) is useless. Especially when development starts upwards of $3000. I’m comparing between developing on Google devices, and non-Google devices.
    - lorenzhs 9y ago
      Your numbers are incorrect, iOS market share is around 22% in the EU5 (GB, DE, FR, IT, ES) according to https://www.kantarworldpanel.com/global/News/iOS-and-Android-push-towards-a-Two-OS-world https://www.kantarworldpanel.com/global/News/iOS-and-Android.... And according to https://www.kantarworldpanel.com/global/smartphone-os-market-share/ https://www.kantarworldpanel.com/global/smartphone-os-market... Windows Phone's market share (as opposed to device sales) in the US peaked at 5.6% in 2013. That's a factor of 4. You may not like iOS, and there's nothing wrong with not considering a platform for ideological reasons, but it's not irrelevant. Also, a used Mac Mini isn't $3000.