Hello dear Android expert friends,
I imagine if you publish a game to Android:
Your studio would need to own a LOT of devices to check that the build works on common devices.
(eg http://forum.unity3d.com/threads/77388-Androids-that-work-with-Unity-Android )
I imagine there are:
shops that do nothing but, own a lot of different Android devices, and test a build for you on all the devices you ask.
Is this the case?
What is the situation ?
Very best
Hello from iOS world! 
Never heard of anything like this. Most people that are small companies just use their own devices and ask friends and family for help. Other companies will save up a bit and buy a few main stream devices, and a few low end devices to see how things work and such.
Caliber, I appreciate the answer, thanks.
I’m surprised, you’d think it is a common need.
Sad only one response! I thought there were a billion Android developers.
I’m gonna ask another question about the common devices right now. Thanks again !
http://forum.unity3d.com/threads/172466-March-2012-which-five-Androids-would-you-choose
A few would be enough really. You can pick a few of the most popular devices for instance, or you pick them based on their chipsets and screen size.
Really depends if you want to go for volume, or overall compatibility.
Hi Shuji - thank for your response - say you wanted to go for volume.
Which five would you choose? Based on today’s market right now. Very best.
See my response in your other topic. Though I don’t consider this something I would do. It’s just based on the popularity of the devices, nothing more. When in fact, you could easily replace two of the phones because there isn’t a whole lot of difference between the galaxy line in terms of architecture (though this may change depending on the S4). I’d probably go for my second statement.
That would ensure volume, as well as overall compatibility. In that way you can support current and last gen devices, and maybe even the one before that.
I see what you’re saying.
Let me put it this way. Say you had to tell the buying public:
“the game works on THESE POPULAR ANDROID DEVICES _ _ _ _”
of course, it might also work on other Android devices. But as I understand it there’s a lot of variance, and you’d only be “guaranteeing” it to work on the top devices, the “usual” gaming android devices.
(As a developer, you might need a whole stack of devices for testing - that’s not so much what I was getting it. It was more a matter of “what would be the common devices, you’d need to guarantee it works on” … if that makes sense.)
If you’re a non-Android developer, you always hear “it’s a big challenge getting it working on all the different devices.” That’s the issue I’m discussing really, if that makes sense. (Hell, maybe that’s all wrong?)
“the game works on THESE POPULAR ANDROID DEVICES _ _ _ _”
Well yes, that is true. You don’t have to explicitly say so though. You can select which devices may download your product on the Play Store. So those with a phone that is simply not able to run it won’t be seeing the game in the Play Store at all. I am not going to make a game compatible with every device, simply because that is what the public would want. You need to know where you are headed with your game before. In my case, I focus on mid to high end devices. Simply because I do not desire to dumb the game down just to make money. I want to ensure the game runs well in all it’s glory and that customers enjoy it like it has been intended. Besides that, people who are still sporting a 400Mhz single core device with 64MB’s of ram, isn’t my intended audience. So why would I support them in the first case?
“If you’re a non-Android developer, you always hear “it’s a big challenge getting it working on all the different devices.” That’s the issue I’m discussing really, if that makes sense. (Hell, maybe that’s all wrong?)”
This pretty much goes hand in hand with what I just said. There are developers who go out of their way to support all the phones possible. This results in the overall quality of your app without it doing any good. Like I said, those people still sporting low-end devices are probably not the people who will do in-app purchases nor are they people who play a lot of smartphone games, or games in general.
Still, developers choose to support them. Resulting in a lot of compatibility issues and also a lot of one-star ratings. I can’t think of any other reason than greed. But I believe such an attitude only results in a bad brand name and an overall lower experience of your app.
So, in my opinion, you should make it clear what your target audience is. If they are true to casual gamers, then you’ll have to look into the mid and high end segment of devices.
Completely understand (and agree)
Do you think it’s worth being EVEN MORE aggressive with the OS level?
So you could say Jelly Bean only. Surely that would weed out more old devices? Or you feel ICS is a good level, today and now?
V best
Well, the problem with that is that it would also weed out devices who are perfectly capable of running your games. There are still some very decent Gingerbread devices out there that might run your game perfectly. So instead of weeding out the majority by going Jelly bean, I’d filter by device. Sure it requires a bit of research on all of the devices, but that’s a one time thing.
The thing is, generally speaking you have to take 3 things into account. Manufacturer update, carrier branded update and ultimately customer update. A good example would be the Galaxy S3. It released with ICS when the phone launched back in May 2012. Currently you can update it to JB, but here’s what I’m trying to say: A device releases with ICS, and will eventually get an update to JB. Now you have to take into account the time it takes for a manufacturer (Samsung in this case) to release it. This takes about 3-4 months. Then the software get’s modified by carriers (Vodafone, T-Mobile, Verizon etc.) and it takes even more months for this update to be pushed to the device. Then finally, after some time it get’s pushed to the consumers. But even then you aren’t really sure when they’ll update it (or not even).
This is a point I’m trying to make. A good batch of Galaxy S3’s may still run ICS depending on some factors. So it isn’t a good idea to filter devices on OS level. Another thing is ICS devices are still regularly being released. There isn’t a really big need to worry about Gingerbread when your making games (apps would be something different of course) but leaving ICS out of the fun would really harm your game availability if you ask me.
It makes total sense. It seems to me ICS is a great basic position, today.