3 ms·

I think the difference in development comes in when you need to do some hardware testing. With apple you need to design for 3 different display sizes, with 2 re

by x0054 13y ago

I think the difference in development comes in when you need to do some hardware testing. With apple you need to design for 3 different display sizes, with 2 resolutions each. You also have to test your product on (at most) 12 different devices, and that's if you really want to cover the field. On the other hand, with Android, if you want to do it right, you need to design for hundreds of different screen sizes, all kinds of different aspect ratios and resolutions, and hundreds of different hardware devices. Or you can forgo all that, and just assume (incorrectly) that majority of the Android users run the latest and greatest Galaxy hardware.

dlhavema 13y ago
there was an article posted here a few weeks ago that talked about an android development company/group that boiled it down to ~10 major categories of devices that they needed to test with to get maximum device acceptance. So for them, unless a whole category of device comes out, like a smart watch for example, they do not need to have every device to test. and this encompassed screen sizes and OS versions.
- prehkugler 13y ago
  > So for them, unless a whole category of device comes out, like a smart watch for example, they do not need to have every device to test. Cue the Samsung Galaxy Gear.
- richardwhiuk 13y ago
  You can test on a limited range, but you will hit handset specific bugs - its just a fact of life on Android.
KirinDave 13y ago
You can shave away a LOT of form factors if you just agree to cut off development before Android 4.0. This is tough for people to resolve to do, but in the US that's 60% of the market. I'd rather ship a good product to 60% of the people in a reasonable amount of time, rather than spending a ton of time working on tiny slices of the last 40% and limiting what the product can do overall. You also end up targeting much newer devices and OS versions, which evades a lot of problems with piracy.
- krschultz 13y ago
  For my market, 90% of my app's users are 4.0+. 97% of all sessions are 4.0+. I still support & test pre-4.0, but I'm starting to consider options for spending less time on Gingerbread support. Progressive enhancement helps (we just don't do a lot of things on Gingerbread at all), but it still clutters up the code. Even a few months ago I couldn't imagine dropping Gingerbread support, but I have seen almost no growth in Gingerbread users over the last 3-6 months.
  - KirinDave 13y ago
    I am glad to hear stories like this. Thank you very much for sharing that.
- rsynnott 13y ago
  > You can shave away a LOT of form factors if you just agree to cut off development before Android 4.0. At that point, though, the idea that Android is a larger market really starts suffering. That cuts out a third of Android users immediately.

by x0054 13y ago

dlhavema 13y ago

there was an article posted here a few weeks ago that talked about an android development company/group that boiled it down to ~10 major categories of devices that they needed to test with to get maximum device acceptance. So for them, unless a whole category of device comes out, like a smart watch for example, they do not need to have every device to test. and this encompassed screen sizes and OS versions.

prehkugler 13y ago
> So for them, unless a whole category of device comes out, like a smart watch for example, they do not need to have every device to test. Cue the Samsung Galaxy Gear.
richardwhiuk 13y ago
You can test on a limited range, but you will hit handset specific bugs - its just a fact of life on Android.

KirinDave 13y ago

You can shave away a LOT of form factors if you just agree to cut off development before Android 4.0. This is tough for people to resolve to do, but in the US that's 60% of the market. I'd rather ship a good product to 60% of the people in a reasonable amount of time, rather than spending a ton of time working on tiny slices of the last 40% and limiting what the product can do overall. You also end up targeting much newer devices and OS versions, which evades a lot of problems with piracy.

krschultz 13y ago
For my market, 90% of my app's users are 4.0+. 97% of all sessions are 4.0+. I still support & test pre-4.0, but I'm starting to consider options for spending less time on Gingerbread support. Progressive enhancement helps (we just don't do a lot of things on Gingerbread at all), but it still clutters up the code. Even a few months ago I couldn't imagine dropping Gingerbread support, but I have seen almost no growth in Gingerbread users over the last 3-6 months.
- KirinDave 13y ago
  I am glad to hear stories like this. Thank you very much for sharing that.
rsynnott 13y ago
> You can shave away a LOT of form factors if you just agree to cut off development before Android 4.0. At that point, though, the idea that Android is a larger market really starts suffering. That cuts out a third of Android users immediately.