How important is it to support older versions of iOS?

Question

Not sure if this is better suited for here or for Stack Overflow - a commenter on a post dealing with similar (though not identical) topic on SO suggested that it would have been better to post it here. If this is wrong then I apologize.

I hope you will excuse this somewhat of a philosophical question. I'm an independent/individual iOS developer. Recently, the one device I keep around for iOS 4.x compatibility testing (an iPod touch) gave up the ghost -- and I am having a devil of a time trying to replace it. Of course new devices all have iOS 5 on them. So one would think to hit up the used market. But alas this is proving fruitless. Even the used market is proving little help, as most sellers think it helpful to upgrade to the latest and greatest before putting them up for sale. And after yet another day spent fruitlessly plowing through eBay search results, I finally had to sit back and think to myself:

How important is it to continue to support iOS 4.x?

It seems (at least if these blog posts are to be believed) that pretty much everybody is on iOS 5 these days, since it is still supported on even fairly old hardware; and that supporting older iOS versions will run you up against the law of diminishing returns, as supporting iOS 4 is both costly and largely unnecessary. At the same time, I look at the new stuff we got to play with in iOS 5 (and soon we'll have even more toys in iOS 6) and I can't help but drool. ARC alone looks like it will save me quite a bit of headache. (Hey there retain/release, don't let the door hit you on the rear on your way out)

So I guess what I'm asking you is, as a user, how important is iOS 4 compatibility? Did you upgrade to iOS 5, or are you still running iOS 4.x on any of your devices? And/or do you know anyone still running iOS 4.x? (The stereotypical tech-unsavvy parent comes to mind) Any and all comments/advice/practical real-world experiences/etc. you can provide is greatly appreciated.

Answer 1

It is indeed more of an opinion question, but here's my 2¢ anyways. I've heard a lot of talk about it, and it's an important question. Before I continue, let me note, that I think after iOS 5, people will be even more likely to upgrade quickly because of OTA updates, and my guess is the absorption rate of iOS 6, for example, will be far greater then previous major releases. This said, chances are, by the time you release your app and iOS 6 comes out, iOS 5 will likely quickly take the back seat, possibly making iOS 4 even less popular (except for older devices).

One key area is the client base that you intend to go for. If, for example, you are building an app that would most appeal to kids in school, or other people that have a higher possibility of being given devices, it would be important to build for iOS 4. This is because my understanding is that many schools are picking up 1st gen iPad, and older iPod Touches, as well as older iPhones. These obviously will only run on iOS 4. I think that is the main thing that will keep iOS 4 on the pie chart - older models.

If you are building a game or something that may appeal more to people that are more technically savvy, it would probably be a much better choice to go with iOS 5. As for other "normal" apps for the general user, valuable resources such as ARC likely will make iOS 5 the better option. I haven't written an app (yet, hopefully), but from what I hear, ARC is hard to pass up, not to mention all the other APIs that you get.

Having talked to developers that still work with iOS 3 and 4, I hear that it is a royal pain, and you will have to expect to spend more time (and consequently money) in the developing stages of the app.

Personally, if I was to program an app, I'm confident I would do iOS 5. From what I hear, iOS 4 is left to people building apps for schools, for example, as well as apps that larger companies that have more cash to work with put out as they obviously want everyone to be able to use their app.

Here is another interesting thought while on those lines - instead of build for iOS 4, why not localize your application, and make it available in some other countries? This would be a way to reach a significant greater number of users.

Again, just my thoughts, but hopefully somewhat helpful.

Answer 2

My take on this is that you must consider a couple of factors in deciding what devices and OS levels to support. The first is that an elderly iOS device still commands a greater level of desirability that for older devices for competitors, and can retain a value at ages where an Android phone will often be recycled or thrown away. As such secondly, there is a large market in "hand me down" devices. I've no figures to back this up, but I am willing to bet that there are way more iPhone 3G devices kicking around than you would expect. These devices often stay in the family, often tied to the same Apple ID, so it can be particularly annoying when an App you already own stops working.

I have iOS 3 and 4 devices (an iPhone 3G and an iPod) that are both in daily use, one as a phone actively, the other as a learning toy for my child with counting/alphabet apps on it (mainly to stop her using my own phone, and because the device had no particular use).

I have found numerous apps that no longer support these devices, when there seems no particular reason for them not to from a functionality point of view, however I agree that if you are producing a new App, I see no reason why you shouldn't choose support the newest OS, especially when you consider that the iOS 6 is going to cover 4 generations (!) of iphones when it is released on the next phone later this year.

But if your App was originally released on iOS 3.0, then I would try to keep it working on 3.0, and if new features require 6.0 APIs etc, then it's time to fork those changes into a new version allowing the old one to still work.