There have been endless articles written debating which platform is more popular with developers, Android or iOS. Some have claimed that iOS is more developer friendly, while the other side claims that Android is bound to win and so developers should focus on that. After analyzing all our data, the verdict is: Android is now more popular than iOS with developers.
It seems to me that previous Stackoverflow statistics posts have made similar errors with the way they state things that they've learned based on traffic to their site - the statement "Android is more popular than iOS with developers" should be followed by the phrase "for visiting www.stackoverflow.com".
I'm not sure how you can extrapolate something as broad as "popular" from what OS is used to visit their website. Popular in what way - what mobile OS they target when building apps? What OS they like to use on their personal devices? Etc.
While I don't disagree that Android is certainly popular and growing ever more so, I don't think the methodology in place really justifies making comparisons of "popularity". This is a place where the homogeneity of the iOS platform really comes in - there's a pretty limited set of devices, which eliminates a lot of the "how do I get Android to do X on Y handheld" questions that would distort their numbers. The Android APIs have certainly been evolving more rapidly in the timeframe under question than iOS, as well - there's a broad base of previously answered questions to draw from without asking another one.
The other way of looking at this data is that Android developers need more help [edit: on StackOverflow] than iOS developers. Which does not imply anything on popularity but tell something very interesting on the quality of the SDKs or developers… (I'm neither an Android nor iOS developer)
Android developers need more help from StackOverflow than iOS developers. Aren't there private forums for Apple developers, that you can only see if you are a registered developer?
Your argument might have some merit, except that you could apply it to almost anything and get non-sensical results. Django for example has close to half of the activity that Ruby on Rails does. Does that mean that Rails has severely worse documentation or APIs? Doubtful.
The more developers use a piece of technology, the more that will have questions about said technology. Poor documentation will certainly lead to more questions, but you can't look at these graphs and have any real data to support your case one way or the other.
Interesting that the author of the article chose to post a link on HN, with a substantially more linkbait title than he picked for the article itself.
The "SO has more activity around Android which proves more developers are using Android" claim is so obviously nonsense that the whole thing just looks like a shameless attempt to get Stackoverflow into news headlines.
I'm not surprised that this is true among the population of stackoverflow users. There are many thousands of devs that have no use for, no time for, and don't care about sites like stack overflow, though. (like me...)
While the methodology used is somewhat lacking (as lots of others have commented on here) I don't think the claim sounds entirely unreasonable either.
Anyone on any platform can develop Android-apps. At least any platform with a JVM, and that's quite a few. Only people who have invested in Mac-hardware can create iPhone apps. That represents around 7% of the machines out there (according to wikipedia :1). Mac- usage may be rising, but Mac is clearly the underdog, and developers are not that different from most people. So the statistics implies that most developers are not using Macs.
So if we accept these terms as reasonable, and they remain reasonably unchanged over time, there being more iOS developers than Android developers would in fact be a very, very strange thing.
I'm not saying this data proves anything, but I don't think it proves anything the other way other as some commenters here have hinted (like the Android SDK being of significantly lower quality than iOS SDK).
I can't possibly be the only one here thinking along these lines?
I don't think we can make an absolute comparison based on those numbers, but the trends should be fairly significant. So it could be safe to say that the number of iOS developers slowed its growth or even stopped, while the number of Android devs is growing faster than before.
Just a personal note here: I just finished developing my first android app using a cross platform solution; I could have started with ios... but I don't have a mac!
Apple has great developer forums at deforums.apple.com.
I suspect most iOS developers go there when they need help. If not, they hang out at any of the dozens of developer forums that started back before the AppStore was created or in the years since.
Further, iOS is a very well designed set of APIs that are very well documented by Apple with very extensive example code.
I've asked a lot of questions about iOS, though mostly about things that are in beta or under NDA, and I've never even considered asking on Stack Overflow.
I agree. A real data analyst or mathematician or anyone collecting survey data for that matter first must study the sample, not the output, which is what has been done in this case. The sample here is stack overflow which is a community of people looking to solve coding problems. There could be a few things at work here:
1. iOS devs hang out elsewhere
2. There are fewer iOS devs, who have tackled and solved the problems already
3. iOS devs don't need a community to solve problems
I'm not suggesting one or all of these are true, just saying these things need to be considered before positing a position based on data. Again, the sample defines the output in a much greater way than the data (e.g., if I survey 100 fisherman about whether they develop iOS apps, 100 will say no)
[+] [-] brown9-2|14 years ago|reply
It seems to me that previous Stackoverflow statistics posts have made similar errors with the way they state things that they've learned based on traffic to their site - the statement "Android is more popular than iOS with developers" should be followed by the phrase "for visiting www.stackoverflow.com".
I'm not sure how you can extrapolate something as broad as "popular" from what OS is used to visit their website. Popular in what way - what mobile OS they target when building apps? What OS they like to use on their personal devices? Etc.
[+] [-] hullo|14 years ago|reply
[+] [-] friggeri|14 years ago|reply
[+] [-] smackfu|14 years ago|reply
[+] [-] Pewpewarrows|14 years ago|reply
The more developers use a piece of technology, the more that will have questions about said technology. Poor documentation will certainly lead to more questions, but you can't look at these graphs and have any real data to support your case one way or the other.
[+] [-] chugger|14 years ago|reply
This is Google we're talking about. Support is virtually non-existent.
[+] [-] ajg1977|14 years ago|reply
The "SO has more activity around Android which proves more developers are using Android" claim is so obviously nonsense that the whole thing just looks like a shameless attempt to get Stackoverflow into news headlines.
[+] [-] qeorge|14 years ago|reply
Anecdotally, we do Android development but not iOS.
[+] [-] acangiano|14 years ago|reply
[+] [-] jcizzle|14 years ago|reply
- Things educated people don't believe
[+] [-] peapicker|14 years ago|reply
[+] [-] josteink|14 years ago|reply
Anyone on any platform can develop Android-apps. At least any platform with a JVM, and that's quite a few. Only people who have invested in Mac-hardware can create iPhone apps. That represents around 7% of the machines out there (according to wikipedia :1). Mac- usage may be rising, but Mac is clearly the underdog, and developers are not that different from most people. So the statistics implies that most developers are not using Macs.
So if we accept these terms as reasonable, and they remain reasonably unchanged over time, there being more iOS developers than Android developers would in fact be a very, very strange thing.
I'm not saying this data proves anything, but I don't think it proves anything the other way other as some commenters here have hinted (like the Android SDK being of significantly lower quality than iOS SDK).
I can't possibly be the only one here thinking along these lines?
:1 http://en.wikipedia.org/wiki/Usage_share_of_operating_system...
Edit: Downvoted? Why? Genuinely curious here.
[+] [-] smackfu|14 years ago|reply
[+] [-] danmaz74|14 years ago|reply
Just a personal note here: I just finished developing my first android app using a cross platform solution; I could have started with ios... but I don't have a mac!
[+] [-] tabbyjabby|14 years ago|reply
[+] [-] r00fus|14 years ago|reply
The article title itself isn't much better (but in the context of stackoverflow blogs, it is a bit more constrained).
[+] [-] nirvana|14 years ago|reply
I suspect most iOS developers go there when they need help. If not, they hang out at any of the dozens of developer forums that started back before the AppStore was created or in the years since.
Further, iOS is a very well designed set of APIs that are very well documented by Apple with very extensive example code.
I've asked a lot of questions about iOS, though mostly about things that are in beta or under NDA, and I've never even considered asking on Stack Overflow.
[+] [-] seanMeverett|14 years ago|reply
I'm not suggesting one or all of these are true, just saying these things need to be considered before positing a position based on data. Again, the sample defines the output in a much greater way than the data (e.g., if I survey 100 fisherman about whether they develop iOS apps, 100 will say no)