Developers Losing Interest In Developing for Android

[Michael Knutson]

<div class='os_post_top_link'><a href='http://macdailynews.com/2011/04/26/apple-ios-shines-in-developer-interest-android-fades-on-fragmentation-weak-tablet-concerns/' target='_blank'>http://macdailynews.com/2011/04/26/...ablet-concerns/</a><br /><br /></div><p><em>"Appcelerator and IDC surveyed 2,760 Appcelerator Titanium developers from April 11-13 on perceptions surrounding mobile OS priorities, feature priorities, and mobile development plans in 2011. The survey reveals that developer momentum is shifting back toward Apple as fragmentation and tepid interest in current Android tablets chip away at Google's recent momentum gains."</em></p><p><img src="http://images.thoughtsmedia.com/resizer/thumbs/size/600/at/auto/1303950490.usr17748.jpg" style="border: 1px solid #d2d2bb;" /></p><p>More interesting statistics, guaranteed to generate at least some some controversy. According to these sources, interest (for Android) is waning in developing cross-platform applications, at least in what may be considered more generic apps. Disappointing and fragmented Android sales have no doubt impacted development for smaller companies that don't have resources to commit to multiple platforms. iOS interest is still high, followed by Android, then, surprisingly, Windows Phone 7. As with all statistics, I'd expect these numbers to fluctuate, and maybe even match up to tablet sales. Competition for the number three position may be the most spirited, with MS, RIM, HP and others lagging far behind Apple and Android, but with possibly compelling products on the horizon.&nbsp;</p>

[Craig Horlacher]

I'm not saying I don't believe the chart but I keep hearing about developers not wanting to develop for Honeycomb and I just find it hard to believe. It seems like every day I get and update for one or more apps to add or improve Honeycomb support.

I also think fragmentation can be a huge benefit but providing different form factors and other features that may be important to one person but not another.

[martin_ayton]

Laridian's latest blog may provide a little insight into this (see http://blog.laridian.com/?p=302). Craig's general points are summed up in this quote:

"After a few more experiments I resorted to Google. I was able to find a blog article describing how someone with the same problem had downloaded the Android operating system source code to figure out what it was expecting. After a lot of research, he came up with an obscure line of XML that caused a progress bar to be displayed. This line of XML wasn’t in any of the documentation. The only way to figure it out was to read the source code for the operating system. Think about this for a minute: Imagine if you had to download and read the source code for Word in order to figure out how to create a table. Or if you had to read the Internet Explorer source code to figure out how to open a new tab. This is insane. But it’s what every single day of Android programming is like.
Finally, consider this: Before starting programming for Android I spent $400 to buy a “developer device”. This is an unlocked Android phone that allows me to install unsigned applications and run them under the debugger. Little did I know that when I bought it in August 2010 it would have Android OS 1.6 on it, ten months after 2.0 was released and 1.6 became obsolete, and three months after 2.2 was released. It’s now eight months after that and there is still no way to upgrade this phone to any version after 1.6. Meanwhile, the iPhone I use for development not only came with the latest iOS version but iTunes automatically updates it every time I plug it in."


That sort of thing would dissuade me from developing for Android pretty quickly. It isn't related to platform fragmentation and it is completely fixable - come on Google, it's time to raise your game here.

[Jason Dunn]

Makes you realize what a hard, uphill battle RIM is going to have getting native QNX apps for the Playbook. And HP's webOS? Yikes. Another hard battle there...and I'm not someone who believes that HTML 5 web apps are somehow going to save the day. Native apps will always be better and offer more, at least for the next few years.

[Jason Dunn]

"After a few more experiments I resorted to Google. I was able to find a blog article describing how someone with the same problem had downloaded the Android operating system source code to figure out what it was expecting. After a lot of research, he came up with an obscure line of XML that caused a progress bar to be displayed. This line of XML wasn’t in any of the documentation. The only way to figure it out was to read the source code for the operating system.

Wow. That's BRUTAL. Google really needs to improve the platform for developers - by making it so hard for them, they won't put money and effort into developing for the platform. And without developers for your platform, you're dead.

[Phronetix]

Seeing this is not shocking. I think it reflects that developers want to move to where the perceived future money will be (another hindrance for the playbook, IMO).

It seems to me that open source development works very well, up to a certain critical point, when a more controlled environment may drive the developer confidence more than being open would.

I wonder if that point is now for the Android OS.

On a side note.... I openly wonder if this development headache, is it temporary, or is there reason to believe it will at some point lead to a user migration away from the Android OS? The reason I say that is because I already see the market publishing market share in such a way as to minimize or ignore biases and variables such as customer satisfaction, customer demographic, number of carriers, device cost and OS generation. And I am sure there are more to consider that would pain a more accurate picture. In the future, a customer backlash against the Android OS could effectively be covered up by the data, and that is not a healthy way to evoke longterm fondness among those whose livelihoods depend on a stable user base. But what else is new. The bottom line is that if the hand that feeds the OS, the developers, are gone, so is the OS. Then so are the customers, eventually. I mean, sheesh, I see the odd Palm device around still. :-)

[Dyvim]

I think it depends on the type of development. If you are an indie shop looking to publish paid apps then I think for now iOS is the proven place to be. But if you are looking to publish a free app and are building some kind of service (where your revenue will come from elsewhere) and you want to reach the widest audience, then I think it pays to develop for both iOS and Android. And of course if you're some big corporation, then of course you need both iOS and Android apps.

I think there's a lot of consulting work for developers to create Android apps for companies that need an app (most likely to be distributed for free) but don't have the in-house expertise to do it themselves. Consulting work of this type can generate a much steadier income when compared to playing the AppStore lottery.

Anecdotal evidence: I went to an NSCoderNight get-together last week (for iOS/Mac developers) and about half the people there were working on Android apps (all consulting jobs not indie apps). And these are dedicated iOS enthusiasts.