By 
I hate studies. No, not that stuff I do at college (though I absolutely hate that too), but rather the alternative dictionary meaning, “ to investigate or examine, as by observation, research, etc.”
I understand that hate is a very powerful word, and I admit that systematic analysis of data can be a very powerful tool, but the unfortunate issue is that more often than not, the “studies” that gain mainstream attention always, always have a company with something to gain funding it. That’s the reason every other news channel you’ll see shouts out that they’re the “Number 1 News Channel in <put your location here>.”
The latest “study” that has ticked me off was conducted jointly by IDC and Appcelerator. You might know IDC as the well respected brand when it comes to such researches, you might not have heard of Appcelerator.
They’re the guys behind the cross-platform development platform Titanium. And according to them,
78.6% of developers were interested in creating apps for Android smartphones during the first quarter of 2012, down from the 83.3% in Q4 of 2011 and down from around 87% in Q1 of 2011.
Now, you might be wondering why Appcelerator would want to show interest in Android development as waning. The reason is that HTML5 has been highlighted a lot, and HTML5 is the foundation for most of the cross-platform tools, including Titanium. The key mention for HTML5 is “78% of app developers surveyed say they will integrate HTML5 in their apps this year.”
The reason given for the decrease in interest? No prizes for guessing: fragmentation. While I do agree that developing on Android comes with its fair share of challenges, there are some fabulous advantages that come with the platform that others cannot offer.
Additionally, former Gartner analyst Mike King and now Appcelerator’s principle mobile strategist, has been quoted as saying developers are lukewarm on Ice Cream Sandwich. And that, by far, is by far the most laughable excerpt of the entire piece.
Every developer I’ve spoken to absolutely loves ICS. Which developer won’t love an OS that allows them to create an app that is optimized for a tablet and a phone, and hell, possibly even a TV, with one single package?
And let’s remember firstly, a developer using HTML5 will face way more fragmentation issues than one developing native, especially when you take into account targeting tablets as well as phones. As I’ve stated in another article before, I have spent the last half year doing some web development and I now find Android development to be a vacation.
Unfortunately, the numbers produced after surveying 2100 developers, which is less than 40% of the expected attendees of the upcoming Google I/O event, have been thrown about as “another sign of weakness” for the Android platform. Just google “Android development news” and you’ll see a herd of sheep reporting on the survey, having eaten up the propaganda.
I wish I had the numbers to back up my arguments, but I just don’t have the resources to accomplish something like that. I did, however, take a survey when I helped kick start a mobile development group in my college, and from about 150 responses, only two individuals showed no interest in developing for Android.
I do accept that certain socioeconomic reasons play a major factor in the overwhelming level of interest Android generates in India, where I live. However, I just don’t see why developer interest would decrease in a platform that could potentially reach a point where a million new devices are being activated each day some time this year.
As we’ve seen in the Mac vs PC fight, developers do go where the numbers are. They’d be foolish not to.
Studies or not, I know there are a few devs that have announced that they don’t want to develop for Android any longer. For the most part though, it seems like iOS and Android platforms are getting the same apps rolling out these days.
Seriously though, this is because of fragmentation. When you make an app for iOS or Window Phone, it works on all the devices perfectly, most of the time. When you make on for Android, you have to address the app to all the different SoC’s.
which is an issue that won’t directly affect a majority of the apps, and will affect even those made with cross-platform tools like Titanium.
You have to make a separate app for iPhone and iPad. If Apple changes screen resolutions you have to redo all the graphics in your app for each version….phone and tablet. If they add features to a new device or OS version how do you handle older devices and versions? Its the same problems as Android. Its just not publicized by the tech media.
I understand not wanting to develop for icecream sandwich. Took twice as long for me to update my app to just work with icecream sandwich. Still, it was worth it, to play my own app on my Gnex. The main problem for me was getting used to the loss of captive buttons, when programming.
I agree, developing for icecream sandwhich was a pain, but yeah its worth it :)
Then you were doing it wrong.
As a developed who uses apcelleratorT, I do see a tapering off of dollar centric developers who use a strategy of android primary, iOS secondary, html5 as it matures. Since webkit browsers have great html5 support, and both mSafari, upcoming releases of chrome and other webkit based browsers for the android SDK are tightly integrated in html5 endeavor, that would be the reason for the metric. It’s easier to dev for ios, ApcelT a port for android (because android users are used to a lesser application Ux, and complain less) and move resources to html5 which directly targets most mOS equally.
I don’t see this Trend changing for a long time, as html5 is still young, maturing quickly, and has support from big money capital investors, and startup sourcing communities.
I would agree with the study, I’ve spent countless hours trying to get my apps working on every device. This could be fixed if Google would start spending more time on improving the developer tools, and stop changing Android in ways that hurt developers (like removing the ability to have separate notification and ringer volumes in ICS). All this and my apps getting pirated all the time pushes me go give up Android development.
Really? What kind of app is it that you can’t get working on all devices? What tips and tricks are you using. In a little under an hour I’ll be back in my code for an app and I want to know what I should be looking out for.
/sarcasm
Like the Galaxy Note, if you have any apps post ICS and have layouts designed for the small tablet 7-9ish range that is what the Galaxy Note uses, which is crazy. A 7-9 inch tablet layout should not be on a 5 in screen it looks really bad. I can get my apps to work it just takes a lot longer than it should. They have increased the developer options in ICS but I don’t think it factors in screen sizes just resolution, a combination would be better.
When I looked at the size of the Note I’d prefer a tablet interface. What your describing though is not really a problem. Google has apparently decided that 5in screens should provide a tablet experience. If users complain then they could change it. If you’ve followed the best practices then you don’t have to do anything and your app would revert to the phone interface. What you are trying to do is go against the grain. But at least on Android you could do that. Apple wouldn’t approve your app.
Just one note: Google makes a decision regarding the OS based on DPI and screen size, and leaves it at that. A manufacturer can manually make changes to affect the display of the OS and the apps. For example, the original Galaxy Tab, based on purely DPI, fell in the MDPI or the LDPI, because of the larger screen size with a lower pixel resolution. However, the resources it used to display the apps were from HDPI, a manual change made by Samsung understanding that DPI alone would create problems in their case.
Hence the final decision of a tablet or a phone experience lies with the manufacturer.
Buy a Nexus phone, problem solved.
Had I known what type of nazi’s were the “crowd control” for developers at Google, I would have never gotten involved with Android at all. Google’s “see-all, do-all, explain-nothing” type of business with developers, is the key thing that keeps me from logging into Google services, and bragging my love for Google ever again. As a huge supporter from the G1 and up, its Google’s actual business strategy of how they deal with their own human developers, that makes me sway from their products.
The problem with these “studies” are that they are literally comparing apples and oranges. Surveying “developers” at various points in time and then comparing the survey results is only good if you survey *the same developers* during each quarter. However, by randomly sampling “developers” in each quarter they fail to account for other factors that may affect their results. As the % of interested Android developers goes down, the assumption is that last quarter Johnny Nerdling wanted to develop for Android, but for some reason Johnny Nerdling changed his mind this quarter and doesn’t want to anymore. Here’s another way they could have (and likely did) get those results:
Q1 – There are 100 developers in the world. They have been around, “developing” for quite a while. They are the most tech savvy and ahead of their curve individuals out there, because they are the FIRST 100 developers to get into the game. Now, 83 of them (or 83%) want to develop for Android.
Q2(a) – Next quarter, a new group of individual cross over and call themselves “developers.” This new batch of kids aren’t as cutting edge and tech savvy. They may not be as interested in cross-platform developing. They are just learning how to “develop” and want to go after one, easy to use, platform, so they are more fragmented. Only half of those 50 new guys (25) want to develop for Android.
Q2(b) – Meanwhile, interest in developing for Android among the first 100 developers has actually increased from 83 to 88! However, once you add in the new batch of guys, you’re looking at (88 + 25) 113 developers interested in developing for Android out of 150 developers. That is only 75.3% – so rather than reflect the GROWING interest among developers to develop in Android, it falsely reflects a shrinking one.
Q3(a) – A new batch of 50 developer freshmen enter the developer world. Now, these guys are special: They never really wanted to be developers. However, they were enticed into it because of a sexy new platform or technology that promised them an easy way to get rich! Also, they saw the results of the “waning interest” in Android development last quarter and thought, “Android is dead – why waste my time?” Out of the 50 new entrants, 45 of them came to the table just because of that new technology and have no interest in developing for Android. The other 5 are a bit different… they just got in the game late, but they are interested in developing for Android.
Q3(b) – Meanwhile, interest in developing in Android among the senior group of 100 developers has now grown from 88 to 95. The sophomore 50 which only had 25 interested in developing for Android – they have now gotten their feet wet in the developing game and are now branching out into multiple platforms. 20 who weren’t previously interested in Android development are now, so that brings the total to 45 out of 50.
Q3(c) Now, there are (45 + 95 + 5) 145 developers interested in developing for Android. Out of a pool of now 200 developers, that is only 72.5%. On no! It appears that interested in developing for Android has dropped even more!
But the truth is, that interest in developing in Android among *new developers entering the marketplace* has been smaller and smaller. And not because Android is worse or losing popularity. It is simply because the newer and newer batches developers are a markedly different type of people and enter the market in markedly different circumstances: There are more platforms to master at once, there are new technologies being released that might entice them into the game. One very likely scenario is that the typical new developer gets into the game and sees a crowded space in Android development, so he is enticed to get an early advantage and carve out a space for himself on a new platform, so he focuses entirely on that new platform first before branching out.
Meanwhile, interest in developing Android has actually risen. The first group went from 83% to 88% to 95%. The second group went from 50% to 90% in a single quarter! But, by mixing their samples, this study seems to tell the opposite story. When, in reality, there is just a different story to tell, not the one they are trying to tell. The most interesting thing about all of this is that the mere act of telling the story they have chosen to tell is likely to influence the outcome of their next sampling. They are, in fact, writing their own narrative and causing it to come true. This happens in political polls all the time.
Very good point. I have long since believed that many of these “devs” actually just learned to develop on iOS or Android. And the ones on iOS I sometimes think are graphic designers that learned to drag and drop their way to an app. Most devs that are experienced are saying what fragmentation….
Anybody that thinks its easier to build a web app that cross browser than it is to build an Android app that works on most phones hasn’t been doing.g any real web development. I welcome you to leave though and leave more room for my apps to make money.
Wtf man you used the meme wrong
Came in to say that
https://plus.google.com/107370779337689020644/posts/8SWfTAtS8LX
Raveesh,
First off you are very wrong about something… HTML5 is NOT the base of Titanium… though that is a platform you can compile your code for, it is also compiled to android and iOS. Titanium DOES NOT use HTML5 for its UI nor does it use a webview for your code to run in.
So that right there trows off your “this is why Appcelerators results would make android look bad” please check your facts next time before you post stuff like this.
The real reason the numbers have slipped? 90% of the people taking the quiz are Titanium users, and the iOS SDK for Titanium is worlds better then android (no fault of their android team) and thus the results you see.