Hacker News Reader

Comment by akurilin | Hacker News Reader

Why not an Android app instead?

To develop an iPhone game you need a relatively modern Apple desktop/laptop (at most 2 year old if you plan to run xcode 4.5, thus fairly expensive) and an iOS device (the cheapest you could go would likely be a 4th gen used iPod, still somewhat expensive) + a yearly $100 license (not sure if this is waived for students?). You also end up learning a somewhat insular language with a ton of idiosyncrasies (Smalltalk on top of C with a somewhat uncommon take on memory management).

I feel like you get a lot more bang for your buck with Java and the ability to develop from any platform out there. An old Ubuntu box will do just fine, or Windows for that matter.

View on HN

Not only that but you can't share the game with your friends (or can you?) Is there a way for them to install the game without also getting a $100 license?

As for Java though, I think kids would do a lot better with JavaScript ;-)

Take this with a grain of salt because I am surely biased; Objective-C is a great language. I really like the Smalltalk method calling style they've used, and you can still use C to mess with pointers if you know what you are doing. If one learned Objective-C first it shouldn't be that difficult to switch to any other object-oriented language, other than memorizing the order of variables in every method. (Also, retain/release is hardly uncommon and with ARC the programmer can basically act like everything is garbage collected anyway.)

That aside, the price of entry to run on iOS is an excellent objection, as is (to some degree) the lack of code portability. I don't know how simple GNUStep is to get running, but it's certainly not as ubiquitous as all the JVMs.

MakeGamesWithUs developer here. Your point about the required computer is fairly valid, but you could get away with having a early-2009-generation computer, or with the cheapest Mac mini ($600).

If you were truly strapped for cash, an iPhone/iTouch device is not actually necessary given the iOS simulator. And we'd totally be willing to test games on our in-house test devices if a developer doesn't have an actual device. Ashu, the guy who wrote the blog post, actually developed his game before he got his first iPhone.

Developers also don't need to worry about the license, as the game is published through us.

I'd disagree with your point about Objective-C. Firstly, automatic reference counting (ARC) mostly eliminates the burden of memory management. Also, learning the language is a requirement if you're at all serious about getting into mobile game development or mobile development in general. Any serious mobile app/game on the market HAS to have an iOS version.

While Android certainly has greater market share than the iOS, the fact is that Android has lots of device/OS version fragmentation. Greater proportions of iOS users, on the other hand, tend to upgrade their devices. This makes us comfortable enough to drop support for devices older than the iPhone 3GS and not have our developers worry about backwards compatibility (the new iPhone 5 screen is a non-issue). People have also found that despite lower market share, iOS versions of the same game offered on Android usually gross more revenue.

That being said, we are definitely moving towards publishing on Android soon.

With respect to device fragmentation on Android: why do you even have to support all the devices with your first game? Thinking of it another way: by developing for iOS, you "dropped" support for even more Android devices (all of them). So why not just develop for the latest Android version, and expand if your game becomes successful.

Is the "iOS grosses more than Android" thing still valid? Have any of your interns made good money with their games?

To clarify, my comment was only related to the didactic aspect of the program with respect to software development.

Regarding everything else, the business side in particular, I completely agree that you cannot do much better in mobile than with iOS today. Perhaps there are some esoteric options with 3rd world cell-phones and selling services through SMS, but let's overlook those for now. That of iOS is a demographic of people who are willing to spend money, and thus it's a good match for your need. As you mentioned, assuming iPhone games, only having to support 1 OS version and 1 resolution with 2 easily bridgeable screen sizes is a big boon.

On a sidenote, I still don't agree that iOS memory management is at all a good entry level paradigm. Reference counting, choosing between strong, weak, copy, assign, unsafe_unretained etc., retain cycles, transferring memory ownership from CoreFoundation libraries are all not trivial concepts for the average 16 year old. I'd be curious to read a blog post from you guys to see what your experience with this will have been.

Using industry-standard game libraries (cocos2d/kobold2d) with ARC, a 16 year old would never need to touch reference counting or the CoreFoundation (the exception is using Box2D physics). It is also not necessary to understand copy, assign, and unsafe_unretained, though it certainly might help in some edge-case situations, with which we would be glad to assist any student developers.

And while it is more helpful to know the difference between strong and weak references, that is IMO a very accessible concept.

so you are advocating that kids should drop +600$ to make games. While they could code game in Lua, python, any other language/script that would be more platform agnostic and probably more rewarding for learning. I would even put that coding in javascript for a game seems cheaper more cross plaform and more rewarding, and allow them to expand if they want to learn more.

Is not clear to me if is only on-site, or for anyone in the world...

Anyone in the world.

Great!

What happend if somebody (me) is already familiar with xcode/obj-c and have created apps but never games?

Sounds like MakeGamesWithUs has installed an Apple Koolaid dispenser in the break room

Nothing wrong with Android - lots of good reasons to build Android apps.

But...I'm running Xcode 4.5 on a 4 year old MBP (late '08) that still performs like a rockstar. And if you are going to build for Android (and take it seriously) you'd need to invest more than the cost of an iPod Touch to have enough devices to cover fragmentation across screen sizes and manufacturers. And I guess (?) you could call Obj-C insular - but learning it opens the door to deliver experiences to a huge chunk of mobile users. Long way of saying, I'm struggling with your "bang for your buck" argument. Especially if you're hoping to monetize your app. Because the primary ecosystem for that still seems to be iOS.

Others seem to agree: http://techcrunch.com/2012/09/25/ios-android-appcelerato/

Not the main point, but I just checked and have Xcode 4.5.1 running on an early 2007 MacBook. So about 5.5 years old.

Of course that still means a MacBook is needed, just not a recent one.

Eclipse is pretty heavy though, and it is my preferred Java/ Android editor.

Could you get away with any of the IntelliJ products instead of Eclipse?

Of course. The Community Edition (free and open source, no field of use restrictions) of IntelliJ IDEA is just fine for Android Application development. You can develop for Android without an IDE at all.

Reply on news.ycombinator.com