On Planning and Reality 3 June 2010Posted by Oliver Mason in Apple, iphone, objective-c, programming.
When I got my iPhone a little more than a year ago, and started developing programs for it, I had a clear idea what my first program was going to be. However, as always, things turn out quite different from how you think they are going to be…
First, it did take me a bit to get used to Objective-C. Not because it is very different from Java (I used to program in C after all before Java came along), but because all the classes in the Cocoa framework need to be learned. There are subtle differences between those and their Java cousins, and after a bit more experience I believe that the Cocoa classes are actually more powerful and easier to use than their Java counterparts.
Some teething troubles, lack of automatic memory management on the iPhone, and a surfeit of squa brackets meant further delays. Finally I had a program written, but it needed more work on the graphics side, artwork and so on. The stuff that really makes a difference, but is very time-consuming and hard if you’re not used to using graphics software. So the easier way out was to write a different program, which is lighter on the artwork.
This then was a todo-list program, which is also suitable for planning small projects. I wanted a program like that, but didn’t want to fork out the money for Things, which also looked a bit like overkill. On the life hack blog I read an article by Dustin Wax on his moleskine setup, and that seemed like something usable, which I then went about implementing as an iPhone app. With a bit of help from a friend with the icon design, and thanks to freely available sound files and icons, ePlanner was born.
In ePlanner I tried out Core Data, which is really a lot easier than messing about with SQLite directly. It uses both tabs and navigation views, and a lot of tables. I found it rather tedious in that all the classes were almost identical, but only almost, not 100%, and it’s hard to see how that could be changed. The behaviour of those classes is ever so slightly different.
The submission procedure was very easy, thanks to a description I found on the web. My app did get rejected, due to a crash on a 3GS; but I don’t have a 3GS, so I could only test it on a 3G and an iPod touch. Thanks to Instruments I could track down the error, which was of course a memory management issue, but one without consequences on the machines I could test it on. After that was changed, the app went through, and has indeed been bought by people all over the world.
It is really a nice feeling to think that someone in Argentina is using my app, as is someone in Hong Kong, some people in the US, Sweden, etc. I used some free Google advertising at the beginning, but that is really expensive, though when I stopped it, sales began to trail off. But that could also have been an effect of it slipping out of the ‘newly released’ slots.
It is indeed not too hard coming up with a program that does sell. The overall process is not too hard, though there were some frustrating moments battling with the various code signing and certificate issues that Apple requires.
I since have bought an iPad, and am thinking of porting ePlanner to this; however, I’ll give it a while so that I get used to how the iPad works. Knowing your way round the platform makes it a lot easier to develop good software, and I am not yet sure how the UI design for the small iPhone screen can best be translated to the iPad’s larger display. But it will come, and I will describe the process on this blog…!
In the meantime, I will re-visit some of my previous program ideas, as it is really not hard to turn them into something that will end up in the App Store, and it is really satisfying to do so.