Disaster averted

These days, the little crane that could constitutes my livelihood. So it would make sense to take very good care of this product. Well, this morning I almost made a monumental mistake that could have done a lot of damage.

When preparing the 1.2 version update, I had put in some testing code that ignores the purchase made by the customer (the game is free, but customers can purchase extra levels). I was about to upload this version to Apple, when I decided to make one more screen shot of the application. I fired up the game on my iPad to make this screen shot, when I noticed that I could not play the level I wanted to, because it was still locked. That's odd, because I thought I already unlocked the content on that device. Then it dawned on me: could the test code still be active?

If this version had gone live, all the paying customers that downloaded the update would have seen their premium content locked again: argh! And I am not sure if Apple's review would have caught this error, as the in-app-purchase got introduced in 1.0 and I am not sure they will retest in subsequent versions. If they do: kudos to them.

Throughout the years, I have learned that paranoia helps when developing software. A paranoid coder is the best coder. And this is especially true during release time. Anyways, let me conclude with what will be coming in the v1.2 release of the little crane that could:

• Big improvement in frames-per-second for smoother gameplay.
• Added level "tower crane".
• Optimized OpenGL ES2 drawing.
• Faster fragment shader (no branching).
• Fixed stutter when first moving levers.
• Fixed crash when quickly changing the level after you fail one.
• Disabled OpenDE asserts for faster physics computations.
• On iPad, prefer interstitials over banners when available.