I am sure with MFA files to demonstrate the errors Andos and Francois will be able to figure out the problems.
I am sure with MFA files to demonstrate the errors Andos and Francois will be able to figure out the problems.
Absolutely Jeff. I'm looking forward to hearing back from them once they've had a chance to troubleshoot the problem. I just wanted to make sure I provided enough information for them to take a look at the problem. I hope my post didn't come off as negative. Other than a few pain points everything is great about the product.
I found the bug
It will be fixed for next release
Great Andos!!!
Okay, I stepped through the Objective-C code for the Advanced Game Board Object and sorted out what is going on. There's a key variable (Looped) used to store the board data being looped through that is not initialized unless you actually put an event into your Frame that calls Set board size to [Width], [Height] at the start of the frame; do this and from that point forward everything works fine. Do not do this and your app with get a Bad Access Error and abend even if you set it on the object in your frame through the UI. The runtime is supposed to derive this from the actual property bag on the object and set it, but it doesn't. I'll let the iOS Device Exporter guys worry about that level of detail. For now everything is fine and working if you do what I describe above.
Well done, Keith! I'll check that on my project later today.
Thanks for the workaround, Keith. I've posted my project here
It's an example on using the Accelerometer object for detecting device orientation by reworking the Brix example for the Move It Object.
It turned out to be a memory corruption bug btw.
(Warning: technical talk)
The "Looped" array (for holding currently looped bricks) were allocated fine, but the size of the game board was incorrect at object creation. It was created with a correct size when using the resize board action which is why the workaround worked.
The wrong size of the board caused values inside the "Looped" array to get overwritten by bogus data.
It was a weird bug to find as the cause was simple (easy to spot when you read the code) but the side effects pointed in a completely different direction
Great stuff Andros. When I created my workarounds for the INI and Advanced Game Board I knew you'd get in there and sort out the root cause. Look forward to the new release when it's ready.
Posting Permissions
