I'm finding a few odd issues which I'll post here for investigation.
In my iOS game I have an object which I create when you collect something. That object has an event to increase the blend coefficient by 5 in an always event. I noticed an odd behaviour where if one instance is active when I create a second instance, the first's blend coefficent seems to be reset to 0 (opaque) and fades out until it goes away.
While this could be a bug in my events (I shall continue to investigate) I have created a test app to try to replicate the issue. See attached.
Firstly, note that I use an ALWAYS event and then an action to set Blend coefficient of object = Blend coefficient of object + 2. I am not using a fast loop to go through each object instance but I'm expecting that the natural behaviour for this event would be to take the Current Blend coffecient value from each instance and add 2 to it.
Secondly, Active objects do not have a built in event condition to test against the current Blend Coefficient value, so I use a Compare two general values condition and in the first box I retrieve the current blend coefficient value. The action I put against this is to reset the blend coefficient to 0.
Thirdly, I create a new instance of the active object 32 pixels to the right of the last one created (using a global value to keep track of the next position)
Now the testing:
In the windows runtime->
1) the objects created every second fade out slowly as I'd expect due to the first event adding 2 to the current value. Each instance seems to behave as I'd expect.
2) The second event (to reset the blend coefficient to 0 when it >250) never seems to be true so the objects fade out and never reappear / cycle around. This seems wrong, but I wonder if MMF is detecting one instance is not >250 and this is causing the event to ultimately resolve false?
In iOS runtime:
1) Not sure if this is working the same, hard to tell. Not been able to replicate/prove the problem in my iOS game (I shall keep trying) but firstly these objects appear to cycle around and once faded out they reappear to fade out again. Not sure if this is due to iOS treating 256 and 0 or if the second event is triggering true.
Also when there are about 30 objects created, no more are created (I assume because too far from frame) and then after a while all the objects blend coefficient synchronise and they all have the same transparency setting (this is different to MMF)
2) As observed above, I am not sure but the second event to test if the blend coefficient > 250 might be working for iOS where it does not for Windows. However based on some other weird behaviour I have noticed in my game I suspect that a blend coefficient >255 on iOS might be wrapped around/treated as a value between 0 and 255 as if it was being MOD'd by 255?
I shall continue to try to prove if the original issue I am trying to track down is a bug in my events or a difference in behaviour between iOS and Windows runtimes however as the example file shows there are some problems/differences highlighted in both the Windows runtime and iOS runtime for you to take a look at.
