I have an idea for what you mean!
This is the way I imagine building maps:
You have graphics, and load them into one actor, call him the map. Then you determine the size of the map (say 10 x
So depending on the size of the game, I say, you clone them either in GE, or through code.
That's for making it visible.
For making it functional, I say you make an array in code, either a 2-dimensional one, or a one dimensional, with width*height elements. To determine a row there: the row is: (index - (index modulo width))/width. That formula first subtracts the modulo, that is, what element the index is in a row: subtracts it to get a number divideable by the width, then divides with the width. The column number is index modulo width. Watch out, that means the lef-most row is row 0!
The array should be an array of either numbers, or whatever you want to use to describe map elements. If that involves more than 1 variable, then create a struct, with different variables, and array the struct.
When creating the map, cloning in runtime (during the game) you can read these values from anywhere, or generate random values through loops.
A number variable can determine the animindex, or animpos.
When you get a function that would make actions specific to one tile, then use the clone number to do stuff... the clone number will correspond to the numbering in the 1-dimensional array!!! That's how I fit everything together...
Hope that helps!
Hope I got your question right