Dinamic position of shape

[sarcasm]

I'm sorry, but your formal specification of that complex problem is too long.

Can you summarise it please.

[/sarcasm]

What do you expect, to randomly rotate shapes around a defined area, until the maximum number of shapes can fit ?

The permutations of position, rotation, size and shape would probably be too great. Also, you'd need to write a generic polygon collision detection algorithm.

If you can't even define your problem, then I seriously doubt that you are capable of solving it.

omcgoverna at 2007-7-15 >

Okay, here's a wild idea, just because the problem has captured my imagination.

You could turn the whole problem into a sort of game of tetris.

When you scan in the shapes, you could pixelate them into oversized pixels blocks.Eg. a triangle might become 3 blocky pixels.

Then start "dropping" the nicest fitting shapes into the box.

When "dropping" the shapes you'll need accurate polygon collision detection.

If this is just a college project, then it should do.

If it's a commercial project, then you may need a more robust solution.

You might randomize the dropping of the shapes a little, and play N games, to work out which one "scored" the most.ie. had the most blocks packed into the defined space. That will be the most efficient packing.

regards,

Owen

omcgoverna at 2007-7-15 >