父子对象之间的双向通信

Question

I've got a Gameboard containing a lot of dices.

Every can be tilted to either direction (limited by the boundaries of the Gameboard and other dices in their way) by dragging the mouse on top of it in on of the four directions. So the dices themself handle the input but before executing a tilt I have to check whether the move is allowed on the gameboard.

Therefore I'd give every dice a reference to the gameboard but is there a better way to get this two-way behaviour?

Answer 1

Instead of handling this logic inside of the dice objects with a reference to the Gameboard, I'd recommend looking at doing the opposite and instead handling this in an an abstracted system (creating a 3rd element that bridges these two components together and handles the required logic), checking whether each move is valid or not.

To do this you'd need to change the mouse input so that it is also handled in an abstracted system rather than on each Dice, so when you touch a Dice object this system is able to check which object was touched, before then running the code to check if the move is valid on that object. This would most likely involve you storing a List/Array/Dictionary of all the Dice objects, so that you're able to check and see if the touched object is contained within and therefore valid. You can then get any properties you need from the Dice object by using touchedObject.GetComonent<Dice>(); (or you can store all these properties separately from the actual MonoBehaviour and instead look them up).

With this method you can check against the Dice object and the gameboard to see if the next move is valid, without having each object being dependant on the other (which would cause circular dependencies). This also gives the added advantage of giving you more flexibility to change or remove this code going forwards.

This is one of many different approaches you could take, another would be to have your gameboard store a reference of each Dice object and have the Dice object fire an event when interact with (that the gameboard is listening for), so the gameboard code is then able run the check - though from experience I'd go with the former approach and look at decoupling as much as possible.