Cheat Engine

[Aviar³]

If so, how did you guys go about implementing transitions?

[InternetIsSeriousBusiness]

shadeslayer

[gogodr]

global variable
int a = 0;

on an event that moves the state to the next one::
- a++;
- if ( a == 3 ) a = 0;

states are 0 1 and 2

machine states are separated by looking at a
if (a == 0){
//do this
}
else if (a == 1){
//do this
}
else if (a == 2){
//do this
}

___
no?

[Aviar³]

Yeah, hardcoded is the simplest way of doing it, though hardly the optimum way I think. After all, how do you reuse the code? How do you minimize the machine? Etc etc. I'm doing it by having a class for State, StateMachine and Transition. A states simple enough, just a name (and optionally onEntry actions and onExit actions represented by a set of Function objects to be executed) and a statemachine is a collection of states and transitions, the current state and the current input. Transitions are simple in the sense of one having an fromState and toState property, but my problem is modeling the input property which represents what input triggers the transition as well as the best way to capture/represent transitions in a statemachine (for when having to search for what transition to take).

Also I don't think your model/implementation captures the possibility of moving from state 0 to an arbitrary state, such as state 2 or n. Also doesn't capture the idea of moving from a state 2 to a state 1 or 0.

[gogodr]

I can't really assist you, I would like to but your saying is too general x.x I can not fully understand the problem you are trying to solve.

what I mean is that a "state machine" is too vague of a concept to start trying to come up with a solution.

[Aviar³]

The problem is representing the concept of statemachines (particularly finite statemachines without epsilon transitions) in a reusable manner. Reusable here means I can, during runtime, add transitions, states and carry out operations on the same (such as minimizing, finding dead states, paths, etc.).

[gogodr]

just so I dont foget.