Something I do not recommend:
Becoming so obsessed with your imagined game that you spend all your time trying to find some way to make the core functionality of the game actually work.
Especially after the original solution seemed so promising but turned out to be incompatible.
Then downloading some kind of Tokamak .dll and fiddling with that until you realize that you can't just call the Scripts, you have to call functions and pass the correct parameters, which you will only figure out through excessive trial and error. And for some reason every time you change the slightest thing the simulation stops working.
So you look for another solution and come upon ODE.
For god's sake, do not spend hours coding in movement and gravity and pondering how to properly use external_calls and even more nebulous parameters. It just isn't worth it. You will never figure out how to stop your player from bouncing and spinning (maybe something to do with torque or angular velocity but there's no way your code will be correct).
And why is it so damn hard to include a simple way to make static bodies?
What you should do is basically lament that 1)The new version of GMPhysics isn't ready, 2)Gamemaker doesn't have a built-in physics simulation, 3)You have to use Gamemaker, 4)There don't seem to be any other physics add-ins that do not require the same amount of knowledge to use as creating one from scratch, and 5)You ever thought that getting a little ghost to push a bunch of blocks around would be an easy thing.
I'm not yet ready to scrap my game idea.
But I'm getting there.