I am in the process of creating a physics platformer. However, I want the player movement to feel very old-school(as in: non-physics like), which means that the character:

• Should always move equally fast horizontally, even on slopes.

• Has a reasonably fast jumping movement. The 'normal'(earthlike) gravity that is the default of Box2d tends to feel floaty.

Now, I have been trying lots of things to create a Box2d player that works this way. Multiple questions over the past few weeks are related to problems I've encountered with that:

• I increased the gravity and the density of the player to ensure they would jump/fall faster. (see this question)

• I worked with the 'unicycle' model(a square connected to a circle) to prevent getting stuck behind small corners. The bad part was that the player was now unable to stay on any slope, because they kept sliding down.

• To prevent that, I learned that I had to keep the density of both objects low.(see this question) This again created the gravity problem. Also, while the player now was able to keep standing on slopes, moving up on them was still extremely slow (and moving down incredibly fast).

I am getting a bit frustrated now over all this crazy behaviour that doesn't work the way I want it to. Now I am wondering:

Is there something wrong with the approach to handle the player's movement outside of Box2d, and after that let Box2d resolve the new collisions with ground and objects?

Or is this going to be extremely unstable and glitchy because I would 'teleport' the player instead of moving them with forces and is there still another, better way to get the movement I'd like?

I am in the process of creating a physics platformer. However, I want the player movement to feel very old-school(as in: non-physics like), which means that the character:

• Should always move equally fast horizontally, even on slopes.

• Has a reasonably fast jumping movement. The 'normal'(earthlike) gravity that is the default of Box2d tends to feel floaty.

Now, I have been trying lots of things to create a Box2d player that works this way. Multiple questions over the past few weeks are related to problems I've encountered with that:

• I increased the gravity and the density of the player to ensure they would jump/fall faster. (see this question)

• I worked with the 'unicycle' model(a square connected to a circle) to prevent getting stuck behind small corners. The bad part was that the player was now unable to stay on any slope, because they kept sliding down.

• To prevent that, I learned that I had to keep the density of both objects low.(see this question) This again created the gravity problem. Also, while the player now was able to keep standing on slopes, moving up on them was still extremely slow (and moving down incredibly fast).

I am getting a bit frustrated now over all this crazy behaviour that doesn't work the way I want it to. Now I am wondering:

Is there something wrong with the approach to handle the player's movement outside of Box2d, and after that let Box2d resolve the new collisions with ground and objects?

Or is this going to be extremely unstable and glitchy because I would 'teleport' the player instead of moving them with forces and is there still another, better way to get the movement I'd like?