Wall running explanation?

[Ikagura Ikagura]

How come some old games had the player going faster when going against a wall at a certain angle?

 

[Boo Man]

when you move diagonally you move one pixel forward and one pixel sideways
instead of just one pixel forward
so it can be faster

 

[χKai]

Have you ever tried wall-running? It works the same way in real life.

 

[Waffles's iconWaffles]

when you move diagonally you move one pixel forward and one pixel sideways
instead of just one pixel forward
so it can be faster

And here I thought it was a quirk of the Doom engine and its endless clones.

Beautiful explanation, man.

 

[Graw]

when you move diagonally you move one pixel forward and one pixel sideways
instead of just one pixel forward
so it can be faster

To add to this. Modern games tend to use linear interpolation to avoid this issue. Basically you find the magnitude of the direction vector with

[I]sqrt[/I](x²+y²)

Then you normalize the vector by dividing x and y by the magnitude calculated above. This gives you a new vector that you would use in your movement code.

For example, if your direction is (1,1) then you'd do.

sqrt(1²+1²)=1.414
1/1.41=0.707
So your new direction vector becomes (0.707,0.707).

 

[MajesticSiN]

Have you ever tried wall-running? It works the same way in real life.

This is true
I've done it exactly three times

 

[Sumea]

How come some old games had the player going faster when going against a wall at a certain angle?

So, I am not expert but what you are describing to me is "surf" that is there especially in Source engine, and has had maps around the "mechanic". To me it feels like a subset of bunny hopping which I will not get into much, since not an expert.

Surf seems to work, as I understand, in manner where the slope is steep enough to cause player character to be pushed down the slope, but also not steep enough so player can cancel the slide down effect with their own player character movement. The engine is likely taking the diagonal (Cancelling slope falling plus forward movement) input from player, and the physics calculation where player is supposed to be pushed into one way, confusing these, ignoring acceleration limits set to player movement as it adds the slope push to the player, and manipulating it with right type of map, the player can, as community named it, "surf" this even in small wavy movements and end up going fast. The other side of coin being you are on super ice physics and also unable to stop easily.

Comparing "surf" to other stuff, Bunny hopping was a common quake engine and it's deriatives "feature" where diagonal movement was faster than forward only as it had both forward and sideways movement to it. The hopping lets player amass much higher speed as deceleration is programmed to happen only on ground, and if you jump immediately after touching ground, or buffer it as the games let you, and move the aim diagonally and alternate left and right diagonal movement forward, you can go in straight line and even learn to control yourself. And here i said I will not get into this one that hard.

One additional cousin to these in my mind at least is "Nintendo did not think you going backwards". Basically, programming in many nintendo games do not suppose that you would move backwards, or, have negative acceleration as the games are programmed to apply only positive speed up value to the character from player input. There is never a time when player character has negative acceleration so there is no reason to program a speed limitation to negative speed value acculumination. Would be silly if someone super backwards jumped through a wall in Mario 64 or super swam through the entire ocean in four seconds in Wind Waker somehow.

 

[Michael Frank]

It looks cool, and that does it for me.

 

[RageBurner]

Probably the power of friendship, the heart of the cards, or something.