Player knockback and add_player_velocity API #8685
Conversation
sfan5
added
@ Script API
Feature ✨
|
Will this PR work server or client side ? |
|
It's client-side, players will have to upgrade. |
sfan5
force-pushed
the
knockback
branch
2 times, most recently
from
4d00a65
to
f30d4ad
Compare
|
Seems a nice implementation. As far as i can see, this adds a velocity to a punched player, with a direction being the line from puncher to punched, and magnitude determined by damage. After the velocity has been added, the punched player's motion is simply affected by the usual player physics. If i'm right this is exactly what i requested in the previous PR. |
function set_velocity(player, vel)
player:add_velocity(vector.multiply(player:get_velocity(), -1))
player:add_velocity(vel)
end |
|
|
||
| core.register_on_punchplayer(function(player, hitter, time_from_last_punch, tool_capabilities, unused_dir, damage) | ||
| if player:get_hp() == 0 then | ||
| return -- RIP |
There was a problem hiding this comment.
Kicking around dead corpses might be fun or even useful. Do we really want to return here?
There was a problem hiding this comment.
All movement is ignored when a player is dead, so this wouldn't work anyway.
builtin/game/knockback.lua
Outdated
| return -- RIP | ||
| end | ||
| if damage == 0 or player:get_armor_groups().immortal then | ||
| return |
There was a problem hiding this comment.
No knockback for immortal players makes kinda sense (maybe there should be an additional group for no knockback).
The damage == 0 should be moved to core.calculate_knockback. Someone might want to make a tool that doesn't hurt but pushes.
Maybe the check for the group should be moved to core.calculate_knockback, too.
There was a problem hiding this comment.
The idea with immortal was the same as with zero damage, so it makes sense to move both.
| local d = vector.length(dir) | ||
| if d ~= 0.0 then | ||
| dir = vector.divide(dir, d) | ||
| end |
There was a problem hiding this comment.
local d = vector.normalize(dir) does the same.
There was a problem hiding this comment.
I still need the length of the initial vector afterwards, this saves one vector.length calculation.
There was a problem hiding this comment.
Oh, sorry, of course.
(Maybe you should return if d is 0.)
sfan5
force-pushed
the
knockback
branch
2 times, most recently
from
62e8895
to
aeca89c
Compare
| @@ -4918,6 +4918,15 @@ Misc. | |||
| of the creative mode setting, checks for "sneak" to set the `invert_wall` | |||
| parameter and `prevent_after_place` set to `true`. | |||
|
|
|||
| * `minetest.calculate_knockback(player, hitter, time_from_last_punch, | |||
There was a problem hiding this comment.
calculate implies that it doesn't modify state. I suggest calling it apply_knockback
There was a problem hiding this comment.
it indeed does not modify state
The intended usecase is knockback, but there's potential for more.
|
Rebased! |
sfan5
force-pushed
the
knockback
branch
2 times, most recently
from
417ef21
to
2b84cfc
Compare
| * Does not apply during free_move. | ||
| * Note that since the player speed is normalized at each move step, | ||
| increasing e.g. Y velocity beyond what would usually be achieved | ||
| (see: physics overrides) will cause existing X/Z velocity to be reduced. |
There was a problem hiding this comment.
Does this mean that add_player_velocity cannot be used to knock a player back with speed significantly higher than walk speed? Also how does this interact with anticheat?
There was a problem hiding this comment.
|
Will this be in the next stable release? |
|
This was merged. So I'd hope it would be 👀 |
implements: #2394 #2960
replaces: #5679 #6489
don't squash
To do
This PR is Ready for Review.
Not implemented:
How to test
Punch another player with different kinds of weapons.