When you slow yourself down in specator mode you'll notice an annoying phenomenon when trying to move sideways. You're not actually moving at the correct angle. Instead it's as if you're rubbing against an angled wall which blocks your movement on one axis.
[media]
This only affects spectator mode (or flying mode in general, but there's no way to change fly speed in creative), as slow movement on the ground behaves as expected with no oddities.
Reproduction Steps:
1:
Go into spectator and set your movement speed to lowest
Look down and have your facing at 42 degrees
Move left
Slightly move your mouse to increase your angle above 42 degrees.
2:
[media]
Create a void superflat world
Run command
/execute in minecraft:overworld run tp @s -7.50 -55.00 -7.5 40.00 90.00Compare your movement in spectator mode at the slowest speed vs regular speed.
Observed results:
1. Your movement is very slanted until your cross 42 degrees and suddenly change direction. It's like a wall was blocking your movement and suddenly disappears.
2. You clearly move at a different angle than your actual facing so you reach a very different position
Expected results:
Your movement would be consistent with your facing angle.
Linked issues
is duplicated by 1
relates to 1
Attachments
Comments 7
Does MC-241951 describe your issue?
No. That issue completely erases the player's momentum on one axis when that component of the motion vector is small, and only when the player isn't pressing the movement keys.
In this issue your motion vector is deflected towards the closest axis, and while the movement keys are pressed.
If anything it'd be a duplicate of the related report (MC-184530). However I don't think that's the case either because this issue is much more severe in it's symptoms. In MC-184530 your movement is only slightly deflected, meanwhile here your movement is grossly off-axis. I think that's because this issue is specific to flying movement and as such it doesn't manifest when crouching/walking.
To me they all actually seem like the same issue, it's just that this and MC-184530 keep the key pressed with low movement speed.
All 3 reports are reporting the same thing: if speed on an axis is low enough, it zeros out, regardless of total speed.
They do, just not every tick; it appears that the motion is not constantly applied, and on most of the ticks, it is now enough to zero out. That's the only difference as a result of keeping the key pressed.
Fixing the early zero-ing fixes all 3 reports.
This was apparently fixed in 25w02a.
Shugoh
Can confirm.