DDraceNetwork

Can somebody explain how this is even possible?

https://discord.com/channels/252358080522747904/295908390956433410/1159586875321897061

I've tried pretty much anything but I am unable to reproduce it even slightly. as you can see in the screenshot screeny somehow ends up at 1154.09 I only get to 1153.78 with the exact same coordinates as screeny. so far it's safe to say that the position from the other tees around don't affect the movement of the tee.. if anybody wants to try for themselves.. Map: Adrenaline 5 y 160.53 | x 1150.00~ lag24 : 1149.31 -2 : 1150.69 screeny: 1152.25 -6 : 1158.91 (edited)

movement.demo

818.8 KB

Steinchen

Can somebody explain how this is even possible?

https://discord.com/channels/252358080522747904/295908390956433410/1159586875321897061

I've tried pretty much anything but I am unable to reproduce it even slightly. as you can see in the screenshot screeny somehow ends up at 1154.09 I only get to 1153.78 with the exact same coordinates as screeny. so far it's safe to say that the position from the other tees around don't affect the movement of the tee.. if anybody wants to try for themselves.. Map: Adrenaline 5 y 160.53 | x 1150.00~ lag24 : 1149.31 -2 : 1150.69 screeny: 1152.25 -6 : 1158.91 (edited)

could you give a clip with the position/speed/angle hud enabled?

Patiga

could you give a clip with the position/speed/angle hud enabled?

2

10:25AM

angle: 82.79 top speed: 15.62

Patiga

could you give a clip with the position/speed/angle hud enabled?

The best thing would probably be to export the movement Any further details you might need are in the file I sent above :D

@Steinchen btw did u try to reproduce it on a ddnet server?

Jupstar ✪

@Steinchen btw did u try to reproduce it on a ddnet server?

Yes

10:26AM

Furthest was .88 by walking from any position

10:29AM

for some reason he just won't get frozen

Steinchen

for some reason he just won't get frozen

why did u remove the speed in the second xdd

10:32AM

that's the interesting part

sec :D

@Jupstar ✪ could the first image be interpolation for the graphics?

Okay so this is actually really weird from the same coordinates as screeny's i get frozen really late

Steinchen

Okay so this is actually really weird from the same coordinates as screeny's i get frozen really late

and you end up in a different spot?

Jupstar ✪

and you end up in a different spot?

Nope, always .78

10:35AM

here also, even later I get frozen

Steinchen

Nope, always .78

i mean, different spot than screeny

Jupstar ✪

i mean, different spot than screeny

Yes, but never anything further than .88

Patiga

@Jupstar ✪ could the first image be interpolation for the graphics?

i guess it'd be the interpolation for prediction then. but if they end up on a different spot something else going on

with 15.62 speed i reach coordinates between .59 and .88

10:43AM

Idk how much it matters but it's always the same coordinates with 15.62 speed .59 ; .62 ; .66 ; .69 ; .72 ; .75 ; .78 ; .84 ; .88 (edited)

@Steinchen on an official ddnet server, right?

Patiga

@Steinchen on an official ddnet server, right?

Yes

@Steinchen very interesting clip, according to my math, this shouldn't be possible Small dive into teeworlds physics:

1 tile = 32 units, so to pass a freeze tiles we'll need to pass at least 33 units, in your case you ended at .09, which implies a distance of 36 units
50 physics ticks per second
max walking velocity = 10 units per tick
friction (applied 1 tick after frozen) halves the velocity each tick
the position is changed by the velocity, but rounded ```

10 -> +10 (tick were we first move into freeze) 10 -> +10 (now frozen, but last movement tick) 5.0 -> +5 2.5 -> +2 1.25 -> +1 0.625 -> +1 => 29 ``` this assumes perfect positioning, and suggests a theoretical maximum decimal of .91, which is 1 unit apart from your .88 record

1:52PM

the speed in the bottom right is in tiles per second. it shows 15.62, and is probably just cut off and it should be 15.625 math: 10 units per tick * 50 ticks per second / 32 units per tile = 15.625 tiles per second (edited)

1:53PM

if you'd have a starting velocity of 11 units/tick:

11 -> +11
11 -> +11
5.5 -> +6
2.75 -> +3
1.375 -> +1
0.6875 -> +1
33

this would be barely enough to pass the freeze tile with perfect positioning, but still isn't enough for the end position in the video, which is 3 units further

1:55PM

starting velocity of 12 units/tick:

12 -> +12
12 -> +12
6.0 -> +6
3.0 -> +3
1.5 -> +2
0.75 -> +1
36 (1 tile, 4 units)

with perfect positioning, the starting velocity 12 would reach exactly the shown position

Steinchen

for some reason he just won't get frozen

this could be normal, in 1 tick, the tee can go up to (10 - 1) * 1/ 32 = .28125 into a tile, before the tee gets frozen

2:01PM

the deacceleration does seem to match my calculations

2:05PM

the hud thing on the bottom-right seems to update every second tick

2:09PM

1153.19: first tick in the freeze
-> ~1152.8775 tick before freeze. 0.8775 = 28 units into tile
28 + 29 should land the tee at 57 units, so 25 or 0.78125 into the tile
HOWEVER, assuming that somehow the tee gets frozen one tick too late, we get another 10 units
28 + 29 + *10* = 67 = 2 tiles and 3 units

2:09PM

2 tiles and 3 units is exactly where the tee landed

2:09PM

my current theory is that the tee is frozen one tick too late, which I have not heard of before

2:13PM

I only get to 1153.78 with the exact same coordinates as screeny.

this perfectly matches my math in the last code block

@ScReeNy

2:44PM

thank you for the detailed answer poggers2

2

2:48PM

Also, if you jump against the roof of the tile you are able to quite easily get to .91 Does that mean that you are able to achieve .94 with perfect setup?

Steinchen

Also, if you jump against the roof of the tile you are able to quite easily get to .91 Does that mean that you are able to achieve .94 with perfect setup?

I don't think jumping helps

Interesting stuff, i think i had a small lag when it happened. I wonder if client lag can provoke it or if its a server lag

Patiga

@Steinchen very interesting clip, according to my math, this shouldn't be possible Small dive into teeworlds physics:

1 tile = 32 units, so to pass a freeze tiles we'll need to pass at least 33 units, in your case you ended at .09, which implies a distance of 36 units
50 physics ticks per second
max walking velocity = 10 units per tick
friction (applied 1 tick after frozen) halves the velocity each tick
the position is changed by the velocity, but rounded ```

10 -> +10 (tick were we first move into freeze) 10 -> +10 (now frozen, but last movement tick) 5.0 -> +5 2.5 -> +2 1.25 -> +1 0.625 -> +1 => 29 ``` this assumes perfect positioning, and suggests a theoretical maximum decimal of .91, which is 1 unit apart from your .88 record

I made a small error in the conclusion here, the maximum distance should be 0.88. that is the distance you managed while testing as well

3:00PM

@Steinchen do you reach 0.91 with normal walking speed?

ScReeNy

Interesting stuff, i think i had a small lag when it happened. I wonder if client lag can provoke it or if its a server lag

afaik, it must be a server lag. the client doesn't have that sort of control over the server

Patiga

@Steinchen do you reach 0.91 with normal walking speed?

Never have

Patiga

I don't think jumping helps

I do get a difference by jumping though

.84 should still be the max

Patiga

@Steinchen do you reach 0.91 with normal walking speed?

I did end up reach .91 with jumping against the roof tho

3:20PM

less friction?

in theory, yes

3:21PM

but I don't think you should actually be in the air

3:21PM

if you have an active hook above you, and go into a horizontal 1-tile gap, you walk around with walking speed while glued to the ceiling

Steinchen

I did end up reach .91 with jumping against the roof tho

do you have a clip of that?

3:22PM

(with position + velocity ^^)

Original message was deleted or could not be loaded.

please don't advertise bot links

Patiga

please don't advertise bot links

i just ping @Steinchen message

My bad. There were bots and the chat was still open

3:24PM

GetSpeed.demo

358.79 KB

Steinchen

My bad. There were bots and the chat was still open

don't understand that yet :/

Patiga

don't understand that yet :/

Might be because of reduced friction?

but friction is massively lower in the air. if you'd get an air tick perfectly you should be easily able to skip the freeze tile

Patiga

but friction is massively lower in the air. if you'd get an air tick perfectly you should be easily able to skip the freeze tile

air tick -> jumping at the latest possible moment?

yes

Sounds easier than expected I wonder whats the reason nobody ever did that previously?

I don't think its possible, my understanding of air physics are probably wrong

Patiga

don't understand that yet :/

possition depends what possition you in and when you start hold you kinda don't get same speed and you go true freeze more blocks less or fall ( with out jumping ) that's why edge jumping true 4 freeze is possible ( or even 5 if perfect possition ) (it's also feels like more space more speed but no ) . try making speed up and freeze go true it from different places , you should get same result every time . here try just hold d and go right you will fall depending where you start running (x.XX ( XX depens and with it you get different speed ))

texnonik

possition depends what possition you in and when you start hold you kinda don't get same speed and you go true freeze more blocks less or fall ( with out jumping ) that's why edge jumping true 4 freeze is possible ( or even 5 if perfect possition ) (it's also feels like more space more speed but no ) . try making speed up and freeze go true it from different places , you should get same result every time . here try just hold d and go right you will fall depending where you start running (x.XX ( XX depens and with it you get different speed ))

please have a look at the write-up/calculations above, I'm aware of position-dependent physics

Steinchen

I do get a difference by jumping though

well it you go different x places in end so it's depens when you jump different effects

could be a random bit flip in data which means there's no way to reproduce this

Try making this with speedup in freeze after you hold and jump and go true this freeze gap sbouldn't possible for randomness

some time when i play the sound of the game like hamer hook didnt work i dont have any sound how fix that pls

8:42PM

i think its not ddnet i see my game volum go to 100% and go to 0% how fix that

LePinpin

i think its not ddnet i see my game volum go to 100% and go to 0% how fix that

Ingame or from the application?

no my windows switch the volum randomly

8:45PM

windows the os

catseyenebulous

Ingame or from the application?

no my windows switch the volum randomly windows the os

Go to the voice&video settings in discord and scroll down until you find this and check if you have it on.

catseyenebulous

Go to the voice&video settings in discord and scroll down until you find this and check if you have it on.

its done i will test thx