DDraceNetwork

louis

maybe a gif can help (edited)

like this?

13:17

is it too small?

Looks like a good video! I think it would be great to include in the wiki.

13:20

Yeah it should be a bit bigger + we use the format mp4 for smaller file size

13:21

https://wiki.ddnet.org/wiki/How_to_contribute#Step_3:_Getting_the_video_into_the_wiki + I'd be interested in the .demo file of that video

Zwelf

https://wiki.ddnet.org/wiki/How_to_contribute#Step_3:_Getting_the_video_into_the_wiki + I'd be interested in the .demo file of that video

if you have good tools you can cut it as well, i'm using blender + i don't have much experience

unfreezeTime_2022-07-06_14-50-34.demo

354.91 KB

Thanks. Will take a look later today. I can cut it down with that

2

Dragere

like this?

Bizarre how the left one unfreezes quicker because of the thing we were talking about yesterday

u video does not exit freeze before timer reset

15:18

so it is not one second difference

jao's hairy kebab

so it is not one second difference

yeah in ur video its much more noticable

it still shows that the freeze time doesn't only depend on when you exit the freeze, but showing the second difference in freeze time is great too.

Learath2

Bizarre how the left one unfreezes quicker because of the thing we were talking about yesterday

There is some float fuckery going on as well. These drops/freeze are all the same height. Only in the second drop both reset

1

Hm float fuckery is unlikely here, this is on a single axis and the timing stuff is all integer arithmetic

15:48

My hypothesis would be that falling from the same height through identical tunnels you should unfreeze at the same time

1

https://github.com/ddnet/ddnet/blob/b93b96d7c921a8df73ae60e94e5e75ad53feacd9/src/game/server/entities/character.cpp#L2146 This is the check leading to the behavior: m_Core.m_FreezeStart < Server()->Tick() - Server()->TickSpeed(). You can't get frozen again, if you are still frozen + the last freeze is less than 1s ago.

Learath2

My hypothesis would be that falling from the same height through identical tunnels you should unfreeze at the same time

But the tunnels are the same (assuming i made no mistake). Still all except the second drop brownbear doesn't reset

Dragere

But the tunnels are the same (assuming i made no mistake). Still all except the second drop brownbear doesn't reset

Right tunnel has a shorter freeze length by 1 tile

Learath2

Right tunnel has a shorter freeze length by 1 tile

yes, but it doesn't change lenght between the drops

?

ill try to make a video

One of them leaves freeze later, thus gets unfrozen later, that's sensible, no?

Learath2

One of them leaves freeze later, thus gets unfrozen later, that's sensible, no?

Bizarre :D

16:15

Can you make the tunnels exactly the same and try that?

16:15

Like same amount of freeze

I made both the same as left and on all except the second drop they don't reset, on second both do. When i shorten the second drop by one freeze both don't reset.

Dragere

if you have good tools you can cut it as well, i'm using blender + i don't have much experience

https://wiki.ddnet.org/wiki/File:Unfreeze-time-1.mp4

File:Unfreeze-time-1.mp4

There are no pages that use this file.

1

Could it perhaps be the client id or the entity id? If you swap the places of the tees does the behaviour swap?

16:37

It's interesting since I don't quite remember anything to cause this

Learath2

Could it perhaps be the client id or the entity id? If you swap the places of the tees does the behaviour swap?

Nope. Same behaviour as far as i can tell

So it is position dependent, so odd especially since you are moving on a single axis

I copied the setup a few times to the right and sometimes it works sometimes it doesnt, so the x-axis seems to be involved. Here is the map in case you want to test yourself

unfreezeTime.map

374.03 KB

ye it's about the ticks

16:55

from what i've seen it's like if the server checked at a certain time for each tee on the map if they are in freeze or not

16:56

if yes it adds 3 seconds to the freeze time

16:56

if no then nothing happens

16:56

with the stars we could see this better

If it was ticks it'd be the same no matter the position

it's not quite well ticks

16:56

it's like

16:57

something else XD

It's float fuckery of some kind, but the y axis is equally dense as he falls, it's quite interesting tbf. Why would x position matter at all in this case

x doesn't

16:57

i don't think it does

Dragere

I copied the setup a few times to the right and sometimes it works sometimes it doesnt, so the x-axis seems to be involved. Here is the map in case you want to test yourself

But it does, as you move in the x direction the behaviour changes

in the case where he falls i mean

Learath2

But it does, as you move in the x direction the behaviour changes

is it consistent?

Idk I didn't check it out yet

T tee

is it consistent?

download the map and test urself, as far as i can tell it is

I guess at a given x position it is consistent

Dragere

download the map and test urself, as far as i can tell it is

then

16:59

it reminds me of this

16:59

https://wiki.ddnet.org/images/9/97/Freeze_corner.mp4

3