DDraceNetwork - off-topic

DDraceNetwork

DDraceNetwork / off-topic

Any languages allowed

Between 2023-06-23 00:00:00Z and 2023-06-24 00:00:00Z

tmnf

poggers2

00:56

Actually, there is even a solo map called trackmania, not that hard to meet tmnf fans

00:56

greenthing

Anime.pdf

tmnf

poggers2

i have tmuf

04:39

so i can flash my mcqueen skin at you

04:39

troll

And then hefest got this run

I like watching documentaries

06:50

About games i have never played

06:51

And tutorials for software i have never used

Voxel

js is only good for the plugin thing adobe animate has and i wish i could copy elements outside of it and then into some pasteboard like layers or frames and shit

Lolol

pilonpl

And then hefest got this run

Trackmania content is so fun to watch. Somebody absolutly needs to make ddnet documentaries about maps

got emudeck on my steamdeck

Image attachment

1

ScReeNy

Trackmania content is so fun to watch. Somebody absolutly needs to make ddnet documentaries about maps

You can't make world record progression because demos aren't public

10:22

Which is very weird

10:22

There is literally no proof

10:23

Admins have access to demos

10:24

Which makes the whole thing even more stupid

@pilonpl admins have access to a special format of demos

10:28

which we dont have yet a mean to turn back into a demo and render

10:29

the special format uses less space

Nice

teehistorian

I just want a proof

of what

Of runs

there is like 1tb or 2 saved already iirc

Wr runs

https://ddnet.org/docs/libtw2/teehistorian/

Ryozuki

there is like 1tb or 2 saved already iirc

But i don't have access to it

me neither

10:31

buts its pointless cuz we cant render it

10:31

but u can make ur own teehistorian demo

10:31

and develop it

10:32

then u give a reason to share

10:32

its also hard to share cuz its lots of data

10:32

but idk ask heinrich

10:32

i dont have access either

mh i think its still possible to get access to most top runs if they are not too old if u just reach out to the players. Another interesting thing to me would be an explaination of skips or sth

I think demos for wr runs should be available to everyone

but generally we are kinda low on footage yes

ScReeNy

mh i think its still possible to get access to most top runs if they are not too old if u just reach out to the players. Another interesting thing to me would be an explaination of skips or sth

They can just refuse to show it lol

pilonpl

I think demos for wr runs should be available to everyone

i dont, but there have been plenty debates already

pilonpl

I think demos for wr runs should be available to everyone

i agree too but its a whole debate xd

i think almost everyone would share if its for a cool video:D

How is that a debate lol

10:33

You need a proof

not rly

so far ddnet is pretty secure

Video proof is pretty standard in speedrunning

as in a finish is a finish, u cant "cheat" the game, (here not including map cheats)

if a rank is suspicious you can ask a team member to verify it

cuz the server is fully authoritive

10:34

@pilonpl this is standard for client side speed runs

yea server side action recording is even safer than video (edited)

here the server knows its a good run

Ok, maybe cheating in ddnet isn't easy

cuz the server is fully authoritive

10:35

this is why its so lag sensitive too

10:35

the client cant do any true prediction, which is where cheats would come in

im a big fan of that tbh

You could technically fake the inputs

there are so many big games that just enable so many cheats/hacks by not doing so

It is hard to do

u can only fake valid inputs

10:36

the server calculates itself the movements based on the input

10:36

the client cant say hey im here

Its been done with demo replaying, but u wouldnt be able to notice that in a video either

the client can only say "hey i pressed right"

Yes but replaying inputs is also cheating

yeah

10:37

but its not the type of cheat i was talking about

10:37

like speedhacks to smth

10:37

or rng manip

In ddnet It is harder to cheat than in most games i suppose but i'd still like to see wr runs

you can find a lot on youtube at least.

i would say its even hard for a human to know if a replay bot was used

From my experience many rank 1's are just very bad (even from good players) and those dont get uploaded unfortunally

simply add some errors here and there

Ryozuki

i would say its even hard for a human to know if a replay bot was used

definitly

also even if u use a replay bot, due to network jitter runs wont be equal always

if the cheating person knows what they are doing, its kind of impossible to make a safe case

yea

You can't really replay the same inputs twice because that's suspicious

Ryozuki

also even if u use a replay bot, due to network jitter runs wont be equal always

actually wouldnt this give an option to catch them? If they need to run a replay multiple times they should have too similar inputs. Im just not rly sure how bad the jitter is

ScReeNy

actually wouldnt this give an option to catch them? If they need to run a replay multiple times they should have too similar inputs. Im just not rly sure how bad the jitter is

The solution is AI lol

10:42

It could change inputs

10:42

So they're slightly different each time

well if the reply bot uses the coords it can add random movement and correct to get to that position with more or less some pixels to not be obvious

10:42

but

10:43

we shouldnt discuss this here kek

10:43

heinrich will be angery

I would like to create tas tools

10:45

But i am not smart enough

well there are some TAS runs

10:46

you could ask them how they did them

Are there any public tas tools?

nothing ddnet specific as far as i know

Can i just send fake packets to the server?

10:47

With python

10:47

Lol

i guess you could

I can ask chatgpt how to do it

if thats so

10:49

projekt is almost finished

10:51

"I'm sorry, but I cannot assist with providing guidance or support for any activities that involve sending fake packages or engaging in malicious actions."

You can just say it's for educational purposes only

10:52

Or say that you are creating a custom client

But if you want some easy start into TAS. Im sure there are some tools for windows that can just emutlate the inputs. Question is how well it would work and how annoying non deterministic stuff in ddnet is(like multiple teleporter tos)

Are teleporters actually random or is it based on some in game variable?

im not entirely sure but i think it prefers the teleporter that is most far away from other tees

That would make sense

but thats just my feeling, no idea how it actually works

I think i could intercept packets with wireshark

11:01

And see how they work

wireshark is definitly a start to get an idea of packets sent and to work with them, lunch is over ill go back to work, gl

idk how to use wireshark

11:47

lol

11:50

actually

11:50

i think it works

11:55

yup

i have no idea whats going on

cooking time

Image attachment

12:28

greenthing

Ryozuki

as in a finish is a finish, u cant "cheat" the game, (here not including map cheats)

please keep all ddnet related topics out of #off-topic

13:25

this is not the correct channel

ezy

please keep all ddnet related topics out of #off-topic

who are you

@Voxel this kind of talk only in dms please

Image attachment

pilonpl

Are teleporters actually random or is it based on some in game variable?

They are pseudorandom, seeded at round start

@Learath2 exams finished?

Just got out of another one. I have 2 more in 2 weeks

16:22

I am dying. My brain physically can not fit any more information. I bet I'm getting close to the shannon limit

u'll forget most stuff u learned again anyway

Exactly, absolutely pointless act

indeed xD

Ampere's law is so inconvenient it can only be reasonably applied to an infinite straight wire. Wtf am I supposed to do with that in a physical actual computer?

Learath2

They are pseudorandom, seeded at round start

waiting for tele prediction

Learath2

Ampere's law is so inconvenient it can only be reasonably applied to an infinite straight wire. Wtf am I supposed to do with that in a physical actual computer?

numerical estimations

louis

waiting for tele prediction

Only 2 problems remain. Giving the client the seed and keeping it up to date on the telecounter

Learath2

Ampere's law is so inconvenient it can only be reasonably applied to an infinite straight wire. Wtf am I supposed to do with that in a physical actual computer?

well you can use the law of biot-savart, but its way more convenient to approximate with amperes when you can

Unless the situation is insanely symmetric how am I even supposed to integrate over the space? Instead of numerically approximating at that point I'd rather assume lumped models :₩

Learath2

Unless the situation is insanely symmetric how am I even supposed to integrate over the space? Instead of numerically approximating at that point I'd rather assume lumped models :₩

I mean there certainly are times where it would be impossible to solve analytically, so you would have to do iterative calculations with a computer

Exported 144 message(s)