DDraceNetwork - developer

What is the reason why one cannot bug through killls sideways?

08:27

Except for ninja

def-

[ddnet-web:master] 1 new commit

ba02b7c Fix mapper link - def-

08:36

def-

[ddnet-web:master] 1 new commit

e59bcaf Fix link in 2birthday - def-

you need to be as fast as around 50 tiles per second

08:45

with killtiles even more, bcs bigger hitbox

def-

[ddnet-web:master] 1 new commit

97669ef Fix mapper name on Tournament #13 - def-

Sideways speed is limited I remember. If you are too fast you slow down again. So you cannot exceed the speed that is needed to bug through kills

i think the clamp is around 180 tiles per second

08:53

have u tried to just give you that acceleration?

That speed is fast enough to get through 52,5% of all kills when falling down but sideways it always killed me

ok

I thought we settled that it depends on where the movement starts whether you skip thru a skippable tile or not

that also

09:00

anyway, the ground movement glitch with speedups is probs the same

09:00

it starts to glitch when you are too fast

09:01

e.g. it does only m_Vel.x = m_Vel.x * RampValue; for the x value

09:01

i guess thats the problem

def-

[ddnet-web:master] 1 new commit

744dfe7 Remove tournaments static page (dynamic now) - def-

Deleted User

anyway, the ground movement glitch with speedups is probs the same

you mean the player not moving when going too fast?

Learath2

you mean the player not moving when going too fast?

yeah

that's a result of the chosen ramp curve approaching 0 as v_x -> infinity

yeah and that happens before you get the speed to skip

hm, we should check sometime the actual amount of speed required and the speed cap

09:12

you just need to be moving faster than 1 tile per tick

with 50 tiles p second you can skip 1 tile, and kill tiles are more than 1 tile, but less than 2 (edited)

mh I think kill tiles are less than a tile

freeze is exactly 1 tile

09:13

and kill tiles are bigger

I guess we have different definitions of "tile"

i mean the mid of the tee

09:13

if the mid moved 1 tile

09:14

and thats exactly happening in freeze

Yeah I see what you mean

Deleted User

it starts to glitch when you are too fast

When velramp*speed hits 75 and you keep adding velocity to it (e.g. by continuing to use jetpack on the x axis) that value will go down slowly again until reaching 0

Learath2

mh I think kill tiles are less than a tile

kill tiles have extra range to avoid skipping them

fokkonaut

kill tiles have extra range to avoid skipping them

only for characters*

Iirc the extra range was just a compatibility thing

the problem is, it depends on how you look at it: e.g. do you say the tee has extra hitbox when hitting a kill tile, then the kill tile is just 1 tile as freeze

09:35

or do you say, i just choose the mid, and assume its always the visual distance

ye the tee has a larger hitbox for killtiles

09:35

Thats probably the best way to say it

09:35

And matches the code

So the reason kill tiles are so damn hard to edge is because they're designed to prevent skipping

... I did wonder if it was just because edging a kill tile should be somehow seen as harder or if there was another reason!

i think it was just random coding

11:00

like most of teeworlds xD

11:00

since kill tiles also exists in vanilla

Lol, its very interesting, i hope it doesn't change because i think kill tiles should be harder to edge... more to lose

i doubt old tiles ever get changed, its more likely someone adds a new kill tile, if ever needed :D, but i guess that also wont be the case so fast

Yeah I don't see a need for it... A kill tile that has the same hitbox as the freeze tile, but then could be exploited if used incorrectly

11:02

Unless thats the purpose of it loool

11:02

The end of the map is just a line of kill tiles and a jetpack gun

11:03

I suppose that transcends beyond Insane 5 stars, lol

If I recall correctly vanilla had fat kill tiles, then they changed collision on them. DDRace wanted to keep them fat

oh really? i used the kill tile collision code for fng and it looks to be the same as ddnet kill tiles

11:33

but didnt check 0.7 if they changed it there

can a discord admin mention https://wiki.ddnet.tw/wiki/FAQ in #questions , even if its uncomplete and probably nobody reads the channel description anyway xD

Added

1

@Learath2 are u Learath2 on the wiki?

yes

he is Learath3, Learath2 was already in use

def-

[ddnet-web:master] 1 new commit

0e4bece Update settings&commands - def-

def-

[ddnet-web:master] 2 new commits

4cf107d Add donation by szaro - def- 4848abb Add UAE, TUR2, ZAF2 - def-

def-

[ddnet-web:master] 1 new commit

95736cd Add 64 € donation by Konsti - def-

projectile.cpp

10.82 KB

15:33

lel discord is cool now

woah

15:49

cpptest.cpp

254 bytes

oh yikes what happend to envelopes? Is there documentation on that?

ChillerDragon ah forgot to ping if you havent seen yet i rlsed a new twmap version 1 week ago or so, not backwards compatible tho

1

18:29

now the parser will now note the version (0.6-ddnet/0.7-teeworlds) that the map was in, so that it can save the map in the original version again

awesome thanks for the note @Patiga i should probably email sub ur repo and cool that u keep pushing updates

18:40

so when i try to move my quads it does some enevelops magic :/ i dont want to downgrade my client pls send brain on how to use the map editor

gerdoe

just notice us where are they :D

ah, oopsie

18:48

@Ryozuki i rlsed a new version of twmap btw ^^

nice

now it has some documentation

18:49

https://docs.rs/twmap/0.2.1/twmap/index.html

@Patiga clippy still returns a shitload of warnings pepe_straight

18:58

.map_err(|err| InfoError::LicenseLength(err))?; cannbe .map_err(InfoError::LicenseLength)?;

18:58

and if u have to map so many errors u can always use thiserror

wdym?

anyway cargo clippy will help u make better code

Ryozuki

.map_err(|err| InfoError::LicenseLength(err))?; cannbe .map_err(InfoError::LicenseLength)?;

thats nice, didnt know that

18:59

@Patiga just do "cargo clean && cargo clippy"

18:59

and u will see all the suggestions

i think it didn't like some formatting decision which i liked

afaik clippy doesnt format anything and i rarely see it say something about formatting

19:01

btrw u can selectively disable warnings

i also still think that Vec<_> makes sense in that context, i don't wan't someone to get weird ideas in that process

u rly should use &[T]

why is that?

its the common practice

19:01

the vec also involves 1 more reference

19:01

as it says

19:02

Requiring the argument to be of the specific size makes the function less useful for no benefit; slices in the form of &[T] or &str usually suffice and can be obtained from other types, too.

but that assumes you want the function to be more useful

i suppose common practice is a valid argument, the speed is rly insignificant, even if it has to follow one more pointer

Learath2

but that assumes you want the function to be more useful

it has no drawbacks

it makes the interface more ambiguous no?

no

i don't wan't someone to get weird ideas in that process

nobody gets weird ideas

yea, that was my point

This way it'd only work with Vec based slices, what if you only want it to work as such?

19:04

the vec derefs to &[T]

you are kinda avoiding our point here