DDraceNetwork - developer

DDraceNetwork

Development / developer

Development discussion. Logged to https://ddnet.org/irclogs/ Connected with DDNet's IRC channel, Matrix room and GitHub repositories — IRC: #ddnet on Quakenet | Matrix: #ddnet-developer:matrix.org GitHub: https://github.com/ddnet

Between 2023-12-29 00:00:00Z and 2023-12-30 00:00:00Z

Fudgy

does ddnet server send a chat message to 0.7 clients when someone finishes a race? I don't seem to receive anything

0.7 has special netmessages for race which are being used

00:07

They send the text clientside

hmm for some reason I think I dont receive any type of message when someone finishes, I get the start message tho

00:08

I will keep looking..

https://github.com/teeworlds/teeworlds/blob/a1911c8f7d8458fb4076ef8e7651e8ef5e91ab3e/src/game/client/components/infomessages.cpp#L103

teeworlds/src/game/client/components/infomessages.cpp at a1911c8f7d...

A retro multiplayer shooter. Contribute to teeworlds/teeworlds development by creating an account on GitHub.

00:15

https://github.com/ddnet/ddnet/blob/ee2dd0ac1c2f5bfa7d13675e97585e256bae0f12/src/game/server/teams.cpp#L786

ddnet/src/game/server/teams.cpp at ee2dd0ac1c2f5bfa7d13675e97585e25...

DDraceNetwork, a free cooperative platformer game. Contribute to ddnet/ddnet development by creating an account on GitHub.

Ok I understand the problem. If I'm on 0.7 client and someone with 0.7 client finishes, it works. But if I'm 0.7 and the one finishing is not on 0.7, I have this message in console instead : [client]: dropped weird message 'invalid' (0), failed on '(type out of range)'

@Learath2

quick question, what is the formula for converting ddnet score points into race timings?

01:30

score + 9999 = milliseconds?

01:30

something along the lines?

hmk, it's just seconds ._.

@Ewan thanks for teaching me english i never get a vs an. Also why do you talk about js not bout error? is the "a" not pointing to error instead of js in the stence "a javascript error"

06:41

@Fudgy are you working on something or just playing? On official ddnet servers? with latest teeworlds client?

Im playing with a 0.7 client from like 3 years ago. It's annoying to not see finish times when playing race with others

poggers2

2

Since other players are not using 0.7, it's sending the 0.6 net message and my 0.7 client cant parse it or smth

Fudgy

Ok I understand the problem. If I'm on 0.7 client and someone with 0.7 client finishes, it works. But if I'm 0.7 and the one finishing is not on 0.7, I have this message in console instead : [client]: dropped weird message 'invalid' (0), failed on '(type out of range)'

Yeah, that's a bug. I'll try to fix it.

heartw

1

lmao ofc furo is fixing it xd

07:11

@furo thanks for writing it. I mean i shouldve guessed it by now that youll fix it xd. But i almost started looking into it xd

:)

ChillerDragon

@Ewan thanks for teaching me english i never get a vs an. Also why do you talk about js not bout error? is the "a" not pointing to error instead of js in the stence "a javascript error"

because it only has to do with the word immediately after it

07:16

"a stack overflow" "an overflow"

07:16

something to do with having to have a *consonent sound to make the words flow smoother (edited)

07:16

because something like "a error" creates an awkward transition from vowel to vowel

07:17

uhh errurr

07:18

https://en.wikipedia.org/wiki/Sonority_sequencing_principle

[ddnet/ddnet] Pull request opened: #7741 Use 0.7 translation layer for Sv_RaceFinish netmessage.

Fixes 0.7 players not being able to receive finishes done by players on 0.6 (and vice-versa).

Checklist

[x] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially base/) or added coverage to integration test
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the...

O thanks

Replying to @Ewan because it only has to do with the word immediately after it

github-merge-queue[bot]

[ddnet:master] 2 new commits

1a94aa6 Use 0.7 translation layer for Sv_RaceFinish netmessage. - furo321 0f1171a Merge pull request #7741 from furo321/translate-racefinish - def-

08:16

[ddnet/ddnet] Pull request closed: #7741 Use 0.7 translation layer for Sv_RaceFinish netmessage.

i can't see what is CNetObj_CharacterCore and it's members

09:28

i just know the core read and write with it

Because the code is generated with python. That confuses most IDEs

09:53

You can open the src folder in your build folder to see its defintion

Replying to @რილია i can't see what is CNetObj_CharacterCore and it's members

chillerdragon

Because the code is generated with python. That confuses most IDEs

thank you that explains

@Jupstar ✪

10:27

It's not called out in the release notes announcement, but rustc will now automatically enable cross-crate inlining for small functions. This leads to both compilation time wins and runtime performance improvements. They are small in this initial implementation, but I expect it to be tuned over time.

Ryozuki

It's not called out in the release notes announcement, but rustc will now automatically enable cross-crate inlining for small functions. This leads to both compilation time wins and runtime performance improvements. They are small in this initial implementation, but I expect it to be tuned over time.

nice

10:28

2024 will be the year of good rust additions

"cargo new" inside a workspace now automatically adds the new crate to the workspace members!

[ddnet/ddnet] Pull request opened: #7742 Optimize allocation of sound sample indices, add missing sound lock usage

Replace linear search for free sample index with free list. This brings sound sample index allocation down to constant complexity independent of the number of allocated sounds. On average the time to allocate sound samples is reduced by around 75% (843µs down to 223µs). For perspective, the time to load all default sounds on client launch is reduced by around 15ms (although this does not significantly affect launch time due to threaded loading). ![sound-sample-alloc-benchmark](https://gith...

is https://github.com/ddnet/ddnet/pull/6914 fully implemented? because i don't see a redirect rcon command anywhere

Allow server to redirect clients by ChillerDragon · Pull Request #6...

This is a preperation for #6754 (redirect players on map vote) And an implementation of #5662 (redirect players) Reopen of #6757 Checklist Tested the change ingame Provided screenshots if it is ...

toneey

is https://github.com/ddnet/ddnet/pull/6914 fully implemented? because i don't see a redirect rcon command anywhere

I don't think there's a command for it, you're meant to manually send the NETMSG_REDIRECT net message in your code, but maybe one should be added?

@toneey yes its fully implemented and working. Ofc not all clients support it yet. You are meant to use this api void CServer::RedirectClient(int ClientID, int Port, bool Verbose)

12:45

@Bejavioue there is no rcon command and heinrich was against adding one so there probably wont be any any time soon

12:45

No, I think that options in the official DDNet server are also important. If people want to use this for moderation, they'd have to patch the source by themselves, which can be enough friction that people might not do it.

@furo when add colored broadcast hehe

12:59

its on my todo so it must also be on urs

[ddnet/ddnet] Issue closed: #4977 When selecting multiple groups, highlight color is not preserved

[ddnet/ddnet] Pull request closed: #7742 Optimize allocation of sound sample indices, add missing sound lock usage

13:49

github-merge-queue[bot]

[ddnet:master] 4 new commits

7a84a74 Add CSample::TotalTime function - Robyt3 87191a3 Use sound lock when setting/getting current time of sample - Robyt3 661339c Optimize allocation of sound sample indices - Robyt3 5873b42 Merge pull request #7742 from Robyt3/Engine-Sound-Alloc-Improvement - def-

GitHub

Click to see attachment 🖼️

@Jupstar ✪ another unsound MT access rust would prevent

int a = 5;
if(a != -1 && (int b = 2))
{

}

why does this throw a syntax error?

why are you doing this

14:13

xd

making things better

14:14

i want "b" only to be in the scope of the if statement and also check if the variable i assign to b is not 0

Ryozuki

@Jupstar ✪ another unsound MT access rust would prevent

too bad that you cannot code it in the next 48hours :/

well it works like if(int b = 2)

14:15

but not with parentheses around it

Teero

i want "b" only to be in the scope of the if statement and also check if the variable i assign to b is not 0

why not simply do a brace around it { int b = 2; if ... }

14:16

then the braces are your scope

int a = 5; if (int b = 2; a != -1 && b != 0) { }

Nouaa

int a = 5; if (int b = 2; a != -1 && b != 0) { }

the 2 in my example might be a function call that i dont want to do twice

compiler expects an expression, not a declaration ig?

Jupstar ✪

why not simply do a brace around it { int b = 2; if ... }

kinda ugly but it was my first guess too

declarations in if's are allowed since cpp17

BlaiZephyr | meloƞ

compiler expects an expression, not a declaration ig?

if(int a = 4) is fine

Nouaa

int a = 5; if (int b = 2; a != -1 && b != 0) { }

ah wait i overlooked it

14:18

mybad

well honestly. you really never need to have a declaration inside a if i bet there is a better way

Nouaa

int a = 5; if (int b = 2; a != -1 && b != 0) { }

works ty

in worst case accept a second if statement

owo

14:18

shit code anyways

owo

BlaiZephyr | meloƞ

owo

see, im big dev

14:21

should get that role at some point

justatest

1

big

14:26

owo

Teero

the 2 in my example might be a function call that i dont want to do twice

couldnt you use structured bindings for that use case? https://en.cppreference.com/w/cpp/language/structured_binding

cpp/language/structured binding

cpp hard

int a = 5;
if (auto [b] = fooBar(); a != -1 && b != 0)
{
}

to assign value to b and then check if b != 0) or smth

BlaiZephyr | meloƞ

int a = 5;
if (auto [b] = fooBar(); a != -1 && b != 0)
{
}

to assign value to b and then check if b != 0) or smth

thats what nouaaa said earlier xd

Nouaa

int a = 5; if (int b = 2; a != -1 && b != 0) { }

.

@Ryozuki now that you llvm and rust core maintainer you are basically professional language designer right? Any chance you get some fixed into the german spec for me?

14:36

for example the german number system is fucked some drunk lang designer messed up ordering of double digit numbers

14:36

s/fixed/fixes/

Teero

.

structured bindings allow you to assign complex types to b, while the latter only allows simple types, but yeah, they basically do the same

wdym simple types

14:39

well it works with CGameContext

14:39

is that a simple type?

simple:

int a = 5;
float b = 3.14;
char c = 'A';

complex:

int arr[] = {1, 2, 3};

14:43

classes are actually considered complex types, because they can hold data structures with multiple values, i dun fckn know >_<

[ddnet-web:pr-funding-2024] 1 new commit

6c08a70 Update funding for 2024 - def-

15:05

[ddnet/ddnet-web] New branch created: pr-funding-2024

15:06

[ddnet/ddnet-web] Pull request opened: #279 Update funding for 2024

15:09

[ddnet-web:pr-funding-2024] 1 new commit

bef1aff Update funding for 2024 - def-

brah, i just found out you guys remove the m_hookTick in Core upon character getting endless hook, i thought the problem was my code for 1 hour justatest

justatest

BlaiZephyr | meloƞ

simple:

int a = 5;
float b = 3.14;
char c = 'A';

complex:

int arr[] = {1, 2, 3};

this makes cpp programmers cry

რილია

brah, i just found out you guys remove the m_hookTick in Core upon character getting endless hook, i thought the problem was my code for 1 hour justatest

justatest

don't fix it tho, i wanna make hook upgrading system with it feelsamazingman

feelsamazingman

16:07

it works fine with prediction

@Jupstar ✪ i broke my promise

16:23

i coded in rust

16:27

i got a minimalistic game idea

16:27

time to try again with bevy

Ryozuki

@Jupstar ✪ i broke my promise

rust makes you weak

Ryozuki

i coded in rust

Did you talk to meloƞ? How did the therapy session go?

NekoEvil

BlaiZephyr | meloƞ

if you need someone to talk - im always here for you

(I'm refering to this btw kek

kek

)

Mr.Gh0s7

Did you talk to meloƞ? How did the therapy session go?

kek

16:45

Ryo if rust keeps you hostage, write unsafe code twice and we'll help you!

@BlaiZephyr | meloƞ xd

hum

ger10 doing some extra shenangians today?

19:02

i cant connect .-.

19:03

Embedded image

use https://ger10.ddnet.org/

did you update something?

19:08

ah nice it works now

19:09

wat did u do

19:09

oh holy shit or does http://ger10.ddnet.org/ not work?

19:09

somehow neither my browser nor the webserver upgrades me to https

19:09

it still says verified but does not seem to work

19:11

@deen if you did not fix something i assume the http version does not work. Maybe you could setup a redirect c:

@Davide ^

deen arent u hosting the web?

yes, but not the way the whitelist works

then you can fix it

ChillerDragon

oh holy shit or does http://ger10.ddnet.org/ not work?

Pretty weird

no need to adapt the whitelist

19:18

just add a http to https redirect

19:19

sounds reasonable anyways

ChillerDragon

just add a http to https redirect

It should be http only as I can remember

19:19

Oh nvm, it's HTTPS too, still, even HTTP whitelists you, both works

well i typed ger10.ddnet.org into my browser arrived at http and it didnt work then i clicked deens https link and i could join

19:19

maybe it was just some other bug then

Let me check

i didnt change anything

19:20

i tried like 3 servers when it didnt work

I confirm it should work even with HTTP

i am not at home. I think at home usually the http also works for me. Yea maybe it was something else.

19:29

Do you need me to test something? You can kick me and i can try http again if you are curious @Davide

ChillerDragon

Do you need me to test something? You can kick me and i can try http again if you are curious @Davide

Sure, send me your IP addr somewhere and I'll remove you from the verified map (edited)

i can whisper u in game

19:30

securest channel

19:30

u online? xd

Wait

19:30

Done, joined

19:30

Username is Davide55

go join me at zillyhuhn.com

19:31

i wanna grab ur ip too xd

Oh, I'm in a random DDNet server

19:31

Just search for my username

bliet ok

19:31

on a random ger10 ? -.

Yes

ok i can confirm http works

Good

still deen i think a https redirect would be cool just for coolness sake

19:34

no idea what happend earlier tho

19:34

thanks for the quick response davide UwU

No problems

19:35

I'll also check the UDP challenge as soon as I can

19:35

Maybe it needs a few updates

i personally do not mind the http request too much since it can even be done via curl

19:35

but i guess the general public would appreciate it

Okok

maybe it also doesnt work because of my hax client

Would be good to implement an automatic curl request before the join

ChillerDragon

maybe it also doesnt work because of my hax client

Oh, yes can be

i send some custom ddnet ex messages

19:36

identifying my client as it self to the server

i think some custom client does the automatic curl within the client already

Just tried with latest version from Steam and seems to work, pretty slowly but it works

ok nice

ChillerDragon

i think some custom client does the automatic curl within the client already

Can be

if it rly is my extra msg that confuses ur prot thats totally fine

19:38

its one alt+tab and ctrl+r ger10 enter

i had the curl running in a loop on my raspi for fokkos server when i was actively playing xd

LOL

worked like a charm

Sadly DDNet is super sensitive, so protections must be extremely aggressive to avoid crashes/lags (edited)

seems fair

19:40

as i said i have no problem but it would be sad if new players or even old but non technical players rq on the issue :c

19:41

but i think everything is smooth and the protection does work really well

19:41

i barley play these days last time i played there were a few lags on ger10 but never longer than a few seconds

19:42

might as well have been non ddos issues idk

I should move the DDNet GER10 VM to the new DC, it would be probably even more stable as connectivity (edited)

19:43

And even less ping (2-3ms less, nothing crazy) (edited)

19:44

Maybe we can schedule that @deen (edited)

what is the best method to give the player msg without spamming his chat ? (don't wanna use broadcast) (edited)

Davide

Maybe we can schedule that @deen (edited)

sure, i can spin down the servers and tell you when it‘s done?

deen

sure, i can spin down the servers and tell you when it‘s done?

Else I can just create you a new VM on the new DC, maybe it's faster? Depends on your install script (edited)

19:49

(Migrating a VM can require a few hours depending on the disk size)

@რილია there is also the message of the day sv_motd and you can also use non important broadcast that will not overwrite important broadcasts in ddnet server code

19:53

you could also use in game stuff like projectiles in the snap for conveing information

19:53

fng even uses it for score numbers or BW for texts on kills

ChillerDragon

fng even uses it for score numbers or BW for texts on kills

yes i seen it

if you need another channel for sending text like custom popups you would have to modify the client

i seen "balance the teams" in fng, the msg shows up on screen, how about that ?

there are a few of those but they are hardcodet

19:55

you can not set a custom message there

feelsbadman

ye its not like in minecraft

19:56

its insane what custom ui a vanilla minecraft client can display

Davide

Else I can just create you a new VM on the new DC, maybe it's faster? Depends on your install script (edited)

yes, that‘s faster. then we can shut down the old after a few hours

ger11 incoming!

20:44

poggers2

20:44

road to 100

i read it as binary ger10 = ger2

20:45

xd

deen

yes, that‘s faster. then we can shut down the old after a few hours

Yes

I'm installing UE5 for the second time today

21:09

21:10

The first was for linux but the performance wasn't very good. Now it's windows time pepeW

pepeW

Mr.Gh0s7

I'm installing UE5 for the second time today

teeworlds ray tracing edition?

im using bevy

21:11

would a teeish vampire survivors be fun?

21:11

xd

i dunno, probably not as much xd

21:11

maybe some tower defense game

21:12

the nade shooter the hooker

Jupstar ✪

teeworlds ray tracing edition?

21:16

ofc

brownbear

21:17

nice

[ddnet/ddnet] Pull request opened: #7743 Fix players on 0.7 becoming invisible when m_HookTick is negative.

If you were to enter a tune_zone that has hammer_duration with a value above 1.25f, m_HookTick becomes negative. Which leads to 0.7 clients dropping the CNetObj_CharacterCore message, because it only accepts values between 0 and int_max. So, I just decided to clamp it at 0 if it's negative when sending it to 0.7 clients. Closes #3915 https://github.com/ddnet/ddnet/assets/141338449/2fa8986a-94aa-41ab-b65d-57110166627a

Checklist

[x] Tested the change ingame
[ ] Prov...

is there some special url format to start the ddnet client and connect to a specific server? (edited)

22:43

(just ping me)

jxsl13

is there some special url format to start the ddnet client and connect to a specific server? (edited)

https://ddnet.org/news/ddnet-links/

jxsl13

is there some special url format to start the ddnet client and connect to a specific server? (edited)

ddnet://host:port (edited)

awesome, thanks

Or https://ddnet.org/connect-to/?addr=38.54.57.140:8321/

1

1

38.54.57.140:8321 is an official DDNet server. https://ddnet.org/connect-to/?addr=38.54.57.140:8321/

github-merge-queue[bot]

[ddnet:master] 2 new commits

a918c61 Fix players on 0.7 becoming invisible when m_HookTick is negative. - furo321 9c8a081 Merge pull request #7743 from furo321/hookduration-07 - def-

23:15

[ddnet/ddnet] Pull request closed: #7743 Fix players on 0.7 becoming invisible when m_HookTick is negative.

23:15

[ddnet/ddnet] Issue closed: #3915 0.7 clients buggy in tunezones?

BlaiZephyr | meloƞ

simple:

int a = 5;
float b = 3.14;
char c = 'A';

complex:

int arr[] = {1, 2, 3};

weird terminology

23:21

they're called primitives

i called em that for ease of understanding, but you're right

i hardly think using the wrong term will help anyone to understand

23:24

shrug

Exported 247 message(s)