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 2025-01-02 00:00 and 2025-01-03 00:00

[ddnet/ddnet-maps] Issue closed: #9 add_path seems available

@Learath2 it's time to change the language's syntax yet another time

08:03

now array size will be before the type itself xd

@deen u are a haskell fan right?

08:14

do u know about granule?

08:14

https://granule-project.github.io/papers/esop22-paper.pdf https://granule-project.github.io/

The Granule Project

The Granule Project: A Research Project Studying the Next Generation of Functional Programming Languages

08:14

(it can be installed via stack, which is a haskell tool iirc)

08:15

its interesting cuz it implemented graded types

08:15

graded modal types

when im finished with my computer science bachelor i will join up in developing (hopefully)

08:15

also: is there a windows rust ddnet build yet (edited)

nice

08:15

not that i think no

unlucky

08:16

would like to give it a shot

why not compile it urself

08:16

install rust

08:16

u are a comp sci bachelor

not yet

08:16

:)

why u limit urself by a title/ study

08:17

learn it urself

real and true

08:17

will do

nice

Fr fr

if u just follow a curriculum u will be average

08:17

or lower

08:17

self learn is the best u can do in this field

08:18

To follow the path: look to the master, follow the master, walk with the master, see through the master, become the master.

08:18

poggers2

08:19

i like this web a lot https://justforfunnoreally.dev/

heartw

1

Learath2

Fr fr

i find it funny how he knows about graded modal types, which is smth rly niche and not about building a rust project

08:20

thats what academia is doing nowadays? justatest

justatest

idk shit about coding tbh xd

08:22

like purely math and network things

xd

08:26

@Learath2 what u think about linear types u would need to write most code like db, result = query(db, blah)

08:26

a good resource https://austral-lang.org/spec/spec.html#rationale-linear-types

08:26

but it doesnt allow u to close a file twice

08:26

or a db

Fluni

also: is there a windows rust ddnet build yet (edited)

https://pg.ddnet.org/downloads/ddnet-playground.exe but not sure if its up to date

Ryozuki

@deen u are a haskell fan right?

I was a Haskell fan around 2007-2013, haven't used it since then

Ryozuki

@Learath2 what u think about linear types u would need to write most code like db, result = query(db, blah)

Hm, it is a smart and correct solution, but it does look very cumbersome to use. Idk, maybe nice?

Learath2

Hm, it is a smart and correct solution, but it does look very cumbersome to use. Idk, maybe nice?

well with borrows maybe its more ergonomic

@deen i registered a server myself and i think i finally fixed it hopefully

Image attachment

ok, trying

crossed_fingers_intensifies

10:25

85/256, seems fine

cool, I'll announce it on Steam now too. Maybe 256 players won't be enough

10:27

we didn't get the registration on GER1 working btw, really weird

probably too old

even with whitelist for old registration servers

10:28

but maybe fine on ger2

if i got motivation i should remake this for next year.

10:29

like with a database or something so maybe we can host multiple ones instead of relying on a gaint one.

would be pretty sick

10:30

i feel bad for those in east asia since you get 250+ ping

honestly, feel fine just by prediction

but with a segmented map its harder to put easter eggs and weapons which (i hope) people find fun

since it is mostly single player

ye maybe

10:31

in my experience jp isnt predicted correctly especially with tunes

it does jitter tho

10:31

ye

10:31

remind me in 10 months or something

10:31

i could remake this

[ddnet-web:master] 1 new commit

663da8d New year post - def-

10:39

[ddnet-web:master] 2 new commits

f841cff fix formatting - def- 59459fb fix - def-

10:39

[ddnet-web:master] 1 new commit

c75e653 fix html - def-

@eidderFsT close to CPU limit

Image attachment

InsaneCat

justatest

justatest

make sv_poses_queue_interval_low higher might help, which is how often it refreshes what player sees on screen. it is already 1500ms tho

10:58

you can also do sv_snapshot_interval 2 if it is not already, or even 3 or 4 until you feel like prediction is not doing fine anymore.

[ddnet/ddnet] Issue closed: #3645 Console font is skinnier when fullscreen is disabled

11:00

[ddnet/ddnet] Issue closed: #9369 Make it possible to copy compatible entities from one entities layer to another

The latter worked

epic

[ddnet/ddnet] Issue closed: #9355 Destructive ctrl+d cant change bind

@deen do you still have 22 23 and 24's poses data

@eidderFsT Where is the source for the server? I can get the http registration in there (edited)

Learath2

@eidderFsT Where is the source for the server? I can get the http registration in there (edited)

github.com/teeworldscn/momentcap

Which branch?

main

Learath2

Which branch?

love you btw.

github-merge-queue[bot]

[ddnet:master] 2 new commits

6e53ebb Set a recursion limit for config files - MilkeeyCat c29ddcf Merge pull request #9447 from MilkeeyCat/pr_recursion_limit_for_config_files - def-

11:26

[ddnet/ddnet] Pull request closed: #9447 Set a recursion limit for config files

When I compile ddnet for linux (ubuntu) cd /root/build/ddnet22/ddnet/ rm -rf build mkdir build cdbuild cmake -DMYSQL=ON .. make -j$(nproc) no error in the console but when I try to change maps in the server it displays this: Segmentation fault (core dumped) (edited)

[ddnet/ddnet] Issue closed: #7920 Recursively executing config files results in a segmentation fault

@eidderFsT do you mind if I break the client updater?

yes

11:29

wait

11:29

no i don't mind

11:30

english is opposite from chinese semantic

eidderFsT

@deen do you still have 22 23 and 24's poses data

i think deleted, why?

i am not

When I compile ddnet for linux (ubuntu) cd /root/build/ddnet22/ddnet/ rm -rf build mkdir build cdbuild cmake -DMYSQL=ON .. make -j$(nproc) no error in the console but when I try to change maps in the server it displays this: Segmentation fault (core dumped) (edited)

run with gdb and do backtrace

deen

run with gdb and do backtrace

okay

i might want to add search, if you have the old one i can update them

11:33

but nvm we could just start adding it this time

nope, removed them

is it planned that super can interact with others when both in solo?

Image attachment

MilkeeyCat

is it planned that super can interact with others when both in solo?

yes

Is that pure ddnet? Or did you make some changes to it?

Replying to @i am not When I compile ddnet for linux (ubuntu with […]

chillerdragon

Is that pure ddnet? Or did you make some changes to it?

i just activated SQL

12:04

and actually i have something like that

12:04

WARNING: MYSQL_OPT_RECONNECT is deprecated and will be removed in a future version. 2025-01-02 12:01:29 I mysql: connection established 2025-01-02 12:01:29 I sql: map info failed: (prepare:stmt:1064): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for t 2025-01-02 12:01:29 I sql: [14] map info failed on all databases

eidderFsT

no i don't mind

I had to send the client back to the middle ages but I got it compiling still atleast

when i disable use sql it works but when i able it it doesn't

Learath2

I had to send the client back to the middle ages but I got it compiling still atleast

justatest

[ddnet/ddnet] Pull request opened: #9454 Allow super users hammer others & fix jumps when jumps value is -1,…

… 0 or 1 Closes #5710

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
[X] Changed no physics that affect existing maps (probably)
[ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://g...

@i am not so you did not edit the code? Which Ubuntu version do you have? Do you use MySQL or MariaDB?

ChillerDragon

@i am not so you did not edit the code? Which Ubuntu version do you have? Do you use MySQL or MariaDB?

i did not, i have ubuntu 24.04 i use mysql

Weird

12:24

Mysql might be untested I think everyone uses mariadb

12:24

You could try that

chillerdragon

Mysql might be untested I think everyone uses mariadb

RARHHHHHHHHH

12:24

okey

12:24

gonna try

12:24

':)

12:24

thx

But ideally it should also work

chillerdragon

But ideally it should also work

yeah but idk... maybe i did something wrong

[ddnet/ddnet] Pull request opened: #9455 Translations for upcoming DDNet 18.9

Checklist

[ ] 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 change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-ad...

12:54

[ddnet/ddnet] Pull request opened: #9456 DDNet 18.9

TODO

[ddnet:18.9-rc1] 1 new commit

71a6cc7 Translations for upcoming DDNet 18.9 - def-

13:10

[ddnet/ddnet] Pull request opened: #9457 Bump nightly version to 18.9

Checklist

[ ] 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 change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-ad...

13:16

[ddnet/ddnet] Issue closed: #3453 queue system when server is full

[ddnet/ddnet-web] New branch created: pr-18.9

13:38

[ddnet-web:pr-18.9] 1 new commit

76a82f9 DDNet 18.9 - def-

13:39

[ddnet/ddnet-web] Pull request opened: #313 DDNet 18.9

chillerdragon

You could try that

with mariadb the data seems to be loading but when i try to change map it say 2025-01-02 13:47:19 E sixup: couldn't load map maps7/Mixi4jkaem.map like why is it loading maps7/ ?

13:48

and not map

[ddnet:18.9-rc1] 1 new commit

aaf817a Translations for upcoming DDNet 18.9 - def-

i am not

with mariadb the data seems to be loading but when i try to change map it say 2025-01-02 13:47:19 E sixup: couldn't load map maps7/Mixi4jkaem.map like why is it loading maps7/ ?

it needs both. You can use the map_convert_07 tool to convert the map to 0.7 format

13:50

if you want compatibility for TW 0.7 players to join your s ervers

deen

if you want compatibility for TW 0.7 players to join your s ervers

and if i don't want

13:51

i can disable it ?

13:51

because i have like 900 maps actually

sv_sixup 0

okkkk thx

13:56

#0  0x00007ffff7950203 in ?? () from /lib/x86_64-linux-gnu/libmysqlclient.so.21
#1  0x00007ffff7952121 in mysql_stmt_fetch_column () from /lib/x86_64-linux-gnu/libmysqlclient.so.21
#2  0x00005555555891d3 in CMysqlConnection::IsNull (this=0x7fffe8000b70, Col=1)
    at /root/build/ddnet22/ddnet/src/engine/server/databases/mysql.cpp:513
#3  0x0000555555625f56 in CScoreWorker::LoadPlayerData (pSqlServer=0x7fffe8000b70, pGameData=0x555555aef820,
    pError=0x7ffff4c59900 "unknown error", ErrorSize=256) at /root/build/ddnet22/ddnet/src/game/server/scoreworker.cpp:241
#4  0x0000555555583315 in CDbConnectionPool::ExecSqlFunc (pConnection=0x7fffe8000b70, pData=0x555555c17390, w=NORMAL)
    at /root/build/ddnet22/ddnet/src/engine/server/databases/connection_pool.cpp:460
#5  0x0000555555582815 in CWorker::ProcessQueries (this=0x5555559ff5f0)
    at /root/build/ddnet22/ddnet/src/engine/server/databases/connection_pool.cpp:320
#6  0x0000555555582559 in CWorker::Start (pUser=0x5555559ff5f0)
    at /root/build/ddnet22/ddnet/src/engine/server/databases/connection_pool.cpp:277
#7  0x00005555556d0857 in thread_run (user=0x5555559fe2d0) at /root/build/ddnet22/ddnet/src/base/system.cpp:820
#8  0x00007ffff6e05ac3 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442
#9  0x00007ffff6e97850 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
(gdb)

what can i do with that....

13:59

when i change maps my server crash

13:59

with "Segmentation fault (core dumped)"

14:00

only when sql is on

14:00

but idk what's wrong with my sql database

Is this current ddnet state?

14:01

I think I've seen this error before and fixed it

deen

Is this current ddnet state?

wdym

deen

I think I've seen this error before and fixed it

ah

14:01

idk

current repository in https://github.com/ddnet/ddnet or an older state?

GitHub - ddnet/ddnet: DDraceNetwork, a free cooperative platformer ...

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

current

from what i can see, is IsNull just not initialized when calling mysql_stmt_fetch_column() ?

14:02

bool CMysqlConnection::IsNull(int Col)
{
        Col -= 1;
        MYSQL_BIND Bind;
        my_bool IsNull;
        mem_zero(&Bind, sizeof(Bind));
        Bind.buffer_type = MYSQL_TYPE_NULL;
        Bind.buffer = nullptr;
        Bind.buffer_length = 0;
        Bind.length = nullptr;
        Bind.is_null = &IsNull;
        Bind.is_unsigned = false;
        Bind.error = nullptr;
        if(mysql_stmt_fetch_column(m_pStmt.get(), &Bind, Col, 0))
        {
                StoreErrorStmt("fetch_column:null");
                dbg_msg("mysql", "error fetching column %s", m_aErrorDetail);
                dbg_assert(0, "error in IsNull");
        }
        return IsNull;
}

my brain isn't braining

[ddnet/ddnet] Pull request closed: #9454 Allow super users hammer others & fix jumps when m_Jumps is -1 or 0

Bind.is_null = &IsNull; will make it be written to

hm, odd

is the problem the sourcecode or my sql db ?

source code

okey

14:06

i can't solve it or i can ?

maybe a compatibility problem with the mysql version you are using

deen

maybe a compatibility problem with the mysql version you are using

hmm probably

which one is it, what operating system are you on?

ubuntu

14:06

22.04 or something

14:06

64 bits

14:06

maybe

can you do mariadb --version (forgot, just mariadb is enough it's a dropin replacement) (edited)

mysql Ver 15.1 Distrib 10.6.18-MariaDB, for debian-linux-gnu (x86_64) using EditLine wrapper huh how the hell is it debian..........

14:07

....

10.6.18 is old - i'm on 11.4.4 - works for me

okkk

14:07

how can i update

14:07

because i installed it today

14:09

Reading state information... Done
mariadb-server is already the newest version (1:10.6.18-0ubuntu0.22.04.1).
The following packages were automatically installed and are no longer required:
  libevent-pthreads-2.1-7 libmecab2 libprotobuf-lite23 mecab-ipadic mecab-ipadic-utf8 mecab-utils
Use 'sudo apt autoremove' to remove them.
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
root@vps385544254:~#

ubuntu has package indexes, the newest version seems to be 24.10 google how to update the package index on ubuntu - the file is located under /etc/apt/sources.list

meloƞ

ubuntu has package indexes, the newest version seems to be 24.10 google how to update the package index on ubuntu - the file is located under /etc/apt/sources.list

okey

14:10

thx

good luck!

thx

meloƞ

ubuntu has package indexes, the newest version seems to be 24.10 google how to update the package index on ubuntu - the file is located under /etc/apt/sources.list

is docker slow

14:11

just curious, cuz usually when i hit these i just use the docker image justatest

justatest

well that just hides the fact that it's borked on their machine but doesnt fix the underlying problem

i'm asking for me tho

14:13

greenthing

kek

- if you're just using docker to build the server, i'd consider it okay'ish

i often keep my servers distro untouched, install dockers and just get my services that way.

that's fair for people knowing how to use docker yea

thx

ReiTW had to sit through 2hours of me suffering with docker

i feel assured

meloƞ

ReiTW had to sit through 2hours of me suffering with docker

what?

14:15

which part

i setup a test environment for blockworlds - and i had to have the DDNet server communicate with an api i wrote

sounds very manageable, maybe i'm too docker coded

looking back it was very very manageable

14:18

but understanding where you are within the docker container etc made my head go brrrrrr

14:18

like.. "okay this file i need to copy into here, this into DAMN IT THAT'S THE WRONG PATH"

justatest

ideally you would never need to go inside a container

14:18

well maybe you are trying to build a image yourself

14:18

then idk, that can be hard

meloƞ

ubuntu has package indexes, the newest version seems to be 24.10 google how to update the package index on ubuntu - the file is located under /etc/apt/sources.list

i don't find it..

14:20

not from my VPS

https://ubuntu.com/server/docs/package-management#updating-the-package-index

we're down to 666 issues justatest

justatest

Image attachment

did melon literally just means apt update

eidderFsT

did melon literally just means apt update

no, updating the actual index by changing the repository within /etc/apt/sources.list

14:25

so he can fetch latest packages

fair

14:26

i also never done that before

i had to do that on my raspberry pi once kek

kek

14:26

broke my entire home assistant with it NotLikeKogasa

NotLikeKogasa

[ddnet/ddnet] Pull request closed: #9455 Translations for upcoming DDNet 18.9

15:03

github-merge-queue[bot]

[ddnet:master] 2 new commits

aaf817a Translations for upcoming DDNet 18.9 - def- 7376184 Merge pull request #9455 from def-/pr-transl-18.9 - Robyt3

15:03

github-merge-queue[bot]

[ddnet/ddnet] Pull request closed: #9456 DDNet 18.9

15:04

[ddnet/ddnet] Pull request closed: #9457 Bump nightly version to 19.0

15:04

github-merge-queue[bot]

[ddnet:master] 2 new commits

4101267 Bump nightly version to 19.0 - def- 0b4b00b Merge pull request #9457 from def-/pr-bump189 - Robyt3

MilkeeyCat

we're down to 666 issues justatest

justatest

robyte actually did it, epyc ddnet maintaining moment

wtf, it's possible to chat with copilot on github -.-

Image attachment

ask him how a rewrite in javascript would look like

Y'all should move to Codeberg, for real

Replying to @MilkeeyCat wtf, it's possible to chat with copilot on github -.-

how fast are their runners?

is there any script to format pyson code in ddnet repo?

that's the only selling argument for me tbh

MilkeeyCat

is there any script to format pyson code in ddnet repo?

i think there is some checker that simply uses pysoron tools

also, am I late to translate?

3h too late maybe

15:28

so no

lmao

Image attachment

i mean that is what you wanted technically xd

15:31

was it to fix the issue

[ddnet-web:master] 1 new commit

826173d Add 25€ donation by Route66 - def-

@Route66 Thank you!

1

On the side note, uh, why am I still here? I don't play DDNet nowadays anyways...

[ddnet-web:master] 1 new commit

c90e3ba 1 € Paypal fee - def-

EGYT

On the side note, uh, why am I still here? I don't play DDNet nowadays anyways...

Nobody can quit teeworlds

false

Replying to @Jupstar ✪ Nobody can quit teeworlds

It's true

false

Replying to @Jupstar ✪ It's true

Prove it

can a developer check my config file pls to see if it works

15:38

to see if the problem comes from there or not*

why don't you let the game check it

15:38

ok go

Me ?

yes

hum

look

Replying to @Jupstar ✪ Prove it

in fact

15:39

I have already tried to look at what it tells me

15:39

but I still have the same problems

tell your problem

the database does not load sqlite data, I wonder if it comes from the connection in autoexec_config

15:40

and when the top5 is displayed it is without nicknames (even though there are basic ones)

15:40

and the /mapinfo displays nothing either even though they are supposed to display data

but you opened the sqlite file and looked if all data was correctly written?

and sometimes it can't even connect to the base

Jupstar ✪

but you opened the sqlite file and looked if all data was correctly written?

Yes

15:40

sql file now

are u on windows?

no

15:40

linx

15:40

linux

15:41

and it's an sql file, I corrected all the errors normally

if writing works and reading not, then maybe it's a bug in the code?

Jupstar ✪

if writing works and reading not, then maybe it's a bug in the code?

what code?

15:41

which

the source code

huh

15:41

I can't do anything about it can I?

you could fire up a debugger and see why reading fails. or add more logs to see if errors occure

Jupstar ✪

you could fire up a debugger and see why reading fails. or add more logs to see if errors occure

ok

15:42

I'm going to see

looks like using a rc version early can get you banned from some servers troll

troll

Image attachment

Really nice! Now the next goal is 666 stars

Replying to @MilkeeyCat we're down to 666 issues

cyberfighter 2

looks like using a rc version early can get you banned from some servers troll

troll

ez

Is MySQL not working a known issue @deen? If only mariadb works and MySQL doesn’t we should maybe mention that in the readme or fix it.

chillerdragon

Really nice! Now the next goal is 666 stars

Hold on let me help

chillerdragon

Is MySQL not working a known issue @deen? If only mariadb works and MySQL doesn’t we should maybe mention that in the readme or fix it.

it's sqlite

chillerdragon

Is MySQL not working a known issue @deen? If only mariadb works and MySQL doesn’t we should maybe mention that in the readme or fix it.

I tried on mariadb too but I was told that I needed version 24.10 of ubuntu to install version 11.4.4 of Mariadb on my server,

i am not

I tried on mariadb too but I was told that I needed version 24.10 of ubuntu to install version 11.4.4 of Mariadb on my server,

and it gave me an error when I tried to change maps (I think it still does)

mysql has been deprecated for a while now with mariadb being the drop-in replacement for it - we could at this point just change mysql to mariadb in the readme (edited)

i am not

I tried on mariadb too but I was told that I needed version 24.10 of ubuntu to install version 11.4.4 of Mariadb on my server,

but I can't install 24.10

can u just start a exec on macos? or is that too advanced?

chillerdragon

Is MySQL not working a known issue @deen? If only mariadb works and MySQL doesn’t we should maybe mention that in the readme or fix it.

should be fixed

Jupstar ✪

can u just start a exec on macos? or is that too advanced?

too hard I think

i think so too

xD

15:48

I'm going to wait for my friend to come back to see all that

15:49

I have never used SQL

chillerdragon

Really nice! Now the next goal is 666 stars

and Route66 renames to Route666

i am not

I have never used SQL

Don't let deen see that, but databases are boring

Jupstar ✪

Don't let deen see that, but databases are boring

fr

cyberfighter 2

looks like using a rc version early can get you banned from some servers troll

troll

that's known

well that happened to me for the first time

@eidderFsT btw i know u asnwered this 20 times, but why exactly do u need a new flag?

ye, it can only be fixed at their end now

like what does the flag tell the client or server

what type of coordinate it is in the input packet

wtf D:

i see

seeing cursors in spec is so cursed

cyberfighter 2

seeing cursors in spec is so cursed

it is as it always should have been

you can turn it off

i also added a dyncam_spec config in ddnet-rs

no menu settings yet, cuz i don't want to touch menus tbh (edited)

Jupstar ✪

i see

any further ideas? cuz usually this convo goes to people trying to figure out ways to do it without a flag in the input.

why ukr server dont work

15:56

?

15:56

Image attachment

wut

Image attachment

eidderFsT

any further ideas? cuz usually this convo goes to people trying to figure out ways to do it without a flag in the input.

In ddnet-rs it's simply solved by the fact that the server knows if u are in freespec or not

zen1xula

Click to see attachment 🖼️

the ukr server is currently under attack. you can check with /ddos here

1

is there a way to revert this

like freecam is a proper concept

15:58

and the clients know about it for other clients too then

brownbear

but i wouldn't say this is better to be clear about it xd

15:59

i actually thought about making it an enum first too

15:59

but i dont think a local player ever wants to follow what the other tee is spectating

@Skeith last chance btw, are you still hating it. and have you convinced lerato to hate it too

I definitely think my suggestion is a massive improvement for the overall feel

where can i read?

16:05

or is there a tl;dr?

we might get more people on yourside during the RC. ill prepare a pr but ill wait for a bit

Jupstar ✪

where can i read?

the skeith hated thing?

eidderFsT

the skeith hated thing?

what he hates and what he wants to improve

16:06

best in 2 sentences xd

https://discord.com/channels/252358080522747904/293493549758939136/1323634633425948763

Jupstar ✪

best in 2 sentences xd

spectating players have a separate zoom value from playing zoom now. because when i implementing the new auto spec cam feature where it sync what spectated player camera setting it is easier to manage how the feature toggles

oh it even copies the zoom?

code wise

yeah that is a no-go imho

Jupstar ✪

oh it even copies the zoom?

yes

zoom is something you set locally

that doesn't bother skeith

16:08

you can turn it off as soon as you zoom once

eidderFsT

you can turn it off as soon as you zoom once

yeah but i think that is a bad default

16:09

it should be opt-in not opt-out

16:09

in 99% of cases u want your zoom

I think it's perfectly fine to see the exact perspective of the person you're spectating, but it shouldn't override your zoom level when you're currently trying to zoom out, I also think it shouldn't have 2 different zoom levels between spectating a person and your own

16:09

these are my gripes with it

mostly the two zoom i assmue

16:10

i do plan on remove the two separate zoom but skeith still want the auto cam thing opt out by automatically turn on via condition which i am opposed to (edited)

in a demo i could understand the zoom thing (maybe) but in a spec i do not tbh

im fine with opt in

It's good information to see what a person is seeing

skeithnerd

1

Skeith

It's good information to see what a person is seeing

how often?

16:11

as moderator maybe

i think if auto cam is off it should never turn itself back on unless when you can see the indicator.

a normal person just wants to spec around

16:12

i'd be surprised if many ppl like it

id be surprised if many ppl hate it either, i think most just wouldnt care.

I think it's just fun in general to see the exact perspective of other tees when you're looking at their POV

Jupstar ✪

i'd be surprised if many ppl like it

btw it was implemented so by default the cursor doesnt go out of your screen too often if you are spectating a zoomed out player (edited)

16:13

which their cursor is futher out

16:14

its really for the spec cursor at least when i implemented it

Also freddie I had a bug with the cursor in a demo I recorded, the cursor was attached to someone else entirely I think? rather than my perspective

will take a look. i could see that happen.

Skeith

I think it's just fun in general to see the exact perspective of other tees when you're looking at their POV

but you don't. or does your screen resolution change too?

there is only one cursor info at all time so in demo it will only work for the tee you are recording.

16:16

wait no

16:16

im lying

16:17

actually i need your demo now (edited)

16:17

or a screenshot

Jupstar ✪

but you don't. or does your screen resolution change too?

Don't think so, but you still know what they're looking at regardless

eidderFsT

actually i need your demo now (edited)

Let me see if I can find the demo

Image attachment

16:18

the heart on the side really doesnt help troll

troll

troll

troll

cyberfighter 2

Click to see attachment 🖼️

it is not toggleable

[ddnet/ddnet] Pull request opened: #9458 Add data to Sv_TeamsState

Closes #8205

Checklist

[ ] 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 change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#...

eidderFsT

it is not toggleable

i think it should be reverted

Skeith

Don't think so, but you still know what they're looking at regardless

I dunno, i am not really convinced. I think we could add a small

icon next to the name in the spectator list and only if someone clicks it, it uses the same zoom too

16:21

I wouldn't want to manually change the zoom all the time, just bcs someone plays on 21:9 TV and zooms out by default

16:21

and Teero that pro plays zoomed in

16:21

that sucks even more

My suggestion makes it only show their perspective at your default zoom, if you zoom out once it breaks it and should never copy their zoom levels until you reset

16:22

I think that's fine

and i dont

It's like spectating someone else in any other game

i think it should never turn back on if you zoomed at all

Yall weird as fuck not gonna lie, but whatever, impossible to convince

16:23

do whatever

Jupstar ✪

I dunno, i am not really convinced. I think we could add a small

icon next to the name in the spectator list and only if someone clicks it, it uses the same zoom too

that might be a good idea but where were you before the RC tho. lol i cant just make a UI change this fast.

Skeith

My suggestion makes it only show their perspective at your default zoom, if you zoom out once it breaks it and should never copy their zoom levels until you reset

So if i have my default zoom, i see Teeros zoom in?

eidderFsT

that might be a good idea but where were you before the RC tho. lol i cant just make a UI change this fast.

but we can still default to off for now

Jupstar ✪

So if i have my default zoom, i see Teeros zoom in?

You'll see exactly what Teero is seeing

16:25

minus resolution difference

Yeah but I dont want to xD

16:25

I simply don't understand why ppl would like that as default behavior

Don't spectate teero then, or zoom out, not everyone's gonna be a weirdo like teero

But if 99% have default zoom then it's even worse

16:25

when suddenly it zooms out or in

Why can't the cursor be accurate regardless of zoom level btw, freddie

yeah who the fuck broke that btw xD

16:27

that is defs the most useless change ever

16:27

the cursor is relative to the tee

Is it a lack of information or is Freddie bad at math?

Skeith

Why can't the cursor be accurate regardless of zoom level btw, freddie

do you mean the zoom level changes consistency?

16:28

i changed it back already. justatest

justatest

if you are talking about aiming, not spectating

Idk, I'm assuming when you zoom it'll just inaccurately show the person's cursor position, if that's not the case then just don't match the person's zoom level at all, it's pointless besides fun / slight information

16:28

I just notice when i zoom out when I see the cursor it glitches their cursor a bit in the process

Skeith

Idk, I'm assuming when you zoom it'll just inaccurately show the person's cursor position, if that's not the case then just don't match the person's zoom level at all, it's pointless besides fun / slight information

oh, we assume cursor is world pos

I'm assuming it breaks it, or is it just recalculating it?

so idk what is broken

16:29

do you have a screenshot comparison or something

yeah but that is just bad design, why should cursor be world pos xD

16:30

if we need to translate cursor pos to world pos for a cmd, then do it on the client

Jupstar ✪

yeah but that is just bad design, why should cursor be world pos xD

i could make it not, but it is nice to see what a cursor is landing on exactly.

the server does need to know the exact screen resolution to work, lmao

aiming is screenpos now

16:30

telecursor is specially handled

ok then let me rephrase

16:31

if it's world pos. then the client should send it like that to the server

16:31

but even that

16:31

should not be used by the server

16:31

it just sounds super inaccurate

16:32

for spec ok, but for calculations, i dunno

are we talkind about aiming or not (edited)

well also about telecursor

16:33

aiming must be relative always anything else must get the world pos from client

16:33

not calc on server

i could remove telecursor chat command and make it a netmsg.

16:34

but we need to remove telecursor fast

16:34

for this version

but i asumme it wasnt u who even added it, or was it?

telecursor? (edited)

16:34

no

Accept friend request so I can send you the zipped demo freddie

16:34

I'm assuming that's why I can't send you this file

just send here

Maybe discord's angry about it being a zip file, 10MB upload limit is amazing

no secrets under tees

16:36

here u have high upload limit too

No thanks, dunno if anything private was said

ok then send here with pw

16:36

and send pw to freddie

16:36

:J

Skeith

Accept friend request so I can send you the zipped demo freddie

one sec. i'm flooded with dms and random things

cursor seems entirely broke in this demo, dunno why

in #bugs we also have a case of wrong interpolation

aiming was inconsistent after someone decided to send world pos, and the world pos is actually wrong if dyncam is on. telecursor position hence was broken under dyncam, and aiming made some old players feel really inconsistent.

[ddnet/ddnet-rs] Issue opened: #5 libvulkan on macOS not found

I have libvulkan in /opt/homebrew/opt/vulkan-loader/lib/, but still seeing:

[2025-01-02T16:41:08Z INFO  graphics_backend::backend_thread] graphics backend thread exited: dlopen(libvulkan.dylib, 0x0005): tried: 'libvulkan.dylib' (no such file), '/System/Volumes/Preboot/Cryptexes/OSlibvulkan.dylib' (no such file), '/usr/lib/libvulkan.dylib' (no such file, not in dyld cache), 'libvulkan.dylib' (no such file)

so i changed it back to screenpos and found other ways to make telecursor work

@eidderFsT Could you just make it to where auto zoom never zooms in past your default zoom

and to make telecursor work client was sending camera info to the server. the camera info is never used for anything apart from telecursor. which lead to spec cursor having the info. which is how auto spec cam was implemented

Skeith

@eidderFsT Could you just make it to where auto zoom never zooms in past your default zoom

i could

16:49

so if we want a rework of telecursor, the whole spec cursor need to be scraped again

Would remove trolling (people zooming in to extreme levels) I really wish we could just have an option to make default zoom auto

16:50

I like my version a lot more than having to toggle it on each time

arch1t3cht

it might depend on your PATHEXT environment variable

I'll give it another try. I'm on a fresh Windows installation at the moment. I encountered this issue on a much older installation before.

Skeith

Would remove trolling (people zooming in to extreme levels) I really wish we could just have an option to make default zoom auto

btw i said i could but im like fixing four or five things at once plus the event server for some reason.

16:52

im currently very overwhelmed. ill try to think more before 18.9 releases, hopefully i have enough time

The current version is fine enough as is, it's just mildly annoying and you should 100% expect people to be annoyed by the double zoom when it hits release

im also rushing to fix chn stuff for websites and bots pepeW

pepeW

idk why it all happened at once lmao

Skeith

The current version is fine enough as is, it's just mildly annoying and you should 100% expect people to be annoyed by the double zoom when it hits release

my goal i think is to remove the two zoom but not the default zoom = auto spec cam

16:55

currently

Add an option for it, #fuckthesystem

btw are we adding cursor to tab?

That whole context menu thing?

ye (edited)

No clue! Would be cool

16:57

(Also don't add an option for it btw, you'll have deen hein and maybe robyt at this point yelling at you, unlucky)

actually. can i add a button in esc menu instead of assuming stuff.

I wish we could have options for things that are used a lot! Options good for convenience :C

is that a place we could use?

I have no clue, I'm just a lowly moderator

like there you can turn it on and off. on is your behavior, off is if you turn it on it auto turns off

That sounds like an option to me, freddie!

16:59

tsk tsk

sure

17:00

so on assmues you want it, and temporarily turn it off. and off assumed you dont want it unless you temporarily turn it on

Skeith

That sounds like an option to me, freddie!

what was the accuracy thing btw

Accuracy thing?

Skeith

Why can't the cursor be accurate regardless of zoom level btw, freddie

this

I think I might've been wrongly assuming something about the cursor

17:02

Is it already accurate to where they're aiming regardless of your own zoom level

Skeith

Is it already accurate to where they're aiming regardless of your own zoom level

is it the thing that become inaccurate at some point

17:03

like a few month ago

Hook collision?

17:04

the laser pointer

hmm

Skeith

Is it already accurate to where they're aiming regardless of your own zoom level

is it #8452

https://github.com/ddnet/ddnet/issues/8452

or are you talking about other things

I assumed it just didn't show an accurate position cause it gets kinda glitchy when you first zoom out when spectating someone, the cursor jumps to a different position, but I think it just corrects itself once

so spectating

Ya, if you're spectating a tee and watching their cursor, if you zoom out their cursor will zoom out as well but then go back to the "accurate" (I'm assuming) location, it corrects itself

assuming specing 18.8

17:08

sounds like a bug

17:09

well more like a compatibility issue but i think i know what caused it

Also I totally do have that hook collision line bug currently, if I zoom out my hook collision shows it hooks through when it doesn't

Skeith

Ya, if you're spectating a tee and watching their cursor, if you zoom out their cursor will zoom out as well but then go back to the "accurate" (I'm assuming) location, it corrects itself

eitherway, ye, i think i touched way too much code for this version, from controls to server cursor to spectating, so it gets really confusing which one when you just say accuracy

Skeith

Also I totally do have that hook collision line bug currently, if I zoom out my hook collision shows it hooks through when it doesn't

i can check that, although i dont think it is related but it sounds like i couldve cause it

Skeith

Ya, if you're spectating a tee and watching their cursor, if you zoom out their cursor will zoom out as well but then go back to the "accurate" (I'm assuming) location, it corrects itself

can you check if it is also like that when specing 18.9

Image attachment

[ddnet/ddnet] Pull request opened: #9459 teleport hookline (only when there is 1 possible exit)

Currently the hookline ignores hook teleports, I think it would make sense for it to pass through the teleport and actually show where the hook will go. This should solve issue #8735 !image

Checklist

[x] Tested the change ingame
[x] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially bas...

Skeith

Click to see attachment 🖼️

Image attachment

17:20

did i miss anything

Don't believe so

eidderFsT

Click to see attachment 🖼️

make community vote if it should be default on

it wont be

17:23

actually idk

17:23

you do that

ok give #announcements permissions

i dont think people know what it does yet tho

give me insanity's role

Yall devs need to not be so afraid of options lmao

17:25

If the feature is used CONSTANTLY and people would obviously be divided on using it, allow an option...

Skeith

Yall devs need to not be so afraid of options lmao

I am afraid of bad default behavior

lol

Jupstar ✪

I am afraid of bad default behavior

Have it defaulted to off, but have an option for people that wouldn't want to have to always toggle it on aka me

17:29

Honestly, I definitely see why people are using the DDNet client less and less

Skeith

Honestly, I definitely see why people are using the DDNet client less and less

More and more ppl use ddnet actually

As a moderator, I don't think this is true

ChillerDragon: does teeworlds 0.6 server acknowledge ddnet packets or just ignores them? i tried ignoring them and ddnet kept sending me them with resend flag xd

i honestly worry people would wonder why they cant see some peoples cursor cuz it was out of screen. but that is also the case for specing old clients so i probably shouldn't

Jupstar ✪

More and more ppl use ddnet actually

nope, more and more people use modified ddnet clients

Skeith

As a moderator, I don't think this is true

But see your argument is two sided. You want new technical features like demos ingame, accounts? Then you cannot spam the client with useless stuff all the time and hope it works forever

Big influx of new players skews data to determine this I guess, but a lot of people have been swapping to tater.

Cellegen

nope, more and more people use modified ddnet clients

See the point is that more and more people are playing DDNet in general, and you missed the joke

Jupstar ✪

But see your argument is two sided. You want new technical features like demos ingame, accounts? Then you cannot spam the client with useless stuff all the time and hope it works forever

Eh, we can certainly be smarter about options, rather than being against having them no matter what

DDNet specific clients, more likely

i should sleep

Good night Freddie

it doesn't mean they are using the original, which is weird, but understandable

haven't been able to for about two nights now lol

Skeith

Eh, we can certainly be smarter about options, rather than being against having them no matter what

I'd say it's a hard topic, I could imagine most options are not used or only by 1-2 ppl

if the topic is "why people use other clients, than ddnet's client" then it makes sense to say that they are not using the vanilla ddnet

Not every russian kid who joins downloads that which must not be named most of them are probably getting vanilla DDNet

But we don't have any stats

Shrug, this feature would be. So I'm definitely advocating for an option here

I didnt say I am against this specific option, I was really more about your generalization

Ya I understand, I believe Freddie tried to add options for this feature when it was originally submitted, and devs opted not to have any. Which is a problem in my opinion

17:36

Ya'll need to think more about having options for major changes, imo

Well maybe you are right, maybe we should be smarter about options. E.g. if a new option is added the dev also has to add tests for it. For the client this probably means taking some kind of screenshots to ensure behavior is as intended

Can probably phase out certain options over time if they're deemed unused

i always have videos in my pr

17:39

well i lied

17:40

but i do care a lot about my pr to have clear enough info imho

Yea even if you do not know the Message id you should acknowledge all vital chunks.

Replying to @Koll Potato ChillerDragon: does teeworlds 0.6 server acknowledge ddnet packets or ju…

im going crazy

17:51

ty

If you really go crazy go allocate some time to setup libtw2 wireshark dissector

17:52

It will keep you sane you can just compare a healthy connection to yours and see exactly where you are going wrong

chillerdragon: i love you

yeah because debugging the same problem for 5 freaking days is not normal

17:52

even more ig

@eidderFsT: u ok?

17:53

Happy new year Freddie

Love you too

how do people sleep when they got tasks to do

17:55

i cant sleep

Pills and alcohol

ah forgot i was asking a german

ChillerDragon: wait are you on arch too?

Yes BTW

is it possible to install older wireshark versions with pacman or i have to build it manually?

Is libtw2 behind? -.-

yeah

17:56

thats why i couldnt set it up the previous time

I often use my own package manager crackman

17:57

Which allows me to get new stuff on Debian and old stuff on arch

17:57

Compiling wireshark is smooth af it just is a fat git repo

but how exactly does it do that

17:57

alright

Then it’s one build command in libtw and moving the .so

17:58

It will take a bit of time but trust me it’s a game changer for your problem

thank you the teeworlds expert

writing cis is harder than writing vk

oh no

Image attachment

change my mind

18:05

well only for windows tbh

18:05

since 2 days i do nothing else lmao

i should stop checking every messages and assume everything is my fault justatest

justatest

eidderFsT

i should stop checking every messages and assume everything is my fault justatest

justatest

nothing is your fault lol

anxiety prob

first of all we have a clash of opinions

no i was looking at general

Jupstar ✪

nothing is your fault lol

oh btw, someone said the auto thing did clamp to their settings already, do you mean clamping to a safer one or is that ok

18:09

ill double check the gl1 handling too tomorrow

eidderFsT

oh btw, someone said the auto thing did clamp to their settings already, do you mean clamping to a safer one or is that ok

mh ok, but what even happens on invalid values anyway

18:09

like what datatype do we even use lmao

int

18:10

ddnet protocol is all int i think?

ig so

int and str

ill double check extreme zoom values too

18:12

str is just an array of uint8, jk

technically int is also array of uint8

I'm unsure why the cursor jump thing happens when zooming Freddie, doesn't happen consistently! gl monkaStop

monkaStop

Skeith

I'm unsure why the cursor jump thing happens when zooming Freddie, doesn't happen consistently! gl monkaStop

monkaStop

are you sure it is not version related

I have no clue, not enough people on nightly to test

do you always see the jump if they dont have an indicator

I tested with Souly he's not on nightly, couldn't get the cursor to jump when I zoomed

18:15

I have no clue what causes it to happen, maybe it's when they're moving the cursor quickly when I zoom

i could remove a piece of smoothing and make sure it always shows the real position

maybe overlay in debug mode like the tee

not that one

18:16

but good idea

18:16

i did turned off interp when debug is on (edited)

18:17

but robyt3 says debug should not toggle features

18:17

so maybe i should make it a overlay

18:18

idk if they buy that i added a debug overlay after the feature is already implemented tho

Koll Potato

oh no

when is this allowed ever? is someone trying to do a fish-style command substitution

eidderFsT

idk if they buy that i added a debug overlay after the feature is already implemented tho

add it for urself for now ig

18:18

if the problem is not consistent then rip

its not the interps fault im pretty sure

I got it to happen while watching cyberfighter, it also appears in the demo. he's on nightly

Ewan

when is this allowed ever? is someone trying to do a fish-style command substitution

i just had a space in the directory name xd

18:19

compiling wireshark now

there was a zoom smoothing handling since server camer info is unsmoothed

should be fine

18:20

chiller not escaping paths??

18:20

justatest

(otherwise it would be a lot of netmsg from client)

nah

18:20

my fault

18:20

i installed wireshark twice

18:20

so the second one got named with (1) in the end

rip

6% compiled in 15 minutes

1

eidderFsT

(otherwise it would be a lot of netmsg from client)

why does server need to know zoom? can’t u send its states only at the start and end of the interpolated window

Skeith

I got it to happen while watching cyberfighter, it also appears in the demo. he's on nightly

He was also zooming when I zoomed, maybe it's due to both people trying to zoom when auto is enabled

Ewan

why does server need to know zoom? can’t u send its states only at the start and end of the interpolated window

e.g. just dont interpolate for the server. but i am forgetting what this issue is all about

Skeith

He was also zooming when I zoomed, maybe it's due to both people trying to zoom when auto is enabled

wait. if auto is off the cursor will always jump, is it like jumping back and forth or just teleporting due to them zooming

Image attachment

the cursor will just teleport when they zoom, maybe you speced right when they were changing zoom and it looked like

That's from me zooming and resetting zoom

18:24

Maybe

only in demo?

Nah, also ingame

they might just be zooming in and out

18:25

it looks like they are just zooming in and out really

so you just want the interpolation so that looks good?

i didnt make interp for that cuz we already had three different interp for zoom.

18:26

im scared of stacking another one

I'll record him zooming out then watch the demo to compare, I guess

do let me know if you are sure it is you zooming causing the jump (edited)

I need a good test subject that stands still xD that's also on nightly. Would a dummy have the same results?

if it is not your zooming, then i guess maybe adding a interp is a better experience. the problem is it will never match camera zoom smoothing so we do need two different smoothing, one for auto, one for nonauto oof

Cursor position is very slightly off btw

it can be

18:33

it is integer

18:33

it will be completely wrong if you zoom way out and point somewhere with a mouse_distance 50 or something

Swapping between dummy and spectating dummy, the cursor is slightly off normal zoom, it's just a pixel or two off, not a big deal

ye, server cursor is different that your mouse

18:34

no way to fix that really

Fair enough, just making sure

well there is a way if i just send another pair of floats but that sounds wasteful

It's barely incorrect, so not a real issue

eidderFsT

it will be completely wrong if you zoom way out and point somewhere with a mouse_distance 50 or something

try this, it can be a lot more wrong if you intentionally make it wrong

Can't test the accuracy with dummy, my zoom resets for both greenthing

greenthing

18:39

oh, 2 clients ig

18:43

Oooo, I can make the cursor jump this way

18:43

It doesn't happen every time either, definitely feels like a bug

18:45

@eidderFsT

Image attachment

18:46

Real POV during the zooming

Image attachment

18:47

real cursor location

Image attachment

18:47

So it's trying to show me the unzoomed cursor location each time I zoom while spectating, neat

18:48

or something like that

18:49

(Nah it definitely isn't, it's just jumping to a different position, idk why)

18:50

Not nerd enough to figure that out, glgl monkaStop

monkaStop

18:52

tl;dr: If the person you're spectating is zoomed out and you change your zoom at all, the cursor will jump.

Skeith

@eidderFsT

noted

18:56

this is much more bug like and i can probably tell why if i look hard enough

18:57

tomorrow

Can I run a pre-compiled version of ddnet on the host?

https://discord.com/channels/252358080522747904/293493549758939136/1324430118596120647

so i installed ubuntu 24.10, mariadb 11.4.3 but it still doesn't work, what can i do ?

Skeith

@eidderFsT

wait did i forgot to track dummy's received camera info separately

That was on 2 different clients

oh

19:00

right

19:00

i should sleep

heartw

1

i am not

so i installed ubuntu 24.10, mariadb 11.4.3 but it still doesn't work, what can i do ?

still the same error?

MilkeeyCat

still the same error?

nope

19:05

like

19:05

no errors on it

19:05

but it's the same problem

19:05

the database "connect" but not the data

19:08

`
2025-01-02 16:23:56 E storage: could not determine user directory
2025-01-02 16:23:56 E storage: could not determine fallback user directory
2025-01-02 16:23:56 I storage: couldn't open storage.cfg
2025-01-02 16:23:56 I storage: using standard paths
2025-01-02 16:23:56 E storage: cannot add path '$USERDIR' because it could not be determined
2025-01-02 16:23:56 I storage: added path '$DATADIR' ('data')
2025-01-02 16:23:56 I storage: added path '$CURRENTDIR' ('/home/ddnet/DDNet-18.8.2-linux_x86_64')
2025-01-02 16:23:56 I host_lookup: host='localhost' port=0 1
2025-01-02 16:23:56 I host_lookup: host='localhost' port=0 2
2025-01-02 16:23:56 I console: executing 'autoexec_server.cfg'

19:08

only error i see

ChillerDragon: i have a massive skill issue with permissions on arch

19:16

ive compiled wireshark but it does not let me capture anything

19:16

"you have no permissions"

19:16

even though i have added my user to the wireshark group

19:21

XDDDD

19:21

im a dumbass

19:21

i was setting up access for /usr/bin/dumpcap

19:21

not the one that i have built

19:22

LETS GOOOO

19:22

it works

19:23

or not

19:23

no teeworlds packets

19:28

nice

Image attachment

huh

19:29

why is my message blocked

19:30

whyyy

19:31

why is it blocked

Image attachment

automod flags it as spam

huh

i am not

why is it blocked

Not right. only two lines required.

Zwelf

Not right. only two lines required.

really

19:34

but which one

Zwelf

ah, and also append 1 for crating the tables (on the first run)

sv_use_sql 1
add_sqlserver r ddnet_server record ddnet 'thebestpassword' 121.0.0.1 3303
add_sqlserver w ddnet_server record ddnet 'thebestpassword' 121.0.0.1 3303 1

these

Zwelf

ah, and also append 1 for crating the tables (on the first run)

sv_use_sql 1
add_sqlserver r ddnet_server record ddnet 'thebestpassword' 121.0.0.1 3303
add_sqlserver w ddnet_server record ddnet 'thebestpassword' 121.0.0.1 3303 1

ohhh ok

19:35

i forgot

19:35

thx

19:37

Jan 02 19:36:25 vps385544254.lowhosting.org DDNet-Server[18894]: 2025-01-02 19:36:25 I mysql: connection established it say that but the data doesn't load

The tables in the database should then have the name record_maps, record_race, record_teamrace, record_points, record_saves. (They get automatically created with setupdatabase 1 parameter.

Zwelf

The tables in the database should then have the name record_maps, record_race, record_teamrace, record_points, record_saves. (They get automatically created with setupdatabase 1 parameter.

yes

19:38

Image attachment

Zwelf

The tables in the database should then have the name record_maps, record_race, record_teamrace, record_points, record_saves. (They get automatically created with setupdatabase 1 parameter.

Those tables were created from the long config above. You can drop them except the ones I mention in this message.

okk

19:39

thx

List of tables: https://github.com/ddnet/ddnet/blob/0b4b00b3d9d16ace40725daa827b886790a58eeb/src/engine/server/databases/connection.cpp

ChillerDragon: basically ten minutes of debugging and i got my server to work, looks like the sequence and ack was correct all along

19:40

i was supposed to send snap single and not snap empty

[ddnet/ddnet] Issue opened: #9460 45 degree aim bind does not work when zoomed in.

It seems that there has been a regression where the 45 degree aim bind does not work when zoomed in. instead it will always shoot upwards. Additionally the hookline is not visible when the bug is happening. I've tested a few versions from the ddnet website and found that 18.0.3 has the bug, while 17.4.2 does not. Here's a video of this bug in action (first half of the video is the old version where it works, second half the one where it's broken) https://g...

Zwelf

List of tables: https://github.com/ddnet/ddnet/blob/0b4b00b3d9d16ace40725daa827b886790a58eeb/src/engine/server/databases/connection.cpp

Image attachment

1

19:41

this is good ?

19:41

okey

19:41

nice

looks good

ChillerDragon: i believe snap empty is never sent to the client xD

Zwelf

looks good

data still doesn't load

Is there data in the tables?

yeah

Are there errors loading the data?

hm

ahh snap empty might be sent when delta size is smaller or equal to 0

the only errors

Image attachment

When you finish a map on the server to add new data, then restart the server. Does the new data appear?

Zwelf

When you finish a map on the server to add new data, then restart the server. Does the new data appear?

nope

19:43

but the /top load

19:43

like if i hadn't a database it would just print an error

19:43

but it doesn't

19:44

Image attachment

19:44

is this ok ?

19:45

Image attachment

yes: https://github.com/ddnet/ddnet/blob/0b4b00b3d9d16ace40725daa827b886790a58eeb/src/engine/server/databases/connection.cpp#L55-L60

ddnet/src/engine/server/databases/connection.cpp at 0b4b00b3d9d16ac...

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

19:46

Does /mapinfo 001 not work?

nope

19:46

doesn't work

19:46

like it say "no map like "001" found

is there any response?

19:46

or does it just wait long doesn't write a message to chat?

Zwelf

or does it just wait long doesn't write a message to chat?

wdym

ah, u gave response in one message above

it says "no map like "001" found in the chat

19:47

yes lol

19:48

and i can't vote another map bc it makes the server crash (only when sql is on)

crash doesn't sound nice. (do you have a stack trace?)

huh

19:49

log file ?

19:49

or gbd thing

dump_sqlservers r
dump_sqlservers w

print the respective database server?

19:49

gdb thingy

Zwelf

dump_sqlservers r
dump_sqlservers w

print the respective database server?

i don't have t hat

in server console (f2)

oh

Zwelf

dump_sqlservers r
dump_sqlservers w

print the respective database server?

i gotta print this in the f2 ?

I think so. Yes.

2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_maps' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_points' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_race' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_saves' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_maps_teamrace' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_points' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_race' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_race_backup' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_saves' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:50:58 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_saves_backup' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: ClientId=0 rcon='dump_sqlservers r ' 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_maps' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_points' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_race' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_saves' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_maps_maps_maps_teamrace' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_points' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_race' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_race_backup' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:51:05 I server: MySQL-Read: DB: 'ddnet_server' Prefix: 'record_saves' User: 'ddnet' IP: XXX Port: 3306

Ah, please restart server so it only has the two entries from the database

ok

19:53

the 2nd

19:53

dump_sqlservers w

19:53

doesn't work

19:53

i think

19:53

2025-01-02 19:52:49 I server: MySQL-Write: DB: 'ddnet_server' Prefix: 'record_saves_backup' User: 'ddnet' IP: XXX Port: 3306 2025-01-02 19:52:49 I server: SQLite-WriteBackup: DB: 'data/ddnet-server.sqlite' 2025-01-02 19:53:35 I server: ClientId=0 rcon='dump_sqlservers w'

hm. should print the same database. Do you have it in the config?

yes

And /mapinfo should start working with the r(ead) database already

okk

19:57

Jan 02 19:56:48 vps385544254.lowhosting.org DDNet-Server[20131]: 2025-01-02 19:56:48 I sql: failed connecting to db: (real_connect mysql

mysql

1045): Access denied> Jan 02 19:56:48 vps385544254.lowhosting.org DDNet-Server[20131]: 2025-01-02 19:56:48 I sql: [6] show top5 failed on all databases

19:57

now

gtg

ok

19:58

bb

ChillerDragon: so i plugged in libtw2 snapshot implementation once again just to figure out that i did not attach the size to the data in the snap single chunk lmao

22:03

it works now though

Image attachment

22:03

ddnet rewrite in rust is real

let's go

whos gonna win? jupstar or koll

or zwelf + patiga

22:08

exactly robyte who starts tomorrow

we all kinda work on different things xd

i think what zwelf is doing is kinda close

22:27

he plans a server impl in rust

oh no

22:31

it took me whole 2 months to just get snapshots and initial connection working lmao

I haven't done any networking yet. Only physics + teehistorian stuff https://gitlab.com/ddnet-rs/twgame

well thats because i had zero knowledge about the protocol before

Zwelf

I haven't done any networking yet. Only physics + teehistorian stuff https://gitlab.com/ddnet-rs/twgame

twsnap?

22:32

or is it just for reading

22:32

not writing snaps

It's a better representation for snapshots in rust (output from twgame physics, input for twgpu rendering) and can be serialized to/deserialized from libtw2 snapshot format. (edited)

22:34

https://docs.rs/twsnap/0.7.3/twsnap/

[ddnet/ddnet] Issue opened: #9461 Map Editor only allowing saving in the AppData/Roaming directory

the map editor shows none of the other standard/added directories nor does it allow for the directory path text to be edited i don't think there is any good reason for it to be this way/hardcoded since it shouldn't cause any issues for the game to save in other folders in the past i remember being able to maptest with a local server easily, not this inconveniently, so maybe this could also be bugged

It's the interop between Patigas and my crates.

sounds amazing

@Koll Potato Do you have a link to your repository?

its a complete mess rn and i currently rewrite most of the code

22:44

even wrote custom derive macros

22:45

you can take a look at it on my github, but its the old version

22:45

and it also doesnt work xd

22:47

and i also need a good name

22:47

twgame, ddnet-rs are taken

22:48

im thinking of teeverse but it sounds too similar to tee galaxy

22:48

alr enough talk i go to sleep

22:48

gn devs

gn

22:51

Not sure if still relevant: If you don't modify the crates that you copied from libtw2, you could also use the version from crates.io: https://crates.io/search?q=pre-rfc3243-libtw2

Koll Potato

im thinking of teeverse but it sounds too similar to tee galaxy

Also to https://github.com/teeuniverse/teeuniverse

22:52

but might still be fine

22:53

Might also resemble teewars

[ddnet/ddnet-rs] Pull request opened: #6 Add basic CI build with artifacts, but without tests.

Also builds ffmpeg feature for Windows using MinGW and lots of hacks. MacOS is build but probably useless (#2) and also the binary only rn.

22:59

[ddnet/ddnet-rs] Pull request closed: #2 ci

23:00

[ddnet/ddnet] Pull request closed: #7886 fix cmake warning when building with -DMYSQL=ON

Oh wow. Who said TeeUniverse? Its client and server code does not really exists but the editor is used for Infclass maps. I maintain a fork at https://github.com/infclass/teeuniverse and last year added a few video tutorials.

Replying to @Zwelf Also to https://github.com/teeuniverse/teeuniverse

23:18

Meh, I somehow misread teeverse. nvm then

i want the teeuniverse org xd

23:21

perfect space to host mods

Yeah, and it was supposed to be something like that. But then the author had lost their energy and disappeared.

I wish we could at least contact him for the name <.<

BTW do you know if there is a place for teeworlds shared knowledge base? DDNet wiki is focused on ddnet (what a surprice). Teeworlds forum is dead, various fandom wiki, etc are dead. I could not find a resource with e.g. list of all known clients.

I am sure ddnet wiki would accept teeworlds content too, but not really, chilllerdragon probs has some repos where list of mods etc. are

They do not respond. It seems that they registered the account (https://github.com/necropotame, and the same @gmail.com) exactly for this project.

Replying to @Jupstar ✪ I wish we could at least contact him for the name <.<

Ah alright

23:33

Too bad

Yeah but those are focused on server mods.

Replying to @Jupstar ✪ I am sure ddnet wiki would accept teeworlds content too, but not really,…

Jupstar ✪

i want the teeuniverse org xd

tee cinematic universe

Exported 842 message(s)

Timezone: UTC+0