DDraceNetwork - developer

[ddnet/ddnet] Issue opened: #8208 Demo shortcuts no longer work when demo menu is closed

b4a22e2feeab15c87630402e3a20cd7bb7926faf is the first bad commit
commit b4a22e2feeab15c87630402e3a20cd7bb7926faf
Author: Robert Müller 
Date:   Tue Feb 6 21:41:59 2024 +0100

    Refactor client state handling in menus

#8156

00:17

dobrykafe

[ddnet/ddnet] Pull request opened: #8209 Allow the vote creator to cancel the vote + sixup vote fix

Vote cancel taken from https://github.com/teeworlds/teeworlds/pull/1436. If you are the vote creator, voting no (pressing f4) will cancel the vote. I didn't include the time limit and added a custom chat response: !vote Also fixed a bug where the server would incorrectly set the fields for protocol7::CNetMsg_Sv_VoteSet. The 0.7 clients would for example receive VOTE_END_FAIL instead of the co...

def-

[ddnet/ddnet] Issue opened: #8210 TCP handshake before UDP connection

Suggested by Davide, apparently FiveM uses this too. First do a TCP handshake (on the same port number) with the server, then connect over UDP.

If you implement the TCP port "punching" and you allow only the IPs who complete the handshake to send packets over the UDP port And you can actually using this system to block small/medium DDoS bypass. For example on a provider like OVH But still, even if the provider doesn't support that, you can run your own UFW/iptables rules

I'm ...

chillerdragon

lmao

damn the embed revealed my un-edited message

Learath2

AVX512 has some real bangers

true

def-

[ddnet/ddnet] Pull request closed: #8206 Change type of m_HookTick to NetIntAny

04:14

github-merge-queue[bot]

[ddnet:master] 2 new commits

32fcd2a Change type of m_HookTick to NetIntAny - furo321 0691d42 Merge pull request #8206 from furo321/rename-hooktick - def-

04:14

def-

[ddnet/ddnet] Issue closed: #8022 CharacterCore::m_HookTick is a duration, not an absolute tick

04:16

github-merge-queue[bot]

[ddnet:master] 3 new commits

b6a1a12 fix votes for sixup - dobrykafe 8ca395c allow the vote creator to cancel the vote - dobrykafe 1b77cbc Merge pull request #8209 from dobrykafe/pr-votes - def-

04:16

def-

[ddnet/ddnet] Pull request closed: #8209 Allow the vote creator to cancel the vote + sixup vote fix

04:16

github-merge-queue[bot]

[ddnet/ddnet] Pull request closed: #8209 Allow the vote creator to cancel the vote + sixup vote fix

ddnet-maps

[ddnet-maps:master] 1 new commit

f1e3c59 A Nagi - ddnet-maps

wii fails at connecting to a server spectacularly

ws-client

<ChillerDragon> also have you thought about live map updates? How hard would it be to build minetee?

chillerdragon: are u on?

06:46

i can give you a hint, you saw live updates yesterday, i compared it against gimp, remember?

Headshot

wii fails at connecting to a server spectacularly

does it get inactive, but still works or does it hang?

it's not frozen, i can move the pointer and click cancel, but it never gets past sending info

06:50

the wii is big endian, could that be an issue?

uff defs

we have some semi good looking big endian hacks in the code

06:51

but i assume they were never truely tested

you can test a big endian thing under qemu-user maybe

06:52

might be easier to debug that way

you can

06:52

xd

06:52

but is it worth your time

06:53

big endian is basically dead nowadays

I think so

+ for rust again, which simply offers to_le_bytes, from_le_bytes

does that just become a nop on le or what?

yeah i guess so xd

06:57

but anyway, headshots problem could also be unrelated

06:57

maybe the compiler is buggy or smth

06:57

or our code lol

true

Headshot

the wii is big endian, could that be an issue?

Did you fix detect.h for the Wii endianness? Last time I saw some log messages from this project it said "unknown endianness"

08:20

New detect.h doesn't allow compiling with unknown endianness

CONF_ARCH_ENDIAN_BIG is defined, yes

what features would make C better and more convenient language in your opinion?

Oop (edited)

1

struct Foo {
    int bar;

    static Foo new(int bar) {
        return {
            .bar = bar
        };
    }

    int gimme_bar(void) {
        return this->bar;
    }
};

int main() {
    Foo foo = Foo::new(69);
    int val = foo.gimme_bar(); // 69

    return 0;
}

i think this would be enough for me xd

MilkeeyCat

what features would make C better and more convenient language in your opinion?

Namespaces, a new stdlib, some sort of metaprogramming capability maybe templating

MilkeeyCat

what features would make C better and more convenient language in your opinion?

RAII

09:02

without RAII a language directly sucks

09:02

I think I'd take defer statements over raii

Learath2

I think I'd take defer statements over raii

does that work together with struct members?

09:04

or is it just an execution model

Deferred statements execute at the exit of a function (or a block if the language wants)

i can't really imagine a use case, looking at the go examples

You can't because you live in a world with raii

09:05

Imagine a world with no constructors no destructors. Then it's useful

09:05

Think of it as explicit destructors that you schedule to run at exit of block (edited)

Learath2

You can't because you live in a world with raii

so basically u evaluate the arguments directly but put the "destructor" the function that handles the arguments at the end?

Learath2

Think of it as explicit destructors that you schedule to run at exit of block (edited)

mh yeah

09:06

i guess can be useful too, since it's more flexible than always surround everything in structs (edited)

When things are evaluated allows for some cute trickery. But it's not the point. It just lets you have something akin to atexit but at a function level (in gos case)

09:08

It would be great in C because properly cleaning up after yourself is the cleanest with ugly goto statements otherwise

09:08

(Or objects and raii, that would also be a solution to the same thing)

well what i like about RAII esp in rust is:

if the "contructor" returned no error, you can assume the object 100% exists, and cleans up itself
^ => if you strictly follow this pattern and try to not use Option<> to often, your inner logic doesn't require so much error handling + i combine it with my hiarc, which basically prevents circling dependencies of struct members

Learath2

When things are evaluated allows for some cute trickery. But it's not the point. It just lets you have something akin to atexit but at a function level (in gos case)

i mean on function level that defs sounds useful maybe sometimes

09:10

it's like RAII for logic, if i imagine it correctly

Raii is definitely more hands off and easier to get right. But you know me, I'm a more natural selection enjoyer. I'd rather have the flexibility and give the user the option to footgun

Learath2

Raii is definitely more hands off and easier to get right. But you know me, I'm a more natural selection enjoyer. I'd rather have the flexibility and give the user the option to footgun

well who knows, maybe u need to shoot into your foot 10000 times to find an ever better approach than anything we know now

FILE *f = fopen(f);
defer fclose(f);
...
...

09:13

This is the usual pattern in a language with defer

ah ok, i thought about it more like an transaction

The footgun is forgetting to defer an fclose, since it's not done for you like with raii through destructors

change values as you want but then revert it under certain conditions or whatever

Jupstar ✪

change values as you want but then revert it under certain conditions or whatever

Ah, no but that does sound interesting. I'll ponder it maybe we can have a new language

Learath2

The footgun is forgetting to defer an fclose, since it's not done for you like with raii through destructors

true, for c that would defs already be an improvement

Robyt3

[ddnet/ddnet] Pull request opened: #8211 Fix demo player hotkeys not working when menu is inactive, fix duplicated rendering of popup menus in demo player

The demo player needs to be rendered also when the menu is inactive, to handle the hotkeys for controlling the demo player and to render the play/pause and speed indicators. Closes #8208. The popup menus are already rendered in the CMenus::Render function, also for the demo player since #8156.

Checklist

[X] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a u...

github-merge-queue[bot]

[ddnet:master] 3 new commits

717cbef Fix demo player hotkeys not working when menu is inactive - Robyt3 9abe501 Fix duplicated rendering of popup menus in demo player - Robyt3 c878cc7 Merge pull request #8211 from Robyt3/Menus-Demo-Hotkeys-Fix - def-

10:24

def-

[ddnet/ddnet] Pull request closed: #8211 Fix demo player hotkeys not working when menu is inactive, fix duplicated rendering of popup menus in demo player

10:24

def-

[ddnet/ddnet] Issue closed: #8208 Demo shortcuts no longer work when demo menu is closed

Robyt3

[ddnet/ddnet] Issue opened: #8212 Resizing window while rendering demo results in broken client rendering and video

With Vulkan I get the log message swap chain image was not ready to be copied and the view becomes weirdly streched over the resized screen when I resize the window while the rendering is active: !screenshot_2024-04-14_13-09-34 This state persists after rendering is finished until the client is restarted. The rendered video is messed up because the image data does not have the expected size. M...

@Ryozuki windows

https://blog.rust-lang.org/2024/04/09/cve-2024-24576.html

Security advisory for the standard library (CVE-2024-24576) | Rust ...

Empowering everyone to build reliable and efficient software.

xd

11:33

old

furo321

[ddnet/ddnet] Pull request opened: #8213 Fix bug where certain tiles would not work in Game tiles menu

Fixed bug where these first buttons wouldn't place any tiles. Most likely caused by #8136 !image

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...

<ChillerDragon> @Jupstar ✪ ah right you said you wanted to keep the in game code close to the editor so in game is also flexible ye that makes sense. nice

13:34

<ChillerDragon> lerato are you a jai enjoyer?

13:36

<ChillerDragon> i saw defer in C code in tsoding videos

github-merge-queue[bot]

[ddnet:master] 2 new commits

2545c14 Fix bug where certain tiles would not work in Game tiles menu - furo321 06f1db9 Merge pull request #8213 from furo321/fix-gametiles-button - def-

13:45

def-

[ddnet/ddnet] Pull request closed: #8213 Fix bug where certain tiles would not work in Game tiles menu

teini94

[ddnet/ddnet] Issue opened: #8214 /timecp name broken

/timecp name show better timecp anymore I assume since atleast around 1-1,5months? but maybe just since around newest version. 2. While fixing it we could also allow to see "/timecp name" without the need of having the map finished.

@Zwelf I'm still getting a bunch of panics:

thread '<unnamed>' panicked at twgame/src/net_msg.rs:470:39:
not yet implemented

thread '<unnamed>' panicked at twgame/src/net_msg.rs:474:38:
not yet implemented

thread '<unnamed>' panicked at validator/src/lib.rs:858:21:
assertion failed: team.load

1

@Robyt3 what is on your UI todo list? inspire me please

I'm currently refactoring the scoreboard

Robyt3

I'm currently refactoring the scoreboard

do you make a awesome fng scoreboard?

16:19

i'll soon have to implement the settings ui for tees i wonder how i want to design it. UI is always so much tryhard lol

For team based modes I took some ideas from 0.7 (colored titles, score in the middle) (edited)

Robyt3

For team based modes I took some ideas from 0.7 (colored titles, score in the middle) (edited)

ah epyc

Eventually scoreboard height could also be adjusted so it only takes up as much space as necessary

16:21

0.7 for reference

Robyt3

Eventually scoreboard height could also be adjusted so it only takes up as much space as necessary

mh, i decided against that for my fng scoreboard, it looks so weird

Robyt3

For team based modes I took some ideas from 0.7 (colored titles, score in the middle) (edited)

16:25

dunno if u ever seen it

16:26

16:27

height basically always stays the same

Robyt3

0.7 for reference

what is the (1) behind the Red team... number of ctf flag captures?

16:28

is that a built-in feature with network support?

16:28

i need that too then

It's the amount of players, XD

Replying to @Jupstar ✪ what is the (1) behind the Red team... number of ctf flag captures?

wtf ok

16:35

why is it in the title lmao

16:35

doesnt sound like a information that has to be highlighted xD

Jupstar ✪

Click to see attachment 🖼️

I think it's would be better if hook and hammer icon on title bar being in middle or after team's name

Peakies

I think it's would be better if hook and hammer icon on title bar being in middle or after team's name

agreed, i might have had it that way, but looked weird too i think. I also don't like the pattern of the title

but I like it

16:47

pattern and corner of info row and column list row gave it a bit of 3D style don't know got my point or not

16:47

but generally noice ❤️

Jupstar ✪

agreed, i might have had it that way, but looked weird too i think. I also don't like the pattern of the title

wanna add scroll, page navigation support for 400+ tee right ?

furo321

[ddnet/ddnet] Pull request opened: #8215 Fix /timecp not working with MySQL

It would previously fail with Unknown column 'hasCP' in 'where clause' on MySQL/MariaDB, but worked fine with SQLite. The query has now been updated to work with both. Closes #8214

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 bo...

Peakies

wanna add scroll, page navigation support for 400+ tee right ?

scroll is no go imo xd would suck hard to having to scroll up and down

18:26

also i think u can easily put more stuff into scoreboard by making font smaller

Jupstar ✪

also i think u can easily put more stuff into scoreboard by making font smaller

well my plan for scoreboard is a bit different having scroll and clickable ui for open up player profile to load that player points, badges, activity point and etc ... (activity point is like xp on my own server)

Peakies

well my plan for scoreboard is a bit different having scroll and clickable ui for open up player profile to load that player points, badges, activity point and etc ... (activity point is like xp on my own server)

bit like csgo scoreboard?

18:48

i dunno if it supports scroll too

Jupstar ✪

bit like csgo scoreboard?

yeah exactly but a bit simpler than csgo scoreboard

i see. mouse support could be cool indeed, if you enable it over a hotkey my fear with scrollbar is simply that it defeats the purpose of just opening it for a short moment. if there is a forced scrollbar you have to use it

well it can have option for old scoreboard ui but plus up and down support key to select player to open it's profile with enter key or something like this

Jupstar ✪

i see. mouse support could be cool indeed, if you enable it over a hotkey my fear with scrollbar is simply that it defeats the purpose of just opening it for a short moment. if there is a forced scrollbar you have to use it

is it possible to open scoreboard by holding down tab button and have this possibility to click or scroll scoreboard item ?

possible, why not

Jupstar ✪

i see. mouse support could be cool indeed, if you enable it over a hotkey my fear with scrollbar is simply that it defeats the purpose of just opening it for a short moment. if there is a forced scrollbar you have to use it

I think the time that player have to spend to find someone name with small size can spend to scroll or by using search box and find that specific player on scoreboard

Peakies

I think the time that player have to spend to find someone name with small size can spend to scroll or by using search box and find that specific player on scoreboard

i dont agree that a scrollbar helps, but yes a search box would help

19:24

but in that case i'd also probably not use the scoreboard

today was the day when chat gpt was finally useful, i had to fill tables in a database with data and chat gpt generated pretty good shit. it's not good at helping with programming but it can generate dummy data

2

MilkeeyCat

today was the day when chat gpt was finally useful, i had to fill tables in a database with data and chat gpt generated pretty good shit. it's not good at helping with programming but it can generate dummy data

it's better than me in pysoron. that's what it is useful for

Jupstar ✪

i dont agree that a scrollbar helps, but yes a search box would help

hmm maybe you right

Jupstar ✪

it's better than me in pysoron. that's what it is useful for

u write pyson?

MilkeeyCat

u write pyson?

no, else i wouldn't need gpt

anybody uses github copilot? 0_o if u do, do you like it? (edited)

MilkeeyCat

anybody uses github copilot? 0_o if u do, do you like it? (edited)

i don't write comments in my code xdd no jk. i used it once, for js it's ok. for other languages only for some basic algorithms

19:28

but been a while since i used it

19:29

do you use it actively?

Jupstar ✪

i don't write comments in my code xdd no jk. i used it once, for js it's ok. for other languages only for some basic algorithms

comments for people who read their code more than once

19:29

it aint me

Jupstar ✪

do you use it actively?

i never used it xd

2

which way is less cursed justatest

MilkeeyCat

which way is less cursed justatest

write proc macro that parses the syntax directly

20:14

without string

Robyt3

[ddnet/ddnet] Pull request opened: #8216 Fix incorrect score kind in demos depending on last server

Clear current server info when playing demos to ensure that the score kind (points, time) is detected correctly. Previously, when playing a demo with points score kind and cl_demo_assume_race 0, the score was not shown as points when the last server that the client connected to had a race gametype. Move the implementation for cl_demo_assume_race together with the rest of the demo server info initialization in the CClient::DemoPlayer_Play function.

Checklist

[X] Tested the ...

MilkeeyCat

which way is less cursed justatest

https://doc.rust-lang.org/std/macro.include_str.html

include_str in std - Rust

Includes a UTF-8 encoded file as a string.

github-merge-queue[bot]

[ddnet:master] 2 new commits

cf65690 Fix /timecp not working with MySQL - furo321 47e4505 Merge pull request #8215 from furo321/fix-timecp-mysql - Robyt3

20:40

Robyt3

[ddnet/ddnet] Pull request closed: #8215 Fix /timecp not working with MySQL

20:41

Robyt3

[ddnet/ddnet] Issue closed: #8214 /timecp name broken

dobrykafe

[ddnet/ddnet] Pull request opened: #8217 Remove unused ConMode function

Probably slipped in during rebase of https://github.com/ddnet/ddnet/pull/7521.

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 [A...

MilkeeyCat

which way is less cursed justatest

mh can't you save this in a file of it's own and include it in here somehow?

Robyt3

[ddnet/ddnet] Pull request closed: #8217 Remove unused ConMode function

21:25

github-merge-queue[bot]

[ddnet:master] 2 new commits

3b7019d remove unused ConMode function - dobrykafe 8a378da Merge pull request #8217 from dobrykafe/pr-remove-conmode - Robyt3

dobrykafe

[ddnet/ddnet] Pull request opened: #8218 Fix alignment of labels in friends list

I'm not sure if this was intended but no variation makes it easier to read. Before: !before After: !after

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 (esp...