DDraceNetwork - developer

DDraceNetwork

Development / developer

Development discussion. Logged to https://ddnet.tw/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 2020-06-09 00:00:00Z and 2020-06-10 00:00:00Z

https://zillyhuhn.com/cs/.1591698785.png why do always the things i need disappear from my history o.O ?

Embedded image

In my case they disappear also from the future

["2020-06-04","�| øø�~D�","status"s"]"[020-06-04","�| øø�~D�","dump_antibot"] this is one interesting entry in teehistorian

is this straight from the file or with some decoding applied?

I'm checking whether the original file has this or it's my script messing up

10:57

hm somehow the users name is not known at this point

rcon commands can be entered before the client enters the game

but not before an auth, can it?

no, not before auth

well actually it could happen before login

well, they can be sent ofc, but should not end up in teehistorian

not sure if th would log yeah

because it only logs actually executed commands

can you take a look at 0357754a-e7b4-46d0-a042-ce496e896709.teehistorian?

11:02

I keep an array strings for each client, on auth init or auth login I set it to t.identity

11:02

https://paste.pr0.tips/7D

11:03

I don't see how a client possibly wouldn't have a name by then, but I guess I should be initializing that vec

11:04

Actually String::new() should take care of it, I thought that'd get me an empty string

String::new gives you an empty string indeed

Uhm I don't get what happened here actually

11:06

this tool does decode this properly

11:07

Šøø™ just breaks this

oh, my fake json isn't jsony enough?

I don't get it, outputting to the terminal it looks fine

11:09

outputting to a file it's broken when I open in vim

11:09

if I reload it utf8 vim says there is a conversion error

11:10

Is that string not utf8?

string is always utf-8

If it was not utf8 how would teeworlds even handle it?

String

11:10

ddnet rejects non-utf8 strings

11:10

not sure what it does about cut-off utf8 though

11:10

perhaps there's a gap

well vim atleast recognizes it's 3 characters when I open with utf-8

11:12

I thought maybe it's my terminal not having the characters but it does output fine if I cat it to the terminal

11:13

what an odd issue, I guess it won't cause issues in the final parser as it'll be js

The String type, which is provided by Rust’s standard library rather than coded into the core language, is a growable, mutable, owned, UTF-8 encoded string type. When Rustaceans refer to “strings” in Rust, they usually mean the String and the string slice &str types, not just one of those types. Although this section is largely about String, both types are used heavily in Rust’s standard library, and both String and string slices are UTF-8 encoded.

11:27

poggers

it's weird that when a mod enforces a vote it seems the commands are executed with their client id but elevated privileges

this is one interesting entry in teehistorian

could it be related to my bind that does "rcon_auth ....; rcon dump_antibot; rcon logout" all in one

....

feelsamazingman

kek

1

thats my real pw btw

nah it was Šøø™s name breaking my vim

ooh lol

now that I think about it we really should have went for bcrypt for our rcon auth system

12:05

I mean it's still not trivial to bruteforce salted md5 but with gpus you can probably generate rather large rainbow tables

yes. there was a problem with using an actual password hashing function though, potential DoS by trying to log in

12:09

also, the checking would have to happen in a background thread

we could have rate limited the logins

12:11

Yeah it'd probably be expensive enough not to do in the main thread though

12:11

but luckily the login wouldn't be too hard to move to another thread

@Learath2 instead of bcrypt, argon2 is better

12:14

https://en.wikipedia.org/wiki/Argon2

Argon2 is a key derivation function that was selected as the winner of the Password Hashing Competition in July 2015. It was designed by Alex Biryukov, Daniel Dinu, and Dmitry Khovratovich from the University of Luxembourg. The reference implementation of Argon2 is released un...

still probably too slow to be ran on the main thread

12:18

(the point of a password hashing function is to be slow)

so put it in a thread

12:19

rate limiting seems good

I guess we can wait until it breaks

todays question, why are C++ compilers and linkers so bad at devirtualisation?

13:36

runtime polymorphism is so expensive without devirtualisation

Image attachment

1

@Learath2 because you didn't mark your methods as final and you can't know if another shared lib will overwrite. i guess

or you use pointers and anyone could redirect them to an object of another subtype

hm, I guess I need to help the optimiser reason about my pointers, it seems it only catches the most trivial cases. After a dynamic_cast it should be safe to assume the type

does using namespace override things in the global scope?

@heinrich5991 Q: How would teehistorian handle 0.7 messages on a 0.6 server?

16:21

The complete data doesn't seem to be enough anymore because Sixup is stateful

record the version on client join, I'd say

@Learath2 ddnet use's LUA?

Nope

18:06

ATH does

Hey can someone help me setup a ddnet server?

sure

double click DDNet-Server.exe

2

DDNet is coded in lenguaje c & c++ right?

c++

& c

lenguaje

If u know c u can use it in c++ code

Exported 79 message(s)