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 2023-01-18 00:00:00Z and 2023-01-19 00:00:00Z

Jupstar ✪

@deen is it not in release client?

I'm not sure what change we are talking about

deen

I'm not sure what change we are talking about

The one with unpack delta error

Wasn't that a server fix?

07:50

The server change for that is not in 16.7 (https://github.com/ddnet/ddnet/pull/6214)

Fix undefined behavior when unpacking snapshot deltas by Robyt3 · P...

From #5646. 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 b...

No it's client side

07:51

Or both

07:52

It's ez to reproduce on Linux. I can later test

i hecking love c++

Image attachment

And this isn't even that bad

tbf u can create such code in most languages

yeye

16:01

its the std code

16:01

thats why its alien

16:01

altho std code in rust doesnt look alien

maybe MSVC standard code looks better xD

u dont need to use __ in names

well u dont need to, but someone defined it as good style xD

#[stable(feature = "rust1", since = "1.0.0")]
    #[inline]
    pub fn binary_search_by<'a, F>(&'a self, mut f: F) -> Result<usize, usize>
    where
        F: FnMut(&'a T) -> Ordering,
    {
        // INVARIANTS:
        // - 0 <= left <= left + size = right <= self.len()
        // - f returns Less for everything in self[..left]
        // - f returns Greater for everything in self[right..]
        let mut size = self.len();
        let mut left = 0;
        let mut right = size;
        while left < right {
            let mid = left + size / 2;

            // SAFETY: the while condition means `size` is strictly positive, so
            // `size/2 < size`.  Thus `left + size/2 < left + size`, which
            // coupled with the `left + size <= self.len()` invariant means
            // we have `left + size/2 < self.len()`, and this is in-bounds.
            let cmp = f(unsafe { self.get_unchecked(mid) });

            // The reason why we use if/else control flow rather than match
            // is because match reorders comparison operations, which is perf sensitive.
            // This is x86 asm for u8: https://rust.godbolt.org/z/8Y8Pra.
            if cmp == Less {
                left = mid + 1;
            } else if cmp == Greater {
                right = mid;
            } else {
                // SAFETY: same as the `get_unchecked` above
                unsafe { crate::intrinsics::assume(mid < self.len()) };
                return Ok(mid);
            }

            size = right - left;
        }

        // SAFETY: directly true from the overall invariant.
        // Note that this is `<=`, unlike the assume in the `Ok` path.
        unsafe { crate::intrinsics::assume(left <= self.len()) };
        Err(left)
    }

16:02

this is the rust std code

16:02

beautifully commented too

16:02

#standards

16:03

u can even learn

16:03

TIL about match reordering

__ in name is not a matter of style, it's a requirement for the std library to be compatible with all possible codes

16:13

standard says that name beginning with _ are reserved. It avoids case where macro would replace code in std headers (edited)

Chairn

__ in name is not a matter of style, it's a requirement for the std library to be compatible with all possible codes

It is a bit of a weird solution making all the code ugly tho

yeah its pretty ugly, but yeah for macros it makes sense tho if they would want to they could add some compiler specific mode that disallows macros, or only allows certain ones

16:24

then they could escape all this trouble probably

16:25

except you do it on purpose lol

16:26

the only use case for macros i currently see is string concat and name of a variable i dunno if there are more, but template lib and constexpr/consteval can replace pretty much everything

Ryozuki

#[stable(feature = "rust1", since = "1.0.0")]
    #[inline]
    pub fn binary_search_by<'a, F>(&'a self, mut f: F) -> Result<usize, usize>
    where
        F: FnMut(&'a T) -> Ordering,
    {
        // INVARIANTS:
        // - 0 <= left <= left + size = right <= self.len()
        // - f returns Less for everything in self[..left]
        // - f returns Greater for everything in self[right..]
        let mut size = self.len();
        let mut left = 0;
        let mut right = size;
        while left < right {
            let mid = left + size / 2;

            // SAFETY: the while condition means `size` is strictly positive, so
            // `size/2 < size`.  Thus `left + size/2 < left + size`, which
            // coupled with the `left + size <= self.len()` invariant means
            // we have `left + size/2 < self.len()`, and this is in-bounds.
            let cmp = f(unsafe { self.get_unchecked(mid) });

            // The reason why we use if/else control flow rather than match
            // is because match reorders comparison operations, which is perf sensitive.
            // This is x86 asm for u8: https://rust.godbolt.org/z/8Y8Pra.
            if cmp == Less {
                left = mid + 1;
            } else if cmp == Greater {
                right = mid;
            } else {
                // SAFETY: same as the `get_unchecked` above
                unsafe { crate::intrinsics::assume(mid < self.len()) };
                return Ok(mid);
            }

            size = right - left;
        }

        // SAFETY: directly true from the overall invariant.
        // Note that this is `<=`, unlike the assume in the `Ok` path.
        unsafe { crate::intrinsics::assume(left <= self.len()) };
        Err(left)
    }

y c++ can be ankward

16:27

rust >>

@Ryozuki if !true || true /* test */ || true {} rustfmt doesnt like to format this as soon as there is a multiline comment inside the line, any command to force it?

Hi , how to install conan ? Visual Studio 2022

Jupstar ✪

@Ryozuki if !true || true /* test */ || true {} rustfmt doesnt like to format this as soon as there is a multiline comment inside the line, any command to force it?

hmm idk

16:52

file a bug

https://datasette.io/

Datasette is a tool for exploring and publishing data. It helps people take data of any shape, analyze and explore it, and publish it as an interactive website and accompanying API.

17:42

i need to try this

17:42

with ddnet

deen

Wasn't that a server fix?

almost forgot

18:18

u defs have to release a new client version with that fix

18:18

its in the release and creates the lag mentioned

18:19

if i rename to name with special chars, all ppl that e.g. have linux are affected

18:19

not only me

18:19

so can make others lag with that

18:20

and the lag is > 1 second, so its more heavy than ddos (edited)

18:21

u can reproduce it with linux on any server by renaming from "nae" to "ᴷᵉᵏᶳ"

Rip Linux users. Windows masterrace!

on windows it might get triggered too

what about macos?

but dunno how to do it reproduceable

18:23

macos probs also affected

Somebody know out of the head where the tees get rendered (in the code)?

render.cpp

templeOS?

::RenderTee

thanks

Pro

18:24

Also players.cpp

Jupstar ✪

its in the release and creates the lag mentioned

what change?

dunno the specific commit it was with the delta unpacking

18:33

u hot fixed it but apparently didnt release a client version with it

18:33

https://github.com/ddnet/ddnet/commit/b62e7f1299fc672e210a761caba7780d6ee9d06a

Quick fix for old client on new server · ddnet/ddnet@b62e7f1

Introduced in https://github.com/ddnet/ddnet/pull/6214/ Not a proper solution yet

But the problem was never in 16.7

18:33

The problem was caused by https://github.com/ddnet/ddnet/pull/6214

Fix undefined behavior when unpacking snapshot deltas by Robyt3 · P...

From #5646. 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 b...

18:33

So how can I remove code in 16.7 that was not added there?

i just tested it with steam client

Jupstar ✪

u can reproduce it with linux on any server by renaming from "nae" to "ᴷᵉᵏᶳ"

maybe those things shouldn't be said here till there's a released fix pepeCorn

pepeCorn

(to avoid people doing it)

Image attachment

18:34

its defs inside the client

ReiTW

maybe those things shouldn't be said here till there's a released fix pepeCorn

pepeCorn

(to avoid people doing it)

this specfic way only affects linux

18:35

windows might have other ways tho

ReiTW

maybe those things shouldn't be said here till there's a released fix pepeCorn

pepeCorn

(to avoid people doing it)

pepe_popcorn

i mean where should i say it else xD

here

18:37

who cares

according to snail this channel is useless anyway

xd

18:38

@snail we useless sadSnail

sadSnail

[ddnet:16.7.2] 22 new commits

ed1bf05 Update translations for upcoming 16.7 - def- 1cdbb1b Version 16.7 - def- 09c21d5 For integrated and discrete GPUs always prefer what comes first in the list - Jupeyy dbfaecc Minor brazilian portuguese fix (by Akari) - def- 8cd3335 Update russian.txt to incoming 16.7 update - lolipodass d4afa55 Update french translations - Chairn 99dae9e Update spanish.txt - n0Ketchp 9ac6b39 Update spanish.txt - n0Ketchp 58558c4 Add error handling when enumerating physical devices - Jupeyy 31a69c5 Followup fixes for physical devices - Jupeyy 0aefd10 Add warning for missing integrated CPU driver - Jupeyy 2c2f319 Update Korean translations by CHaBek - cwh7435 47cb543 Add new french translations for 16.7 - Chairn 3cfefd1 Update brazilian_portuguese.txt - rffontenelle 50cefec Add ui_smooth_scroll_time variable to adjust smooth scrolling - Robyt3 5d04254 Update simplified_chinese.txt - Cheeser0613 9a882b4 Update traditional_chinese.txt - Cheeser0613 c5b020d Missing parameter in error list - Jupeyy d00fee3 Show current memory usage in assert too - Jupeyy 750129c Version 16.7.1 - def- 8a75365 Quick fix for old client on new server - def- 8d04415 Version 16.7.2 - def-

@Jupstar ✪ can you test if 16.7.2 tag fixes it?

maybe make a hidden chan for trusted devs

deen

@Jupstar ✪ can you test if 16.7.2 tag fixes it?

sure

epic github bot

deen

@Jupstar ✪ can you test if 16.7.2 tag fixes it?

seems fine

Great, thanks

18:42

Should be released in ~1 hour

btw how u decide when a new major

18:42

we can do like kernel

you mean 16->17?

"when it feels the minor is too big in number"

18:42

ye

When a big feature comes or when we are at 16.9 already

18:43

16.10 would look stupid

xd

18:43

true

18:43

or maybe we could go with the date

18:43

its unfun but rly ez to pinpoint

18:43

23.01

18:43

23.1.0

18:43

*

Some people complain about 16 being a too large number for the development progress we have, but I don't care, they can just imagine a 0. prepended

18:44

Date I don't like. It only works when there is consistent amount of work going in

18:44

What if everyone is busy and we have no new features for 6 months? That should be 1 digit higher, not 6

18:44

when we release again

well why?

anyway, release numbers don't matter much, so not worth putting much thought into it

but ye i was just saying it for fun xd

I think the next release should increase 1 of the numbers by exactly 1 , otherwise feels like there could be versions inbetween that you misseds

true

[ddnet-web:master] 1 new commit

90ebbbd Version 16.7.2 - def-

19:46

[ddnet/ddnet] Pull request opened: #6289 Fix `sv_vanilla_antispoof` for vanilla 0.6.5 clients

This fixes issues that prevented vanilla 0.6.5 clients from joining DDNet servers with sv_vanilla_antispoof 1 (closes #2074). The dummy map was not valid. The size of the tiles and game layers was calculated incorrectly. The last member variable included in CMapItemLayerTilemap version 2 should be m_Data. Previously only the size of the member m_aName was subtracted from the total size, which was resulting in an incorrect item size, as the size of the following 5 members also needs...

[ddnet:master] 4 new commits

64578d4 Improve str_hex documentation - Robyt3 0e59b84 Add str_hex_cstyle to print bytes as a C style array - Robyt3 1a9357b Fix sv_vanilla_antispoof for vanilla 0.6.5 clients - Robyt3 0ba6c1c Merge #6289 - bors[bot]

22:16

[ddnet/ddnet] Issue closed: #2074 Vanilla Client 0.6.5 can't connect to DDNet servers?

22:16

[ddnet/ddnet] Pull request closed: #6289 Fix `sv_vanilla_antispoof` for vanilla 0.6.5 clients

22:20

[ddnet:master] 1 new commit

f3867d6 Version 16.7.2 - def-

[ddnet:master] 2 new commits

65cb2ed Ensure integer wrapping instead of preventing overflow/underflow - Robyt3 4a4d6c5 Merge #6284 - bors[bot]

23:11

[ddnet/ddnet] Pull request closed: #6284 Ensure integer wrapping instead of preventing overflow/underflow

Exported 118 message(s)