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 2022-10-07 00:00:00Z and 2022-10-08 00:00:00Z

[ddnet/ddnet] Issue closed: #5910 Time score field in http register

https://github.com/NixOS/nixpkgs/blob/nixos-22.05/pkgs/games/ddnet/default.nix#L73

06:26

oh ddnet exists

@Ryozuki agree ventoy is awesome

[ddnet-web:master] 1 new commit

6d0edf8 Moker -> Vanishing - def-

[ddnet/ddnet] Issue closed: #5899 Question when searching server lists

@deen saw u forked nixpkgs

13:17

Btw they say we use apache 2.0

13:17

I think its wrong xD

I see mention of ddnet.tw, I make PR

Good thing about nix is its not bound to nixos, you can use it on any distro

@deen do you know if ddnet logs are always valid utf8?

[ddnet/ddnet] Pull request opened: #5913 Only use pthread_attr_set_qos_class_np if it's available

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...

Jumping through the code it could print invalid utf8, we don't validate the output

14:16

some places have validation, but probably not all

14:17

You could add something like str_utf8_check in log_log_impl

14:18

But I think it's more important to get any output at all than to only output utf8, or to skip non-utf8 chars

[ddnet:master] 2 new commits

c4463a3 Only show ingame hint in Game tab - def- 96a3e55 Merge #5912 - bors[bot]

14:40

[ddnet/ddnet] Pull request closed: #5912 Only show ingame hint in Game tab

I see

14:42

Just wanted to know to make a pretty parser

If nothing went wrong we probably don’t output invalid utf8, but there is nothing that guarantees it, so if you are doing sth in rust, you should properly check it

14:52

It happened before, e.g. names with utf8 in them getting truncated wrong

14:53

Teeworlds not being properly designed considering utf8 from the start is a shame, bolting it on always has bugs

Yeah im doing it in rust thats why i asked

15:01

To know wether to make the parser using [u8] or str

15:02

I can always convert it to str using lossy

15:02

It doesnt even allocate iirc

u8 and lossy yeah

I wanna install gentoo and go back to contributing more to ddnet

15:05

Tomorrow im getting my super fast usb

15:06

400mb/s

Putting gentoo on a usb?

Im gonna make a multiboot usb

15:06

So i dont have to bother again

Ryozuki

ill buy a 64gb usb3.2 and install this https://github.com/ventoy/Ventoy

This

15:07

Ill install gentoo on my system ofc

You might want a ramdisk or ssd partition to put your portage tmp in. Compiling everything from scratch implies a lot of IO

Yeah

15:07

Nixos had a option to copy all to ram before install

Oh btw, gentoo has no installer still to this day iirc

Yeah no probs

15:08

Like arch

15:08

Well arch now has smth i think

15:08

Optionally

The handbook is excellent, though not as good as the arch wiki, still very complete

Ill embrace rust fully on my new system

15:09

Ill use leftwm probs

15:09

A rust wm

15:09

Alacritty for term

If you use genkernel to configure the kernel for you instead of configuring it yourself you can probably get it all set up within an hour or two

And rust binutils if possible

15:10

@Learath2 when i tried to install gentoo on a vm i think i used genkernel

Ryozuki

And rust binutils if possible

You mean coreutils?

Yeah

Uutils does work on gentoo last I checked

https://github.com/uutils/coreutils

GitHub - uutils/coreutils: Cross-platform Rust rewrite of the GNU c...

Cross-platform Rust rewrite of the GNU coreutils. Contribute to uutils/coreutils development by creating an account on GitHub.

15:11

nice

15:11

but appart from that i alias ls to exa

15:11

and i use fd instead of find

Exa?

exa is a better ls

15:11

way better

15:11

https://edgarluque.com/blog/modernize-your-tools/

Modernize your linux workflow with Rust

Various tools and software that will modernize your workflow

15:11

kek

Ooh, i thought you have an alias for ls called exa

nono xd

15:12

sd for sed too

15:12

and ripgrep for grep

Is there a syslog written in rust? Traditional system logging and rust would have polar opposite fanclubs I imagine

15:13

Maybe some cutting edge binary format, elastic search on redis syslogger

xd

when wine on rust?

i see there is lot of competition vs redist

15:14

for a redist with less mem usage but as fast as it

15:14

redis*

15:14

https://rocksdb.org/

RocksDB | A persistent key-value store

RocksDB is an embeddable persistent key-value store for fast storage.

I should find some use for a key value store so I have an excuse to mess around with redis

15:15

I just love traditional rdbms I never get to use web scale things like redis and nosql mongodb

web scale kek

15:16

u dont need mongodb ever

15:16

if u need dynamic-ish data but u probs will have some sort of structured data

15:16

jut use postgresql with a json column

But is postgres web scale?

@Learath2 i think its cuz premature-scale is not worth it, and you/we never built anything that needs to scal to the point of needing a shared cache using redis

@timakro Can you support https://trashmap.ddnet.org and redirect https://trashmap.ddnet.tw/ to it?

Learath2

But is postgres web scale?

Did you just make up a new slogan for the company I work?

15:18

at*

because i doubt redis is faster than a hashmap using a fast hash embedded into the app if you dont need to share the store

deen

Did you just make up a new slogan for the company I work?

do you make web scale postgres?

postgresql + rocksdb = yugabytedb

15:18

yes

deen

postgresql + rocksdb = yugabytedb

oh i seen yugabyte mentioned on a hn comment today

15:18

thats why i said the rocksdb thing

15:19

https://news.ycombinator.com/item?id=33117650

indogooner

Apache Pegasus - A horizontally scalable, high-performance key-value store

15:19

here

15:19

this new shiny thing came out

15:19

it uses rocksdb

Ryozuki

because i doubt redis is faster than a hashmap using a fast hash embedded into the app if you dont need to share the store

Well ofc. But if you didn’t need to share it, why would you be looking at an external key value store anyway?

tmikaeld 8 hours ago | prev | next [–]

>

Will be interesting to see the benchmarks!

>

There's a lot of KV engines that uses RocksDB now, like CockroachDB (Forked into PebbleDB though), YugabyteDB and TiDB.

>

Those are all many times slower than Redis though, so having a middle-ground aimed to be similar to Redis, that doesn't eat all RAM, is very exciting!

>

reply

Learath2

Well ofc. But if you didn’t need to share it, why would you be looking at an external key value store anyway?

cuz you can say you use redis?

15:19

hip hip

15:20

BASED

Also, just how bad is postgres/mariadb are at being a KV store that people created specific engines for these uses? Indexes improve lookups significantly

i think its too overkill for key val

That's not the main reason for these KV stores. The main reason for is because you can distribute them easier

yeah

Sure but say rest of your backend uses postgres, instead of hosting one more piece of software with it’s maintenance overhead, why do companies prefer a dedicated KV store to just another table?

You are making up more arguments for YugabyteDB, it's a single DBMS with Postgres, Cassandra and Redis interface

deen

That's not the main reason for these KV stores. The main reason for is because you can distribute them easier

Hm, idk enough about how the internals of a rdbms compare to a kv store to know for sure, but wouldn’t replication be solved in a similar manner?

whats Cassandra

15:23

so many software

15:23

oh its apache take on mongodb

if you only need one writer and have the replica as reader, it can work. But if you want lower latency for a DB distributed around the world, it's a bit more complicated

deen

You are making up more arguments for YugabyteDB, it's a single DBMS with Postgres, Cassandra and Redis interface

That sounds like a very useful piece of software actually. Solves all your data storage problems, can you refer between the 3 distinct(?) types of data?

Nope, unfortunately not (yet?)

Good thing about nix is its not bound to nixos, you can use it on any distro

15:25

@Ryozuki that sounds nice

15:26

i ended up writing my own package manager for debian to get updated software and was close to hacking something together that can install aurs on debian :D

yuga is free right?

It’d actually be super cool if a column in a postgres table could refer to a nosql document or a kv store key. You could have best of all worlds

The DBMS itself is free and open source, if you want a fancy web interface to orchestrate it, that costs

i see

but we have open source users who are happy to write a few scripts instead

is it worth for a small or medium scale app

deen

The DBMS itself is free and open source, if you want a fancy web interface to orchestrate it, that costs

Oh, btw is it just 3 different pieces of software under the hood or do you guys implement each part?

what does cloud native mean?

Ryozuki

what does cloud native mean?

Web scale

YugabyteDB uniquely combines enterprise-grade relational database capabilities with the horizontal scalability and resilience of cloud native architectures. YugabyteDB is like PostgreSQL reimagined for a cloud native world.

I actually don’t know these new terms

me neither xd

Learath2

It’d actually be super cool if a column in a postgres table could refer to a nosql document or a kv store key. You could have best of all worlds

Didn't want to say anything before, but found an open issue for, so it's planned: https://github.com/yugabyte/yugabyte-db/issues/830

[YSQL] Access YCQL tables via the YSQL API · Issue #830 · yugabyte/...

Jira Link: DB-699 Is it possible to access a cassandra table (created using cqlsh) via the postgres api?

15:29

via https://docs.yugabyte.com/preview/faq/compatibility/#can-i-insert-data-using-ycql-but-read-using-ysql-or-vice-versa

FAQs about YugabyteDB API compatibility

Answers to common questions about YugabyteDB API compatibility.

is there a rust lib for yugabyte

15:29

doesnt look like

Sure: https://docs.yugabyte.com/preview/develop/build-apps/rust/cloud-ysql-rust/

Build a Rust application that uses YSQL

Build a small Rust application using the Rust-Postgres driver and using the YSQL API to connect to and interact with a YugabyteDB Managed cluster.

oh

You can just use rust-postgres, we are wire-compatible

Learath2

Oh, btw is it just 3 different pieces of software under the hood or do you guys implement each part?

I don't know about the redis and cassandra interfaces, the postgres one is just Postgres source code adapted to our rocksdb layer

15:31

which makes sense because Postgres is huge and we want 100% compatibility

I like your companies project and how it’s open source

yeah looks like a fun project to be in

15:33

@Learath2 lets make the ddnet foundation and fund ourselves BASED

BASED

so we can be in fun projects too

Well we are in the most fun project ever, ddnet itself

it is fun to code in ddnet yeah but imagine working full time on it

I think I’d be okay with that for an average wage

funding just 1 dev would cost a lot anyway

15:36

there is truly lot of free will work here

15:37

and most ppl dont even notice xd

15:37

maybe we could apply to some open source projects grants

15:37

and do a per issue funding

Yeah, no way this project could sustain even one full time dev

15:38

Not with this business model. We need to start selling nfts and hats and skins

and not a js dev, a c++ dev

15:38

the avg wage is way higher

Let’s port ddnet to electron and js, then we can hire a js dev instead

based

15:40

https://opencollective.com/

Raise and spend money with full transparency.

Open Collective is a legal and financial toolbox for groups. It’s a fundraising + legal status + money management platform for your community. What do you want to do?

[ddnet/ddnet] Issue opened: #5914 Menu open message overlaps with debug mode message

[ddnet/ddnet] Pull request opened: #5915 Don't overlap menu + debug messages (fixes #5914)

!Screenshot 2022-10-07 at 19 03 03

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 base/) or added coverage to integration test
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physic...

heinrich5991

fokkonaut has tested it now, it seems to work

@heinrich5991 I can agree, it works fine. Thanks for your attitude!

[ddnet:master] 2 new commits

317b2df Support multiple values to search/exclude (fixes #4119) - def- c5401b8 Merge #5879 - bors[bot]

17:16

[ddnet/ddnet] Issue closed: #4119 exclude multiple words

17:16

[ddnet/ddnet] Pull request closed: #5879 Support multiple values to search/exclude (fixes #4119)

@Ryozuki just build trade stonks bot that prints moni to fund ddnet devs

[ddnet-web:master] 1 new commit

afb7249 Update index.php - snailx3

deen

@timakro Can you support https://trashmap.ddnet.org and redirect https://trashmap.ddnet.tw/ to it?

done

@heinrich5991 is there a reason you do python style imports on rust?

18:56

when most ppl group by crate

18:56

btw format can do it for you

18:57

Image attachment

18:57

so much lines for nothing

it plays nicer with diffing tools

idk about that

it's not really python-style though, is it? ^^

it is

18:58

cuz in python they recommend one import per line

from decimal import Decimal

18:58

no, only one module per line, I think

well yeah not exactly equivalent if you wanna be pedantic

Ryozuki

idk about that

I thought this was a pretty objective advantage

i dont think so

in what way does it not play nicer with diffing tools?

i prefer smaller size of imports

18:59

as in code wise

18:59

it looks rly ugly imho

we seem to talk past each other already

19:00

I'm saying it plays nicer with diffing tools, and I thought you disagreed with that

19:00

but apparently you only disagree with the worthwhileness of that, which I didn't even intend to argue about

diffing tools c an perfectly show the diff

19:00

idk

heinrich5991

I'm saying it plays nicer with diffing tools, and I thought you disagreed with that

Well naive diffing tools yes, but honestly most modern diff tools will mark just the changed part

i like t his more

Image attachment

19:01

my whole screen is imports xd

Image attachment

19:01

also why are we bound to rust 1.48 or smth

Learath2

Well naive diffing tools yes, but honestly most modern diff tools will mark just the changed part

I have yet to see a tool that supports the list style. note that it's being line-wrapped so imports moving to a different line will screw up word-based diffs

Ryozuki

my whole screen is imports xd

my solution to that is not looking at the top of the file

I also prefer the braces, but honestly I'm just a weirdo in that aspect, I have a thing for shorter source

heinrich5991

I have yet to see a tool that supports the list style. note that it's being line-wrapped so imports moving to a different line will screw up word-based diffs

Oh, I usually append them to the end, but I guess people also sort them, that might indeed make it more annoying

if you do happen to have such a diff tool, I'd be happy, because I sometimes have to look at these kinds of diffs

in all rust code i seen

19:03

you are the weirdest somehow heinrich

19:03

xd

19:03

not in a bad way, but you are the most picky

19:03

https://github.com/tokio-rs/axum/blob/main/axum/src/json.rs

19:03

everyone uses this style

19:03

also since rust automatically detects if you are using a import

19:04

diffing such a thing is not that hard requirement

Ryozuki

also why are we bound to rust 1.48 or smth

because I like being able to compile it on the debians we use for hosting the game servers, without managing another tool like rustup

since cargo c will tell you

19:04

unlike python where unused imports dont make warnings

19:04

diffing imports is not a thing u rly need (edited)

translates to "I don't see why you need to diff imports"

yeah i dont

19:05

just cargo c

19:05

u should also do cargo clippy

19:05

while at it

I did cargo clippy on the masterservers because you asked for it

19:05

and implemented all non-nonsensical suggestions

you didnt fix the issues

heinrich5991

I have yet to see a tool that supports the list style. note that it's being line-wrapped so imports moving to a different line will screw up word-based diffs

This is what I meant, it can handle the case of adding things to the end and this pretty decently

Image attachment

they arent nonsensical

nonsensical to me. what are ones that you'd like to see fixed for reasons other than "clippy says so"?

the pages on clippy explain the reasoning

19:06

well the ones u were missing was doing stuff like Ok(someresult.map_err(blah)?) when someresult.map_err(blah) does t he trick

19:07

Image attachment

I don't see the readability improvement there

i see it

19:07

urs is less readable

19:07

actually

19:07

if you know rust u know map_err returns a result already

Hm, I like clippy's suggestion aswell, though I don't find either more readable. I just have a thing for less characters in source code

its nonsensical to wrap a result in a result again

[ddnet:master] 2 new commits

774f7a6 Don't overlap menu + debug messages (fixes #5914) - def- 665f829 Merge #5915 - bors[bot]

19:08

[ddnet/ddnet] Issue closed: #5914 Menu open message overlaps with debug mode message

19:08

[ddnet/ddnet] Pull request closed: #5915 Don't overlap menu + debug messages (fixes #5914)

unless you need to rely on the into() from ?, which isnt the case here

anyway, despite different opinions, I think we focus on the wrong thing

19:09

what's much more important is that there's not enough documentation on how the protocol works, e.g.

im just sad we gotta take ur weird takes on rust

19:09

you go against all the norms

19:09

xd

19:10

why cant we be normal

Ryozuki

you go against all the norms

So? Normality =/> Good

a code that looks like most other rust codebases attracts more rust devs

It's normal to beat women in saudi arabia

i look at our rust code and find it ugly

Learath2

It's normal to beat women in saudi arabia

ok

I rustfmt-ed the source code even though it made stuff look worse IMO, and implemented clippy suggestions that sounded reasonable to me (for reasons other than "clippy says so")

19:11

I even reported bugs to clippy where it suggested worse things

well i just want the imports grouped

19:12

they look better and diff tools can handle it

they handle it worse than line-based imports

fuckyousnail

heinrich5991

I rustfmt-ed the source code even though it made stuff look worse IMO, and implemented clippy suggestions that sounded reasonable to me (for reasons other than "clippy says so")

It seems the argument here is that your opinion doesn't matter, it's classic appeal to authority, clippy and rustfmt know better than you, they are sanctioned by the language

rustfmt doesn't even have an opinion on grouped vs non-grouped imports

Learath2

It seems the argument here is that your opinion doesn't matter, it's classic appeal to authority, clippy and rustfmt know better than you, they are sanctioned by the language

yes, cuz this way everyone can read the code like they are used to

if you have a tool that can actually show me the semantic differences between two imports, tell me

and clippy is reasonable most of the time, maybe there are bugs yeah but most of the suggestions i found made sense

19:14

the rly opinionated ones are disabled by default

I mean it is right in that you can also write the code another way

heinrich5991

if you have a tool that can actually show me the semantic differences between two imports, tell me

Your desire for nice diffs doesn't matter either, he said as much, "diffing such a thing is not that hard requirement"

it's kinda like the clang-tidy checks that made the code look non-uniform

heinrich5991

rustfmt doesn't even have an opinion on grouped vs non-grouped imports

yeah sadly it doesnt, which is why we got a discussion here

19:15

a formatter shouldnt have options and enforce a consistent style

19:15

so we dont lose time

Learath2

Your desire for nice diffs doesn't matter either, he said as much, "diffing such a thing is not that hard requirement"

it is not, when u have a tool to check for unused imports

Ryozuki

a formatter shouldnt have options and enforce a consistent style

I'll go hard disagree on that chief, but you do you

i think im being reasonable there

I guess most people do all the time

19:16

that's like not even a data point

xd

Ryozuki

i think im being reasonable there

Your opinion is reasonable, your assertion that it's objectively true is what's not reasonable

e.g. they also work better with merge conflicts (or even create fewer)

anyway yeah, im rly opinionated in that i want to use the consistent most used style across rust codebases

Learath2

Your opinion is reasonable, your assertion that it's objectively true is what's not reasonable

ok yeah thats true

19:17

assume when i say "should" as "in my opinion it should"

I'm not sure that your preferred import style is the most used one btw

19:17

i.e. the grouping of imports from different modules

rand, syn, libc use it

19:18

the top 3 ones

19:18

im sure if i look at the others too

I just wing it. If it looks good to my eye and doesn't waste too many characters I'll ship it

19:19

I think my behaviour would piss off @Ryozuki the most

No rules

yeah we are oppsite

19:19

i like consistency

I think @heinrich5991 might be with you on that one

so far rustfmt is only bad for big arrays formatted in a way like matrices, there its fine to ignore the line

can you link a file with that import style?

in my opinion

a random one from the rand crate didn't have it: https://github.com/rust-random/rand/blob/766c7eccd73a1f2768f7ce2a4469005a65f7f9a2/src/prelude.rs

https://github.com/rust-random/rand/blob/master/src/rng.rs#L12

heinrich5991

a random one from the rand crate didn't have it: https://github.com/rust-random/rand/blob/766c7eccd73a1f2768f7ce2a4469005a65f7f9a2/src/prelude.rs

this one is literally using it?

this does not group imports from different paths together

look at the {}

like you suggest

19:20

it only groups imports from the same path together

i guess

heinrich5991

i.e. the grouping of imports from different modules

like I said 3 minutes ago

yours doesnt even do that

Image attachment

Ryozuki

so far rustfmt is only bad for big arrays formatted in a way like matrices, there its fine to ignore the line

Here is one change rustfmt does that I don't like

Image attachment

@Learath2 remove the {}

19:21

i think it works

There are others too, but I just remembered this one

yes, I'm just saying I'm not sure you follow "the most popular import style" either

hmm maybe

19:22

thats why i wish it had a default grouping

19:22

instead of doing nothing

19:22

rustfmt

19:22

i think they discussed it some time ago

19:22

@heinrich5991 how did u check the msrv

19:23

i did cargo update, which updates the lockfile

19:23

it should keep the msrv i think but idk

it doesn't

meh

19:23

there is some urge in me about disliking outdated deps (edited)

19:23

xd

19:24

meh i wont even pr

19:24

this isnt worth

delete the lockfile if you want everything as up-to-date as possible

19:24

the lockfile helps you with a reproducible build

cargo update does that

19:24

it updates the lockfile based on the cargo.toml

19:24

im not talking about the custom cargo upgrade

what's the custom cargo upgrade?

19:24

ah

19:25

it updates cargo.toml?

cargo upgrade does ye

19:25

cargo update updates the lockfile based on semver from cargo toml

19:25

or smth

heinrich5991

what's the custom cargo upgrade?

https://crates.io/crates/cargo-edit

19:28

https://github.com/ddnet/ddnet/issues/4415

Report player skin data to HTTP master · Issue #4415 · ddnet/ddnet

This would allow to show the player skins in the ddnet website /any custom website that uses the http master to get information. I remember discussing that it shouldn't be a problem to add ...

19:28

wanted to do this and the other

19:28

shouldnt be hard

19:28

which is why i looked at mastersrv and noticed that

19:28

and lost my time xd

seems like the best improvement would be stopping to lose your time over that

19:29

I had a rust crate that used (in my opinion) horrendous tabs whereas I thought rust had standardized on spaces

19:29

but I still managed to contribute to it even though I considered pinging the author about it (which I didn't, in the end)

19:29

btw, the master server isn't involved in your server info

19:30

you just give it free-form data

yeah i know

19:30

xd

free-form json data

@heinrich5991 well

19:31

atleast fix the test on mastersrv

19:31

its missing a import

19:32

RegisterAddr is not imported

I mean I'm probably gonna fix it

19:32

but your attitude that I must fix it is wrong

ok sry

19:32

ill fix it

ah thanks. I'd have also fixed it, whatever you like

[ddnet/ddnet] Pull request opened: #5916 fix import on mastersrv

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...

ty

does the r come before p

19:34

its probs not ordered

19:34

oh it is

Ryozuki

atleast fix the test on mastersrv

I just saw a bad pattern here, but it was probably not meant as such from you, as I haven't seen that from you so far, sorry for being rude in response

19:34

your alphabet is correct

19:34

A < P < R

nah it was my bad

19:35

i was rude xd

heinrich5991

free-form json data

by free-form you mean that for example, on the players array, i can add a new field to each player?

yes

19:41

well, I don't mean that by free-form

19:42

but yes, you can

19:42

the masterserver doesn't care

but it still reports it right

19:42

to the clients

yes

ok

but it'll also report {} to the clients

19:42

if you send that

19:42

it's the clients that do the filtering

19:42

but they don't care about extra fields they don't know

19:43

I could probably look at the rust PR again

i didnt make a pr

19:44

but if you want

I'm talking about the "add rust to ddnet PR"

19:48

did you want to make additions there?

ah maybe i said smth before

19:48

but i should recheck

19:49

hmm since UpdateRegisterServerInfo is on the server engine and not gameserver

19:49

i wonder how should i get the player info xd

19:50

guess i can move it

you could theoretically ask for extra data from the game server

true

19:51

RegisterExtraPlayerInfo

19:51

but gotta take care to format the json correctly

[ddnet:master] 2 new commits

5cfec69 fix import on mastersrv - edg-l 31b0c0a Merge #5916 - bors[bot]

19:52

[ddnet/ddnet] Pull request closed: #5916 fix import on mastersrv

[ddnet/ddnet] Issue reopened: #5899 Question when searching server lists

@heinrich5991 i guess lan server list uses the old model right

yes

20:16

haven't thought about what to do there yet

20:17

it would be nice if we could get the json server info there, too

yeah

if you have a design, tell me

20:17

or if you want to discuss one

20:17

(even before having one)

maybe we can discuss it tomorrow

20:18

i wanna see if i can finish the skin thing

20:18

@Learath2 do u know how to know if a teeinfo skin is 0.6 or 0.7

20:18

should i bother to tell both formats or just convert to 0.6

both

20:19

in the future, we might want to support 0.7 skins in the client

20:19

e.g. https://github.com/ddnet/ddnet/pull/5660

[DRAFT] Start working on 0.7 skins menu by ChillerDragon · Pull Req...

I started working on this a while ago and it is still far from being finished. Just wanted to share that I am working on this and get some feedback if it even has a chance to be merged when finishe...

1

i see

idk why http://ddnet.org/ doesnt load on my browser

20:55

maybe my internet is acting up

doesnt load for me either

[ddnet/ddnet] Pull request opened: #5917 report extra player info to master http

couldn't test yet this doesnt show extra info on the client server list yet, just makes the server report it

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 exist...

this is so odfd

20:58

https://ddnet.org/downloads/ loads

20:58

but https://ddnet.org/

20:58

doesnt

20:58

@deen something is up with ddnet.org

What do you mean it doesnt load

20:59

It loads for me

both work for me

20:59

can you check devtools?

Maybe CTRL + F5 could help

it works now

21:02

but it didnt before

21:02

and i did ctrl f5 obviously

1

1

1

Ryozuki

@deen something is up with ddnet.org

Strange, nothing suspicious in logs or graphs

i guess it was my net

[ddnet/ddnet-discordbot] Pull request opened: #34 Fix twmap-edit binary call

Missed it when I updated the other two calls some time ago

Jupstar ✪

@Patiga where is the tool?

https://gitlab.com/Patiga/twmap-py/-/blob/6d556b62a7763a13c9a0befdebc5421bf85a7ef3/scripts/menu_background.py here it is :)

I have released twmap-py 0.4.0 now. The biggest change, besides the license change to AGPL, is that 1 = 1 tile/100% everywhere now, so the fixed point numbers can be used intuitively now (quad position x = 1.5 is now 1.5 tiles right of the map origin, before you'd need to do int(1.5 * 2**15)) I also pushed two fun scripts I created :) client-uptime.py creates the client-uptime map that displays how long the client in the style of a digital display menu-background.py converts an image/gif/video into a menu background map

Hey guys I just saw that when you click the button "Join Tutorial Server" on the welcome page it tries to join an ipv6 address

Exported 432 message(s)