Guild icon
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 2022-12-01 00:00:00Z and 2022-12-02 00:00:00Z
Avatar
Oke looks like i successfully woke up for advent of code greenthing
Avatar
Avatar
heinrich5991
the code looks severely broken
yeah it is
04:50
i started rewriting it at all and just pushed half of code that works
04:51
so now its broken
04:51
especially async version
04:51
and also logging has been changed so parser is also broken
Avatar
Is there any way to use 2 differently sized images for one use case? Like, if there are 2 images, but one of them is merely an extension of the other one? If I want to add extra emotes I want there to be backwards compatibility with the old format.
Avatar
Avatar
Voxel
Is there any way to use 2 differently sized images for one use case? Like, if there are 2 images, but one of them is merely an extension of the other one? If I want to add extra emotes I want there to be backwards compatibility with the old format.
I don't think we have support for that yet
Avatar
Best would be to make a new image so the old client just ignores the new one and keeps using the old
Avatar
Avatar
Alexander
Best would be to make a new image so the old client just ignores the new one and keeps using the old
So then would I need to create an entirely new enum for that?
Avatar
Well I hardly know anything about the inner workings of ddnet so I don't know what an enum is
07:18
I'm just saying it's better to create it as a seperate feature
07:18
Or integrate it into the existing
Avatar
I'd integrate it into the existing image
07:23
but I think best would be to ask for feedback from other developers @Voxel
Avatar
Don't integrate stuff into emotes. Just use a new image
08:07
Plan carefully which emotes and fill all space
Avatar
Avatar
heinrich5991
I'd integrate it into the existing image
I'd seriously consider the amount of work required to do this bloat. It's currently allowed to have different height and width. So we'd need to detect at runtime if there are new emotes. Simply using a new image is much easier and we can abstract our current logic to work the same for both emote files
Avatar
Does anyone here know how to use HttpPostJson() to make an http request? I'm trying to make an in game translator using a free api I found, but am unable to get a response and just get an error
Avatar
What is the error
09:20
Also is the api even using json
Avatar
https://libretranslate.com/ is the api I'm using, and the state of the response is always error
Avatar
No logs in console?
09:32
E.g. why it is error
Avatar
I think i might be formatting my request wrong and its giving an error as a response
09:34
if I try to unpack the values without checking if the state() is HTTP_DONE then I get an error in console
Avatar
I did have the data in my request formatted incorrectly, I figured it out 😄
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:16:21Z
@Patiga any idea why ipython doesnt wanna show the help page :(
10:17
10:18
Avatar
what about help twmap.Map
10:18
i actually never used help xD
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:18:38Z
na none work :(
Avatar
help(twmap.Map)
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:18:59Z
na
10:19
what about python
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:19:11Z
something is cursed w iypthon
Avatar
not ipython
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:19:20Z
python works fine
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:19:27Z
but ipythons autocompletion is cool
Avatar
then ipython suxs
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:20:32Z
got it
10:20
its twmap?
10:20
trol
10:22
ok ipython is trash w twmap :(
10:24
omg breakpoint() also eats the help page watafak
Avatar
Nice
Avatar
Avatar
Skystrife
I did have the data in my request formatted incorrectly, I figured it out 😄
👍
10:25
Interesting project btw.
10:26
Often heard about it. With an open tool we could maybe integrate it optionally in ddnet.. tho reading tos is hard
10:26
Since our player base is relatively huge we might create too much traffic^^
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:26:33Z
yo @Skystrife if you want some free api keys for libretranslate lemme know i run https://trans.zillyhuhn.com/
Avatar
chiller lgbt
Avatar
Ah u can self host?
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:26:57Z
da
Avatar
well its libre
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:27:14Z
used on foxxonauts tw server for chat translate
Avatar
so if its gpl
10:27
u should be able
10:27
agpl
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:27:20Z
but its not very good
10:27
its pretty bad actually xd
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 10:27:29Z
but super funny
Avatar
any erlang experts here?
Avatar
Avatar
Jupstar ✪
I'd seriously consider the amount of work required to do this bloat. It's currently allowed to have different height and width. So we'd need to detect at runtime if there are new emotes. Simply using a new image is much easier and we can abstract our current logic to work the same for both emote files
I'd disallow images that don't have the fixed aspect ratio instead. the problem with using more images is that it becomes harder to share emotes
10:35
but that's basically why I asked for a discussion issue
Avatar
I'd say it's harder if we don't split it. Bcs all current images are basically invalid then
10:36
Or not useful have to be extended
Avatar
all current images can continue to be shared
Avatar
Plus the extra code not nice
Avatar
yes, I see that downside
10:36
there's extra code involved
10:36
that's bad
Avatar
Avatar
heinrich5991
all current images can continue to be shared
Yes but e.g. if u want image for new emotes and like ur current. U have to extend ur current
Avatar
yep
10:38
but I would find it unlikely that this happens, they're usually in the same art style
10:38
all of the emotes
Avatar
But the new are more ddrace related or not?
10:39
Do different art style maybe not bad
Avatar
ChillerDragon: does help(twmap) work?
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 11:08:37Z
no
11:08
only twmap? but its way less feature complete
Avatar
Avatar
[quakenet] ChillerDragon
yo @Skystrife if you want some free api keys for libretranslate lemme know i run https://trans.zillyhuhn.com/
Please change trans to translate, i nearly added it to firewall banmonkalaugh
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 11:40:51Z
hrhrhr
Avatar
Avatar
[quakenet] ChillerDragon
yo @Skystrife if you want some free api keys for libretranslate lemme know i run https://trans.zillyhuhn.com/
that'd be awesome!
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 12:15:02Z
how can i send you stuff without leaking it here? :D
12:15
@Ryozuki u here?
12:19
@Skystrife patiga should have sent you the key. It has 1028 requests per minute :)
Avatar
Avatar
[quakenet] ChillerDragon
@Skystrife patiga should have sent you the key. It has 1028 requests per minute :)
damn nice 😄
Avatar
wowo, its finally time for the bash-madness again .__. #!/bin/bash # Advent of Code 01 while read -r l; do if [ -z "$l" ]; then nm+=( "$c" ) c=0 else c=$(( c + l )) fi done < "$1" nms=$(xargs -n1 <<< "${nm[@]}" | sort -n) x1=$(tail -n1 <<< "${nms[@]}") x2=$(tail -n3 <<< "${nms[@]}" | xargs | tr ' ' '+' | bc) echo "$x1 $x2"
12:36
leaking code
Avatar
Hi @Index , nice bash 😄
❤️ 1
12:52
That looks like something I could have written, haha
12:52
and annoyed the other ddnet admins with
Avatar
My shell solution: paste -s -d+ input|sed -e"s/++/\n/g"|bc|sort -n|tail -n1
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 13:03:46Z
bc is non standard tho is it?
Avatar
bc(1) is compliant with the IEEE Std 1003.1-2017 (“POSIX.1-2017”) specification.
13:09
I made a repo for the answers, not sure yet if I'll participate for more than a few: https://github.com/def-/adventofcode-2022
Avatar
My solution is dirty c++ code with stl and repeating codemonkalaugh
Avatar
more important that it works
Avatar
So true
Avatar
Avatar
deen
I made a repo for the answers, not sure yet if I'll participate for more than a few: https://github.com/def-/adventofcode-2022
tf single line tryhard
Avatar
Last time I participated I made a point of using a different language for each solution: https://github.com/def-/adventofcode-2020 😄
13:13
Was so annoying towards the end when I ran out of good languages
Avatar
im gonna do advent in erlang
13:41
cuz i need to learn it
Avatar
Is there a fastest code wins category?
Avatar
Avatar
Jupstar ✪
Is there a fastest code wins category?
no, its only about who coded it faster on the leaderboards
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 14:19:10Z
meh patiga making numpy fast is hard i wanna iterate over all tiles and do some comparisons and it is sooooooooooooooo slow
14:19
been reading numpy doc and SO for the whole day and aint getting any further
14:29
interesting read
14:29
erlang is about letting things crash xd
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 14:30:55Z
tr '\n' ' ' | sed 's/ /\n/g' | sed 's/ /+/g' | sed 's/+$//g' | awk '{ print "echo $((" $0 "))" }' | bash | sort -n | tail -n 1
14:30
pog
14:31
basically giving aoc RCE on my machine axaxax
Avatar
Avatar
deen
Last time I participated I made a point of using a different language for each solution: https://github.com/def-/adventofcode-2020 😄
You used swipl in both 1st and 19th day, not cool feelsbadman
Avatar
Avatar
Jupstar ✪
Is there a fastest code wins category?
we can do that between ourselves tho
14:42
the best would be to have a machine to run the code which we share
14:42
kek
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 14:45:14Z
what about running it in a pipeline?
14:45
should be benchmarking with same specs for all i hope
Avatar
Mh that's sad. So if u have no time u loose xd
Avatar
Avatar
Ryozuki
the best would be to have a machine to run the code which we share
Yeah and no asm xd
Avatar
if u mean a ci pipeline no
14:51
they have lot of noise
14:51
maybe a RPI with a bare bones linux
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 14:52:45Z
How does one properly benchmark? Inst there a way to limit cpu speed to a fixed rate? Or count cpu cycles for a program?
Avatar
No smt and constant frequency sounds sane enough
14:55
10 runs per program
14:55
Rest is margin of error
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 14:59:10Z
lmao deen urs throws a error xd
14:59
ah hm its the empty line at the end of my input
15:01
shieet deens is way faster than mine xd
Avatar
deen wizard
15:02
i mean urs runs sed 3 times
15:02
vs 1
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:02:32Z
ikr
Avatar
bash for calculation is probably slower than bc
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:02:56Z
i assume its the seds
15:05
ayo at least im faster than @Nana ナナ :p
Avatar
wrong index lol
Avatar
ChillerDragon: what kinds of comparisons? you usually do comparisons with the whole matrix element-wise, or with slices of it
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:07:31Z
i want to get the diff of layers
15:07
so i did a iteration and if statements and its unusably slow
15:07
with numpy magic i couldnt figure it out
Avatar
something like layer1 == layer2
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:08:08Z
no
15:08
i want all the things that are different
Avatar
this should give you a matrix where each element tells you if the tile at this position is different
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:08:34Z
like a list of tiles with their flags and coordinates that are diffenrnt
15:09
wait == doesnt return a bool but a array?
Avatar
iirc, yes
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:09:12Z
wotefek
Avatar
yup, tested it
15:11
with numpy, try to always make operations with the entire matrix or slices of it if possible. if you iterate over a numpy array manually for some matrix-wide operation, you are probably doing it sub-optimal
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:11:45Z
okay yea that array of bools is kinda cool
15:12
how do i get the only the elements that are true or false instead of the true and falses
15:12
yea i understood that much but idk how to operate on the whole matrix
Avatar
that is mostly just internet searches for me ^^
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:13:19Z
didnt work for me :(
Avatar
a, its tiles1[tiles1 == tiles2]
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:14:10Z
tried that
15:14
looks weird
Avatar
it should give you the elements
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:14:41Z
[0 0 0 ... 0 0 0]
Avatar
right, you have a 3-dimensional array
15:15
and want a 2-dimensional array at the end
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:16:13Z
yea 2d would be cool but it needs to have all the info x,y,tile,flags
Avatar
that would be a combination of numpy.indices, numpy.equal, numpy.concatenate and that mask-indexing thing
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:19:23Z
yikes wat
Avatar
at least that would work
15:21
first do the equal things and figure out how to tell numpy to stop at the second axis to get the mask then take the tiles you want and concatenate that matrix with the indices matrix then index into those tiles with the mask
15:22
I think that would work, I don't have the time rn to figure each step out, hopefully the documentation helps you enough now ^^
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 15:22:26Z
oof
15:22
that sounds so complicated
15:25
i think i rq for today but thanks a lot the == thing i would have never figured out
Avatar
ah wait
Avatar
Erlang uses the actor model, and each actor is a separate process in the virtual machine. In a nutshell, if you were an actor in Erlang's world, you would be a lonely person, sitting in a dark room with no window, waiting by your mailbox to get a message. Once you get a message, you react to it in a specific way: you pay the bills when receiving them, you respond to Birthday cards with a "Thank you" letter and you ignore the letters you can't understand.
too real
Avatar
you can do tiles1 == tiles2.nonzero() to get the indices
15:27
then you could manually index into the matrix. if the diff isnt huge, then this should be fine
15:29
eh not sure, try it out I guess, gtg
15:29
and gl :)
15:49
i found some old hardware i had
15:49
the ram is 512MB xD
15:52
back when the ram didnt have fancy gamerz heatsinks
Avatar
poor pins
16:02
its broken
16:03
its stuck to the thermal paste xD
16:03
probs a pentium
16:03
its curious it doesnt have pins in the center
Avatar
no, that's a socket 754 amd cpu
Avatar
Avatar
Ryozuki
i found some old hardware i had
What year is that? I took apart a 2004 dell dimension and it looks similar. Only one stick of 512mb of ram☠
Avatar
I have 128mb and 256mb ram sticks sitting around somewhere kek
Avatar
Avatar
funman
What year is that? I took apart a 2004 dell dimension and it looks similar. Only one stick of 512mb of ram☠
maybe same xd
16:54
it was from a pc tower i had
16:54
well the family had
Avatar
For consistency, since some forks will use it, some won't, leading to inconsistent workflow results

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
  • ...
Avatar
Posted by Jeffrey Vander Stoep For more than a decade, memory safety vulnerabilities have consistently represented more than 65% of vulne...
19:51
This was very interesting to solve with javascript
19:54
19:54
*happy noises*
Avatar
Avatar
Patiga
you can do tiles1 == tiles2.nonzero() to get the indices
parentheses missing
👍 1
Avatar
Avatar
Alexander
*happy noises*
cool 🙂
Avatar
Posted by Jeffrey Vander Stoep For more than a decade, memory safety vulnerabilities have consistently represented more than 65% of vulne...
Avatar
"As Android migrates away from C/C++ to Java/Kotlin/Rust" remove Java & Kotlin and im fine
20:48
bcs thats the biggest trash about android
20:48
also sad that they didnt include java & c in older versions
20:49
bcs the advanced usage of selinux is only android 8 thing iirc
20:50
to me the stats look like the number of vulnerables decreased faster than usage of rust. so there must be smth to it different than languages used, or not?
20:53
anyway i think its a bad blog tbh i prefer clear data
Avatar
Does anyone know how to make an async http request? I'm currently doing it like this, but it freezes the game to make the request which is not ideal
Avatar
this looks like marketing and i'm not really biggest android fan
Avatar
Avatar
Skystrife
Does anyone know how to make an async http request? I'm currently doing it like this, but it freezes the game to make the request which is not ideal
just look for other httppostjson jobs are called
20:54
u are litteraly doing it sync
Avatar
happen to have a blog post on such a topic with clearer data? I'd be interested
Avatar
no, but do u feel like this blog is good?
Avatar
Avatar
Jupstar ✪
u are litteraly doing it sync
I know, I'm just a novice at c++ and I saw it done this way elsewhere in the code but didnt see an async way
Avatar
this does not look like it could only relate to languages used, except they purposly made the usage of rust where the vulnerables are
20:56
^ tho that would be surprising still
Avatar
Avatar
Skystrife
I know, I'm just a novice at c++ and I saw it done this way elsewhere in the code but didnt see an async way
the else where is async
20:57
its not even just async
20:57
it runs in a different thread
20:57
i dunno where u come from
20:58
its not like async in javascript
Avatar
ye ii feel the blog is vague too
Avatar
there is a job manager, basically a thread pool
Avatar
and idc about java either xd
20:58
android shoudl use wasm
20:58
ez sandbox
Avatar
no they should go rust
20:58
but not kotlin
20:58
this trash xd
Avatar
+ older distro have more vulnerabilities cause we had more time to find them
Avatar
but they need a sandbox
Avatar
and i want clear facts
Avatar
wasm is perf
20:59
and rustcompiles to wasm
Avatar
i like stats but only if they well described
Avatar
more time and more advanced tools
Avatar
analytics
20:59
a wasm runtime
20:59
made in rust
Avatar
no hate but asm is still slow
20:59
its basically like java vm
20:59
better stay for browsers with asm
20:59
use native rust
21:00
wasm*^^
Avatar
Avatar
Jupstar ✪
its basically like java vm
ok ofc not that bloated
Avatar
idk if its that slow
Avatar
thats ofc impossible xd
Avatar
Avatar
Ryozuki
idk if its that slow
if it has to run in an interpreter/vm
21:01
it must be slow by design
21:01
with very good jit u might get ok perf after a while, but first runs are usually slow
Avatar
tbh idc about android
Avatar
wasm is still only around js level ofc wasm isnt as mature yet, it will defs beat js totally soon
Avatar
i just want browsers
21:02
to use wasm by default
21:02
and deprecate js
Avatar
yes
21:02
me too
Avatar
im sure the code quantity would decrese by millions
21:02
XD
Avatar
i hope more frameworks shift to wasm compatible languages
21:03
and best ofc in system level languages
21:03
bcs they best
Avatar
Avatar
Chairn
correlation is not causation
I think this is a lazy meme
Avatar
Avatar
Ryozuki
im sure the code quantity would decrese by millions
tru xd
Avatar
tell us what specifically you didn't like about the analysis
21:03
they didn't just post the graphs and say "clearly, this is the cause"
Avatar
See runs at https://github.com/ddnet/ddnet/actions/runs/3596433433/jobs/6057097318 and https://github.com/ddnet/ddnet/actions/runs/3596433429/jobs/6057097200 ``` Run sudo apt-get install libmariadbclient-dev libwebsockets-dev mariadb-server-10.3 -y Reading package lists... Building dependency tree... Reading state information... Package mariadb-server-10.3 is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is on...
Avatar
(also, did you know that causation does not cause statistical correlation? and causation is a weird thing to define, anyway)
Avatar
didn't read the blog, just saw the graph @Jupstar ✪ posted
Avatar
well they basically write the same in the blog
21:05
"While correlation doesn’t necessarily mean causation, it’s interesting to note that the percent of vulnerabilities caused by memory safety issues seems to correlate rather closely with the development language that’s used for new code."
21:06
"We believe Android’s ongoing shift from memory-unsafe to memory-safe languages is a major factor." We believe
21:07
i just wonder, doesnt a big company like google can give better statements than that
21:07
anyway
21:08
i just hope the ndk wont suck anymore in future xDDD
21:08
i hate so much to call java code for everything
21:08
their whole security model is wrapped by java api
21:08
or kotlin
Avatar
as i said, more time for older distro
Avatar
i mean its inherently difficult to provide "truth" claims on this maybe
Avatar
and probably more used than most recent ones
Avatar
Avatar
Chairn
as i said, more time for older distro
do u mean old code is more safe?
21:12
i dont directly udnerstand that
Avatar
Avatar
Jupstar ✪
do u mean old code is more safe?
its not xd
Avatar
no, just more time to investigate old code
21:12
i don't think android 13 is also more used than said android 11
Avatar
ah
21:13
now i get it
Avatar
lots of phone still relying on older release, so they are more interesting to investigate
Avatar
true thats indeed also a problem
21:13
well anyway, if they use safe rust it must be memory safe its just about the amount of rust they use yet
21:14
and yes ur are right too^^
Avatar
we gotta wait a few years to be affirmative on this anyway
Avatar
but then again
Avatar
but i do hope rust will fill its promises 🙂
Avatar
why didnt they include android 6+
21:15
would be much more interesting
Avatar
rust is safer
Avatar
it's google, obviously they are part and judge on their paper
Avatar
but i guess u mean the promise that using c++ ultimately leads to memory unsafety even with experts
21:15
and that rust solves that
21:16
Rust prevents out-of-bounds memory accesses and buffer overruns via runtime bounds checks - what’s the cost of those bounds checks for a real-world, production application?
21:16
interesting btw
Avatar
mh yeah, but here again would be more interesting in a big code base. But i tell ya, even if rust would loose 10% perf they would gain it, if we wouldnt design CPUs with rings xD
21:19
so stop this micro kernel bs and lets go smth new supervised
Avatar
they actually gain perf here
Avatar
alone the static linking
21:20
all the dead bloat code
21:20
gone
Avatar
the bounds check added by rust enable some optimizations in this case
21:20
xD
Avatar
actually, new special loop instruction could be nice as well
Avatar
all drivers in every app xDDDDDD
Avatar
set register to boundary, and check that all access don't overflow at the hardware level
Avatar
we shoudl advance the x86 arch to atleast skylake or smth
Avatar
the world
21:21
well
Avatar
guys, the world is not european or american. lot of people rely on old hardware
Avatar
who cares
21:22
@Jupstar ✪ if we compiled official ddnet binaries with a more modern march would we increase overall perf for players?
21:23
x86-64-v4
Avatar
Avatar
Ryozuki
@Jupstar ✪ if we compiled official ddnet binaries with a more modern march would we increase overall perf for players?
depends
Avatar
lets distribute in x86-64-v3 or x86-64-v4
Avatar
generally probably a bit yes
Avatar
but e.g. c++ standard which we link might already be insanly optimized
21:24
so all calls to those might not really improve
Avatar
Avatar
Ryozuki
Click to see attachment 🖼️
can't you target specific instruction set features like AVX2 only?
Avatar
yeah but most probably if the cpu has avx2 its v3
21:32
this baseline is established with intel and amd
21:32
In 2020, through a collaboration between AMD, Intel, Red Hat, and SUSE, three microarchitecture levels on top of the x86-64 baseline were defined: x86-64-v2, x86-64-v3, and x86-64-v4.[39][40] These levels define specific features that can be targeted by programmers to provide compile-time optimizations. The features exposed by each level are as follows:[41]
Avatar

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...
Avatar
[quakenet] ChillerDragon BOT 2022-12-01 21:45:09Z
cursed but working interactive fifo prompt to administrate servers in 50 lines of bash with 0 planning or using any brain https://zillyhuhn.com/cs/.1669930903.png
21:45
you can't tell me any programming language can compete with that
Avatar
Python
22:14
monkalaugh
Avatar
Game-playing artificial intelligence (AI) systems have advanced to a new frontier. Stratego, the classic board game that’s more complex than chess and Go, and craftier than poker, has now been mastered. Published in Science, we present DeepNash, an AI agent that learned the game from scratch to a human expert level by playing against itself.
22:16
did you see the diplomacy AI the other day?
Avatar
Man AIs are getting so fucin interesting lately
Avatar
man AIs? soon AI AIs?
Avatar
Hopefully
Exported 364 message(s)