DDraceNetwork - developer

ddnet-maps

[ddnet-maps:master] 1 new commit

b29a9a6 M Yoko 2.1 - ddnet-maps

can u pls make an account system

07:04

i dont like that people can just steal my name and finish maps for me that i wanted to finish myself

07:04

its so dumb

07:05

i really like ddnet and all

07:05

but man

07:05

make a god damn account system xd

-1

@Iza not as easy as you might think

uh oh

got discussed many times already

time for debate #53

https://github.com/ddnet/ddnet/pull/53

Editor: Scale quad point selection indicator with zoomlevel by cina...

1

make a ddnet login system

07:07

everyone has their own userid

07:07

solved

nice

07:07

sounds easy, why dont u go make a pr

@Iza what if I just register my acc before Aoe does, then steal his name?

07:08

as an example

no

07:08

because i will do it before u

fokkonaut

@Iza what if I just register my acc before Aoe does, then steal his name?

ip, mac adress

can change

and people can proof themselves

07:09

i dont think ur mac adress changes weekly

so many players, we cant check it for everyone

its just stupid that people can finish maps for me

07:10

idk what u people think about it

I think nobody wants to finish maps for you, you should have just chosen a more uncommon name

07:11

DDNet warns you about that in the beginning

i could choose the most uncommon name in the world, but that doesnt change the fact that people still can pretend to be me xd

07:12

i didnt saw that warning tbh

finishing for others is good in 90% of cases, faking isnt tho

07:12

i think accounts is good but its not so simple to implement, it would need a lot of planning. also this discussion happens like every week here

well these discussions dont happen without reason

It doesnt help if we discuss the same things withe different people every time, because people dont understand that its a big act

#43411

07:14

#3411 *

https://github.com/ddnet/ddnet/issues/3411

Accounts · Issue #3411 · ddnet/ddnet

If anyone wants to summarize the discussion, feel free to. I just wanted to create the issue now.

If there is no clear goal, no one will code this. If we don't decide on the scope before someone codes it it will stay as a pr being discussed forever

you gotta plan something

07:15

ik its harder than said

07:16

but as you can see alot of people bring up discussions about it weekly

I can't plan anything myself. This is a huge feature and we don't get to change much about it after we roll it out

It has to be perfect on the first try

So people need to contribute to the discussion on github, where we intend to come up with a plan

At least the core

At this point I mostly need devs talking but even that is hard to get

:D

07:17

Learath are you here for a second?

i'd suggest to take inspiration from other game devs

07:18

how they solved problems like these

@Iza thats not too easy, because we still have the problem with old players and a lot of ranks

07:18

We cant just randomly give the old points to anybody who claims to be that guys

07:19

There are players with the same names, who just both play and get points

its just a random idea and idk how well it works out, but how about taking ip & mac adress "captures" from players like daily or weekly and then with that data finding out which ip adress or mac adress uses the account most frequently

@Iza feel free to join the discussion on github and present us your plan, maybe you can bring it into a new direction :)

im not a github person xd

time to become one

07:48

its fun easy to use place

07:48

its run by evilcorp tho

ChillerDragon: you are really cool you know that?

i feel offended

ChillerDragon: troll

ChillerDragon: cuz you chill

09:11

i'm sorry

https://github.com/ecriminal/Exploit-Discord-Cache-System-PoC did you see this yet?

ecriminal/Exploit-Discord-Cache-System-PoC

Exploit Discord's cache system to remote upload payloads on Discord users machines - ecriminal/Exploit-Discord-Cache-System-PoC

10:49

Not very interesting, abuses the fact that discord doesn't sanitize pngs

ehm what?

10:50

is that just saying that discord doesn't re-encode PNGs?

10:50

doesn't seem like a vulnerability to me

10:51

I guess it would be defense in depth to re-encode them

injects a payload to the end of a png, which you add to a server as emoji, then you send said emoji and the cache downloads it, triggering the AV

ah, it's AV being stupid

10:51

hm

it's not an actual RCE but did make me worried for about a second why there is a threat detected in my discord cache

10:52

I would truncate any PNG I host at IEND tbf

you can surely embed stuff that AV detects in the image part of a PNG

10:53

doesn't sound like a fix to me

10:53

have you tried disabling your AV? it lowers performance and increases your attack surface

10:54

I guess re-encoding is the only proper solution

10:54

I got a fairly trivial payload in mine, dunno which discord server I got it from

what's the 'payload'?

10:54

a zip file?

10:55

with a malicious executable?

http://codepad.org/IWiDTA7i

10:55

VBScript

10:55

It's kinda hopeless though, I don't see how he could ever get someone to run this

10:56

I guess you could send a python file along with it, saying "xDe check out this free discord nitro hack abusing the cache <insert payload emoji>"

just execute my payload directly already

and someone who thinks they know a lot about computers but don't coooullld be fooled

10:57

yeah, overall meh attack

the only impact I can see is the AV detecting this and the user getting scared

yeah, I can't see node js or discord being stupid enough to execute their cache

I wonder why web browser caches never trigger such false positives

they probably do something like encrypt it or encode it in a way that AV can't access it?

11:08

I can see the AV then trying to crack the encryption

AV means anti virus?

yes

That stuff windows ppl use

11:09

11:09

Another discussion about accounts monkalaugh

Iza

ip, mac adress

https://wiki.archlinux.org/index.php/MAC_address_spoofing

MAC address spoofing

not really relevant, we could only see the MAC address from the same computer and then also see the real one

fokkonaut

Hey, I have moved my 0.6 whisper part in my mod to PreProcessMsg(), no logic change, just changed the place where the whisper is split into parts and where we get the message out of the /w command. https://github.com/fokkonaut/F-DDrace/commit/cb5622b612da54c43e0ab2ba2a96c67d2e74f133?w=1 somehow, on the way from preprocessmsg to onmessage again, the m_pMessage gets (invalid utf8), and i dont know why

@Learath2 do you have any idea about this one? or @heinrich5991 maybe

debug print it on the way and find out what kind of corruption occurs and where

when returning its all fine, but when using the m_pMessage in the main function it returns invalid utf8

11:14

the rest of the packet works (target, mode..)

11:14

i dont know what i should debug else, otherwise i wouldnt ask

11:14

i checked it right before the return and directly inside of the CL_SAY part of OnMessage

what kind of corruption occurs? where in the string?

it couldnt have been modified

11:14

?

you say it's invalid

11:15

what do you mean by invalid? what kind of corruption?

yeah, i return it as a string and it returns (invalid utf8)

11:15

and when sent to the client it just doesnt display anything

11:15

print*

11:15

not return

11:18

i dont get it, without the change it all worked, also the m_pMessage did not fail

11:18

i really dont get it xD

if you say '(invalid utf8)', then that string needs to come from somewhere

11:18

try grepping the source for it

(invalid utf8) comes from the formatting

11:19

not from the tw source

so from src/base you mean?

i will check it

11:21

https://github.com/fokkonaut/F-DDrace/blob/F-DDrace/src/base/system.c#L129

fokkonaut/F-DDrace

F-DDrace is a server mod for Teeworlds 0.7 developed by fokkonaut. - fokkonaut/F-DDrace

11:22

only appearance

I guess that's where it comes from then

yeah, doesnt help me any further i think, how can it break on the way from returning to the main function

helping you is annoying for me

1

you seem to assume that you know everything already, when the fact that something does not work as you think it should, suggests that you need to rethink an assumption

11:27

debugging means "I don't understand why the code is doing what it does, so I can't be sure about my assumptions"

11:27

"I try to find the point where my understanding is wrong"

You know where the print that causes (invalid utf8) happens, I'd suggest breaking there

11:29

looking at what exactly is broken about the string if anything obvious

11:30

Then if it's not obvious, I'd disable threaded logging and step into the actual logger code where the string occurs, to see why it's thinking that

(invalid utf8) seems to only exist on windows, no threaded logging there

even easier

11:31

@heinrich5991 we should try to reenable aio on windows

yes

11:31

maybe we should add a config option to disable threaded loggin

11:31

g

We don't have one?

look at me suggesting a config option

11:31

no, can be annoying for debugging

ah actually it's not really easy to add a config option for that

11:32

system.c has no idea what a config is

engine can do that on setup

11:32

but I guess we get a bootstrapping problem

I guess we could have the client call sth like dbg_enable_treaded_log

yea

well @heinrich5991 i am sorry but my knowledge is far away from analyzing a character or a string like this to see what breaks

11:38

what i know is that it shouldnt break from just returning, and nothing more, thats why i have no idea about it

So your assumption that you are "just returning" it is wrong

11:41

maybe you are returning a misaligned pointer, maybe you are accidentally returning a pointer to a pointer, maybe the memory you are returning is now out of scope

oh, yeah, that might be it

11:42

thanks

You are doing the very last one...

yes

11:42

the whispermsg

It would be so obvious if you'd just put a breakpoint on where you return from preprocessmsg

why?

because then you could probably observe the string changing

as i said, i put a dbg msg right before the return

11:43

and it was all okay

print aWhisperMsg before the return would show you the correct string

11:44

print aWhisperMsg after would show you garbage

yea, thats what it does; invalid utf8

At that point I would print &aWhisperMsg

11:46

Anyway, you are just being lazy with your debugging imo. You made the correct observation but didn't look into it's consequences

trust me i did everything i can, i just didnt know more

A return indeed shouldn't be changing a string. So what does a return change? The only thing it changes is the stack frame

and i dont even know what that is

Well, now you know \o/

thanks

sourcecode - compiler - objectcode - linker - library - executable programm

12:31

not sure about lib and linker tho

yeah idk what a linker is

is it static or dynamic lib?

12:32

it might also be lib last if its dynamic lib i guess

12:32

the linker links your obj files and libarary together to a executable

12:32

the compiler turns the code into obj files

12:33

is this ur homework or something? Am i supporting you to cheat a exam? please dont get me in jail

Nah some homeschooling

source code - compiler - object code

ye thats what i said heinrich

12:35

also nice ip grabber link @NeXus

then object code + libraries using the linker transforms to executable

12:35

I think

@ChillerDragon what where

yea is it linker or libs first

12:35

12:35

this

that's a discord link

thats what i said

12:35

:D

I think they want to see that the linker combines libraries and object code to an executable

yes

12:36

get creative NeXus

thanks

you need more than - to seperate them :D

the image shows that this is possible ChillerDragon

little bug, you can type x_ skins in here (in game you're just default)

not really a bug

12:47

client says to server i want to be x_spec, but server disallows it

12:48

some servers might allow that

it's the client that forbids that btw

but it probs shows other tees faking x_spec as default

12:57

right?

yes, it's the client that does this replacement

but only if the server tells the client to do that

yes, we have a server flag that disables it

1

2

any chance we can get a command to set view pos in ddnet? To bind camera positions in maps? Might as well ask upstream before i slap it in my fork

That seems a bit problematic, not everyone has the zoom level. What do you want to use it for?

write a bind that teleports me in a specific position in the map

13:52

also as spectator you can quickly toggel between hotspots of a map

13:53

i do not see how it relates to zoom

I assumed you want to change the view position during playing, not spectating

clientside^

For freeview spec sounds fine to me, even though I'm not sure it's that useful

clientside in spec sounds okay for me

He wants it for a custom client xd

i dont want it in my custom client id prefer it in ddnet

13:56

yea i thought about spec

13:56

inagme it sounds like infinite dyn bind

13:57

oka lets see if i can code it xd

13:57

then ill pr against ddnet

a couple of years ago i think ddnet had the feature that when you watch a player in spec you rather follow his client and not only the tee. so when he watches something in spec your camera automaticaly follows his nd if he switches dummy so does your cam... is it just a false memory or why was it removed?

TIL spec char is a skin

Didn't even know this was a possibility

for me it was 10min

14:11

I was quicker googling

14:11

than to wait for it

There is a way to get past it?

I had ssh access

14:12

plus you can get past with a live medium

Funnier thing is I actually logged in, then I tried to sudo and made a typo 3 times

14:12

So now I have a logged in tty that I can't escalate to root

14:12

And another I can't log in on at all

do you know the root pw? su might work

There is no root pw

fair

Atleast iirc, that's how I usually set up my linux systems

14:14

Anyway only need to wait an hour to update my system

66min might be enough to try the live medium

14:14

ah, you have a logged in session and can work as usual?

14:14

then all is fine

wait so if i fail sudo 3 times i get timeout on ddnet?

14:15

scary shit

why do you have access to ddnet servers? ^^

oops

14:15

i mean

14:15

yikes

14:16

no u troll i mean if i am connected with my tw client to a server and then on a terminal type sudo test and fail

14:16

will it close my session and kill my client to server connection

no

is this a bug

14:27

feature

can assets break ninja skins too

the katana skin could break it but i think it only uses the x_ninja skin

aren't katanas behind tees?

yes

weird

idk just make sure ur x_ninja skin is good first but it might be smth else

couldn't get more info out of the guy having the problem tbh

14:31

actually katanas are drawn just fine lol

14:32

and it is kinda weird that the eyes are fine if ninja skin is broken

m true

@deen CHN1 have no vote options

myb perma server?

no, the vote options are empty, perma server have some vote options at least i think. and it wasn't perma i'm pretty sure

ChillerDragon https://github.com/edg-l/shorust

edg-l/shorust

A simple url shortener server made in rust. Contribute to edg-l/shorust development by creating an account on GitHub.

ok the votes appear after manually switching to a map using /map

15:04

i /map'ed every empty chn1 servers, now most of them are kobra now.

15:04

vote works now tho

@Learath2 I even added sha256 support to teehistorian2demo now

15:07

https://github.com/heinrich5991/libtw2/commit/be7ff357b1567f923f9d168c59b94c872106d4d1

cool Ryozuki

15:10

i never really felt the need for a urlshortner tho

15:10

do you have a running instance i can use Ryozuki ?

i can set it up fast

don't use URL shortener please, they contribute to link rot

put one in readme that people can test its nice

15:11

link rot?

i made it mainly to practise

link rot = links 404ing after some time

oh

without a link shortener, the original site can go down

15:11

with a link shortener, the original site or the link shortener can go down

ye

i mean

15:12

ppl use link shortener services to track info

15:12

xd

yes, people do plenty of nasty stuff

but i wanted to make some quick app to learn stuff

15:12

and i indeed learning, the error wrapping stuff

15:12

learned*

sorry, I didn't want to complain about your project

15:13

just about ChillerDragon wanting to use a link shortener

@Deleted User are CUIElements supposed to be self contained?

heinrich hater

@heinrich5991 oh, I forgot to deploy a newer version of teehistorian2demo :/

now the newer version isn't even a downgrade in any respect

(edited)

15:17

Ryozuki: not sure if intended, the db included in the git repo has some link to some youtube video

oh

15:17

it included the db

15:18

didnt meant to lmao

.gitignore!

yea, forgot

15:19

ah it has a link to some 80's japanese music i found cool

15:19

nothing to worry about xd

Ryozuki: i made a url linking to ur url shortener using my url longener https://paste.zillyhuhn.com/zu

15:19

just to trigger heinrichs dead link concern

ChillerDragon i think ur paste can be predicted

15:20

does it follow the abc?

no its random

random with 2 chars?

yes

not very securi iirc

but fast to type

15:21

dont hate on pstd

15:21

its best paste service

(&mut rng)

15:21

.sample_iter(Alphanumeric)

15:21

.take(6)

15:21

.map(char::from)

15:21

.collect()

15:21

rust best

15:21

ez to make a random readable toen

15:21

token

echo $RANDOM

15:21

shell better

Can I pay one of you to port CScrollRegion over to ddnet?

how much

15:22

u can pay in gpus btw

15:22

a 3070 is enough for me

xd

15:23

id take a 3090

thats 2k€ sir

Are you insane? I would pay at most 1/10th of a gpu for a feature already coded

xd

gönn doch

Let's start the bidding at 20 euros

not worth

if you pay in btc

15:23

i take 20

15:24

will be 2k when i finished coding

kappa

15:24

i hear its going down rn

15:24

from 30k to 26k €

15:24

yep

buy the DIP!

I'll pay you the 20e and you can get bitcoin with it?

ChillerDragon buy TRX

Trx scam coin haha

i made a gamble that day u said

15:25

and bought 10€ worth of TRX

15:25

https://coinmarketcap.com/currencies/tron/

TRON price today, TRX marketcap, chart, and info | CoinMarketCap

Get the latest TRON price, TRX market cap, trading pairs, charts and data today from the world’s number one cryptocurrency price-tracking website

15:25

@Ama why is it a scam xd

Its not scam but it has too many coins

15:26

And produces more every month

https://coinmarketcap.com/headlines/news/tether-usdt-routed-tron-ethereum/

More Tether (USDT) is Now Being Routed To Tron than Ethereum | Head...

There is a spike of USDT transactions in Tron, a dPoS network, based on data from Coin Metrics, a blockchain analytics firm, a CoinDesk report on Jan …

So it will never be Worth much i think i cashed out 15k 3 years ago when it went to 2000 satoshi

i wonder what this means

I'll up my offer to 30e, you can get a reaaally nice bottle of gin for that

doesnt it have a cap?

I bought at 300 satoshi

15:27

And sold at 1800

15:28

At 1 point i had 1 btc with an 2k usd investment

15:28

Too bad i didnt hold on to the btc haha

hehe

Yes, we all have cryptoregrets, can we get back to the topic on hand?

15:29

I require CScrollRegion ported

i would consider at 100€, im just in a rly lazy mod for c++ rn

15:30

rust is too great

15:30

ChillerDragon go code it

15:30

u dont contribute since long

15:30

nobo

100 is a little too rich for my blood, two entire bottles of very decent whiskey

Learath2

@Deleted User are CUIElements supposed to be self contained?

dunno what u mean

Do they render themselves?

no, they basically only a state object

So they are more tracking of elements than actual elements?

yeah

What is m_OwnUIElements for and what is m_UIElements for?

15:36

(obv CUI::)

ui elements maintained by CUI class

15:37

like, deleted

ChillerDragon here the url shortener https://noby.world/

Url shortener | Shorust

A url shortener made with rustlang.

15:37

and yes, its a domain i own xd

15:37

https://noby.world/Wvl9MA

edg-l/shorust

A simple url shortener server made in rust. Contribute to edg-l/shorust development by creating an account on GitHub.

What does it mean for an ui element to be maintained by the CUI class?

that it is managed by the cui?

15:39

e.g cui says when its rendered

ah nice noby.world url shortner i love it

and passes events to it

15:39

idk

15:39

xdd

https://noby.world/rAyPg7

chillerbot

chillerbot has 8 repositories available. Follow their code on GitHub.

15:40

love it

nice

maybe i should host a instance

It seems to me if you create a CUIElement it self inserts into m_UIElements, if you get one thru GetNewUIElement it's also inserted into m_OwnUIElements

@Learath2 nothing special, it just means u dont need to deallocate it

and then buy facebook ads to promote it

ChillerDragon see, this one is hardly predictable

15:40

ur 2 letter id is

then sell out the url shortner and redirect everything to some ads and get rich

ChillerDragon shh

15:41

actually smart

the 2 letter is awesome

15:41

i can remember them and use it to share pastes across devices

@Deleted User what are some elements using this new api?

server browser

ChillerDragon

[ddnet/ddnet] Pull request opened: #3531 Add set_view camera position command

Allows to bind spectator camera positions. Can be used to write teleport binds for test server or quickly spectate specific spots of the map.

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 if it works standalone, system.c especially
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that aff...

So this is a class that keeps track of a vector of quads(?) and uses quad containers to make sure they aren't re rendered all the time?

15:49

(I know I should have asked these before merging)

it keeps track of specific attributes the streamed functions use, e.g. color of text, the text itself and checks if any of them changed

15:50

its still streamed, just cached

So it can handle only text?

and ui rects

How does it decide if a ui rect changed?

how can i record all gameplay but only save when i want to

with common attributes color size

Im 'corneum

how can i record all gameplay but only save when i want to

if u mean demo record, there is cl_replays

If you mean actual video your gpu is much better at that then we could ever hope to be, take a look at ReLive for AMD and ShadowPlay for Nvidia

gpus are fast and all but i feel like recording tw demos is less overhead than an actual screen record

16:24

and demos can always be tunred into video aftwards

16:24

but i think classic corneum is here for replay bots and wants to hax some ranks xd

no i just dont want gigabytes of demos piling up

oh nice

16:24

so legit

but still keep cool moments

ye cl_replays is ur friend

cl_replaybot

and turn auto demo record off?

16:25

@noby i actually had conflicts when i tried to merge with upstread because some replay functions already existed

16:25

:justatest:

idk probably auto demo off give it a try or ask the math brain

16:26

or wait how did i use to call him

16:26

forgot his nickname

ChillerDragon: Problem is recording videos or tw demos on ourside has "some" overhead while recording a video for gpus have very negligible overhead as they have hardware for this specifically

did u ever run fraps on windows

16:27

xd

16:27

rip framerate

fraps is not the same as gpu accelerated recording

yea fair i guess

obs can utilize some of the hardware capabilities now, so that's a more fair comparison

yea obs performs quite well

16:28

still feels like more overhead than tw demos

@Learath2 i read always everywhere that cpu encoding/recording has more quality always

i never noticed a difference when recording tw demos

well ofc

how do you think they "accelerate" the process?

16:29

silicon real estate is premium, can't waste too much

https://en.wikipedia.org/wiki/Jiuzhang_(quantum_computer)

Jiuzhang (quantum computer)

Jiuzhang is a 76-qubit photon-based quantum computer prototype developed by the University of Science and Technology of China (USTC). It is named after Jiuzhang suanshu, an ancient Chinese mathematical classic. On 3 December 2020, USTC announced on Science that Jiuzhang has successfully conducted a Gaussian Boson sampling task in 200 seconds, wh...

16:30

i wonder about quantum computers

16:30

china supposedly made this

nvenc is quiiite good though, still not perfect but at that speed it's much better quality than I ever expected

thats 2.5 billion times faster than their own supercomputer (edited)

16:31

The Chinese team claims it to be 10 billion times faster than Google's superconductor-based Sycamore processor, and as such, becomes the second computer to attain quantum supremacy

16:31

they also claim its 10 billions time faster than google quantum computer

16:31

xdd

america should drop a bomb on that thing before they start to decrypt all our communications

i mean

16:32

if quantum computers exist

16:32

they already decrypted all

16:32

i would only assume that murica army is already using them

16:32

there is quantum proof cryptography tho

quantum computers exist, they just weren't strong enough yet

cana quantukm computer run tw with 20x zoomout tho

"post quantum cryptography" is a really interesting field

cant they uise the quantum computer to find a super good covid vaccine thanks bye

16:33

imagine using science for human needs and not info warfare or wars

16:33

we would be in another place rn

16:33

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

Quantum supremacy

In quantum computing, quantum supremacy or quantum advantage is the goal of demonstrating that a programmable quantum device can solve a problem that no classical computer can solve in any feasible amount of time (irrespective of the usefulness of the problem). Conceptually, quantum supremacy involves both the engineering task of building a powe...

16:33

Ryozuki

imagine using science for human needs and not info warfare or wars

thats kumbaya

I think google with their quantum computer already demonstrated solving some insane problems

16:36

I wonder if they have enough qubits to break rsa with any meaningful key size

16:40

Apparently around 6k qubits are needed to break AES-256, not even close

16:41

oh and that's logical qubits not physical ones, so multiply that by 12 or so

ah

16:46

then makes sense

@Learath2 source? I didn't know you could break stuff like AES with quantum computers

https://arxiv.org/pdf/1512.04965v1.pdf I found it in some guys comment in stackexchange but I closed the tab

16:56

I guess it doesn't give significant advantage for AES?

16:56

Oh, it should actually, you can browse the search space more efficiently

https://crypto.stackexchange.com/questions/6712/is-aes-256-a-post-quantum-secure-cipher-or-not

Is AES-256 a post-quantum secure cipher or not?

We know Grover's algorithm speedup brute-force attacks two time faster in block ciphers (e.g brute-forcing 128 bit keys take $2^{64}$ operations not $2^{128}$). That explains why we are using 256 ...

16:58

this stuff says you can "only" reduce the search space by sqrt()

16:58

means that 256bit AES is still secure, 128bit is not

1

17:02

funny comment

famous last words before the ASIO builds a quantum computer the size of jupiter to spy on australian citizens

bors[bot]

[ddnet:master] 2 new commits

23c145d Improve ClShowDistance a bit - def- 77a9f41 Merge #3530 - bors[bot]

17:28

bors[bot]

[ddnet/ddnet] Pull request closed: #3530 Improve ClShowDistance a bit

@Learath2 would you mind taking al ook at https://github.com/ddnet/ddnet/pull/3492 ?

Support F-Client and replace non-extended extension messages by hei...

This continues what #3439 started, and also replaces non-extended messages that are DDRace extensions with extended messages. The DDNet server still sends the old message to old clients and the DDN...

bors[bot]

[ddnet:master] 2 new commits

cd8e67d Handle switch number 0 - def- 93acedf Merge #3523 - bors[bot]

17:41

bors[bot]

[ddnet/ddnet] Pull request closed: #3523 Handle switch number 0

bors[bot]

[ddnet:master] 2 new commits

199da24 Prevent OpenSSL SIGPIPE (fixes #3513) - def- 989d26b Merge #3527 - bors[bot]

17:50

bors[bot]

[ddnet/ddnet] Issue closed: #3513 Client crashes sometimes on player join

17:50

bors[bot]

[ddnet/ddnet] Pull request closed: #3527 Prevent OpenSSL SIGPIPE (fixes #3513)

def-

[ddnet/ddnet] Pull request opened: #3532 Fix crash on namechange (fixes #3444)

Checklist

[x] Tested the change ingame
[x] Provided screenshots if it is a visual change
[x] Tested in combination with possibly related configuration options
[x] Written a unit test if it works standalone, system.c especially
[x] Considered possible null pointers and out of bounds array indexing
[x] Changed no physics that affect existing maps
[x] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...

bors[bot]

[ddnet:master] 4 new commits

9196a9a Fix crash on namechange (fixes #3444) - def- a99f4f1 Add str_utf8_truncate test for size 0 - def- 281492c Fix memory leak in Thread.Lock unit test - def- 7f4ec45 Merge #3532 - bors[bot]

18:59

bors[bot]

[ddnet/ddnet] Pull request closed: #3532 Fix crash on namechange (fixes #3444)

18:59

bors[bot]

[ddnet/ddnet] Issue closed: #3444 Client crashing when changing name

Does the client or server get weird if you run it for a long time without restarting?

shouldn't

19:00

if it does, it's a bug

Well maybe you can recreate it. I used cheat engine to speed hack and after 14 hours some bullets stopped moving and my grenades don't rotate. Even after reload

2

(sounds plausible, these kinds of bugs)

Do the official servers restart when empty?

@Ryozuki investing in warefare is important when space invaders come and attaxx us

19:10

and like zombies

ChillerDragon we can find any signs of life by ourselves, we wont be able to protect if they come anyway

19:10

cant*

@Im 'corneum yes, daily

well only zombies then

Maybe we can try out https://github.com/wolfcw/libfaketime or something similar to see if we can reproduce these physics problems

wolfcw/libfaketime

libfaketime modifies the system time for a single application - wolfcw/libfaketime

What does this do compared to speedhack?

19:21

Also, can u host servers that run on 50% speed so I can play Gores better troll

it's for linux instead of windows

def-

[ddnet/ddnet] Issue opened: #3533 after 14 hours some bullets stopped moving and my grenades don't rotate

As reported by Im 'corneum: Well maybe you can recreate it. I used cheat engine to speed hack and after 14 hours some bullets stopped moving and my grenades don't rotate. Even after reload Maybe we can try out https://github.com/wolfcw/libfaketime or something similar to see if we can reproduce these physics problems

@Im 'corneum hehe I remember running a speedhack on both game and server cause I was too bad to test my parts years ago

Yeah it's a useful tool... Didn't know you found that before me saddo

yeah in like 2016

You could have told me(

is there any trigonometry function using time as input?

19:26

might be the problem for rotation stopping

I think I found it even before I played on ddnet, messed around with it on vanilla

Chairn

is there any trigonometry function using time as input?

You mean the hack itself caused it?

i remember long ago having a bug because trigonometry function didn't work out when input was too big

19:28

that's just an idea because you said you used speedhacks

heinrich5991

[ddnet/ddnet] Issue opened: #3534 Server crashes if there's a MySQL error during initialization

``` [2021-01-23 00:02:50][sql]: MySQL Error: MySQL server has gone away free(): invalid pointer Thread 2 "DDNet-Server" received signal SIGABRT, Aborted. [Switching to Thread 0x7ffff6743640 (LWP 2903048)] 0x00007ffff75e8615 in raise () from /usr/lib/libc.so.6 (gdb) bt #0 0x00007ffff75e8615 in raise () from /usr/lib/libc.so.6 #1 0x00007ffff75d1862 in abort () from /usr/lib/libc.so.6 #2 0x00007ffff762a5e8 in __libc_message () from /usr/lib/libc.so.6 #3 0x00007ffff763227a in mall...

heinrich5991

[ddnet/ddnet] Issue closed: #3534 Server crashes if there's a MySQL error during initialization