Teeworlds - bridge

Teeworlds

IRC / bridge

One-way IRC channel bridge. If you want to be able to send messages to IRC, contact @Dune or @heinrich5991. https://www.teeworlds.com/?page=docs&wiki=rules/irc_rules

Between 2019-10-09 00:00:00Z and 2019-10-10 00:00:00Z

Dune: 'ello dünê

hello däëy

Hello guys. I have a little problem. I cannot compile the newest (released) source of Teeworlds 0.7.3.1. My compiler gives me following error:

13:01

[109/127] [4] c src/engine/external/wavpack/wputils.c src/engine/client/backend_sdl.cpp:18:21: error: function 'void glTexImage3D(GLenum, GLint, GLint, GLsizei, GLsizei, GLsizei, GLint, GLenum, GLenum, const GLvoid*)' definition is marked dllimport GLAPI void APIENTRY glTexImage3D(GLenum target, GLint level, GLint internalFormat, GLsizei width, GLsizei heigh

13:01

t, GLsizei depth, GLint border, GLenum format, GLenum type, const GLvoid *pixels)

Hey Jay

13:27

That is odd

13:27

What is your system?

13:27

What version of SDL do you have o n it?

win 10, newest update

13:32

SDL 2 (i guess)

13:32

64 bit

13:34

I use the official sources and compile with bam5. No changes at all

Did you install yourself or did bam do that for you?

13:35

SDL2 I mean

bam downloaded it

13:35

(it said that)

It says that when it attempts to iirc

13:35

Can you run bam config

13:35

And see what it says on the sdl line

AUTO sdl using supplied win32 libraries

Hmm hold on

Jay: can you check other/sdl/? Is there something there?

ye

13:40

bunch of *.txt files, and 2 directories (include and windows)

13:40

and 1 sdl.lua

Hmmm

but thos files get created after launching bam5 client once

13:41

before that they are not in there

13:42

(relaunching bam5 client does not change the error)

13:43

i do can use my VisualStudio 2017 project generated by CMake but its full of potential errors ( because every single code base is splitted in different projects ....)

13:44

so this is not really comfortable to code with :/

yes that is normal, it is the files that are downloaded by bam from https://github.com/teeworlds/teeworlds-libs

teeworlds/teeworlds-libs

dependencies used by teeworlds. Contribute to teeworlds/teeworlds-libs development by creating an account on GitHub.

13:46

you mean compiling with VS17 works?

ye this works

13:47

but VS17 still outputs 271 error and this count increases with more files opened

13:47

its annoying :C

yeah I bet

13:47

I'm scratching my head

hmm. I do can compile the game when i remove that GLAPI from GLAPI void APIENTRY glTexImage3D(...)

13:49

but every graphic is actually an error graphic then

you mean by editing backend_sdl.cpp:18?

yes

can you check https://github.com/teeworlds/teeworlds-libs/blob/master/sdl/include/SDL_version.h ?

teeworlds/teeworlds-libs

dependencies used by teeworlds. Contribute to teeworlds/teeworlds-libs development by creating an account on GitHub.

13:51

should be in other/sdl/include

13:51

especially line 62, #define SDL_PATCHLEVEL 8

its 8 on my side

13:53

this is the full output of the bam5 error

13:53

https://pastebin.com/tzwrQbrx

[1/2] [1] c++ src/engine/client/backend_sdl.cpp src/engine/client...

13:53

dunno if it helps a bit more

so you have the right SDL2 version. looks like a bug somehow, but I'm not knowledgeable enough

13:55

you might want to open an issue there: https://github.com/teeworlds/teeworlds/issues

teeworlds/teeworlds

A retro multiplayer shooter. Contribute to teeworlds/teeworlds development by creating an account on GitHub.

13:55

it is strange that it is bam specific though, and not a problem with cmake

i got that problem somewhere too in 0.6.4. DDNet never had this issue at all

13:57

iirc

really? never heard of this issue

yea a programme named Henningstone helped me out pretty fast

13:58

but i lost contact to him and seems like he is not developing anymore

13:58

programmer*

huh, he is usually in this channel, seems not anymore

isn't it related to Assa message ?

what message?

bridge | [teeworlds] <Assa> Teeworlds uses legacy OpenGL

14:05

bridge | [teeworlds] <Assa> I will do a pull request, because a cmake flag is missing

14:05

bridge | [teeworlds] <Assa> you can't compile with a newer opengl

huh, downside of the double build system, requires double the fixes :/

14:08

indeed https://www.khronos.org/opengl/wiki/Common_Mistakes#Legacy%20Generation

Common Mistakes

14:08

glTexImage2D is deprecated

uhm? my error is happening with glTexImage3D

14:11

not 2D

ph

What do you guys think about not counting a wrong rcon password twice? Imagine some has a rcon_auth bind and unluckily presses it multiple times he gets a ban. But if we ignore duplicated wrong passwords that could be saver

my bad Jay

14:13

@ChillerDragon: rcon auth was used as a way to crash-flood servers in the past iirc, sounds like you might open an attack vector?

yes for sure sounds like an attack vector

14:18

but maybe it can be implemented save idk just an idea probably not worth the effort

These flood detection bans are so annoying. some servers ban you for a minute just for trying to connect twice in quick succession~

yes, that is the default

14:44

https://github.com/teeworlds/teeworlds/issues/1631

Make the "stressing network" automatic ban less trigger-happy · I...

I experience getting banned regularly when restarting the client for a setting to take effect. Would it be possible to make this condition more gentle?

the thing is you can prime it by hitting enter just two times

14:50

the next connection attempt will then end you

imo the server can stay as trigger happy but the vanilla client should simply sanitize user input and even if the user spamms enter/esc or what ever limiting the connection attempts

problematic is the "joining a full server" scenario.

14:58

i usually deploy the "keep enter pressed" ddos approach :P

xD

hello world

@Dune i got it to work now

16:25

I just deleted the GLAPI from GLAPI void APIENTRY glTexImage3D and renamed the datasrc folder to data now its working

for the cmake deprication: just set the CXXFLAG -DOpenGL_GL_PREFERENCE=LEGACY

Glad to hear! Not sure why you would rename the datasrc folder though, it is copied to data as part of the build process

16:54

@Assa if this cxx flag is added to cmake it should be added to bam too

in my buildfiles i just added it to the environment, I may need some help to modifiy the cmake/bam files

17:00

settings.cc.flags_cxx:Add("-DOpenGL_GL_PREFERECE=LEGACY")

17:00

for bam, but for cmake no idea

17:03

cmake_policy(SET CMP0072 OLD)

17:08

@Dune this is a cmake flag, you don't need to set it in bam - bam has no problem when old and new OpenGL stuff is installed

@Assa yet #2263 is a bam issue

@Dune no Data folder is outputted after the build process

17:52

only the exe and an objs folder

cmake ?

i am talking about another issue

bam

try : bam conf=release content

18:03

Assa: isn't it weird that the issue occurs with cmake but not with bam ?

18:04

we might need some guru :3

guide to break cmake:

18:06

$ sudo apt-get install libglu1-mesa-dev freeglut3-dev mesa-common-dev

18:06

done

seeing what happened to Jay, I'd wager it occurs to bam too...

well rand answer helped me

Jay, no data folder in build/[arch/arch] ?

18:33

even though build completes?

it putted the dll's and data in the release folder

18:33

well in the debug folder there is no data and dll's at all (after a successful build)

that's strange, it should be

just the exe itself and an folder called "objs"

unless you compile for server or client only

i do bam5 client

18:34

is that wrong?

ah, so you compile for client only, maybe that doesn't do the contents

18:34

lemme check

i used to build teeworlds like bam5 client_debug and it worked back in the 0.6.4 days

it doesn't

bam5 content should

I never specify targets so I don't know :|

18:36

great I broke my build then :D

bam conf=debug <target>

how do I keep breaking my builds

18:40

windows is a mystery

windows doesn't even have a command line tool to zip stuff

18:41

i used the "jar" command from java to do it

can't link to SDL2 anymore :|

18:41

and the CI stuff uses cmake only

did you do my "guide to break cmake"?

no, I just tried building teeworlds after two weeks of not doing so

18:42

(on windows)

18:42

seems my build configs naturally deteriorate over time

maybe the new windows update broke it?

I don't have internet, so it can't be an update

the only thing harder to compile is mac

ah, bam auto setup back to mingw, explains

should mingw be added to the CI, too?

Put away sharp objects in the room if you attempt that

puts away huge knife

Add bam first, there were problems that went undetected for a long time on windows

can teeworlds register for github actions so we can get rid of appveyor, travis, and circleci? I promise I will setup a workflow for each OS you want

See https://github.com/teeworlds/teeworlds/pull/2194

automatic latest builds/trunk builds for testing new features by A...

Added deployment to appveyor, circleci and travicci to enable automatic trunk builds. os: win64 win32 unix osx (mac) It adds 4 releases (for each os), which are automatically updated and overwrit...

why do you sent my my own PR? That has nothing to do with what i said

Hmm, doesn't it make the PR obsolete then?

yes ... and no

18:56

it outdates it: Github actions is (as the name suggests) A github CI/CD service and provides a way easier deployment

18:56

I tested it in my catch64 repo

it's great :0

it really is

but it doesn't fix my windows build :'(

and suddenly, crash

what crashed, rand ?

a zcatch server

19:43

guess

19:45

other guess, leave&kill

which one?

ah, same as the issue

20:46

jsxsl you pulled the latest master recently, didn't you?

I guessed wrong apparently, server still only

21:22

online

21:22

but, it did hang at some point

21:23

and disappear from server list

merged master

Exported 168 message(s)