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 2021-03-17 00:00:00Z and 2021-03-18 00:00:00Z

@Learath2 new f-client has timeout code which also gets sent, can you look into making timeout protection work for 0.7?

@deen i also had issues on amd igpu, dunno is it related. but it's most likely driver issue bcs my gpu and xserver crashes, or just freeze

11:12

is ddnet client has jetpack prediction as well?

nori

@deen i also had issues on amd igpu, dunno is it related. but it's most likely driver issue bcs my gpu and xserver crashes, or just freeze

had = it works now?

hard to reproduce, it's random

11:19

usually i can play

11:20

sometime happens and i get dozens of amdgpu error on dmesg, but strangely only experienced it on ddnet

could you post those errors as a file?

https://gitlab.freedesktop.org/mesa/mesa/-/issues/4465

[renoir] system freeze randomly when playing the game (#4465) · Iss...

Brief summary of the problem: System hangs/freeze while playing the game randomly. I only experienced it few times while playing DDRaceNetwork...

i mean i also have an amd gpu, and basically never had a crash, atleast not on the stable driver

11:29

cant be so insanly hard to reproduce xd

i play everyday ddnet and it happened like 2 or 3 times in 3 weeks :D

i often switch between mesa 20 and 21 and both run completly stable for me. indeed the only crash i had recently was outside of ddnet(probs not even opengl)

11:32

but when i first got my gpu the driver were more unstable too, but that was also the time amd went opensource

i had some freezes, but only the game

13:12

and after some seconds it came back

@Deleted User oh, ddnet used to evaluate opengl-on-vulkan: https://www.phoronix.com/scan.php?page=article&item=zink-january-2021&num=2

Is that like a translation layer to run ogl on devices that dont have support for it but do have support for vulkan?

13:25

Sounds like an extremely small subset of devices :P

fokkonaut

@Learath2 new f-client has timeout code which also gets sent, can you look into making timeout protection work for 0.7?

Is it broken? I thought I made this even work across versions

didnt test it, but there was a chat message

13:28

indicating that 0.7 client wont be able to load

deen

@Deleted User oh, ddnet used to evaluate opengl-on-vulkan: https://www.phoronix.com/scan.php?page=article&item=zink-january-2021&num=2

xD he should have used the GPU benchmark tho

13:36

but really cool

Learath2

Is that like a translation layer to run ogl on devices that dont have support for it but do have support for vulkan?

i think its more to replace opengl drivers, because 20% perf drop can be compansated by new processor generations

13:37

just like DXVK etc.

13:37

d9vk even outperforms the native radeon driver

fokkonaut

indicating that 0.7 client wont be able to load

Ooh, I remember some issue with the stupid shuffling around if connect and disconnect messages I didnt bother finishing

13:38

Yeah it needs a massive revamp :/

so it works in theory?

The 0.7 client cant change its clientid

13:38

It only works if you modify your client to let you accept changes to client id

13:39

We need to do it the other way around

13:39

Instead of changing the client id we need to "reparent" the tee

well, on my server i use idmaps due to the 128p support anyways

13:44

thats how i got it working there

Uh, I will take a look, if it doesnt look to hairy I'll try to get it in but honestly I've been in a shitty mood this entire month so it's kinda uncertain I can get to it

13:49

If you want it certainly done you should PR it yourself :P

i dont need it in very fast

13:51

xd

13:51

im kinda unmotivated the last 1-2 months too

[ddnet/ddnet] Issue opened: #3714 make it possible to copy text from the console

Should we release another DDNet version? I don't remember if we had any big changes

15:59

https://github.com/ddnet/ddnet/compare/15.3.2...master

DDraceNetwork, a cooperative racing mod of Teeworlds - ddnet/ddnet

i dont know anything noticable either

Mh, just a minor bump if we do a release

Can release another version with accounts

1

[ddnet/ddnet] Pull request opened: #3715 Add dummy_reset and remove +resetdummy

It's works now for both on console/cfg and as bind. Also tried to fix annoying dummy hammer issue when reset dummy

Checklist

[X ] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ X] 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 affect existing maps
...

https://github.com/ddnet/ddnet/pull/3654 Any idea why the style check/clang might be failing me? I've run the clang format locally and it doesn't change anything. I've overwrote the ddracechat file with the master version but no changes were detected

Added swap functionality to server client: by houseme-kyle · Pull R...

Added a new chat command /swap based on the discussion in the associated issue: #1103 Swaps the character's position with another tee using the existing Save/Load functionality. Restriction...

@Zodiac probably you have a different clang-format version locally, it's quite annoying. Just ignore it I guess, I can fix it later

1

@Zodiac you need clang-format 10, I think

I've got clang 11.0 it seems. Will look to downgrade for the next PR.

Do you use ArchLinux?

19:18

I just installed clang-format-static-bin from AUR and then used /opt/clang-format-static/clang-format-10

Going to do another check-in now. Downgrading the version.

I pushed already

19:19

it should be fine

19:19

I hope it was okay that I pushed into your repo

Yeah ofc - I see there is another file that is the issue. I am making it part of my build process to prevent this again (edited)

19:20

I'm using Windows and Visual Studio which likely part of the reason

Don't put effort into getting clang-format right, we need to find a better solution that uses a fixed version. For example we could just do the clang-format as part of the merge into master

Too late

Figured it out. Should be smooth sailing from here

Exported 62 message(s)