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 2020-01-05 00:00:00Z and 2020-01-06 00:00:00Z

@heinrich5991 outputting DOT seems to be the easiest way

[ddnet/ddnet] Pull request opened: #2028 Simplify anti stuck code more

@Learath2 example pls

12:57

input and output

@heinrich5991 https://gist.github.com/Learath2/120269494b3bd42c786fe3c55687503a this is the code

13:20

Depends on the renderer ofc

$ ./a.out 
fish: './a.out' terminated by signal SIGSEGV (Address boundary error)

huh, compiles and runs for me

(gdb) bt
#0  btree<int, std::allocator<int> >::Insert (this=0x7fffffffe390, Data=@0x7fffffffe388: 383) at a.cpp:59
#1  0x000055555555537a in main (argc=1, argv=0x7fffffffe488) at a.cpp:121

13:23

(gdb) print (*pRoot)->m_pData
$3 = (int *) 0x1

Oh cmon, I was sure I got allocators figured out

13:24

does pRoot point to valid memory?

yes

pRoot I mean

13:24

*pRoot...

yes

13:25

otherwise the expression wouldn't have worked

13:26

it's the first insertion that fails

oh, I think that new is wrong

smells like uninitialized variables(?)

13:26

food

can you try something real quick?

13:28

I updated the gist, I think that should work

Is there a way to automatically join spectators after connecting to server? And I don't mean server or map settings, but something I could add to client settings

@Soreu I don't think so

meh

13:49

Wanted to make my life a lil bit easier on android xd

Should be pretty easy to code though, if you are up for it

I don't understand a single line in TW code

13:53

Wanted to also update DDNet android version entities_clear.png to be able to watch with entities on, but new HT entity made it impossible :F

It can be a little tough to get into the tw codebase

Image attachment

13:54

instead of placing new HT in the empty space next to kill tile, it replaced unhook, and the unhook was moved

13:55

and that single thing makes it impossible to watch with proper entities_clear on android version xd

uh, why?

you mean why it was placed that way, or why do I even bother trying to make it work? xd

I don't see why just replacing entities_clear doesn't help

because either everything is unhook, or new HT

14:00

Image attachment

14:00

Image attachment

doesn't the second one look right?

normally, whenever new entity is created, it was placed in blank place, but new HT moved existing already unhookable

14:00

1st is android with updated entities

14:00

2nd is current DDNet on pc

14:01

to compare how it looks and how it should

I don't get how that ends up wrong. The client doesn't do anything to the indexes the map has

14:03

let me check the code

something was certainly changed after introducing new HT

14:05

Image attachment

I remember there was something weird with the unhook, it got moved yes

14:05

i dont know how and why tho

ofc having android version compatible with SDL2+ would be far better

14:05

but i'm working on what exists

14:06

maybe for compatibility with vanilla?

Android Support is gone anyway now, whole android code got removed

I know

I would love to get android working again

\cries in spectator mode during work** (edited)

xd

I'm just too unmotivated though. Maybe next year

said in january sounds sad xd

14:07

but I understand

What was the setting that toggled entities_clear?

14:08

I just am not doing too well overall, and I doubt it's getting better within this year

cl_overlay_entities 100

14:08

hm

14:09

don't bother. I just understood that even after fixing it, that wouldn't work

14:09

on android

14:09

or, at least, I think...

14:10

does server send info about entities index, or client knows it already?

@Learath2 still crashing

14:12

run your program through valgrind pls

@heinrich5991 could you? valgrind doesn't work on macOS, some update broke my clang so I cant use the ASan either

let's kill IRC

And the only gcc I have access to is 4.9 so no c++17

https://gist.github.com/heinrich5991/e6260630d6e7a59877860b3ae0ef7d84

@heinrich5991 btree() : m_pRoot(nullptr) {}; inside the tree class should fix the use of uninitialized value

I'd suggest node *m_pRoot = nullptr; instead

14:19

works now

I guess my compiler was initializing it to 0x0

nah, I guess your allocator happened to hand out stuff with zeros inside

14:24

@Learath2 how do I render this now?

14:24

spoonfeed pls

https://dreampuf.github.io/GraphvizOnline/

14:25

paste there

nice

14:25

ty

It's nice that dot doesn't care about the order of things so it's relatively easy to output

for posteriority: piping into dot -T png or dot -T svg also works

14:28

I actually needed that at the last ctf

14:28

a good graph renderer

[ddnet:master] 2 new commits

f5b8bbc Simplify anti stuck code more - fokkonaut 77957c0 Merge #2028 - bors[bot]

15:05

[ddnet/ddnet] Pull request closed: #2028 Simplify anti stuck code more

does antiping get deactivated when you use pause? when im freezed and i pause for a short moment and then go back and try to move then i t wobbles again until i can move again

[ddnet/ddnet] Pull request opened: #2029 Properly display hdiutil status in CMake

(cherry picked from commit cc2be456d4d95bb3196e8201a06ce19403dfc3c9)

@Im 'corneum yes

why

No need

i have a kp_2 bind and it doesnt work its just russian text

23:47

bind kp_2 say ....

23:47

kp_1 works

Exported 101 message(s)