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 2017-09-25 00:00:00Z and 2017-09-26 00:00:00Z

<heinrich5991> Ryozuki: on undefined behavior: https://kristerw.blogspot.de/2017/09/why-undefined-behavior-may-call-never.html

Why undefined behavior may call a never-called function

My twitter feed has recently been filled with discussions about the following program #include typedef int (*Function)(); s...

[ddnet-maps:master] 1 new commit

b69310c A gendum - def-

[ddnet:master] 1 new commit

5b05e55 Fix grammar (thanks noby) - def-

[ddnet-web:master] 1 new commit

6ac4145 Revert "fix css-sorter.css" - def-

<Ryozuki> i see

https://www.youtube.com/watch?v=2dTpLEUnLmk

Eric Idle - "Always look on the bright side of life"

09:09

listen while coding hard stuff

Whats the reason for tuning values being 100 times bigger than i would expect them to be (100* the default values, and the values listed in the documentation) https://imgur.com/nkfU2jj

Embedded image

10:35

also why are tunings updated in spec and pause

<deen> floats being stored as an int?

10:54

<deen> with a few digits

i would have thought, but every time they are used they are added as they are

10:56

Embedded image

it has an automatic conversion from CTuningVariable to float

11:07

look for operator float() in the CTuneVariable class

K got it

11:09

what about spec&pause

what about spec and pause?

what about spec and pause? What are you talking about

tunings are sent when you are in spec/pause and you mouse over a new tune zone

11:14

why

where?

nMessage in gameclient.cpp receieves a NETMSGTYPE_SV_TUNEPARAMS

11:15

On

where is it sent?

uhhhhhhhhhhhh

11:19

CGameContext::SendTuningParams

11:21

you want stack trace?

/*
    Function: str_comp
        Compares to strings case sensitive.

12:20

shouldnt it be 'two' ?

yes

u can do a quick edit? line 922 in system.h

12:21

`````

12:22

/*
    Function: str_comp_num
        Compares up to num characters of two strings case sensitive.

12:22

maybe who wrote this was missing the w key?

12:22

ill make a commit fixing this.. (edited)

this one is correct, no?

oh

12:23

yeah i read it wrong

[ddnet/ddnet] Pull request opened: #878 Fix little typo

^

[ddnet/ddnet] Pull request closed: #878 Fix little typo

12:25

[ddnet:master] 1 new commit

f472b02 fix little typo - Ryozuki

thanks for the easy PR ❤

check https://github.com/ddnet/ddnet/pull/874 also

Add instructions to run tests in README by Ryozuki · Pull Request #874 · ddnet/ddnet

(title)

generally you want to capitalize the first letter of the commit message btw

oh :o

but nvm for those

[ddnet/ddnet] Pull request closed: #874 Add instructions to run tests in README

12:31

[ddnet:master] 3 new commits

5e815cf add instructions to run tests - Ryozuki f4515ac tests instructions are only for debian/ubuntu - Ryozuki f2d9c40 Merge pull request #874 from Ryozuki/patch-2 - def-

@Dr. Jekyll they are also send in spec/pause as otherwise prediction breaks

what you think about server side localisation? specially meant for server messages like "finished race", "you now have endless hook.." with a client option to select the server messages language (edited)

no thanks

13:01

learn english

xD

(same for client imo)

ok nvm then

@HMH prediction of what, if you are in spec/pause you cant do anything

well, my opinion is not so important

imo it is

teeworlds looks like game for childrens

@Dr. Jekyll projectiles for example

so imo, it could be good

13:04

x

13:04

xd

13:04

a lot young people

and teeworlds helps them learn english also

13:04

xD

like every game

13:04

13:04

and yes! i am the example

13:05

i couldn't say any word 3 years ago

13:05

:<

13:05

but then i joined #ddnet, AND FUCK YEa. i started to type

13:05

and type, and type

13:05

now i am a bit better, but still

why is where im pointing my spec camera going to affect the behaviour of proectiles?

prediction of what you see

13:06

it really makes things smooth

13:07

e.g a projectile is easy to predict

Good morning (sorry for my bad english!)

but the tuning the client receives is for the quare you are looking at, not the squares that other tees are moving in

so?

so how is getting tuning tuning for specced zone helping prediction

imagine a other tee is in a tune zone, and has less gravity, you need to know there is tune zone there to predict that tee movement

but client doesnt recive new tuning for other tees

13:10

other tee moves into a low grave zone i dont recieve new tune

I could imagine it helps for projectiles since they're entirely predicted

13:10

if you look at some point of the map and the tune zones are big, there's a chance that the tune zone you get corresponds to the projectile

13:11

but that sounds like an area that needs improvement

yes, without correct tuning the trajectory is completely off

hmmmm so you oculd actually be seeing incorrect prediciotn?

13:11

hmmm

yes

13:11

AFAICT

this is the best you can get for vanilla client as it can handle only one set of tunings at once but ddnet clould handle and predict all of them

i would like to make a poll and see how much people use vanilla to play ddnet

13:13

vanilla always causing trouble

I used it for quite some time

On ger:

> grep "player has entered the game." *.log* | wc -l
21396
> grep "using Custom Client" *.log* | wc -l
17869

$ python
Python 3.6.2 (default, Jul 20 2017, 03:52:27) 
[GCC 7.1.1 20170630] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> 17869/21396
0.8351561039446626

13:24

16% use non-ddnet clients

nice

sweet, python is 3.6.2

@heinrich5991 expect 10% to be botters tw_twinbop

tw_twinbop

they also use ddnet version things IIRC

huh, never noticed it before https://imgur.com/F1VRn3a

that's exactly what I was talking about ^^

@heinrich5991 there are many non-ddnet clients that are based on ddnet or parts of ddnet

<heinrich5991> but they would send ddnet versions, wouldn't they?

Version | Occurences
10086 | 9939
10071 | 1861
10067 | 886
708 | 847
10085 | 479
10043 | 475
10083 | 387
903 | 368
10063 | 302
10042 | 232
1003 | 227
10050 | 221
10062 | 219
10060 | 173
10035 | 164
900 | 146
10084 | 110
10061 | 99
10070 | 97
10064 | 94
10082 | 92
1002 | 71
705 | 67
10034 | 55
10066 | 47
10041 | 44
1 | 42
10065 | 35
902 | 33
703 | 29
10080 | 29
999999999 | 27
801 | 26
602 | 23
1001 | 21
800 | 17
10081 | 17
1337666 | 16
16 | 12
502 | 11
605 | 8
709 | 6
704 | 4
308 | 2
10068 | 2
702 | 2
707 | 2
604 | 1
706 | 1
2 | 1

2

999999999

13:47

1337666

no 997

13:48

:/

13:49

OR EVEN 112

997?

:"/

997 is police number in

1337666 is SP Someone

banhammer

2

<deen> 1 is old DDRace client

112 is the emergency number in all of the EU

yes

here in catalonia too

<deen> people don't like updating...

and tw vanilla has something?

the old one didn't have autoupdate no??

@Ryozuki "in all of the EU" ^^

ah

<deen> these ones are not that old, they have autoupdate. people just don't care to update

13:56

<deen> maybe we should have kept the annoying message in the middle of the screen

13:57

<heinrich5991> do we force a restart on update?

I believe sushi's client is 1 too

1

1

<heinrich5991> is the update UI modal?

13:57

<heinrich5991> mh... maybe I should eat our own dogfood to make the update process easier ^^

https://forum.ddnet.tw/viewtopic.php?f=48&p=57785&sid=25944b88f8a2157be2f56769fce4e272#p57785

14:01

best post

14:01

no irony

hahahah

remove topic?

Love deen's response

14:02

14:02

We faked people on polish fanpage on facebook, 2 years ago

14:02

as well

moved it to offtopic

can you remove it? (edited)

14:03

I think leaving it there helps people find bots

no one gave a bot link

14:03

i dont see a prob (?)

they gave a skype ID to contact with bots

oh

14:04

let's send him bot with data encryption wirus or smth

14:04

:v

u bad person

no

just joking

(deleted the post)

thanks

14:05

you could also tell the user that bots aren't supported in this forum ^^

14:05

via pm or so

14:05

(don't remember the name)

Private MEssage?

private message in the forums

wait can someone help me identify some of those version numbers that @HMH posted?

15:11

1337666 is sp | someone's custom client

15:11

9999999999999 is headshot

15:11

605, 602 are zclient (cheat)

fury

1

15:11

308 is hclient

15:11

708 is fucking baumalein >_>

15:12

502 is fclient

15:12

if anyone knows what the other remaining "short" version numbers are it would be helpful

15:13

also 0 should show on this list, some clients actively send version number 0

i get a error compiling on windows with mingw

17:27

src\engine\client\backend_sdl.cpp: In member function 'virtual int CGraphicsBackend_SDL_OpenGL::Init(const char*, int*, int*, int*, int, int, int*, int*)':
C:\Users\Edgar\Desktop\ddnet\src\engine\client\backend_sdl.cpp:561:41: error: 'putenv' was not declared in this scope
     putenv("SDL_VIDEO_WINDOW_POS=center"); // ignore_convention

(edited)

17:29

On Windows, it's named _putenv

17:29

i think

17:30

from SO

On Windows, it's named _putenv. The leading underscore correctly signals that it's not a standard C function but an implementation extension. You might be able to get away with /Dputenv=_putenv

can you put the error into `` (instead of the quote :P) (edited)

uh, why it resends the message to irc

17:32

this is a bot bug

https://github.com/42wim/matterbridge/issues

42wim/matterbridge

matterbridge - bridge between mattermost, IRC, gitter, xmpp, slack, discord, telegram, rocket.chat,hipchat (via xmpp), steam and matrix with REST API

17:32

if you have a github account, you could report a bug

no i dont have one :)

17:33

let me see if happens again -edit- (edited)

it happened for my edit as well

what bot version?

the interesting question is: what should happen instead?

nothing

so I can say A on Discord an B on IRC?

u cant edit on irc

maybe an "edit to message from 12:34"

with git diff

17:34

:>

17:35

1. - a
1. + b

17:35

or smth

diff isn't optimized for single lines I think

Aren't synchronised envelopes supposed to reset to time 0 when you cross the start line? is there more steps to using them than just ticking the box?

they're nto synchronized to race time anymore, but to map time

Alright

17:38

my real problem is why non-syncronised envs dont reset to 0 when you disconnect/reconnect or when you rcon reload

17:39

my guess is its because s_Time is static in CMapLayers::EnvelopeEval

17:39

but i dont get the logic why its static

sounds like a good idea to fix

17:41

you can report and issue or make a PR if you want

[ddnet/ddnet] Issue opened: #879 CMapLayers::EnvelopeEval s_Time is static

Disconnect/reconnect OR rcon reload doesn't reset non-sync envelopes to zero time static float s_Time = 0.0f; Then it calculates a delta between s_LastLocalTime and Client()->LocalTime() and incrememnts s_Time by the delta Unsure of the logic of this, ultimately it should result in the same value. Seems to work fine if s_Time is simply Client()->LocalTime() but might break sync envelopes. Also unsure how s_time being modified for sync envelopes and non sync can co-exist with the same s...

@Ryozuki matterbridge version v1.2.0-12-gb49fb2b

Exported 209 message(s)