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 2021-08-02 00:00:00Z and 2021-08-03 00:00:00Z

there is a cheat in my game to whom should i complain @Akari

o.O

01:15

what do you mean a cheat in your game

aim bot

01:18

man uses aim bot

Deleted User

man uses aim bot

use #reports next time and tag @ moderator

01:50

then we will come and banhammer them

can someone recommend me DAC sound cards and TN monitors? (edited)

03:51

i hate my sound card and my VA monitor

03:52

i used to have a 144hz TN and it was way more clear than this

03:52

currently lookin at this sound card https://www.amazon.com/dp/B07T9YYVV6

Creative Sound Blaster AE-7 Hi-Res Internal PCIe Sound Card, Quad-C...

Creative Sound Blaster AE-7 Hi-Res Internal PCIe Sound Card, Quad-Core Processor, 127dB DNR ESS SABRE-class 9018 DAC, Xamp Discrete Custom Bi-amp, Discrete 5.1/Virtual 7.1, Dolby, DTS Encoding (Black)

[ddnet-maps:master] 1 new commit

55d8253 M Inspire, M Squadron, M PreciousDream, M Caverun - ddnet-maps

@lynn Don't get a TN panel, horrible colors when you're not perfectly centered. For a DAC it's commonly recommended to get an external one, the FiiO E10K is popular here and worked well for me. (But with my new motherboard the onboard sound sounds exactly the same, can't tell it apart, so I sold mine again)

[ddnet/ddnet] Issue opened: #3980 dnsbl could use manually set dns server?

Instead of the system one. Some hosters really don't like me changing the DNS server and keep fighting me.

deen

@lynn Don't get a TN panel, horrible colors when you're not perfectly centered. For a DAC it's commonly recommended to get an external one, the FiiO E10K is popular here and worked well for me. (But with my new motherboard the onboard sound sounds exactly the same, can't tell it apart, so I sold mine again)

I mean, i always used a TN and im used to it, maybe in that case ill get an IPS? idk, i fucking hate color bleed and VA is the worst with that

07:49

like i want a high refresh rate monitor, and VA is kinda the worst with that

07:50

oops i said the same thing

deen

@lynn Don't get a TN panel, horrible colors when you're not perfectly centered. For a DAC it's commonly recommended to get an external one, the FiiO E10K is popular here and worked well for me. (But with my new motherboard the onboard sound sounds exactly the same, can't tell it apart, so I sold mine again)

it also seems like Creative has an external one? check out the sound blasterx g6

07:50

and it's a higher bitrate too

I used to have a few Creative sound cards, they were ok, but pretty expensive

07:51

the nice thing was the DAC needed no drivers, weird windows-only setup stuff etc

yea, i just dont use the software on windows

07:51

and it works but for whatever reason the audigy fx is really quiet

07:51

much like my onboard audio

07:52

so for now im connected to a fiio a3

07:52

which connects to the audigy but really the fiio is doing most of the work lol

07:53

@deen if you have any specific monitor recos as well let me know

07:53

i might try to get a 1080p 240hz IPS with good reviews

07:53

since there are a lot of those it seems

High bitrate audio I also tried, and could never tell the difference in a blind test. See https://web.archive.org/web/20190209231201/http://people.xiph.org/~xiphmont/demo/neil-young.html for better arguments

i always noticed a difference on 32bit/196khz on pulseaudio in linux

I'm using an old TV with horrible latency as a screen, so better not use that

compared to the 24bit i have on windows

07:54

maybe there is something specific to linux that made the audio more precise i dont know

07:55

this is a cool write-up tho

sampling rate, not bitrate

07:57

The writeup by Monty (who made Vorbis) is nice, not sure why it's down now. He also had some videos explaining A/D basics

#off-topic dude

10:06

stop spam here

Any idea how to pass multiple files to a unix pipe without breaking it? I want to tail a logfile and when the server restarts and creates a new logfile i want to pick it up and tail the new one. Currently have some bash script that picks up if there is a new log file and kills the old tail and replaces it with a new tail. That works but when i try to read that thing from python i get broken pipe error

                uuid="$(generate_uuid)"
                current_file="$(get_latest_logfile)"
                show_latest_log "$follow" "$logpath" --id="$uuid" &
                while true
                do
                        if [ "$current_file" != "$(get_latest_logfile)" ]
                        then
                                pkill -f "id=$uuid"
                                pkill -f "tail.*$current_file"
                                current_file="$(get_latest_logfile)"
                                show_latest_log "$follow" "$logpath" --id="$uuid" &
                        fi
                        sleep 1
                done

why you want to tail a logfile

13:23

https://xyproblem.info/

Asking about your attempted solution rather than your actual problem

to pipe it into python

13:23

dont ask :/

13:23

ik my setup is janky

chillerdragon: tail -F?

14:00

and just pipe that into your python script

yes

14:26

but how to switch out the pipe without breaking it @deen ?

14:28

if i pkill the tail and run tail again it breaks the pipe

don't kill it

14:29

tail -F will automatically switch to the new log file

dfq

14:30

no way

14:30

so i give a directory?

14:30

and it checks for new files?

14:30

magic

no, you move the old log file away and make your server always use the same name

14:31

logrotate style

hrmpf

14:31

i do not want to change my whole logging process just for this

it's how every sane logging system works though

do ddnet servers log like this?

yes

the official ones?

yes

how?

14:32

systemd?

14:32

hmm

Nah, just like this: https://github.com/ddnet/ddnet-scripts/blob/master/servers/run64.sh

ddnet-scripts/run64.sh at master · ddnet/ddnet-scripts

The magic scripts that make DDNet run. Contribute to ddnet/ddnet-scripts development by creating an account on GitHub.

wait all official servers run with systemd?

14:32

multiple servers?

14:33

i dont get it

log file of currently running server on port 8303 is always 8303.log

yes

14:34

but how do you have multiple servers running with systemd?

then when you rotate it away on server restart you rename it to 8303.log.20210802 or whatever

isnt that a mess?

no

14:34

Nah = no

so do you use systemd at all?

14:35

go leak some infrastructure

only rudimentarily to start everything

14:35

but not for logs etc

how are the individual servers run then?

14:35

with scripts that do the rotating for you?

Yes, I linked you the run64.sh script, which is run by this: https://github.com/ddnet/ddnet-scripts/blob/master/servers/run-all.sh#L7-L9

ddnet-scripts/run-all.sh at master · ddnet/ddnet-scripts

The magic scripts that make DDNet run. Contribute to ddnet/ddnet-scripts development by creating an account on GitHub.

so they run in nohup i see

14:36

yea ok nice TiL good logging style is rotation

14:36

but I will not implement that any time soon i think

14:37

any idea how I can fix my issue with updating a pipe to a new file?

14:37

or recover from broken pipe in python?

14:37

update stdin?

don't use stdin, but write to a named pipe

14:38

aka a fifo file

oof

14:39

if i do it crazy like that might as well just connect to econ from python directly

14:39

and skip the pipe

I don't know how to solve it in python, in shell script you can easily do sth like this: (while true; do tail -f $file; file=my_next_file; done) | python3 my_python_script.py

that would work in python too i guess

14:40

the issue is that its not tail -f; tail -f; but its tail -f & and pkill tail

yeah, don't do that

maybe i can send a more chill signal?

14:44

the while loop would not reach the 2nd iteration because "tail -f $file;" does never finish

Those pkills are pretty horrible too. Better to store the pid in a variable when you fork it

oh thats a cool idea

14:50

SIGTERM fixes the broken pipe crashes but then it just does not recieve the new pipe anymore

14:51

i somehow have to highjack the old pipe

14:53

tail -f log.txt > /proc/$tail_pid/fd/0 that feels wrong

Exported 109 message(s)