Guild icon
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 2020-11-03 00:00:00Z and 2020-11-04 00:00:00Z
Avatar
GitHub BOT 2020-11-03 07:10:27Z
Author icon
TsFreddie
[ddnet/ddnet] Pull request opened: #3232 Added a Hint for Disabling Debug Mode
Added a line at the bottom of the screen when cl_debug is 1 !screenshot_2020-11-03_15-09-52

Checklist

  • [ ] Tested the change ingame
  • [*] Provided screenshots if it is a visual change
  • [ ] 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 ...
Avatar
GitHub BOT 2020-11-03 07:31:00Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3233 Fix clang-tidy in text.cpp

Checklist

  • [ ] Tested the change ingame
  • [ ] Provided screenshots if it is a visual change
  • [ ] 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
  • [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...
Avatar
GitHub BOT 2020-11-03 07:43:16Z
Author icon
def-
[ddnet/ddnet] Issue opened: #3234 gfx_opengl_major 1 by default on Intel GPU on Windows (with bad driver)
Seems to have helped multiple people, for example on Discord and Steam: https://steamcommunity.com/app/412220/discussions/0/3003298578159903103/?tscn=1604354661 Not sure if we know the exact bad driver versions, but at least we can do it for Intel + Windows.
07:47
Author icon
bors[bot]
[ddnet:master] 2 new commits
996e519 hint for disabling debug mode - TsFreddie e256b11 Merge #3232 - bors[bot]
07:47
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3232 Added a Hint for Disabling Debug Mode
Avatar
Ryozuki 2020-11-03 08:31:04Z
today vim is 29 years old yay (edited)
Avatar
[quakenet] ChillerDragon BOT 2020-11-03 08:51:03Z
lol nice hacktoberfest reward can be used to plant trees :D
08:51
planting trees is such a meme
08:55
seems like planting a tree requires a shipping address? How should i understand that? xd
Avatar
[quakenet] ChillerDragon BOT 2020-11-03 09:10:49Z
@Дядя Женя can i get you to look into compose binds in vanilla as well? it sadly is even more broken there :/ thanks for the ddnet fix btw
Avatar
[quakenet] deen BOT 2020-11-03 09:11:28Z
https://en.wikipedia.org/wiki/Netsplit
Netsplit
In computer networking, specifically Internet Relay Chat (IRC), netsplit is a disconnection between two servers. A split between any two servers splits the entire network into two pieces.
Avatar
GitHub BOT 2020-11-03 11:59:29Z
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3233 Fix clang-tidy in text.cpp
11:59
Author icon
bors[bot]
[ddnet:master] 2 new commits
2762583 Fix clang-tidy in text.cpp - def- 643bbb5 Merge #3233 - bors[bot]
Avatar
GitHub BOT 2020-11-03 12:10:31Z
Author icon
bors[bot]
[ddnet/ddnet] Issue closed: #3207 Why does kicking someone from your team kill the entire team
12:10
Author icon
bors[bot]
[ddnet:master] 2 new commits
2c98f2b Kick-voting someone from team doesn't kill locked team (fixes #3207) - def- 9b64b9b Merge #3222 - bors[bot]
12:10
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3222 Kick-voting someone from team doesn't kill locked team (fixes #3207)
Avatar
Patiga 2020-11-03 12:13:45Z
i think the new novice maps should be moved into the perma server and replace 'Basics', 'CandyLand', 'Pados', 'Kobra 2'
Avatar
TsFreddie 2020-11-03 12:27:44Z
f3
Avatar
GitHub BOT 2020-11-03 12:52:55Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3235 Make bors also look at clang-format & clang-tidy

Checklist

  • [ ] Tested the change ingame
  • [ ] Provided screenshots if it is a visual change
  • [ ] 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
  • [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...
Avatar
deen 2020-11-03 13:04:04Z
@Patiga which one specifically?
Avatar
Patiga 2020-11-03 13:06:13Z
linear, beyond dreams, inure
Avatar
jao 2020-11-03 13:06:53Z
not linear
13:06
thats more solo than novice
Avatar
deen 2020-11-03 13:07:10Z
ok, added Beyond Dreams and Inure
Avatar
jao 2020-11-03 13:07:26Z
f3
Avatar
GitHub BOT 2020-11-03 13:12:21Z
Author icon
bors[bot]
[ddnet:master] 2 new commits
c8efa7d Make bors also look at clang-format & clang-tidy - def- 1ae1204 Merge #3235 - bors[bot]
13:12
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3235 Make bors also look at clang-format & clang-tidy
Avatar
GitHub BOT 2020-11-03 13:20:02Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3236 Fix grammar in autoexec_server.cfg (Credits to Blade)

Checklist

  • [ ] Tested the change ingame
  • [ ] Provided screenshots if it is a visual change
  • [ ] 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
  • [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...
Avatar
GitHub BOT 2020-11-03 13:29:33Z
Author icon
bors[bot]
[ddnet:master] 2 new commits
8fd33f0 Fix grammar in autoexec_server.cfg (Credits to Blade) - def- c1c838f Merge #3236 - bors[bot]
13:29
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3236 Fix grammar in autoexec_server.cfg (Credits to Blade)
Avatar
Ryozuki 2020-11-03 13:36:54Z
im a nobo at this kind of stuff, but i made a blog post about ddnet source code https://edgarluque.com/blog/intro-to-ddnet.html (i plan on extending this to explain how to do more stuff) justatest
An intro to the DDraceNetwork game source code - Blog | Edgar Luque
An intro to the DDraceNetwork game source code, explaining where most of the stuff is located and what it does.
Avatar
heinrich5991 2020-11-03 13:42:54Z
nice 🙂
13:43
src/base is also an abstraction layer over the platform APIs
Avatar
Ryozuki 2020-11-03 13:43:23Z
👀
13:43
" it also contains most platform specific code"
13:43
well maybe i worded it bad
13:43
xD
Avatar
deen 2020-11-03 13:50:26Z
Nice, linked in https://forum.ddnet.tw/viewtopic.php?f=16&t=2420
Avatar
Ryozuki 2020-11-03 13:50:34Z
poggers
Avatar
heinrich5991 2020-11-03 13:50:40Z
Zwelf: could also be linked in the wiki mb
Avatar
Ryozuki 2020-11-03 13:50:56Z
can u send zwelf link
Avatar
heinrich5991 2020-11-03 13:51:10Z
Zwelf is on IRC, shold see pings
Avatar
Ryozuki 2020-11-03 13:51:19Z
i mean
13:51
what did he do
13:51
i want to see it
13:51
OH
13:51
im retarded
nouis 1
13:51
i understand what you mean now
13:51
xD
13:52
for some reason i read as in zwelf did a article too
Avatar
GitHub BOT 2020-11-03 14:11:47Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3237 Create fallback wordlist in sourcecode (fixes #3206)

Checklist

  • [ ] Tested the change ingame
  • [ ] Provided screenshots if it is a visual change
  • [ ] 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
  • [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--u...
14:12
Author icon
def-
[ddnet/ddnet] Issue opened: #3238 Crash when > 60 players join
Reported by HamidReza: `` (gdb) bt #0 GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51 #1 0x00007ffff63f042a in GI_abort () at abort.c:89 #2 0x00007ffff642cc00 in __libc_message (do_abort=do_abort@entry=2, fmt=fmt@entry=0x7ffff6521fd0 " Error in %s': %s: 0x%s \n") at ../sysdeps/posix/libc_fatal.c:175 #3 0x00007ffff6432fc6 in malloc_printerr (action=3, str=0x7ffff6522028 "munmap_chunk(): invalid pointer", ptr=, ar_ptr=) at malloc.c:5049 #4 0x0000555...
Avatar
Ryozuki 2020-11-03 14:24:31Z
It shall be safe to destroy an initialized mutex that is unlocked. Attempting to destroy a locked mutex results in undefined behavior.
14:24
how do u make sure a mutex is not locked?
14:25
https://github.com/ddnet/ddnet/blob/4abf699726a23936bd21db7a9a29d57da7bd349d/src/engine/client/sound.cpp#L373
ddnet/ddnet
DDraceNetwork, a cooperative racing mod of Teeworlds - ddnet/ddnet
Thumbnail
14:28
is it safer to acquire the lock and then unlock it and delete it?
Avatar
deen 2020-11-03 14:28:29Z
that doesn't sound correct
Avatar
Ryozuki 2020-11-03 14:28:35Z
yeah
14:28
https://stackoverflow.com/questions/46345261/how-to-safely-and-correctly-destroy-a-mutex-in-linux-using-pthread-mutex-destroy
How to safely and correctly destroy a mutex in Linux using pthread_...
I read about APUE 3rd, 11.6.1 Mutexes, there is an example about lock and unlock a mutex in this chapter: struct foo { int f_count; pthread_mutex_t f_lock; int ...
Thumbnail
Avatar
deen 2020-11-03 14:28:43Z
between the unlock and delete someone else could lock it
Avatar
Ryozuki 2020-11-03 14:30:35Z
ah, from what i see u usually join the threads and then destroy the lock
14:30
atleast from code im seeing elsewhere
14:31
idk
Avatar
deen 2020-11-03 14:31:43Z
yes, that's safe
14:32
or get all threads into a state where you know they can't aquire the lock
Avatar
Ryozuki 2020-11-03 14:32:14Z
i see
Avatar
GitHub BOT 2020-11-03 14:33:26Z
Author icon
def-
[ddnet/ddnet] Issue closed: #3238 Crash when > 60 players join
Avatar
Learath2 2020-11-03 14:38:49Z
The combined bind pr we merged concerns me. Given it looked way too simple. Whenever I thought there was a simple fix/improvement in that bind code I broke something else that I didn’t consider. Just the fact that how combined chords behave w.r.t simple binds is not well defined really makes it harder than one would expect
Avatar
Ryozuki 2020-11-03 14:45:56Z
sound code is hilarious
Image attachment
14:45
"the real shit:"
14:46
BASED
Avatar
heinrich5991 2020-11-03 14:46:46Z
I don't know how the sound code works
14:46
I think we mix sound on our own?
Avatar
Ryozuki 2020-11-03 14:46:56Z
yea
14:47
i got curious so im kinda learning about it
14:47
well i also wanted to know if #3167 was obvious
Avatar
DDNet BOT 2020-11-03 14:47:34Z
https://github.com/ddnet/ddnet/issues/3167
Turning off sound and on again duplicates sounds · Issue #3167 · dd...
So probably the stopping doesn't work, only seems to affect Windows or some?
Thumbnail
Avatar
Ryozuki 2020-11-03 14:48:33Z
we could use SDL_mixer
14:48
but its another dependency
14:48
not that its a big one tho
Avatar
heinrich5991 2020-11-03 14:48:56Z
if sound mixing is simple, we can also do it ourselves, no?
14:49
I guess it's just adding stuff together?
Avatar
Ryozuki 2020-11-03 14:49:08Z
is it rly that simple?
Avatar
Learath2 2020-11-03 14:49:48Z
It should be tm
Avatar
heinrich5991 2020-11-03 14:50:11Z
idk, I'd guess you add configured_volume * current_displacement of all sound sources
Avatar
Ryozuki 2020-11-03 14:50:16Z
btw we dont have "surround" sound right? if someone hammerhits on ur right u hear it on the right?
14:50
i dont remember right now xd
Avatar
heinrich5991 2020-11-03 14:50:31Z
we do have sounds at specific positions
Avatar
Ryozuki 2020-11-03 14:51:04Z
2.0.2:
  • Implemented 24-bit and surround sound support for FLAC audio files
14:51
sdl mixer has it out of the box
14:51
poggies
Avatar
Learath2 2020-11-03 14:52:17Z
We could already be doing it. I don’t really remember
Avatar
heinrich5991 2020-11-03 14:52:41Z
we do stereo
14:52
I guess we don't do surround
Avatar
Learath2 2020-11-03 14:54:06Z
Honestly, the simple mixing we do is simple on circuit, it should be trivial in code
Avatar
Ryozuki 2020-11-03 14:55:39Z
well its true it looks simple
Avatar
Learath2 2020-11-03 14:56:41Z
If we wanted eq per channel or something like that, it might be worth looking into a library that probably has vectorization which would help with performance
14:57
Oh, and that loop that looks slow also probably vectorizes great
14:58
It might even get auto vectorized if iirc and that version of gcc and clang are out
Avatar
Ryozuki 2020-11-03 15:02:51Z
you know
15:02
SDL_MixAudioFormat
15:02
exists in sdl base
15:03
https://wiki.libsdl.org/SDL_MixAudioFormat
15:03
This is provided for convenience -- you can mix your own audio data.
15:03
weird
15:04
lets use openal BASED
15:05
Distance attenuation, doppler shift, and directional sound emitters are among the features handled by the API. More advanced effects, including air absorption, occlusion, and environmental reverb, are available through the EFX extension
15:05
imagine all those sound effects in ddnet
15:05
monkalaugh
Avatar
Learath2 2020-11-03 15:26:27Z
Doppler shift lol, tees might actually be moving fast enough for that to matter
Avatar
TsFreddie 2020-11-03 16:03:51Z
@Learath2 it is possible to have cursor data stored in demo
16:04
ideally for spectating too. (edited)
16:08
i'll just ask about demo, because i have no idea how demo works.
Avatar
[quakenet] ChillerDragon BOT 2020-11-03 16:08:08Z
@Ryozuki nice blog post more pls
16:08
also where can i like :D
16:09
you rly work on a portfolio latley hm?
Avatar
TsFreddie 2020-11-03 16:09:31Z
thonk also i probably need a lesson on the difference between netobject and netmessage
Avatar
[quakenet] Ryozuki BOT 2020-11-03 16:11:55Z
ChillerDragon yeah gotta have stuff i guess
16:11
ChillerDragon nice to know u like it
Avatar
Ryozuki 2020-11-03 16:14:04Z
this dude keeps messaging me with this
Image attachment
16:14
i wonder what it means
Avatar
TsFreddie 2020-11-03 16:14:22Z
thinking emoji
Avatar
Ryozuki 2020-11-03 16:15:08Z
:thonk: also i probably need a lesson on the difference between netobject and netmessage
@TsFreddie i think a netmessage is more like a event
Avatar
TsFreddie 2020-11-03 16:15:47Z
so like netobject is intended to be packed into snaps?
Avatar
Ryozuki 2020-11-03 16:16:54Z
i guess
16:16
heinrich is the expert in this
Avatar
TsFreddie 2020-11-03 16:18:16Z
EeveeShy
Avatar
heinrich5991 2020-11-03 16:18:38Z
netobj is in snaps
Avatar
TsFreddie 2020-11-03 16:18:44Z
k
Avatar
[quakenet] ChillerDragon BOT 2020-11-03 16:25:36Z
@heinrich5991 when do you open a school or a 4 weeks bootcamp to learn tw coding?
16:25
i would pay 5 bucks lmao good price
Avatar
Deleted User 2020-11-03 17:16:56Z
Any intel user on windows already made the october update from windows?
Avatar
deen 2020-11-03 17:18:12Z
@Deleted User what do you want to check?
Avatar
Deleted User 2020-11-03 17:18:21Z
if the intel driver is updated by it
Avatar
deen 2020-11-03 17:18:33Z
How do I detect if I have the october update?
17:18
I get "Feature update to Windows 10, version 2004"
Avatar
Deleted User 2020-11-03 17:19:10Z
19042.610
17:19
https://docs.microsoft.com/de-de/windows/release-information/
Windows 10 – Versionsinformationen
Informationen zur Freigabe von Windows 10-Versionen
Thumbnail
17:19
i dunno they say 20H2
17:21
appearently its still a manual update yet, but will rollout soon
17:21
"manually seek to “Check for updates” via Windows Update"
Avatar
deen 2020-11-03 17:26:54Z
Yeah, I did that, installing now, but it's called 2004 for me, not 20H2
Avatar
Deleted User 2020-11-03 17:27:04Z
ok, dunno xD
Avatar
deen 2020-11-03 17:27:09Z
https://support.microsoft.com/de-de/help/4562830/feature-update sounds like the same
17:29
super slow for me, I'll check back tomorrow
Avatar
Deleted User 2020-11-03 17:29:48Z
i dont even know if windows updates intels drivers anyway
17:30
was just an assumption bcs it was so sudden with all the problems
Avatar
TsFreddie 2020-11-03 17:31:04Z
Image attachment
Avatar
GitHub BOT 2020-11-03 17:36:58Z
Author icon
def-
[ddnet/ddnet] Issue closed: #3199 Feature Request: /timesnear
17:37
Author icon
def-
[ddnet/ddnet] Issue closed: #3172 Pair up players automatically?
Avatar
GitHub BOT 2020-11-03 17:48:23Z
Author icon
l-ouis
[ddnet/ddnet] Issue opened: #3239 deepfly bind in controls
just a suggestion - maybe add a place for deepfly in controls? It's a pretty common bind.
Avatar
TsFreddie 2020-11-03 17:49:21Z
thonk any idea why other players' gun look so shaky during fast movement?
17:49
with both antiping on and off
17:51
it's like the interpolation is wrong or something
Avatar
Ryozuki 2020-11-03 17:52:22Z
https://www.dsogaming.com/pc-performance-analyses/source-code-watch-dogs-legion-leaked-560gb/
John Papadopoulos
The source code for Watch Dogs Legion has been leaked, weights arou...
The source code for Watch Dogs Legion is around 560GB, and it's already available for download on various private trackers.
Thumbnail
17:52
560gb
17:52
of sourcecode
17:52
insane
17:52
thats more than my ssd
Avatar
heinrich5991 2020-11-03 17:52:54Z
probably also assets?
Avatar
Ryozuki 2020-11-03 17:53:07Z
yeah
17:53
but its still a lot xD
17:53
it will be interesting to see whether a DRM-free version performs faster than the protected version. After all, Watch Dogs Legion has ridiculously high CPU requirements on the PC.
17:54
this means watch dogs legion will have mods
17:54
maybe
Avatar
TsFreddie 2020-11-03 18:09:45Z
Avatar
Deleted User 2020-11-03 18:10:38Z
@TsFreddie subpixels
Avatar
TsFreddie 2020-11-03 18:11:02Z
huh?
Avatar
Deleted User 2020-11-03 18:11:15Z
if u exactly between two pixels its blurrier
18:11
than when u close to one pixel
Avatar
TsFreddie 2020-11-03 18:11:25Z
also should IntraGameTick ever go beyond 1.5f?
18:11
no. it is shaky
18:11
not blurry
Avatar
Deleted User 2020-11-03 18:12:16Z
its dummy copy or what
Avatar
TsFreddie 2020-11-03 18:12:24Z
spec other clients
Avatar
Deleted User 2020-11-03 18:12:30Z
then its just the network which is limited to 25 ticks
Avatar
TsFreddie 2020-11-03 18:12:31Z
(1)TsFreddie is another client
Avatar
Deleted User 2020-11-03 18:12:41Z
u can enable high bandwidth
18:12
then its 50ticks
Avatar
TsFreddie 2020-11-03 18:12:57Z
thonk so the intragametick doesn't adapt to 25 ticks?
Avatar
Deleted User 2020-11-03 18:13:39Z
well it cannot predict the future xd
18:13
its like teleporting with antiping
Avatar
TsFreddie 2020-11-03 18:14:13Z
it is not future tho?
Avatar
Deleted User 2020-11-03 18:14:18Z
how not?
Avatar
TsFreddie 2020-11-03 18:14:28Z
how is interpolation predicting?
Avatar
Deleted User 2020-11-03 18:14:36Z
not like a cursor has some kind of velocity
18:14
how should it predict something
Avatar
TsFreddie 2020-11-03 18:14:45Z
it is just lerping between two snaps
Avatar
Deleted User 2020-11-03 18:15:17Z
what else do you want todo xD
Avatar
TsFreddie 2020-11-03 18:16:10Z
idk interpolation should at least make it smooth, but it didn't.
18:16
isn't that wrong?
Avatar
Deleted User 2020-11-03 18:16:29Z
mhh, don't think so
18:16
backward interpolation would just delay it
18:16
and you don't know where the cursor is next tick
Avatar
TsFreddie 2020-11-03 18:17:38Z
i'm doing backward interp for spec
18:17
Image attachment
18:17
how is this forward
Avatar
Deleted User 2020-11-03 18:18:19Z
ok but spec is probs less critical
18:18
then ingame
18:18
but still, this might look strange
18:18
tho i wonder if this cant cause issues
18:19
when following someone for example
Avatar
TsFreddie 2020-11-03 18:19:49Z
btw the video is not using intragametick for cursor
Avatar
Deleted User 2020-11-03 18:20:11Z
well as said, the cursor has no velocity or smth, it cannot predict anything
18:20
if the client has some interpolation its probably bad anyway
Avatar
TsFreddie 2020-11-03 18:20:29Z
thonk no even backward?
Avatar
Deleted User 2020-11-03 18:20:42Z
well this will create delay
18:20
as you interpolate the past
Avatar
TsFreddie 2020-11-03 18:20:55Z
why do you care about delay when spectating?
Avatar
Deleted User 2020-11-03 18:21:16Z
well i assume the cursor is wrong too if you dont spectate?
Avatar
TsFreddie 2020-11-03 18:21:27Z
huh?
Avatar
Deleted User 2020-11-03 18:21:38Z
i mean the weapon movement
18:21
of the other tee ofc
Avatar
TsFreddie 2020-11-03 18:21:51Z
yes. it is wrong for all other clients
Avatar
Deleted User 2020-11-03 18:22:08Z
so your idea is to delay the whole spec by 1 tick?
18:22
then you can interpolate everything
Avatar
TsFreddie 2020-11-03 18:23:00Z
thonk i don't get it. i don't care about other tee for now. i just want to know why backward interp is buggy. (edited)
Avatar
Deleted User 2020-11-03 18:23:16Z
well as said, its interpolating the past
18:23
its not really nice to have wrong timescales
Avatar
TsFreddie 2020-11-03 18:23:47Z
i'm not saying you should change it to backward or something
Avatar
Deleted User 2020-11-03 18:23:56Z
example: you start a hook, but the cursor is still somewhere else
Avatar
TsFreddie 2020-11-03 18:24:01Z
the interpolation is janked even if i do backwards interp
Avatar
Deleted User 2020-11-03 18:24:10Z
ok
18:24
then you code is wrong or what do you want to say xD
Avatar
TsFreddie 2020-11-03 18:24:27Z
thonk
18:24
i just want to say
18:24
intragametick probably shouldn't go beyond 1.5f or something
18:24
it is probably a bug
Avatar
Deleted User 2020-11-03 18:25:27Z
yes
Avatar
TsFreddie 2020-11-03 18:25:43Z
🤨 so any idea why?
Avatar
Deleted User 2020-11-03 18:26:03Z
is that always or really just iinside this single calculation?
Avatar
TsFreddie 2020-11-03 18:26:57Z
definitely not always always. since it depends on the frame time. but intragametick is just a variable, if it is wrong in this routine then it is probably wrong everywhere
Avatar
Deleted User 2020-11-03 18:28:55Z
can u quickly enable high bandwidth and see if it happens there too
Avatar
TsFreddie 2020-11-03 18:29:08Z
sure
Avatar
Deleted User 2020-11-03 18:29:41Z
i'd assume that this happens bcs tw is 50 ticks but the snaps can get as old as 2 ticks
Avatar
TsFreddie 2020-11-03 18:30:52Z
thonk but i guess if we really don't care about how other tees looks then we can just leave it as is. (if it is the tick issue)
Avatar
Deleted User 2020-11-03 18:31:20Z
well maybe we should use the prediction intra tick for most stuff
Avatar
TsFreddie 2020-11-03 18:31:22Z
i think in 50ticks server other tees looks fine
Avatar
Deleted User 2020-11-03 18:31:29Z
but yeah for the weapon movement this probs doesnt work
18:33
if tw wouldnt have this cancerous snapshot model we could just send individual packets every tick
Avatar
TsFreddie 2020-11-03 18:34:03Z
huh
18:34
with high_bandwidth enabled it feels like it is under interpolating now.
18:34
thonk let me check the intragametick values just to be sure
18:37
the values looks fine
Avatar
Deleted User 2020-11-03 18:37:27Z
yeah
18:37
makes sense then 😄
Avatar
TsFreddie 2020-11-03 18:37:46Z
thonk won't fix?
Avatar
Deleted User 2020-11-03 18:37:59Z
well as said, for cursors probably unfixable
18:38
all other stuff should just use the prediciton intra tick
Avatar
TsFreddie 2020-11-03 18:38:17Z
sad noises
Avatar
Deleted User 2020-11-03 18:38:38Z
the best would be to update cursors every tick
18:38
that would require a unique packet
18:38
independent of snapshots
18:39
we just need to overtake vanilla
18:39
and fix everything
Avatar
TsFreddie 2020-11-03 18:39:35Z
thonk maybe just do some animation style interpolation instead of keeping accuracy.
Avatar
Deleted User 2020-11-03 18:39:36Z
go @TsFreddie get vanilla admin
Avatar
Learath2 2020-11-03 18:39:38Z
what are you trying to do?
Avatar
TsFreddie 2020-11-03 18:39:47Z
spec cursor
Avatar
Learath2 2020-11-03 18:39:53Z
spec cursor?
Avatar
Deleted User 2020-11-03 18:40:03Z
the weapon movement of others tees
18:40
cursor movement of other tees
Avatar
noby 2020-11-03 18:40:12Z
ohh cool
Avatar
TsFreddie 2020-11-03 18:40:13Z
show cursors of spectating player
18:40
i guess
Avatar
Learath2 2020-11-03 18:40:23Z
hm, isn't that already saved tho?
Avatar
noby 2020-11-03 18:40:25Z
a better implementation of the thing i made
Avatar
TsFreddie 2020-11-03 18:40:33Z
you mean in the demo?
Avatar
noby 2020-11-03 18:40:38Z
no it isnt saved in demo
Avatar
Deleted User 2020-11-03 18:40:44Z
would BE AWESOME to have higher precision too
Avatar
TsFreddie 2020-11-03 18:40:49Z
over the protocol only angle is sent
Avatar
Deleted User 2020-11-03 18:40:49Z
e.g. the hookline is always wrong
18:40
for the other tee
Avatar
TsFreddie 2020-11-03 18:40:56Z
from server to client that is
Avatar
noby 2020-11-03 18:41:02Z
i think he means like this except actually done correctly rather than drawing a floating tee
Avatar
TsFreddie 2020-11-03 18:41:04Z
client do sent cursor position to server.
Avatar
Learath2 2020-11-03 18:41:31Z
Why don't you forward it along to the client?
Avatar
TsFreddie 2020-11-03 18:42:00Z
huh? i was forwarding it with a separate netobject
Avatar
noby 2020-11-03 18:42:01Z
whats the best way to do that
18:42
should it be snapped since its updated so frequently
Avatar
Learath2 2020-11-03 18:42:20Z
oh, exactly that, I guess I misread your conversation
Avatar
TsFreddie 2020-11-03 18:42:36Z
ye.
18:42
the problem is the interpolation is jank
Avatar
Learath2 2020-11-03 18:42:49Z
What seems to be wrong? Bad interpolation? 😄
Avatar
TsFreddie 2020-11-03 18:42:51Z
and i have no idea how demo works
Avatar
Learath2 2020-11-03 18:43:07Z
Does the demo show cursors even?
Avatar
TsFreddie 2020-11-03 18:43:23Z
i mean it should
18:43
well not should
Avatar
Deleted User 2020-11-03 18:43:35Z
tho demos will always have the inaccurate netobj cursor
Avatar
TsFreddie 2020-11-03 18:43:39Z
probably nice to have let's say
Avatar
Learath2 2020-11-03 18:43:50Z
If it's not part of the snap it's not sent
18:44
s/sent/saved
Avatar
TsFreddie 2020-11-03 18:44:21Z
ye, it is not. also it's not in localplayer as well.
Avatar
Learath2 2020-11-03 18:44:24Z
Maaaybe for the local tee I could see the entire input being saved, though I don't remember an override there
Avatar
TsFreddie 2020-11-03 18:44:43Z
i'm only sending spectating players' cursor alongwith spectatorInfo (edited)
Avatar
Learath2 2020-11-03 18:44:46Z
So the demo is probably cheating using the angle if it shows the cursor at all
Avatar
Deleted User 2020-11-03 18:44:59Z
it doesnt
Avatar
TsFreddie 2020-11-03 18:45:00Z
it doesn't
Avatar
Deleted User 2020-11-03 18:45:10Z
and if it could, the precision is bad
Avatar
TsFreddie 2020-11-03 18:45:39Z
so i was trying to ask you if there's a hacky way to just store local players cursor info
18:45
in demo*
Avatar
Deleted User 2020-11-03 18:46:06Z
well invent a new netobj extension
Avatar
noby 2020-11-03 18:46:14Z
could it not be added to snapshot instead
Avatar
Deleted User 2020-11-03 18:46:16Z
that would be nice for ddnet anyway
Avatar
TsFreddie 2020-11-03 18:46:24Z
does everything in demo have to be network based?
Avatar
Deleted User 2020-11-03 18:46:36Z
it current is using the snapshots
18:46
to render stuff
Avatar
Learath2 2020-11-03 18:47:17Z
You can do it the same way 0.7 injects the netmessages into the demos
Avatar
TsFreddie 2020-11-03 18:47:43Z
thonk i mean if the interpolation could be better it probably won't matter too much. even csgo only has 20tick demo.
Avatar
Learath2 2020-11-03 18:48:04Z
There will always have to be interpolation
Avatar
Deleted User 2020-11-03 18:48:06Z
but tw theoretically has 50ticks
Avatar
Learath2 2020-11-03 18:48:12Z
Demos are still 50 tps
Avatar
Deleted User 2020-11-03 18:48:14Z
and 50ticks is ok
18:48
its really just that snaps are only send every 2nd tick
Avatar
TsFreddie 2020-11-03 18:48:37Z
but the interpolation is over interpolating in 25ticks, and a little bit under-interpolating with high_bandwidth enabled
Avatar
Deleted User 2020-11-03 18:48:54Z
well for demos you could just use the next tick for interpolation
Avatar
Learath2 2020-11-03 18:49:24Z
For live you'll need to predict sadly
Avatar
TsFreddie 2020-11-03 18:49:34Z
let's not talk about predicting anymore, spectating player's cursor is fine with backward interp thanks.
Avatar
Learath2 2020-11-03 18:49:43Z
backward interp?
Avatar
TsFreddie 2020-11-03 18:49:52Z
the term @Deleted User keep using
Avatar
Deleted User 2020-11-03 18:49:55Z
interpolating the past XD
Avatar
TsFreddie 2020-11-03 18:50:05Z
from prev to cur tick
Avatar
Deleted User 2020-11-03 18:50:05Z
i dunno if there is a term xD
Avatar
Learath2 2020-11-03 18:50:24Z
Oh yeah, that's fine
Avatar
TsFreddie 2020-11-03 18:50:32Z
but that's wrong currently lol
Avatar
Learath2 2020-11-03 18:50:41Z
Wrong how?
Avatar
TsFreddie 2020-11-03 18:50:53Z
prev to cur interp is
is over interpolating in 25ticks, and a little bit under-interpolating with high_bandwidth enabled
Avatar
Learath2 2020-11-03 18:51:09Z
Oh wait, I don't think that works
18:51
We render at intraticks
Avatar
TsFreddie 2020-11-03 18:51:25Z
intraticks is the interpolating stuff no?
18:51
(now-prev)/(cur-prev)
Avatar
Deleted User 2020-11-03 18:52:03Z
but as said for 25 ticks
18:52
its 2 ticks per actual tw tick
18:52
like 50/25
Avatar
Learath2 2020-11-03 18:52:23Z
Well, you don't know what you are interpolating towards at an intratick
Avatar
TsFreddie 2020-11-03 18:52:57Z
what? you are interpolating from prev tick to now based on cur tick. both prev and cur is already received
Avatar
Learath2 2020-11-03 18:53:15Z
You'll always be a tick behind that way
Avatar
Deleted User 2020-11-03 18:53:43Z
intratick is cur tick of cur time - cur tick
Avatar
Learath2 2020-11-03 18:53:43Z
Atleast I think so, I haven't looked at the clientside rendering for a while
Avatar
TsFreddie 2020-11-03 18:53:52Z
so what's the current snap when you calculating intraticks?
Avatar
Deleted User 2020-11-03 18:53:53Z
thats why it gets 1.5f
Avatar
TsFreddie 2020-11-03 18:53:55Z
Image attachment
18:54
it is a fake one that's in the future?
Avatar
Deleted User 2020-11-03 18:54:39Z
lets say it like this: tw is always 50 ticks but snaps are only 25 ticks so u get snap u wait 1 tick, dont get snap u wait 0.5ticks u are at 1.5f intraticks still didnt get a new snap
Avatar
TsFreddie 2020-11-03 18:55:24Z
i know that already. learath2 seems to having trouble with intratickthonk .
18:56
also isn't there a way to tell whether the server is sending snaps every 2 ticks?
Avatar
Deleted User 2020-11-03 18:56:49Z
well normally u just use the prediction intratick
Avatar
TsFreddie 2020-11-03 18:56:53Z
have a delayed intratick might help with adjusted range.
Avatar
Deleted User 2020-11-03 18:56:53Z
that solves all issues
Avatar
Learath2 2020-11-03 18:57:14Z
Well we do indeed run a tick behind everywhere. I didn't know that
18:57
How silly 😛
Avatar
Deleted User 2020-11-03 18:57:29Z
its not really behind
Avatar
TsFreddie 2020-11-03 18:57:32Z
thonk it is the most smooth way i suppose
Avatar
Deleted User 2020-11-03 18:57:35Z
its just less often sent
Avatar
TsFreddie 2020-11-03 18:57:48Z
antiping should take care of the other way
Avatar
Learath2 2020-11-03 18:57:49Z
No it is really behind, even if you send 50tps
Avatar
TsFreddie 2020-11-03 18:58:03Z
not the interpolating itself
Avatar
Deleted User 2020-11-03 18:58:18Z
the prediction clearly uses a tick from the future
Avatar
Learath2 2020-11-03 18:58:22Z
The client state at any given time is always behind the state of the current snap
Avatar
TsFreddie 2020-11-03 18:58:29Z
but intratick is not a prediction
Avatar
Deleted User 2020-11-03 18:58:40Z
but intraticks are only rarely used
Avatar
TsFreddie 2020-11-03 18:58:45Z
it always wasn't
Avatar
Learath2 2020-11-03 18:58:53Z
Yeah, I know
Avatar
noby 2020-11-03 18:59:11Z
what is intratick used for
18:59
just rendering?
Avatar
Learath2 2020-11-03 18:59:16Z
I just assumed we were interpolating between current and the next predicted tick
Avatar
TsFreddie 2020-11-03 18:59:25Z
also aren't intratick used everywhere when antiping is off
Avatar
Deleted User 2020-11-03 18:59:31Z
it does, just not for eveyrthing
Avatar
Learath2 2020-11-03 18:59:43Z
Intraticks make things actually look smooth
Avatar
TsFreddie 2020-11-03 18:59:55Z
and not it is not smooth even with intraticks
18:59
here
19:02
i mean if intraticks is cur to next before, you'd thought vanilla would be looking like having antiping on without actually compensating for network latency lol.
Avatar
Learath2 2020-11-03 19:02:24Z
It can't be not smooth, you are doing something wrong then
Avatar
TsFreddie 2020-11-03 19:02:41Z
lets say it like this: tw is always 50 ticks
>
but snaps are only 25 ticks
>
so u get snap u wait 1 tick, dont get snap u wait 0.5ticks u are at 1.5f intraticks still didnt get a new snap
Avatar
Learath2 2020-11-03 19:02:44Z
It can be over/under shooting
Avatar
TsFreddie 2020-11-03 19:02:48Z
this is probably the reason
Avatar
Learath2 2020-11-03 19:02:49Z
It can't be not smooth
Avatar
TsFreddie 2020-11-03 19:03:14Z
the under shooting might be me imagining
19:03
but over shooting in 25 ticks is definitely a thing.
19:03
well 25 snaps.
Avatar
Deleted User 2020-11-03 19:04:13Z
yes, just rewrite most stuff with the prediction intra ticks
Avatar
TsFreddie 2020-11-03 19:04:20Z
no
19:04
let's just do the behind thing but have it adapt to 2 ticks spacing
19:04
intratick should not do prediction
Avatar
Learath2 2020-11-03 19:04:58Z
Let's rewrite this whole mess to be prediction agnostic
Avatar
Deleted User 2020-11-03 19:05:18Z
i dunno @TsFreddie i think its cancerous to have it behind
Avatar
Learath2 2020-11-03 19:05:23Z
A function to mix between two snaps and you just pass the snap state you want rendered
Avatar
TsFreddie 2020-11-03 19:05:31Z
thonk it was always that way
19:05
aiming would be a bit off if you change it
Avatar
Learath2 2020-11-03 19:06:38Z
I actually love that idea. I'm adding it to my list of TODOs
Avatar
TsFreddie 2020-11-03 19:06:50Z
but of course you can have an option alongwith the fixed old ones.
Avatar
Deleted User 2020-11-03 19:06:52Z
yes @TsFreddie bcs the current code is bad too xD
Avatar
TsFreddie 2020-11-03 19:07:12Z
thonk please don't replace it
Avatar
Deleted User 2020-11-03 19:07:31Z
it be already better if it just assumes that the prediction tick is as recent as the current tick
Avatar
TsFreddie 2020-11-03 19:08:27Z
it's like giving the antiping high ping shaky effect to high velocity tees with no antiping as well.
Avatar
Deleted User 2020-11-03 19:09:01Z
well dunno what u imagine
19:09
i'd just imagine a faster interpolation and would still look the same
Avatar
TsFreddie 2020-11-03 19:09:31Z
probably not if any prediction is involved
Avatar
Deleted User 2020-11-03 19:09:42Z
yeah i mean without a real prediction
19:09
just not that it interpolates over 2 ticks
Avatar
TsFreddie 2020-11-03 19:09:56Z
but you can't interp into the future without prediction
Avatar
Deleted User 2020-11-03 19:10:02Z
it bascially assumes the prediciton tick is the prev tick
19:10
not the real prev tick
19:10
which is 2 ticks behind
Avatar
TsFreddie 2020-11-03 19:10:15Z
oh that
19:10
thought you mean changing the entire intratick to forward interp (edited)
19:10
from cur to next
Avatar
Deleted User 2020-11-03 19:10:54Z
nah i won't start a teewar(s) rn
19:11
haha very funny
Avatar
TsFreddie 2020-11-03 19:11:01Z
thonk k
19:11
i don't get it
19:11
that's a lie
Avatar
Deleted User 2020-11-03 19:11:18Z
teeworlds was named teewars in the past
Avatar
TsFreddie 2020-11-03 19:11:23Z
ye
19:11
just remembered
19:12
😄 k sleep time
Avatar
Deleted User 2020-11-03 19:12:12Z
gn8
Avatar
TsFreddie 2020-11-03 19:14:00Z
A function to mix between two snaps and you just pass the snap state you want rendered
@Learath2 can you do it for vanilla as well if you really do it.
Avatar
[quakenet] ChillerDragon BOT 2020-11-03 19:55:53Z
hm so either im stupid or i lost a demo ._. is it possible to playback a demo while it is being recorded and does it stop recording?
Avatar
Souly 2020-11-03 20:02:33Z
shouldn't stop recording
20:02
and you should be able to play it while recording
20:03
at least on 0.7 justatest
Avatar
deen 2020-11-03 20:22:02Z
@Deleted User ok, i have the update now. Intel UHD Graphics 620 driver date: 21.01.2020, driver v ersion: 26.20.100.7810
Avatar
BlackTreez 2020-11-03 20:22:31Z
guys
20:22
i need aip-race map
20:22
and standard infection mode without classes
Avatar
Deleted User 2020-11-03 20:24:26Z
@deen that sounds bad 😦
Avatar
deen 2020-11-03 20:25:37Z
windows update finds no further updates, so that's it
Avatar
Deleted User 2020-11-03 20:25:59Z
yeah
20:26
then rip intel users
Avatar
deen 2020-11-03 20:26:30Z
Can we just blacklist this one driver version, switch to gfx_opengl_major 1 for it and show a red warning on startup telling them to update driver?
20:26
better to look bad on it than crash in the middle of a long run
Avatar
Deleted User 2020-11-03 20:28:42Z
if i get a clear sign from @louis that new drivers don't crash bcs of this bug anymore, yes probably.
Avatar
louis 2020-11-03 20:29:29Z
which new drivers
20:29
the newest one had a separate bug
20:29
(not unique to ddrace client), so i don't want to use it
Avatar
deen 2020-11-03 20:30:34Z
what bug?
Avatar
louis 2020-11-03 20:31:30Z
when i tabbed into/from any fullscreen window there would be a black screen for a second
Avatar
Deleted User 2020-11-03 20:36:59Z
ok
20:37
just important that we dont blacklist random drivers
Avatar
louis 2020-11-03 20:37:39Z
ill try again when i have some time
20:37
probably tonight
Avatar
[freenode] Zwelf BOT 2020-11-03 20:42:45Z
[ddnet] <heinrich5991> Zwelf: could also be linked in the wiki mb
20:42
Yes, thanks for pointing it out
Avatar
heinrich5991 2020-11-03 20:45:25Z
maybe also make sure that all the links are archived on web.archive.org
20:45
(or archive them yourself 😛 )
Avatar
Ryozuki 2020-11-03 20:51:07Z
i wonder how many gb of data web.archive.org has
20:51
amazing work tho
20:52
but nothing beats https://libraryofbabel.info/
Library of Babel
A project towards a universal library. By this art you may contemplate the variation of the 23 letters.
20:52
it has every possible book text article ever written
20:52
BASED
Avatar
heinrich5991 2020-11-03 20:53:14Z
As of September 2018, the Wayback Machine contained over 25 petabytes of data.[34][35]
20:53
As of December 2014, the Wayback Machine contained 435 billion web pages—almost nine petabytes of data, and was growing at about 20 terabytes a week.[13][31][32]
Avatar
<BµmM> 2020-11-03 20:53:33Z
how big is 1 petabyte
Avatar
Ryozuki 2020-11-03 20:53:36Z
(i heard of library of babel from this https://www.youtube.com/watch?v=GDrBIKOR01c) rly interesting stuff
Vsauce
Messages For The Future
Avatar
heinrich5991 2020-11-03 20:53:57Z
I think it's 1024 terabyte, 1 terabyte is 1024 gigabyte
Avatar
<BµmM> 2020-11-03 20:54:04Z
ahh
20:54
wow
Avatar
GitHub BOT 2020-11-03 22:03:06Z
Author icon
Zwelf
[ddnet/ddnet] Pull request opened: #3240 Fix UUID in test case comment
for teehistorian-ddnetver-old@ddnet.tw

Checklist

  • [ ] Tested the change ingame
  • [ ] Provided screenshots if it is a visual change
  • [ ] 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
  • [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://githu...
Avatar
GitHub BOT 2020-11-03 22:13:55Z
Author icon
bors[bot]
[ddnet:master] 2 new commits
5914dbb Fix UUID in test case comment for teehistorian-ddnetver-old@ddnet.tw - Zwelf f8da42a Merge #3240 - bors[bot]
22:13
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3240 Fix UUID in test case comment
Avatar
Ryozuki 2020-11-03 22:57:57Z
https://github.com/github/dmca/pull/8148
Update 2020-10-23-RIAA.md i fixed it by resynth1943 · Pull Request ...
here is my picture: do you like it.
Thumbnail
22:58
https://github.com/github/dmca/tree/416da574ec0df3388f652e44f7fe71b1e3a4701f github itself has a copy of youtube-dl
22:58
hilarious
22:59
https://twitter.com/t3rr4dice/status/1320660235363749888
Author icon
Tzlil (@t3rr4dice)
so the CEO of GitHub popped in #youtube-dl a few hours ago, looks like they are trying to get youtube-dl back, i hope this works out
Retweets
410
Likes
1590
Image
Footer icon Twitter • 2020-10-26 09:35:11Z
22:59
👀
Avatar
Learath2 2020-11-03 23:39:23Z
I highly doubt youtube has copyright on the rolling cipher used on youtube lol
23:39
This is such a classic example of copyright abuse
Avatar
GitHub BOT 2020-11-03 23:52:49Z
Author icon
bors[bot]
[ddnet/ddnet] Issue closed: #3206 Support starting without wordlist.txt
23:52
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3237 Create fallback wordlist in sourcecode (fixes #3206)
23:52
Author icon
bors[bot]
[ddnet:master] 2 new commits
06b534c Create fallback wordlist in sourcecode (fixes #3206) - def- dcfdd05 Merge #3237 - bors[bot]
Exported 471 message(s)