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 2020-10-05 00:00:00Z and 2020-10-06 00:00:00Z
Avatar
GitHub BOT 2020-10-05 00:15:53Z
Author icon
bors[bot]
[ddnet:master] 2 new commits
d5fc613 Revert "Show File button in demo player" - def- 870003d Merge #3011 - bors[bot]
00:15
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3011 Revert "Show File button in demo player"
Avatar
Deleted User 2020-10-05 03:09:12Z
winter night is ok, jungle night is nice. but i'd have no problem with replacing heaven, and winter day/jungle day
Avatar
GitHub BOT 2020-10-05 04:30:51Z
Author icon
ddnet-maps
[ddnet-maps:master] 1 new commit
02dc718 A GuggPaloczi - ddnet-maps
Avatar
deen 2020-10-05 08:48:23Z
@Pepe it's written out on stdout, so not so easy to find with steam 😄
08:48
@Pepe I'd suggest making a "myServerconfig.cfg" and adding a sv_rcon_password there
Avatar
mind 2020-10-05 11:52:34Z
Hi, I have found a bug in the automapper when using the NoLayerCopy function.
11:53
At first here are the rules.
test.rules
620 bytes
Avatar
fish and chips 2020-10-05 11:53:15Z
g
Avatar
mind 2020-10-05 11:53:36Z
test will map this pattern
Image attachment
Avatar
fish and chips 2020-10-05 11:53:48Z
woahhh
11:53
i never tried editor
11:54
this looks hard
Avatar
mind 2020-10-05 11:55:07Z
and test2 will randomly change some of the stones with ID 34.
Image attachment
11:56
test3 is just test combined with test2, but it breaks the pattern?
Image attachment
Avatar
Pepe 2020-10-05 12:10:14Z
How about with the non steam version? I just configure the server.cfg and it works @deen ?
Avatar
Pulsar 2020-10-05 12:26:49Z
@mind not a bug, it's just kinda hard to understand it sometimes. Basically it will place 34 and immediately after it may change it to 33. It makes the first rule valid again and you get those repeats. Should work if you check 33 as well NOTINDEX 33 OR 34. (edited)
Avatar
mind 2020-10-05 12:29:32Z
Ah ok, I always had the impression that it would handle one Index rule after the other.
12:32
Wait, so it's placing ID 34 before checking the conditions for ID 34?
Avatar
Pulsar 2020-10-05 12:37:07Z
no, it checks the conditions first, but they stay valid, if it places 33 and you don't check for 33
Avatar
mind 2020-10-05 12:39:00Z
Ok now I undrestand what's happening, thank you very much.
Avatar
deen 2020-10-05 12:44:08Z
@Pepe for the non-steam version you could check in standard output of the client (but then you have to run it from command line). I think adding your own cfg is always easier
Avatar
Patiga 2020-10-05 12:57:53Z
is there any idea behind the physics layers (game, front, ...) having an automapper each?
12:58
(i've already asked that a few days ago but that discussion ended rather abrupt)
Avatar
Soreu 2020-10-05 12:59:10Z
what automapper?
12:59
you mean if you set any other tileset layer as game tiles?
Avatar
Patiga 2020-10-05 13:00:01Z
they are not visible in the editor, but in the map format the game layer, front layer, tele layer, etc each have an automapper
Avatar
Soreu 2020-10-05 13:00:30Z
.-.
Avatar
Patiga 2020-10-05 13:00:44Z
which i would remove if they truly dont do anything and have no purpose whatsoever
Avatar
GitHub BOT 2020-10-05 14:02:29Z
Author icon
def-
[ddnet-web:master] 1 new commit
99dcf00 Add autoreload - def-
14:03
Author icon
def-
[ddnet-web:master] 1 new commit
13e1972 correct dir - def-
14:06
Author icon
def-
[ddnet-web:master] 1 new commit
03b546b upd - def-
Avatar
Deleted User 2020-10-05 14:11:49Z
Is there 0.7 skins to DDNet ones convertor? I could write one.
14:12
In my mod are predefined skins for zombies which looks different at 0.7 and 0.6 protocols as skin systems differs.
14:12
I am thinking about making a tool that would convert 0.7 skins (json) to ddnet ones (png).
14:13
That way my 0.6 players will get the same experience and will finally see 0.7 skins in my game.
14:13
This sed thing doesn't work
Avatar
Ryozuki 2020-10-05 14:35:10Z
tee_thinking <meta http-equiv="refresh" content="5">
Avatar
deen 2020-10-05 14:46:20Z
@Ryozuki i used that before, but wanted a way to turn it off, so there's now a timer and button on https://ddnet.tw/status/
Avatar
GitHub BOT 2020-10-05 14:56:54Z
Author icon
bors[bot]
[ddnet:master] 2 new commits
50b20fb Make animations more consistent with high float values - Jupeyy 1b253bf Merge #3002 - bors[bot]
14:56
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3002 Make animations more consistent with high float values
Avatar
Deleted User 2020-10-05 15:02:17Z
I hope this query is cached for 5 seconds. Or else that's a nice DoS loophole.
15:04
Sorry if that will sound like I am cryticizing you. I am not. But do you really need users to see those stats with a 5 sec resolution?
15:05
If you use that dashboard for yourself only, try nito.net it's realtime system monitoring and agregator. And it's free plan is very nice.
15:06
Also that public dashboard is a nice way for attacker to know how successful are his attacks.
15:07
@deen
15:07
Image attachment
15:09
I'd considered every column except the first 4 as a security risk. That's why status pages even on very big services are laconic.
15:09
(in your status page)
15:11
Also nito.com SaaS is so nice that it require about 1-2/ minutes per machiine to set it up and start receiving graphs on metrics and alerts. (edited)
Avatar
GitHub BOT 2020-10-05 15:17:15Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3018 Editor: Improve usability of server settings
  • Add Mod button
  • Show Del button only when it makes sense
  • Update command input text when clicking on an existing command
  • Make command clearable
As requested by Cøke
heartw 1
Avatar
Ryozuki 2020-10-05 15:20:24Z
"free plan" sounds like something ill never use then
15:20
monkalaugh
Avatar
deen 2020-10-05 15:20:56Z
@Deleted User we reloadit with 120 seconds, not 5 seconds
15:21
NITO looks suspiciously like Grafana 😄
15:22
I tried some such solutions before and they always had quite high resource impact, which I found ironic. The current solution is low-resource enough to run on all our servers
15:22
I'm aware that attackers can also see if their attacks are successful, but this also means players can see if server is currently under attack. I wouldn't want to remove it.
15:24
@heinrich5991 @Learath2 91 alerts from the code scanning, haven't looked into them yet: https://github.com/ddnet/ddnet/security/code-scanning
Avatar
Ryozuki 2020-10-05 15:24:40Z
👀
Avatar
deen 2020-10-05 15:25:13Z
oh well, mostly just our multiplications happening before conversion to larger type
Avatar
Ryozuki 2020-10-05 15:25:45Z
that python thing tho
15:25
well i guess it doesnt matter
Avatar
heinrich5991 2020-10-05 15:27:04Z
localtime is thread-unsafe apparently
15:27
according to man 3 localtime
Avatar
Ryozuki 2020-10-05 15:27:31Z
yeah
15:27
i think localtime_s
15:27
is thread safe
Avatar
heinrich5991 2020-10-05 15:27:59Z
_r
15:28
but what systems support it? does windows?
Avatar
Ryozuki 2020-10-05 15:28:36Z
ah _s is windows only
Avatar
heinrich5991 2020-10-05 15:28:39Z
ah yea
Avatar
Ryozuki 2020-10-05 15:28:54Z
or not?
Avatar
heinrich5991 2020-10-05 15:29:03Z
https://en.cppreference.com/w/c/chrono/localtime
localtime
15:29
or not
Avatar
Ryozuki 2020-10-05 15:29:06Z
The implementation of localtime_s in Microsoft CRT is incompatible with the C standard since it has reversed parameter order and returns errno_t.
Avatar
heinrich5991 2020-10-05 15:29:07Z
since c11
Avatar
Ryozuki 2020-10-05 15:29:14Z
its implemented differently
15:29
in windows
Avatar
heinrich5991 2020-10-05 15:29:17Z
eh baah
Avatar
Ryozuki 2020-10-05 15:29:19Z
cuz they are smart
Avatar
heinrich5991 2020-10-05 15:29:30Z
probably because they had it before the standard(?)
Avatar
Ryozuki 2020-10-05 15:29:34Z
yeah
Avatar
heinrich5991 2020-10-05 15:29:59Z
localtime_r will be available with C2x, but god knows when MSVC will support that standard
Avatar
Learath2 2020-10-05 15:43:45Z
We can do an ifdef
Avatar
Ryozuki 2020-10-05 15:47:22Z
https://devblogs.microsoft.com/cppblog/c11-and-c17-standard-support-arriving-in-msvc/
Elnar Dakeshov
C11 and C17 Standard Support Arriving in MSVC | C++ Team Blog
Please see our Visual Studio 2019 version 16.8 Preview 3 release notes for more of our latest features. Our team is happy to announce that C11 and C17 are becoming supported language versions in the MSVC compiler toolset starting with Visual Studio 2019 version 16.8 Preview 3!
Thumbnail
15:47
@heinrich5991 this is a blog from 14th
15:47
september
Avatar
Learath2 2020-10-05 15:52:41Z
"While there is currently no support for any C11 optional features, we are committed to providing the most impactful optional features in future releases." hopefully one day we'll have standard threads
Avatar
Ryozuki 2020-10-05 15:56:07Z
For many years Visual Studio has only supported C to the extent of it being required for C++. Things are about to change now that a conformant token-based preprocessor has been added to the compiler. With the advent of two new compiler switches, /std:c11 and /std:c17, we are officially supporting the latest ISO C language standards.
15:56
monkalaugh
15:56
@Learath2 > Atomic and threading support are on our roadmap.
Avatar
Learath2 2020-10-05 15:58:00Z
Which insane developer at microsoft thought they should do a character-based preprocessor instead of a token based one? 😄
Avatar
Ryozuki 2020-10-05 15:58:13Z
monkalaugh
Avatar
GitHub BOT 2020-10-05 15:58:57Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3019 Fix centering texts by making TextWidth accurate
Previously the spacing after the last character was added to the width, leading to all centerd texts being slightly too far right. This was especially noticable for the x in clearable buttons, but affected everything. Thanks to @edg-l for bringing this up on Discord. Look at the buttons at the top as well as the clearable "Commands:" edit box: Old: !screenshot-20201005@175704 Ne...
Avatar
Ryozuki 2020-10-05 16:02:16Z
👀
Avatar
Deleted User 2020-10-05 16:10:49Z
@deen what do you mean 'reloading in 120 seconds'? graphs are changing every 2 seconds now
16:10
here https://ddnet.tw/status/
16:11
isn't that the same for you?
Avatar
heinrich5991 2020-10-05 16:11:21Z
I'd guess whole page is reloaded every 120s, updating the player statistics below the server stats
Avatar
Deleted User 2020-10-05 16:12:03Z
I've talked about CPU graph and everything about your system internals
16:12
it seems that this endpoint is not rate limited curl 'https://ddnet.tw/status/json/stats.json' \ -H 'authority: ddnet.tw' \ -H 'accept: application/json, text/javascript, */*; q=0.01' \ -H 'user-agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/85.0.4183.102 Safari/537.36' \ -H 'x-requested-with: XMLHttpRequest' \ -H 'sec-fetch-site: same-origin' \ -H 'sec-fetch-mode: cors' \ -H 'sec-fetch-dest: empty' \ -H 'referer: https://ddnet.tw/status/' \ -H 'accept-language: en-GB,en-US;q=0.9,en;q=0.8,ru;q=0.7' \ -H 'cookie: __cfduid=d519efb3471ac69be67074582ad22e73a1600530880; style=null' \ -H 'if-none-match: W/"5f7b45a0-2787"' \ -H 'if-modified-since: Mon, 05 Oct 2020 16:11:12 GMT' \ --compressed
Avatar
deen 2020-10-05 16:12:14Z
I haven't changed anything in the top part
Avatar
Deleted User 2020-10-05 16:12:18Z
as I get new values every time
Avatar
deen 2020-10-05 16:12:38Z
of course, what good would updating every 2 s be if the server doesn't allow it? 😄
Avatar
Deleted User 2020-10-05 16:12:41Z
and I think there is a bash script involved that starts a shell
16:12
that's my concern about DoS
Avatar
deen 2020-10-05 16:12:54Z
no idea what you mean
Avatar
Deleted User 2020-10-05 16:13:51Z
I mean are those values on CPU time and each are coming from machines directly?
Avatar
deen 2020-10-05 16:14:22Z
each machine runs a python script, that connects via udp to our ddnet.tw main server, and updates the values every 2 seconds
16:14
gets written into a .json file on disk every 2 seconds as well and nginx serves that through cloudflare
Avatar
Deleted User 2020-10-05 16:14:45Z
nice
Avatar
deen 2020-10-05 16:15:00Z
so if you request it more often you'll only stress cloudflare and maybe nginx, nothing else
Avatar
Deleted User 2020-10-05 16:15:02Z
That I was asking about proxying it.
16:15
Yes, thanks!
16:15
*about caching
16:15
Sorry if that wasn't useful. You have nice architecture already
Avatar
deen 2020-10-05 16:16:15Z
if you look through our website, you'll notice that much of it works that way, just creating static pages. some really don't like this approach, but I'm obviously a fan
16:16
no problem, could've been that I missed something
Avatar
Deleted User 2020-10-05 16:17:43Z
I am a huge fan of prerendering heavy pages too.
16:18
I mean backend-heavy ones.
16:18
Nice, thanks (edited)
Avatar
GitHub BOT 2020-10-05 16:22:07Z
Author icon
def-
[ddnet:15.1-devel] 1 new commit
423eea4 Autumn theme fix cloud animation (fixes #2953) - ChillerDragon
16:22
Author icon
def-
[ddnet:15.1-devel] 1 new commit
294501d censorlist without case - def-
16:23
Author icon
def-
[ddnet:15.1-devel] 1 new commit
a314442 Scale auto mapper text size (partially fixes #3013) - def-
16:23
Author icon
def-
[ddnet:15.1-devel] 1 new commit
5570da6 Allow quitting instantly with SDL_QUIT when non-warning dialog showing - def-
16:29
Author icon
bors[bot]
[ddnet:master] 2 new commits
b7dd750 Autumn theme fix cloud animation (fixes #2953) - ChillerDragon e913c7c Merge #3004 - bors[bot]
16:29
Author icon
bors[bot]
[ddnet/ddnet] Issue closed: #2953 Clouds in autumn theme jump around too often
16:29
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #3004 Autumn theme fix cloud animation (fixes #2953)
16:36
Author icon
bors[bot]
[ddnet:master] 2 new commits
54bbf69 Fix laser prediction for non ddrace mods - Jupeyy 75bdcce Merge #2980 - bors[bot]
16:36
Author icon
bors[bot]
[ddnet/ddnet] Pull request closed: #2980 Fix laser prediction for non ddrace mods
Avatar
GitHub BOT 2020-10-05 17:04:58Z
Author icon
def-
[ddnet/ddnet] Pull request opened: #3020 Fix all Multiplication type alerts by CodeQL
Multiplication result converted to larger type
A multiplication result that is converted to a larger type can be a sign that the result can overflow the type converted from.
Example: https://github.com/ddnet/ddnet/security/code-scanning/17?query=ref%3Arefs%2Fheads%2Fmaster
Avatar
Deleted User 2020-10-05 17:39:59Z
Image attachment
17:40
@louis hello, how can I get my ddnet like this?
Avatar
Soreu 2020-10-05 17:47:26Z
@Deleted User #announcements , Release Candidate beta version
Avatar
louis 2020-10-05 17:48:55Z
@Deleted User it's a beta version that's out for testing. it'll be in the official steam release soon
Avatar
Deleted User 2020-10-05 17:49:55Z
And could we get that assets?
Avatar
GitHub BOT 2020-10-05 17:51:51Z
Author icon
Patiga
[ddnet/ddnet] Pull request opened: #3021 only load + save auto mappers for tile layers, not physics layers
If I'm not mistaken, loading a map without auto mappers, then saving would've caused uninitialized values to be written to the auto mappers of physics layers.
Avatar
louis 2020-10-05 17:52:59Z
those are the default assets in the picture AbyssY
Avatar
Deleted User 2020-10-05 17:53:56Z
Oooh ok i get it
17:54
Thanks
Avatar
Patiga 2020-10-05 17:54:49Z
regarding https://github.com/ddnet/ddnet/pull/2973: those values should never be read anyways, however i just found out that the 'standard values' for those should probably be 0, instead of 1 (at least that is the case for when those layers are loaded)
17:55
i used -1 because it was often used to indicate that a pointer references nothing in the datafile
17:55
*instead of -1
17:56
if anyone bothers, i would create another pull request to set them to 0 instead of -1
Avatar
heinrich5991 2020-10-05 17:57:59Z
why 0 instead of -1?
Avatar
Patiga 2020-10-05 17:58:45Z
in editor/io.cpp they were used to check if it was a specific tilemap type by checking if(that_value)
18:01
my small problem with 0 is that it could actually be an index
Avatar
heinrich5991 2020-10-05 18:01:34Z
yea, I think I like -1 better ^^
Avatar
Patiga 2020-10-05 18:04:30Z
trying to fix the clang style problems manually, what does the last one want from me? https://github.com/ddnet/ddnet/pull/3021/checks?check_run_id=1210725166
Avatar
GitHub BOT 2020-10-05 18:04:56Z
Author icon
def-
[ddnet/ddnet] Pull request closed: #3017 Save config file directly (fixes #3012)
Avatar
deen 2020-10-05 18:07:19Z
@Patiga please don't fix them manually. that's the opposite effect of what the auto-formatting should have 😄
18:07
@Patiga can you run python scripts/fix_style.py?
Avatar
Patiga 2020-10-05 18:07:47Z
okay :)
Avatar
deen 2020-10-05 18:08:44Z
Here's the diff
message.txt
2.98 KB
Avatar
Patiga 2020-10-05 18:09:02Z
thx, got it
Exported 155 message(s)