21/Bokujishin/Cyril Bissey/France/no team

I found some time to work on text encoding, once again shamelessly adapting/stealing code from Flame CZE and adapting it to Godot.
I have released version 0.8.1 on GitHub containing this feature as well as fixes for several other issues with various packets.


I should now have everything I wanted to add (unless I'm forgetting something), so will bump the version to 1.0 after some more testing and bug fixing.

Some projects I am working on with this tool:

• Telemetry recording and display
  
• Live race stats (minimap with driver names, standings with tyre compounds, pit stops, time deltas, etc.)
  
• PTH and SMX file parsers to display them in 3D
  

From the set pitstop suggestion thread:

Quote from Scawen :

We can set number of laps and a penalty, so I guess the number of pit stops done is the one missing thing.

I noticed that the PEN packet has a spare byte, could it be used to set custom penalties in addition to predefined ones? e.g. 5s, granularity of byte allows for single second penalties.

These penalties could be easily adjusted, though I must say I'm not sure how it should be presented to the players or actually set as part of the PEN packet - alternatively, add a new packet for this, such as PET (PEnalty Time)?

The command to set the penalty could take the format /penalty USERNAME TIME.

#21

21/Bokujishin/XRR

Quick nitpick in IX.4.a), I believe "all other teams" should be "all other drivers".

yes/21/Bokujishin

In the Signups post, it says we need an S2 license, but isn't Rockingham S3 only?

21/Bokujishin/Cyril Bissey/France/no team/cyk774

Session AND SERVER of Incident: Main, Race 2
Lap AND MPR timecode of incident (or session time or UTC Time of Day): lap 2, MPR 4:48
Car(s) involved: 21, 35, 27
Location of Incident: turn 5
Brief Description of Incident: 35 spins 21 into the gravel, may have been "helped" by 27 (not sure if sliding on their own or following contact with 27).

For anyone using the Godot game engine, I released an InSim library written in GDScript, which includes support for InSim, OutSim OutGauge and InSimRelay.

Website with class reference, guides, demos
GitLab repository (releases for v3.0+ (MIT license)
GitHub repository (releases up to v2.0) (MIT license)





The library supports InSim version 9 as of LFS version 0.7F.

What is included:

• All InSim packets are implemented.
  
• All text input is done from UTF8 and converted to LFS format, all packets convert text back to UTF8.
  
• OutSim can be used from the configuration file.
  
• OutGauge can be used from the configuration file.
  
• InSimRelay is available.
  
• Requests to the REST API are implemented (requires authentication).
  
• Multiple file formats can be read and written (SMX, SET, PTH, LYT, CAR_info).
  
• There are utility classes for unit conversion, time string manipulation, Godot camera handling.
  

Feel free to try it if you're interested!.


Special thanks:
Flame CZE and his Node InSim library, where I found how to parse car names for mods and took inspiration for text conversion.

Last edited by Bokujishin, Thu 31 Jul 2025, 18:02. Reason : Updated GitLab links

Updated #21 with some more logos.

Quote from mbutcher :

We (Sim Broadcasts) also noticed some frame drops during our broadcast of round 1 last Sunday. We'll be trialling a new AMD CPU with a 50% higher clock speed for round 2 to try and combat this (3 cheers for AWS ) - Your experience would have been smoother than ours because you were driving, not observing 35 cars at full LOD all at once.

I can assure you my experience was not that smooth as my game actually *crashed* on both occasions (that's why I dropped out and had to start both races from the pitlane), though that was due to my Linux/wine not being properly configured and should not happen next Sunday (I hope, I'm not planning to install Windows ever again on my PC...).

The T1 incident with the Ferrari mod looks horrendous though, I've never seen such frame drops (or such T1 carnage ), it's definitely necessary to have sensible limits to avoid this kind of situation.

@Scawen I suppose the freezes occur while the game is loading the assets (car model and/or skin), would it be possible to load those on a non-blocking thread instead? And either keep the car invisible while assets are still loading, or load and display them from lowest LOD to highest, so that the game could still display something while it loads the higher detailed stuff, and avoid stuttering/freezing?

I know you don't agree with many modders about triangle count, and I understand the reasoning, though I don't personally agree that 65k triangles is *too much* - however I do agree that getting 65k for the sake of getting them, even if they don't bring anything visually, is a waste of resources. I just think that some other popular sims (think iRacing or ACC) manage grids larger than LFS, with cars that I believe are probably all more detailed than LFS cars (and probably LFS mods as well), and they seem to work pretty well; the engine running the games is different though (Unreal for ACC, not sure about iRacing), so while I agree modders should not abuse the limits, and I understand why you don't want to increase those limits, maybe there is still something you could do to alleviate this particular issue?

I'm not sure what the threshold is where the problem appears, I know some people talked about it after the NDR GT4 Sprint Cup with the N.400S GT4 (there were 35 cars racing + safety car and rescue vehicles), I didn't personally notice any stuttering but I do have a rather powerful computer (Ryzen 7900X, AMD 7900XT, 32GB of RAM, NVMe SSD).

Last edited by Bokujishin, Tue 16 Jan 2024, 22:13.

yes/21/Bokujishin

yes/21/Bokujishin

Here is #21.

I am making an InSim/OutSim/OutGauge library using the Godot game engine, so I went through the documentation as provided in docs\InSim.txt, and found the following mistakes, as of version 0.7E:

• lines 450-451: TINY_PING is in the wrong section, the same lines appear in the proper section at lines 487-488.
  
• lines 1043-1046: TINY_VTC is in the wrong section, and is already documented at lines 894-897.
  
• line 1769: typo: TINT_NCN instead of TINY_NCN.
  

I believe I went through the entire file and found nothing more, though I did not read all parts in full detail (such as AutoX).

21/Bokujishin/Cyril Bissey/France/no team

Great job on this update! I hope you take a well-deserved break, and feel free to make the next series of test patches as early as reasonably possible so we can test the tasty new physics and graphics

My understanding is that the rotator controls the angle (just like the slider controls the position), whereas the spinner controls the speed of rotation, or angular velocity.

Quote from Scawen :

The temporary download is http (not https) it's actually fine, no-one will hack your communications while you download this file. It's not a sensitive communication, just a publicly available file so if you can tell your browser not to worry about it that will be fine.

Yeah, I wasn't too worried, just thought it was unusual; anyway, in-game update works just fine, even on my Linux machine.

When I click on the link, Firefox tells me it didn't download because of a potential security risk, this never happened with the previous test patches.

Edit: In-game update worked perfectly fine.

Popup lights do not deform with the rest of the car. Yes, I drove straight into a wall, twice, to get this kind of damage. Probably not too important since wheels can appear to not be physically attached either.

Quote from ScammerGamerLOL :

Will there be graphics and night mode in the 0.7E? and how many days left until release?

No, 0.7E will only have what we have in the current test patches, new graphics/lighting is (hopefully) what the next series of test patches will be for.
I believe Scawen said he wanted to release 0.7E sometime next week or the week after.