- Welcome to Stunts Forum.
Recent posts
#11
Stunts Reverse Engineering / Re: Removing limits on Stunts ...
Last post by HerrNove - January 18, 2026, 10:04:20 PMGiven the voting trend I don't think this poll needs to continue for one full month as I planned. I'll keep it open till next Sunday in case new insights arise, then I will wrap up the new SuperSight edition.
#12
Stunts Reverse Engineering / Re: Removing limits on Stunts ...
Last post by Overdrijf - January 18, 2026, 06:57:14 PMThere were some later projects around changing visual elements
https://forum.stunts.hu/index.php?topic=4056
https://forum.stunts.hu/index.php?topic=2753
Ironically that last thread already has me complaining that I can't find the most impressive example anymore.
But yeah, physical models are a different thing entirely. They're a much more integral part of the program. We can color the road on the inside of a tunnel white (which will affect all tunnels in that copy of the game, we can't create extra elements, even if they're a copy of an original element), but we don't have a way to alter the grip of that road, not as far as I know.
.
Now that I'm here anyway: No, I would not consider that cheating. I think you could already do the thing you're talking about if you wanted to. There's not a lot on the forum about it, but the Restunts project (contact the members involved for more information) is at this stage kind of an HD replay viewer, with loads more camera options than the game itself.
https://forum.stunts.hu/index.php?topic=3714.0
You could theoretically import all your replays in there to do these measurements (I think), it just takes a bunch of time. In practice, if it's useful to compare section times down to the 20th of a second you're probably looking at two very similar lines, which means you'll be jumping at a very similar height, which means I expect the advantage gained to be very minimal.
https://forum.stunts.hu/index.php?topic=4056
https://forum.stunts.hu/index.php?topic=2753
Ironically that last thread already has me complaining that I can't find the most impressive example anymore.
But yeah, physical models are a different thing entirely. They're a much more integral part of the program. We can color the road on the inside of a tunnel white (which will affect all tunnels in that copy of the game, we can't create extra elements, even if they're a copy of an original element), but we don't have a way to alter the grip of that road, not as far as I know.
.
Now that I'm here anyway: No, I would not consider that cheating. I think you could already do the thing you're talking about if you wanted to. There's not a lot on the forum about it, but the Restunts project (contact the members involved for more information) is at this stage kind of an HD replay viewer, with loads more camera options than the game itself.
https://forum.stunts.hu/index.php?topic=3714.0
You could theoretically import all your replays in there to do these measurements (I think), it just takes a bunch of time. In practice, if it's useful to compare section times down to the 20th of a second you're probably looking at two very similar lines, which means you'll be jumping at a very similar height, which means I expect the advantage gained to be very minimal.
#13
Stunts Reverse Engineering / Re: Distorted explosion sprite...
Last post by HerrNove - January 18, 2026, 06:34:41 PMThat is one of the bugs, the one responsible for the rectangular shape. After fixing that I got the picture of yesterday with a round but off-center explosion.
The other bug, harder to see (without ChatGPT I would have never been able to catch it) is the one before, where the code omits a dereference and treats a int** as if it were a int*. The C language has defences against this kind of mistake, but in Assembly it is very easy to fall for that. Which strongly hints that this part of the engine was directly written in Assembly.
The other bug, harder to see (without ChatGPT I would have never been able to catch it) is the one before, where the code omits a dereference and treats a int** as if it were a int*. The C language has defences against this kind of mistake, but in Assembly it is very easy to fall for that. Which strongly hints that this part of the engine was directly written in Assembly.
#14
Stunts Reverse Engineering / Re: Distorted explosion sprite...
Last post by Foxpup - January 18, 2026, 06:25:02 PMQuote.x2 is written to, but it should be y2Oof. Yup, pretty sure that's the bug right there. In what perhaps should have been obvious from the start (but hindsight's 20/20), bugged explosions are indeed distorted they way they are because one of their x coordinates is erroneously set to the sphere's y coordinate.
#15
Stunts Reverse Engineering / Re: Distorted explosion sprite...
Last post by HerrNove - January 18, 2026, 05:28:27 PMHere the vanilla game with binary patch.
#16
Competition and Website / Re: New features 2026
Last post by dreadnaut - January 18, 2026, 04:29:02 PMI'm making some sweeping changes to some back-end code 👀 Let me know if anything looks odd.
#17
Stunts Reverse Engineering / Re: Removing limits on Stunts ...
Last post by HerrNove - January 18, 2026, 04:20:22 PMQuote from: Victor Narl on January 18, 2026, 09:06:36 AMBut I have an unusual suggestion for your mod: Add the ability to create trick elements in the custom track editor not only on asphalt but also directly on dirt or ice surfaces (for example, it would be cool to see slalom on ice or dirt). This would be a real breakthrough in track construction and would take the game to a new level!
Eh, custom track elements would be the Holy Grail of modding Stunts. They are a very hard challenge because Stunts stores them in a very complex way, where the visual and the physical properties of each obstacle are codified in different places.
As far as I know, the only successful attempt to edit the visual model was in @Overdrijf 's Primal Carnage, which features a green road. Nobody, to my knowledge, has ever succeeded in editing the physical properties. @Duplode had left some notes about this topic in 2009, but apart for a brief exchange last year the conversation has not moved.
#18
Stunts Reverse Engineering / Re: Distorted explosion sprite...
Last post by HerrNove - January 18, 2026, 03:48:09 PMWell, I managed to implement a quick fix in the assembly.
I found out the exact point where the 2D on-screen bounding box of the 3D primitive are calculated, and disabled the calculation for spheres, since it is bugged. Given that the explosion rectangle is computed by combining the bounding boxes of all the car's primitives, this results in a correct explosion sprite as long as spheres are not part of the external hull, or at least they don't extend the silhouette considerably. If we need one day to have the Death Star as custom car, we might need more sophisticated machinery.
Here the diff of the patch.
I'll add it to the next SuperSight release. The good of my simple fix is that it can also be applied as binary patch in the original exe file.
Here the Williams. Each experiment used up a fresh one, because loading replays is boring. I probably destroyed more racecars than the pilots of Williams GP did in its 49 years of history.
And here the Lola against Default's windmill, as homage to the post that started the discussion in 2022.
I found out the exact point where the 2D on-screen bounding box of the 3D primitive are calculated, and disabled the calculation for spheres, since it is bugged. Given that the explosion rectangle is computed by combining the bounding boxes of all the car's primitives, this results in a correct explosion sprite as long as spheres are not part of the external hull, or at least they don't extend the silhouette considerably. If we need one day to have the Death Star as custom car, we might need more sophisticated machinery.
Here the diff of the patch.
I'll add it to the next SuperSight release. The good of my simple fix is that it can also be applied as binary patch in the original exe file.
Here the Williams. Each experiment used up a fresh one, because loading replays is boring. I probably destroyed more racecars than the pilots of Williams GP did in its 49 years of history.
And here the Lola against Default's windmill, as homage to the post that started the discussion in 2022.
#19
Competition 2026 / Re: Cars and rules for 2026
Last post by Duplode - January 18, 2026, 03:13:12 PMQuote from: alanrotoi on January 18, 2026, 02:53:52 PMI don't know. The core of the rule is to help in the same level to top pipsqueaks and also those who follows. Top pipsqueaks will have 2 points more often than the rest so the whole PTB wouldn't have mean to be.
That is a factor too. More points granted overall means bigger gaps between the involved pipsqueaks. Another aspect of the system for which this concern is relevant is carryover. One of the reasons carryover stops at PTB +1 is that it is primarily meant for those fighting for the lower PTB positions, rather than for the lead.
#20
Stunts Reverse Engineering / Re: Crash explosion sprite iss...
Last post by HerrNove - January 18, 2026, 03:12:28 PMThe topic was revived here: https://forum.stunts.hu/index.php?topic=4549.0