Forum
Questions, builds, help and the odd argument. The mods and guides are out front.
I will report updates made to SlrrExhaustive (which will only concern bugs reported or ones I find) here.
Also if or when I find something more to add to the Native function thingmod
Or if I find something interesting to share like the MapRPKDecode or the RPKMergeTool
I will also try and relay bugreports here.
Concerning SlrrExhaustive I will keep an update link this one:
SlrrExhaustivePatch
That will contain all the patches made from the start merged together but I will update the main file under the link from the SlrrExhaustive page as well.
Added 2 minutes later:
I got a bugreport for SlrrExhaustive that can crash the carlot (consistently if you have won the ROC and have the "secret" car) the patch for that has been merged to the main download link and to the patch download aswell.
Added 24 minutes later:
I also have some more offset definitions for the Native functions (setI,setF...) in the RawEdit class (under LINK) I did not update the download on the page but if some one wants to here are some more values:
public static IntValueAddress Address_I_bDrawHeadlightRays = new IntValueAddress(0x006187BC);
public static IntValueAddress Address_I_bDrawHeadlightFlares = new IntValueAddress(0x006187C0);
public static IntValueAddress Address_I_ShadowSize = new IntValueAddress(0x6187B8);
public static IntValueAddress Address_I_Shadows = new IntValueAddress(0x6187B4);
public static IntValueAddress Address_I_TextureSize = new IntValueAddress(0x6188B4);
public static FloatValueAddress Address_F_cfg_ShadowDetail = new FloatValueAddress(0x65AED0);
public static FloatValueAddress Address_F_cfg_ObjectDetail = new FloatValueAddress(0x65C424);
public static FloatValueAddress Address_F_cfg_VideoGamma = new FloatValueAddress(0x00617728);
public static FloatValueAddress Address_F_WheelGroundFeedbackFactor = new FloatValueAddress(0x60E020);
public static FloatValueAddress Address_F_WheelBreakeFactor = new FloatValueAddress(0x60E024);
public static FloatValueAddress Address_F_LastFrameDrawTime = new FloatValueAddress(0x63C528);
public static FloatValueAddress Address_F_GameTimeSeconds = new FloatValueAddress(0x0063C530);
MMMh this is something i'll gladly follow 😃
I pulled another suspension kit (a properly adjustable one) to the mod (only to the patch for now) to be able to create something like this:
:wtf: :wtf: :confused2: This is :RICE:
Yup... maybe it is overthetop a bit 😛 but whatever this was my childhood fantasy 🙂
I can confirm now that the InnerPolyEntry.unkownCount1 field in an RPK is indeed not used Slrr skips this field when loading models from an RPK.
(the call that returns the value of this field is made at 0053E40E and the result is thrown away at 0053E417 with another call after skipping the value in the buffer just above]) I did not find other code that accesses this part in the file buffer.
Okay.
Hey, I managed to get the game to work, but I have 1 problem. As soon as the effects of ReShade have completed loading, the overall UI disappears, I can still click on the icons, but I cant see them.
Do you know why? Mind you I dont have the most powerful PC if it has to do with that, but it normally runs my own modded SLRR at 1080p/45fps
hmm this sounds like reshade didnt find the speed variable, and it thinks it is some large value and blurs out everything it happened to me aswell during testing. please dont paste it here but maybe on pastebin can you share the d3d9.log to verify it finds the float?
I have updated the patch link (LINK) the tournament generation did not work as intended after you became one of the top 5 racers (I wanted to generate random racers instead of getting them from the ladder because it would be dull racing against the same cars over and over only because you cant improve your position any further.)
I updated the above link(LINK) again. It turns out the followSplineTrack() function of the Bot class can crash the game if unknown(for me) circumstances are met.
I used this function extensively; upon completing a lap all tournament opponents chose a new spline to follow if there was more than one defined, this sometimes resulted in silent crashes during tournament races. This error in the engine can still crash the game on tournament start but it is far too rare to properly debug and fix in the engine so this will remain a "half-workaround"...
Wouldnt you be able to let it pick two splines with a random seed and get midway from them? This would kind of make "infinite" variants of the same driveby
This followSplineTrack() call boils down to a command on a native object of type sl:0x0000006Er which refers to the native bot class defined in the exe, this is the source of the crash; the game processing the "AI_spline" command on a bot so until someone finds out what are the special circumstances that make the game crash I think the best is to avoid issuing more than the absolute minimum of AI_spline commands to the engine.
On the other hand if and when this gets fixed your idea is totally sound and would be interesting to see the randomised routes the cars take on a track sometimes using a shortcut othertimes not sometimes messing up a curve and the like, would surely make the races more lifelike.
On Valo City, will the AI re-route if they are knocked off course, or will they try and follow their original path? I have to wonder because I am considering trying to make the
as a working SLRR map.Well Valocity is different from other tracks I noticed that even if you drop an AI racer somewhere far from his intended spline it will navigate the track relatively competently meaning it wont collide with buildings and will try to avoid traffic, and will eventually get on-track and follow that, nevertheless if the spline goes through a building the car will try to do the same. This is probably done by utilising additional data telling the cars where the roads are (this should be the city.trc).
On regular tracks the AI will calculate the nearest point on the spline and full throttle straight to that point and after reaching it follows the spline as usual. There is another thing worth mentioning the AI does not care for the order of the control points they do not act like checkpoints; at a point where the spline is self-intersecting the cars momentum will decide which "branch" will be taken (thou that usually will be the logical one) if you have two segments running parallel to each other but with opposite normals (like a very narrow oval track) it can easily happen that the AI chooses the wrong branch turns around and simply ignore the other.
However you can use this behaviour as a "feature" aswell, in one of the races in SlrrExhaustive there is a jump over a fence that not all ai-cars can make on the first try mostly because they interfere with eachother, the ones that cant make it crash to the fence but I made a loop in the spline so the ones not making the jump will in time inch there way nearer to the other "branch" and will chose that to follow which will loop them around to try the jump again. The point here is that the loop(a part of the otherwise continuous spline) is not taken by all the cars (ignored by the ones that jump over and closer to a "later" part of the same spline).
Added 38 minutes later:
I watched the video only now realised that its a link
This is possible in Slrr without decoding the trc file format but would need the script of the track to manage the "traffic" cars only keeping in the game the ones near the player but they would need to be full AI racers "sitting" in traffic cars and that would mean numerous AI_spline commands which is unstable so it would be better with a .trc and regular traffic code in the background.
