openmultiplayer
Incorrect vehicle model size in GetVehicleModelInfo
Describe the bug If you use GetVehicleModelInfo with VEHICLE_MODEL_INFO_SIZE you will obtain for most of vanilla vehicle models that size is invalid. Moreover: there is also problem that Axis Aligned Boundary Box is not centered at vehicle position. So it would be better if GetVehicleModelInfo can give center of boundary in addition to correct sizes. You can watch this video that demonstrate incorrect boundary: the white-black spheres are boundary by GetVehicleModelInfo, the red ones - by my parsing results (you can find it there. NOTE: there are only hull/chassis boundary (without wheels, damageable & moving parts).
To Reproduce
- GetVehicleModelInfo(modelid, VEHICLE_MODEL_INFO_SIZE, size_x, size_y, size_z);
- Make min & max for every axis with min_x = -0.5 * size_x; max_x = 0.5 * size_x;
- Attach some objects at corners.
- Compare with visible object.
Screenshots and/or logs Video.
Release version This problem begins in samp and still appears in omp up to v1.1.0.2612.
What determines what "correct sizes" are? Seems to me they can be added as a separate function since you're just getting different data - the collision bounding box - while the samp ones take in account mirrors etc
What is the definition of "( vehicle size )"? What is the definition of "model info size"?
Why "vehicle size" includes very long water waves behind? Why "vehicle size" includes 1-2 meters of nothing under bottom of chassis (for example: Enforcer, Securicar, Bus)? Why "vehicle size" includes 3 meters of nothing behind truck trailers? Why "vehicle size" for some models includes wheels and others not? Why "vehicle size" includes for Packer ~3 meters of nothing in front of engine?
Should it named as size of whole .dff model?
For starters, this data is what samp has been using for over a decade. So changing that is not possible since it can (and will) break people's scripts. But to ensure data is invalid, can you share your test script?
For newcomers kalkor said that he never fix something that can be fixed with scripts.
Im not so sure someone using cases of invalid data. Here is your script to observe every model fuckup.
There is also bad wheels position for forklift
Also you can find out that Cargobob have valid wheels position Another plane that have wheels position is RC Baron ... In the same time most of planes and helicopters that have wheels (can only react to surface & cant be hidden) there are no data at all about wheels (models: 417 (Leviathan), 425 (Hunter), 511 (Beagle), 512 (Cropduster), 513 (Stuntplane), 563 (Raindance), 593 (Dodo)).
P.s.: Hideable wheels more complex case.