I just had this interesting thought. What if you could make your own cars from the endless amount of parts there are collectively from all the vehicles, including mods? You would choose the frame, wheels, body, steering wheel, etc. What do you guys think? How else could you see this working?
Every Car would have to be re-jbeamed to a universal standard, but it can be done. Heck, to a smaller extent I did it with one of my mods
I can see it working, but the problem occurs when you realize that all the parts are designed to specifically fit on one car in one spot, when you start mixing them, there will be overlaps and empty spots...
Well, it will work, but it won't be pretty, what if I put the hood of a T Setires on a Barstow? Or a D Series cab on a pigeon? It couldn't work because of... 1. Overlapping materials 2. Not fitting right 3. Unable to have part placed in car Look at what happend with the off-road bumper and the bull bar, that is what would happen when people decide to screw around with it
Parts such as engine, wheels, drivetrain, aero kits and sometimes suspension can be swapped but i takes effort because they all have to be connected using beams otherwise they will just fall off on spawn which has happened to me countless times. --- Post updated --- Baiscally anything besides panels can be swapped unless you take the time and effort to make it fit. Eg daniel398767 and shelb ye (sorry i forget his name but you should know who i mean.
One time i wanted to start modeling a car for beamng. Idea was to get the model to the standards of the early 2000s like a car from GTA 3. Would have been more realistic to actually get something done as a complete starter. Plus it could have been a bless for everyone with weak hardware. But i had to admit to myself that i had not the time to start another hobby. So i never begun working on it. Point of this post is with a lowpoly car in the fashion of GTA 3 it is the most realistic approach for an open plattform with a lot of interchangeable parts because of its simplicity compared to todays standards. Also you'll forgive some oddities more likely on a comicy lowpoly model. Most of GTAs cars of this era had a chassis, doors, front fenders, hood, trunk, bumpers and wheels without modeled suspension. When we addionally seperate the engine, back fenders, the roof and the pillars from the chassis we're getting somewhere. But there are still a lot of problems to solve. A single chassis cant work for every class. So we'll need chassis for compacts, sedans, wagons, pickups, suvs... Also it will never be possible to have both huge variety and great interchangebility with most of the parts. When we start to strictly standarize the transitions of the different bodyparts( eg hood to a-pillar to roof etc) for interchangebility the car wouldnt look very distinct with different trims. With a loose standardization most of the parts wouldnt be compatible with each other. In my head it just works out great. Having 5 to 10 different chassis which are having a lot of different body parts to choose from. addionally different simplified suspensionsystems with standarized suspensionpoints to be interchangeable with the different bodys and powertrains. Of course only jbeamed with no visible model to keep it simple. Could be a great startingpoint for everybody interested in carmodding. Modeling and jbeaming a lowpoly fender would be much easier and is keeping motivated because you'll have a not to distant goal. Also there would be a lot of simple learning material. But who really wants to have cars in beamng which look like from the early 2000s? Who is going to determine the standards? Plus everything written here are uneducated guesses from myself so i'm not really aware of a lot of other problems.
It's certainly a very interesting idea. Like a kit car, only probably more difficult to conceive and put into practice. If you could get the parts to automatically "glue" themselves to the appropriate area, that would eliminate a lot of the blank spots and poorly fitted parts, but the jbeam might be compromised at that point. I honestly have no idea. I was just thinking of Little Big Planet at the time and their mechanics of parts-fitting.
The idea of "individually" jbeamed parts isn't always accurate at all. The parts in the game are basically jbeamed in convenient or workable ways - ways that are not always compatible with a reality-based perception of "parts." Most jbeams are also done with a lot of absolute assumptions going on... eg with Part A assuming that Part B will be at exactly XYZ coordinates. Lots of 'parts' are actually big lumps of parts - and they get jbeamed as a lump. The game also lacks more basic parts systems at the moment, so this suggestion is way out of reach IMO. nodeOffset can't be cumulative. That's why the pickup and the extended pickup have different rear suspensions. "props" such as gauge needles, steering wheels, pedals, etc are not affected at all by nodeOffset. (spotlight props are though) Those two are fixable and I think it's reasonable to assume that we'll see improvements in the future. Another big example of the limitations are the wheels and tires (and wheel spacers). The game simply does not have an understanding of parts - everything is with these items is done in a painstakingly manual way. That's why in the base game you'll often find hubs, wheels, and tires which physically would fit together, but the game does not allow you to combine them. Offsetting the nodes, meshes, etc.. that's done mostly by hand. For that reason lots of stuff ends up left out. I think that it's conceivable that we could see some improvements in certain areas such as wheels, but nothing as broad as what you describe. There wouldn't be much point without procedurally generated parts. The bug shield from the D-Series won't fit the T-Series - either visually (mesh) or the skeleton (jbeam). Same goes for other stuff.
Too bad it isnt possible at the moment. But i really appreciate your write-up. It is always good to know why exactly an idea isnt going to work out.
Are you sure, im pretty sure you could scale the mesh if need to make it fit the t-series hood. As well as relocated the mesh but as well as this you need to move the jbeam so that the movement of the actual jbeam doesnt bug out the mesh. Such as if the mesh 1 m away from the jbeam structure it is allocated to and the jbeam pivots slightly from going over a bump the mesh will flail flail around becuase it movement is made much bigger becuase of the distance away from the Jbeam
While I don't think it's very germane to the discussion... Yes, I'm sure. They don't even use the same attachment method. Take a close look at both. Hypothetically if they did, I doubt that a simple scaling operation would get the angles to line up correctly. If the scaling operation would line things up correctly... how would you parametrically determine the correct scaling operation to apply to each accessory? The vehicle width alone certainly wouldn't produce the correct results. How would the game know which X, Y, and Z scaling(s) to apply? Putting all that aside - while it's true that you'd have to move both the mesh and the jbeam (which is what nodeOffset does)... the effects of not moving both would not be what you think. Moving a mesh completely off of the nodegroups to which it is assigned will result in the mesh failing to spawn (inability to map to any nodes).
