I'm a beamng's entusiast. I hope you are developing a "traffic system" with traffic rules, similar to CITY CAR DRIVING, that will be different from spawning vehicles with A.I. (beamng's roads are too desert). I hope you are developing a "pedestrians system", maybe with dummies (it will be great to put some dummies also into cars as passengers), maybe powered by an engine as EUPHORIA, for more unexpected accidents. I hope you are developing a interstate map with a large "roads system" that will let to make trips longer than single scenarios, as in EURO TRUCK SIMULATOR. I hope you are developing a "first person immersion" that will signify a better simulation of driver's head and body movements into the car and the possibility to leave the car and explore the scenario by feet. I hope you are developing "changing wheather conditions" with changing grip conditions of the road during the same game session and without manual map modding.
Oh look another first-time poster that hasn't read the forums and the FAQ. Even the most powerful of PC's have trouble running 8 vehicles at 60FPS. Imagine 16 cars would be like 20 FPS. The same thing as above. Not possible due to an engine limitation. Cars will shake tremendously even while they are only 20 Kilometers away from center point. Likely not going to happen. May or may not happen.
there is no need to be so negative about everything. the game will eventually need more than a few vehicles at once in order to be function as a proper driving game. the only other solution i can imagine would be creating a rally-game (which doesnt seem very likely considering the content). while it might seem impossible now im sure there will be a solution to the issue at some point in the future. for example having traffic cars with a much less detailed jbeam structure or jbeams that only load in when you get close to them (similar to what rigs of rods did).
I agree only for traffic but two thing can make it impossible 1. Multiplayer (if A day he come) 2. Bad fps
As others have stated, performance is too much of a problem to allow for large-scale traffic. More than 4 vehicles on the road at any point is unlikely for the next year or so (due either to optimizations or Moore's Law). And unlike as stated, vehicles cannot switch to simpler physics skeletons in real time. Moreover, traffic AI is difficult to program in a way that will work as well as a real driver would without making the job of the level/vehicle designer much harder. Pedestrians are an absolute no-go. It would add a violent component to the game, increasing the rating, creating controversy, and attracting all kinds of the wrong crowd. Humans are notoriously difficult to create without making it either creepy or stylized. And, to be interactive, the human characters would need to be given a physics skeleton of their own, which would mean creating a realtime locomotion system. The calculations needed to do so would bog down any computer without having to simulate two+ cars at the same time. As for Euphoria, I would advise reading this: http://blog.wolfire.com/2009/11/why-we-are-not-using-euphoria/ Euphoria isn't some animation method or open-source engine, it's a costly contract-locked service that would at least double the development costs. Traffic is an eventual maybe, pedestrians are a stern no. Some code exists to suggest that some scenarios can take place across multiple maps, using loading zones, though I'm not sure if it's implemented. Otherwise, there is a map currently in development that will be 4x bigger than any official map we've got so far. Accessibility and motion sickness would be too much of a problem. After all, this is a game about cars, not the people driving them. The ability to leave the car has been passively mentioned a handful of times, but I wouldn't count on it. Features have been put in place that should allow dynamic weather in the future, and near as I can tell, it's planned. I doubt that it's at the top of the developers' to do list.
That won't happen... Here's why: - It could cause lag. - They would have to push their age ratings into the red with all that blood and gore. But I can see your point though...
I have not read all of this post but for the pedestrians part and the blood and gore stuff. Just take the Midnight Club 3 route of people just jumping onto walls or dodging you.
Oh look thats one of those guys that is the first one to get an erect*** when a mod releases that has all of those things the OP mentioned.
id love a traffic mod and this i agree 100% its desserted i find it annoying even if one or two car spawn in near the play and make it seem some what alive it doesnt have to be all over the map just one specific area near the player
There are no impossible tasks, just incapable people. If you want to do something bad enough you can do it. As for traffic that doesn't seem nearly as impossible as it was made out to be here. Here's how I'd go about doing it: Traffic cars doesn't have to be simulated with the full beamng engine all the time. They could be imitated by a basic script that moves them around on set routes around the road network. Or a few placed as parked cars around towns. 99.99% of the time this basic script would be sufficient. And only switch to full beam simulation mode, when there is a chance of interaction between the player vehicle and the traffic vehicle. Meaning when it gets within a certain distance of it. The tricky part is making the transition between placeholder physics and full simulation smooth. I agree that is a challenge. But even if initially there is a small lag when this happens I'd still be very happy with it compared to no traffic and life on the maps at all. This way it could be possible to keep track of dozens of vehicles on a map. And if you take out a traffic car by crashing into it, it could be simply removed, after you move away certain distance from it. (Which obviously should be a much larger distance in the 400-1000 feet range) Also if you're concerned because you have a very slow computer, there could be a traffic density slider in the game, with an outright off position even.
I'm not on the physics team for the game, and am not the most code- or physics-literate guy in the world, so take this as the semi-educated words of a fellow player, not as the 100% representative words of a developer. Your suggestion isn't impossible. In fact, there's a name for it in the game development world: physics sleep. Unfortunately, Beam doesn't support physics sleep. Part of the reason why is the need to define whether a node or beam is currently active or not bogs down the physics engine a bit (and when you're trying to squeeze every last calculation out of the hardware, you need to prioritize a LOT). But the bigger issue is that 'waking' an object from sleep will always have to stop the game for probably a second or two. Remember that at the rate that Beam's calculations are performed (2000Hz), you can't just add objects to the simulation at runtime like you can with most 60FPS rigid-body physics options. There'd be no difference between waking a vehicle and spawning one fresh, and that's just an unacceptable gap in a game where the player is moving at these speeds. As for the philosophical part of your post: no, nothing is impossible. But with such a small company, you can only really get away with doing one impossible thing, and that impossible thing is the physics engine itself.
