I know I said in the past I'd make Ninkeo Kart my final game in 3DRAD, and look how that turned out... I never finished it because of a lack of motivation with that particular idea. I did have a unity prototype going but I reckon it doesn't have that same 'feel' the old 3drad versions did.

BUT, I did learn one hell of a lot about the scripting engine and how it interfaces with objects. I feel I'm ready to take the next step on my games to make them mostly scripted instead of dnd objects.

So I'm here to ask you guys, what would make a PERFECT racing game made in RAD? Like something you'll be proud to show your friends.

From my research, these are the most popular things / most requested items that I see come up the most often:
- Timer (including lap and split times)
- Minimap OR progress bar (similar to rally games)
- Decent and not totally brain dead AI
- Somewhat decent graphics for RAD standards (including GUI design)
- Decent track selection
- Decent car selection

To help you come up with some ideas, the theme I've decided on for this game is junker style. Basically rat rods and mad max like cars going round a track. Track being ambiguous, it could be a normal racetrack or even just a narrow goat track dirt road.
For better or worse, I aim to re-use some of the new ninkeo kart tracks I never used, there's quite a few of them I think would be great re-skinned and tuned slightly.

Looking forward to what you guys reckon

I think once you get a video or demo to see or play with, then maybe some ideas will form.. right now i think you've got it covered, more specific game ideas would be more related to what you have designed already.

what about adding power ups? like not necessarily making it into a mario kart game but putting speed boosts and weapons here and there to make things more varied since this is a mad max kinda game
hell, maybe even do them randomized to add replayability to the tracks
you do you man, i've been trying to make a 3d rad racing game for a long ass time, but i just dont feel confortable with how the pc cars feel

Another idea to add onto that is to have hidden traps around the track, whether they activate or not is totally random. That could be quite easily done too.

Yeah car handling isn't something that can really be done properly without a script, since the pcar seems to benefit the most from dynamic updates (soften the suspension going around a bend, nitro boosts etc)

In my opinion, GUI is at least 40% of graphics, maybe more. I take extra care now with GUi, make it as professional as I can but at the same time make it fit with the game. It's actually quite hard to get it right straight away.

I'd say, give the player a sense of urgency, get their heart going, let them know that if they don't do X within X time, s**t will go down

I'll be honest with you @ghgames, GUI design isn't really my strongpoint at all. If it's harder than ms paint I'm pretty much useless at it hahaha.

*whispers* neither is car design but I've got a buddy that helps me out

Based on my experience, here are the stuff I would suggest you to go for...

Go with low-poly style for graphics as these are the only ones that require little to no modelling skill and have low performance cost. 3D Rad doesn't give you the tools to make your graphics look good. You always have to get away with workarounds in order to get something that looks acceptable. For example, what I used to do is split my track models into 3 - 4 skinmeshes, so I can use a different shader on each. Bad idea as it definetly doesn't help 3D Rad's engine.

You definetly should spend some time on the UI. If you go with low-poly, UI wouldn't be such a problem for you. If you have enough knowledge in MS paint, then it's pretty much gonna work . Just google "simplistic UI" or "minimalist UI" to get some ideas. They pretty much work in low-poly.

@nuskadamo, I have tried making a racing game with powerups. It does work, but as always: Performance sucks. This is due to inefficient usage of objects. It's just how 3D Rad is built. If you want to have each racer have 4 powerups (assuming that they require 4 objects), then for 6 racers you already have 24 objects. And ignoring performance cost, this is ridiculous to manage. Every time there is a bug, you have to replicate your code on 6 different places.

Decent car selection is somewhat possible. I recommend using the iObjectRefresh function instead of having each car have its own skinmesh. But decent track selection... Oh, boy...

Decent track selection requires you to multiply the already stacked replication that you have to do by the number of tracks you're gonna have. Simply because each track has to be its own project. Multiply that by the ridiculously slow loading of 3D Rad, and you're already frustrated enough to quit forever. What you can try is again utilize the iObjectRefresh for changing the track. Not sure how well it'll do for rigidbodies though.

For the AI, well... Not totally braindead, but remember that you're only exposed to a limited amount of external info. For example, it's not very possible to improve the AI's avoidance, because it would require 4 different scanner objects for each car.

Minimap is quite a simple one, but you can't mask parts of it. I assume it would be just fine though.

In conclusion... Question is: do you really want to create a racing game in 3D Rad? Are you willing to go through the frustration? If not, just don't waste your time and use Unity

@powersupersport I would have used Unity, though I just wanted to give the old girl one last go before moving onto it for all my future projects. Who knows, I might showcase this old engine on the Knockout forums and get some guys interested in pushing it to the limit again.

@nuskadamo 3DRad's really best suited to simple arcade style racers, as that's where most of the objects are aimed towards. The difference is whether you use the stock assets (eg. the buggy and the highway kit) or go all out and model everything from scratch.

@drninkeo, definetly would be interesting to push the engine to the limits, but certain games are just outrageous to make. Games with repeating entities are just a burden to make due to no possible duplication of objects runtime. The closest you can get is using imposters, which can move or do stuff with the help of a script, but they are incredibly linear, and cars cannot be imposters.

If you remember my project Geometry Dash replica http://3drad.boards.net/thread/383/geometry-dash-3d-beta-released, I think this is the kind of project that has some potential in 3D Rad. For some odd reason, this is just it. Here's why...

For each block in the level I create some amount of imposters. Then I put these imposters where they have to be. It's just convenient and close to what you'd do in Unity. You'd make your object and you'd make a prefab out of it. Then you'd spawn that wherever needed. Works here, because each block has only two objects that need to be instantiated - a rigidbody for collision and a skinmesh for visuals.

The game worked. It still works and you can play it. The reason why I left it is because apparently 3D Rad has a limit on how many objects can be linked to a script. You can only link up to 90 objects (reach up to OBJ_1980). The rest are ignored. I had a huge script of 2000 lines (not that many for my standard now , but back then it was a lot). So when I split the script in two, being careful to replace the references correctly, the game just didn't work anymore. So I gave up.

This is generally why you probably shouldn't make a game in 3D Rad, or at least not a racing one. After GD I made another project, called Beating Power (https://www.youtube.com/watch?v=st7gc_C_wik), which had music involved. This is also a project that has potential in 3D Rad, because again, the modules that I use are not more than 2 objects per module. Therefore I can create imposters for them and their management will be easier. But for a racing game, you can't even clone the cars to allow for multiple racers. And loading tracks is just impossible without duplicating the project for each track. The management is just too prone to error and ridiculous to fix if a bug occurs.

Anyways, that's just my opinion in the end. If you want to push the limits of 3D Rad and make a decent game, then maybe it shouldn't be a racing game.

Honestly, at the end of the day I totally agree with you @powersupersport, it's kinda hard to recommend using 3DRad for anything beyond basic messing around with. Really, Unity is the next step after finishing with 3DRad. There's nothing wrong with sticking with the old engine until you reach the limits with it, or if your project isn't going to be complex to the point that it starts breaking it at the seams.

You've convinced me lol, I'm going to see how feasable it is to start on Unity with the workflow I have since I still use sketchup for modelling everything atm.

@drninkeo, don't forget I was a huge patriot of 3D Rad. I even made a poll back in the day asking people which is better - 3D Rad or Unity, hating on Unity in a way. That was obviously a terrible and redundant poll in a 3D Rad sided community, but I was 16 back then .

I've explored quite a lot of its features. I even tried making a minecraft replica, which is 100% impossible. I was getting poor framerate with very low render distance and only one layer of ground. Here's a link: http://3drad.boards.net/thread/116/craft-terrain-generation-system-test

So yeah, for 2 years of 3D Rad and maybe 3 years of Unity I can safely say that you should switch to Unity as soon as possible. Transition won't be easy, just own it . But once you get in the basic workflow, you'll love it