Crashing Cars

[MeanMichael]

Ok, i just read the guidelines and hopefully this fits them.
Somewhere I saw that a Dev said that car driving is a long ways away. but when it happens one of the first things you should be able to do is run your masterpiece into a wall. I think that safety testing is a engineering concern in real life so safety testing in game should be looked at.

whoever looks at this thanks for looking, And any nonsensical thing i just said blame it on me being 14.

[Daffyflyer]

Safety ratings for cars will be a thing, but there won't be any actual visible crash tests, that one is REALLY hard

[Kubastach]

Safety ratings for cars will be a thing, but there won't be any actual visible crash tests, that one is REALLY hard

Actualy making realistic (in some sort) damage model isnt that hard, c'mon even carmageddon got realistic model. i made one in about 3 hours,there is special program to this (i didnt remember its name), you import car model, then mark areas that should be able to be damaged (eg. doors, mask etc.), choose durrablity previev effect and program generate damage model (file) and code that you need to input in game. Its very easy

[Killrob]

Yeah just as easy as making diesels for the game, and making I5 engines when you already have I4 engines. I mean come on, it's just increasing a number by one, right?

[adamd]

you import car model, then mark areas that should be able to be damaged (eg. doors, mask etc.), choose durrablity previev effect and program generate damage model (file) and code that you need to input in game. Its very easy

I'm no expert, but even assuming that your program gave a damage model that was usable, and assuming it was realistic and could account for all situations, you're still only halfway there. It would have to tie in with a real-time physics engine that had proper collision detection, and could account for every possible scenario of impact angle and speed (even though no 2 real life crashes are ever identical, in a computer simulation, same inputs should yield same results). I spent a few weeks this year trying to create a basic real-time physics engine, with just basic collision detection on spheres and cubes with reaction forces. When the months were up I had some balls and boxes that bounced off each other. That's it.

Like I say, I'm no expert and have no idea how it would work in Automation, and as Killrob says, it's not as easy as just pressing a button and having it all work. Hopefully that, combined with what I said above, gives you some idea of how big a task it would be.

[Kubastach]

you import car model, then mark areas that should be able to be damaged (eg. doors, mask etc.), choose durrablity previev effect and program generate damage model (file) and code that you need to input in game. Its very easy

I'm no expert, but even assuming that your program gave a damage model that was usable, and assuming it was realistic and could account for all situations, you're still only halfway there. It would have to tie in with a real-time physics engine that had proper collision detection, and could account for every possible scenario of impact angle and speed (even though no 2 real life crashes are ever identical, in a computer simulation, same inputs should yield same results). I spent a few weeks this year trying to create a basic real-time physics engine, with just basic collision detection on spheres and cubes with reaction forces. When the months were up I had some balls and boxes that bounced off each other. That's it.

Like I say, I'm no expert and have no idea how it would work in Automation, and as Killrob says, it's not as easy as just pressing a button and having it all work. Hopefully that, combined with what I said above, gives you some idea of how big a task it would be.

i know, this program make only destruction model, not working without proper engine, i used it in another (this same producers) program to made something i call "crash test simulator", this program let you only decrease time of making model from 5 weeks to about 2-3. you think that i am so stupid that "one click" and destructive "thing" is ready? This program make it just easier. (making damages models for 13 cars took me 3 hours when i already got engine and models).

[adamd]

i know, this program make only destruction model, not working without proper engine, i used it in another (this same producers) program to made something i call "crash test simulator", this program let you only decrease time of making model from 5 weeks to about 2-3. you think that i am so stupid that "one click" and destructive "thing" is ready? This program make it just easier. (making damages models for 13 cars took me 3 hours when i already got engine and models).

I wasn't meaning to make it sound as if I was calling you stupid so I apologise, I was just trying to explain that as it's from another publisher, it could be in an entirely different format, and making it work in Automation would still require many, many hours of work.

[Kubastach]

i know, this program make only destruction model, not working without proper engine, i used it in another (this same producers) program to made something i call "crash test simulator", this program let you only decrease time of making model from 5 weeks to about 2-3. you think that i am so stupid that "one click" and destructive "thing" is ready? This program make it just easier. (making damages models for 13 cars took me 3 hours when i already got engine and models).

I wasn't meaning to make it sound as if I was calling you stupid so I apologise, I was just trying to explain that as it's from another publisher, it could be in an entirely different format, and making it work in Automation would still require many, many hours of work.

Actualy it can create models in like 13 formats, maybe at least one would be proper format

[Daffyflyer]

Just because it's possible, doesn't mean it won't still add a massive set of new problems for us to solve and a whole lot of extra work. We need to pick our priorities to ensure that our tiny team can actually finish a game. I'm sorry that's the case, if we had a much bigger team and budget we might consider it (although even then there would be bigger priorities first)

[Kubastach]

but, maybe some time ?

[Daffyflyer]

Probably not, no.

[MeanMichael]

i hope my new post doesn't start a fight like this...

[rammstein]

i like this idea