Page 2 of 2

Re: Is there a way to contact Urre?

Posted: Thu Feb 07, 2019 11:02 pm
by toneddu2000
Mail sent!

Re: Is there a way to contact Urre?

Posted: Thu Feb 07, 2019 11:19 pm
by Urre
I gave what you sent a quick look, and I'd recommend you use the car game source as your base instead, since it is more up to date and has more features. Mainly it has support for directional friction, which was required to make the car. The car is 6 particles, 4 for the wheels, 1 for bomb and 1 for car roof. The wheels have directional friction, which is what makes the car able to turn when the wheels turn.

Re: Is there a way to contact Urre?

Posted: Fri Feb 08, 2019 9:03 am
by toneddu2000
Ah, so Strapbomb car code was practically Twig plus car physics? Didn't know that! Well, in this case, I'll use the code you already released. Thanks a lot Urre!
Can I ask you where did you learn math for physics? Do you have some links with good documentation to share?

Re: Is there a way to contact Urre?

Posted: Fri Feb 08, 2019 11:02 am
by Urre
Math is my achilles heel when it comes to gamedev actually. All the more sophisticated math in Twig is by LadyHavoc, I can only use premade functions because I know what they do, but not how they work. Like if there's a matrix multiply function available, or quaternion related functions, I can use those to achieve what I need, but I cannot figure out how they work. It's mostly because you do get by as a gameplay programmer with only using premade functions + vector math, so I've rarely been hindered too much in my day to day work by this shortcoming. Another thing I don't understand is why certain approaches to physics integration are more stable than others. I sort of get why basic euler doesn't work, but I don't get why euler works when you do half steps like in Twig (courtesy of LadyHavoc ofcourse). I write a new physics engine every year or so, based on experience I've gathered about what works and doesn't, not actual math proficiency, which means I'm always a bit behind others who write physics engines. I basically look at Erin Catto's presentations, and do what he says he did two years ago or whatever :D

The list at the bottom of this document has some good resources: https://docs.google.com/document/d/14Iw ... rmr4Q/edit

Re: Is there a way to contact Urre?

Posted: Sat Feb 09, 2019 3:37 pm
by toneddu2000
LOL! Well, this makes me feel not alone! :biggrin: When I write and read physics code I feel like a boy tinkering with adult stuff..

Thanks for the lecture, Urre, I'll definately take a look!