That was also my first guess, and it makes me quite annoyed at his claim that "no other game that he is aware of does it this way", because I'm fairly certain that there are plenty of situations where closed-form particles are also the more performant option.
As the same time, to be fair, plenty of people do not realize closed form formulas are the best approach here. I'd say that requires a level of intuition for the physics and maths involved that is not strictly necessary to make functioning particles the "regular" way.
Anyway, to switch the topic to something more fun: do you think one could make Age of Empires 1 and 2 perfectly rewindable?
Not the OP but I would guess it's not that it was unique or surprising, but that it was didactically/pedagogically nicely explained. Experienced devs might find it superfluous since they've already guessed the solution from as soon as the problem is stated, but this is something I could share with a CS 100 student and they'd get a lot out of it.
Right. I also doubt this approach is that unique since closed form computation is also a desirable property for simulating the particles entirely on the GPU.
Comments
To save you a click: “ However, the particles in Braid don't affect gameplay in any way; they're just cosmetic. ” so he used deterministic functions.
So pretty much exactly what you would expect someone to do.
But he got to get his clicks somehow!
That was also my first guess, and it makes me quite annoyed at his claim that "no other game that he is aware of does it this way", because I'm fairly certain that there are plenty of situations where closed-form particles are also the more performant option.
As the same time, to be fair, plenty of people do not realize closed form formulas are the best approach here. I'd say that requires a level of intuition for the physics and maths involved that is not strictly necessary to make functioning particles the "regular" way.
Anyway, to switch the topic to something more fun: do you think one could make Age of Empires 1 and 2 perfectly rewindable?
[0] https://www.gamedeveloper.com/programming/1500-archers-on-a-...
I preferred his own explanation, to be perfectly honest.
What did you find unique or surprising about it?
Not trying to start an argument, just good to know.
Not the OP but I would guess it's not that it was unique or surprising, but that it was didactically/pedagogically nicely explained. Experienced devs might find it superfluous since they've already guessed the solution from as soon as the problem is stated, but this is something I could share with a CS 100 student and they'd get a lot out of it.
My initial thought was recording the position in each frame.
Obviously after reading it, it seems obvious. And if I were implementing it, I am pretty sure I would have came to the same solution.
But it was not immediately obvious to me from the title.
Yeah it's a bit obvious, isn't it? It's the first thing I thought before I clicked through.
Well Braid came out in like 2008, and it was not obvious to me as a noob game dev when I first read it.
... It does seem real obvious _now_
Right. I also doubt this approach is that unique since closed form computation is also a desirable property for simulating the particles entirely on the GPU.