Joined November 2015
- Tweets 2,677
- Following 0
- Followers 228
- Likes 5
209 Photos and videos
Jun 11
I mean I can basically use the post as a script more or less. The graphics would be animated of course.
1
59
Jun 11
Remembered that at small scales, parabolas and circles approximate each other, so this isn't really a concern.
(Did you know that if you throw something, it doesn't follow a parabola but a very stretched out circle? Orbits are elliptical, not parabolic.)
Jun 8
We can trace along a parabola, that is a mostly solved problem. What is not quit solved yet, would be to do actual circle segment traces. Parabolas would perfectly represent motion like in jumps, but I am not sure it applies well to tickless airstrafing, for example.
1
135
Jun 11
What I am saying is this: It doesn't matter much because we are still dealing with small timesteps. Yes, there will be some error, but the error will be pretty mild and chasing perfection is unreasonable.
79
Jun 8
Collision traces in games are usually done along a line.
This puts a constraint on the timestep length that is actually usable. Collision in 16 tick would feel fucked up, even if head motion was using hermite- instead of linear interpolation (like in CS:GO).
1
1
196
Jun 8
We can trace along a parabola, that is a mostly solved problem. What is not quit solved yet, would be to do actual circle segment traces. Parabolas would perfectly represent motion like in jumps, but I am not sure it applies well to tickless airstrafing, for example.
1
1
219
Jun 1
Its interesting to me how much hate Windows 11 is getting.
I do have occasional issues with it, but nothing major. It just kinda works?
I will eventually go to Linux, for many reasons, but I don't feel pressured to do so at this point, because, again, it works
1
1
157
May 30
So apparently hooking client functions with S2 Metamod is possible.
So...I can make...tickless subticked airstrafing in CS2?
1
1
192
May 28
And what if he was wrong?
(realised something while doing the surf stuff, not that it matters basically anywhere)
4 Nov 2023
Are you referring to the case where you jump and bump your head during the jump?
That case is indeed a little inconsistent. Is there a case where that is relevant to gameplay?
It could be made consistent but it would cost fps that I suspect users would rather have.
3
280
May 25
This graph makes nonsensical assumptions.
It assumes >1 engine lost = mission failure.
It also assumes that engine fratricide leads to mission failure (a given, since >1 engine lost = mission failure).
May 23
The math looks like this.
You want a rocket that captures the positive of being able to handle a single engine failure.
Without having too many engines so it fails on propagating failure risks.
7/
1
2
369
May 25
The values on the graph match 0.99^n (probability of one engine lost) * 0.9. The reality is that even if we just assume we can lose 15% of engines, we could lose 5 engines and still make it. For 33 engines that makes the success probability go to 96.7% instead of 93.3%.
1
1
95
May 25
And the fratricide = mission failure assumption is probably wrong. Assume for a second that one engine, on average, takes out 0.5 more engines. In this case, you have exponential decay on failed engines and that means more engines make your success probability approach 1.
76
May 23
Smaller tweaks to tickless surf possibly playing a bit of surf locally? (depends on how easy it will be to set up the plugins)
twitch.tv/H7pernerd
144
May 22
It also showed me again, how dank this shit is.
The surfing acceleration is so comedic that you can't really play surf like before.
I think I should work on a way to dynamically limit acceleration while still keeping the tickrate independence.
1
1
125