Thanks, but I'm terrible at math, well except vectors and matrices. I looked around a few places, but people seem to say it very complicated to do... Also the mathlib doesn't seem to handle Bezier curves. Anyways, if anybody has any idea how to simplify all this, feel free to share it.
On Fri, Jul 22, 2011 at 7:46 PM, Marek Sieradzki <marek.sierad...@gmail.com>wrote: > On Sat, Jul 23, 2011 at 1:30 AM, Psy_Commando <psycomma...@gmail.com> > wrote: > > I've been trying to get my entity to follow a path made of path tracks, > but > > the movement is either too stiff, or "bouncy" when reaching a node and > > heading on the next path. My goal is to simulate a fixed wing flying > > aircraft, not very accurately but just enough so its convincing. > > I'm using the code from the physflyer npcs to compute my velocity. But > those > > just float around, and strafe left or right on sharp turns, while I'd > want > > it to roll towards the turn, pull up, and turn less abruptly. > > > > This far I've considered a few options: > > - Building Bezier paths in-between each sets of 3 nodes, using one point > as > > a control point. However, I have no ideas how to compute the closest > point > > to the Bezier curve. And the junction of each group of 3 nodes would > still > > probably be very sharp... > > Can't you just solve 2 neighbor Beziers analytically to get equal > (0th,1st,2nd...) derivatives? I wrote some bits that did so like 3 > years ago but I remember it wasn't that hard. Someone must have solved > it in some library or even inside Source's mathlib. I have the > equation for getting B-spline with equal derivatives but it's long and > book is in Polish. :P > > > - Following less rigorously each nodes, and requiring the plane entity to > > actually need to point in the direction it wants to go. But then I'd need > to > > add in some lift and drag effects, if I want the turning to work > properly. > > > > Any suggestions ? > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > >
_______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
