Hi Yvonne and all others, I just created a patch to have the navigator reacting as we said, so forward and right return the actual step (I want in the next step not just to refuse the step but walk the maximum possible step until a certain distance (what about groundDistance?) is reached.
There are two new keys in OSGNavigator.h: LEFTROT and RIGHTROT, mapped left and right of the forward (in the OSGSimpleSceneManager.cpp), which makes them 't' and 'u' (by the way, most people will override this anyway, like me with my qwertz keyboard ;-) I place the patch on our university webspace: http://www.cs.uni-magdeburg.de/~dwalz/files/walknavigator.patch And also I want to try to do this intersection with a volume, is there already a method? Thanks and regards, Daniel On Mon, Sep 13, 2004 at 08:22:16PM +0200, Yvonne Jung wrote: > Hi, > > Hi again, > > thanks for the hint. > > > > I want to have a walknavigator with a slightly different behaviour, > first I want it to give a return value in case the person hit the wall > and second I want to move with the function right aside instead of > rotating (or have a different method like "aside(Real32)" or s.th. like > that) the code is the same as in "forward(Real32)" except for that mv and > sv have to be exchanged (and sign I don't know yet). > > > > > just check out a version from one year ago in order to get the > [walk|fly]navigator moving instead of rotating. > But in most cases it seemed to be quite unhandy to me to have a walking > mode that only allows walking around like a crab :-) > > > > You are right, it might be restricting to have only the possibility to > move into a direction you don't see... > but to rotate you also use the mouse already, so this is redundant in the > user interface, which is not as bad as it sounds in my opinion, because > you might have situations, where you don't want to grab the mouse or the > other way round. > > I don't want to make this big, because I supouse you had some discussions > already, when you changed this... > > But what about having just two keys more, like: > ROTLEFT | WALK | ROTRIGHT > LEFT | BACK | RIGHT ? > > > Ok, that sounds better than only having this or that, perhaps PgUp and > PgDown? > > And what about the second question, are there objections against having a > return value in case of collision or gravity influence? > > > Don't think so, this would only mean to exchange the return type from > void to e.g. int. > > And last but not least, I have a model where the ground is not at one > piece, so there are holes behind walls and it happens some times, that > the person gets in an area, where the gravity check skips the action and > the user can't move back. > In my opinion this should continue at this point, because the movement is > not directed upwards, so there is no reason for doing so, or did I get it > wrong? > > > > Concerning gravity and collision the WalkNavigator is designed quite > simple. The detection of smaller obstacles like bars is almost random, > because momentarily only the old IntersectAction is used which in > addition is not so fast. It would be better to check the geometry > against a timely moved hull of the user. > > Thanks for your opinion, > > regards, Daniel > > > > Regards > Yvonne > ------------------------------------------------------- This SF.Net > email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins > to receive an Apple iPod Mini FREE for your judgement on who ports > your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. > 13. Go here: http://sf.net/ppc_contest.php > _______________________________________________ Opensg-users mailing > list [email protected] > https://lists.sourceforge.net/lists/listinfo/opensg-users ------------------------------------------------------- This SF.Net email is sponsored by: thawte's Crypto Challenge Vl Crack the code and win a Sony DCRHC40 MiniDV Digital Handycam Camcorder. More prizes in the weekly Lunch Hour Challenge. Sign up NOW http://ad.doubleclick.net/clk;10740251;10262165;m _______________________________________________ Opensg-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/opensg-users
