From: Sander v. D. <sgv...@gm...> - 2011-03-21 23:40:32
|
Hey all, I have committed the extended rules. The way they work: - 'touch groups' are created. A touch group is a group of agents that touch each other (either directly, or through other agents in the same touch group). These are similar to ODE's 'islands'. - If an agent is in a touch group with more than 2 agents (including himself), and he wasn't in a group the timestep before, which means he is the last to join the group, he is relocated outside of the field. - If it is not clear which agent joined the group last, e.g. when 3 players were all separate at time t, but were all touching each other in time t + 1, an agent is chosen at random. This seems to work well enough, though two things I am not sure about: - Relocating the agents completely outside of the field fits the other rules, may be a bit harsh? Perhaps just beaming them FreeKickDist meters away would be enough. Although that could be a beneficial/defensive position. - The way it currently works, agents will not be relocated when several touch groups join, e.g there are 2 touch groups of two, that in the next timestep form a touch group of 4. However, I am not able to produce such a situation and think it will be really unlikely. Also, I don't have a good intuition which agent to throw away in such a case anyway. Sander On Mon, Mar 21, 2011 at 3:40 PM, Hedayat Vatankhah <hed...@gm...>wrote: > Hi Sander, > Thank you for the reminder :P. I'll try to do some tests in the mean time. > > Thanks > Hedayat > > *Sander van Dijk <sgv...@gm...> <sgv...@gm...>* wrote on > 03/21/2011 8:15:30 AM +0350: > > Hey, to keep you posted, I am implementing a new touching rule as we have > discussed: if 2 are touching and a third joins, he is moved away. I think I > have a working setup for this, just fighting with boost 'smart' pointers > (apparently smarter than I am). Hope to have a working version in the > repository tomorrow. > > Sander > > On Sat, Mar 19, 2011 at 6:41 AM, Hedayat Vatankhah <hed...@gm...>wrote: > >> Hi all, >> We should release a new version of simspark/rcssserver3d very soon. Please >> check the latest SVN and let us know if there are any problems. It'd be nice >> if you can test with the multi-threaded ODE prepared by Sander. >> >> Thanks >> Hedayat >> >> >> ------------------------------------------------------------------------------ >> Colocation vs. Managed Hosting >> A question and answer guide to determining the best fit >> for your organization - today and in the future. >> http://p.sf.net/sfu/internap-sfd2d >> _______________________________________________ >> Simspark Generic Physical MAS Simulator >> simspark-devel mailing list >> sim...@li... >> https://lists.sourceforge.net/lists/listinfo/simspark-devel >> >> > > > -- > Adaptive Systems Research Group > Department of Computer Science > University of Hertfordshire > United Kingdom > > -- Adaptive Systems Research Group Department of Computer Science University of Hertfordshire United Kingdom |