> One thing that would be great would be a gradient algorithm that supports > nydus channel, i.e. some teleportation points. *That* would affect gameplay.
I wrote this a while ago: I think teleporters would only work as buildings that handle unit allocation. The alternatives: 1.One field size transporters: They will be bottlenecks to probably popular places. You can simulate this by separating to parts of a map with a solid wall which has only one gap (1 field wide) 2.Many field size transporters: Globs will run by in too use it as a short cut from one edge of the building to another and therefore go weird paths. That is unless the pathfinding uses a penalty for transporter paths. I forgot the reason why I don't like this. 3.One field size transporter which allows units to walk on top of each other in its proximity to overcome the bottleneck. But when I read Leos and Bradleys mails, I feel that transporters and tunnels aren't such a good idea after all. We really should try to make glob2 more based on local actions. Transporters and tunnels are counter productive unless we overdue work that enhances local issues. One quick idea: Starcraft-Zerk/Warcraft-Undead-style areas. The user has to place a flag or something like that and can construct buildings only near to that spot. The range can be increased with upgrades. The general pathfinding and unit allocation is done for each "flag-regions" individually. Exceptions are a few units/buildings that have to have global pathfinding/allocation for gaming reasons. Pathfinding between those "flag-regions" can be more expansive. We could limit the number of those flags and upgrades (each upgrade counts as an other flag). Then they user has to decide whether he wants big towns, or many towns (or intermediates or mixtures). -- Kai Antweiler _______________________________________________ glob2-devel mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/glob2-devel
