wowgetoffyourcellphone Posted February 20, 2017 Report Share Posted February 20, 2017 2 Quote Link to comment Share on other sites More sharing options...
jeffnz Posted February 20, 2017 Report Share Posted February 20, 2017 It doesn't say much about pathfinding though. Quote Link to comment Share on other sites More sharing options...
Loki1950 Posted February 20, 2017 Report Share Posted February 20, 2017 The whole video is about pathfinding for massive armies and they don't go into to much detail so they can maintain a commercial advantage it's not an Open Source project Enjoy the Choice 3 Quote Link to comment Share on other sites More sharing options...
wowgetoffyourcellphone Posted February 21, 2017 Author Report Share Posted February 21, 2017 Thread doesn't have to be about 1 video, guy. Quote Link to comment Share on other sites More sharing options...
sphyrth Posted February 21, 2017 Report Share Posted February 21, 2017 The guy in the video said that each unit uses its own unique pathfinding. They don't do group paths if I understand it correctly. His previous pathfinding algorithm was "From point A to point B."... doesn't pyrogenesis have long and short pathfinders? Some commenters said that it's fake... claiming that the video was only the sped up version of the original. Only you guys can say something about that... but the arrows did seem too fast for real-life. Quote Link to comment Share on other sites More sharing options...
sphyrth Posted February 21, 2017 Report Share Posted February 21, 2017 Here's a little quote from a little search that I happen to notice with 0 A.D.'s pathfinding (Alpha 20 at least): Quote A part of many RTS games is unexplored terrain. Since you can't see the terrain, your units shouldn't know where to walk either, but often they do anyway. (emphasis mine) 2 Quote Link to comment Share on other sites More sharing options...
jeffnz Posted February 23, 2017 Report Share Posted February 23, 2017 These videos are less "showy". There's no need to show off graphics and gameplay when demonstrating pathfinding See also, Retro Ahoy on Youtube and his video about Pacman's ghost ai https://youtu.be/15dxuAbTC0A?t=7m48s 3 Quote Link to comment Share on other sites More sharing options...
Wesley Posted March 13, 2017 Report Share Posted March 13, 2017 (edited) Combine that Kynapse AI solution with some grouped units and a solution for the trade wagon conga line problem and we have a game. Edited March 13, 2017 by Wesley Quote Link to comment Share on other sites More sharing options...
Radagast. Posted March 26, 2017 Report Share Posted March 26, 2017 (edited) mhm. From this video nothing is evident. Even if not using group pathfinding it could still cache the found path of one unit and reuse it (at least for the long pathfinder that is how it looks like). (No information about grid resolution is given either.) Also the enemy units are fully static. (Apart from that it's a pure murder video without much value.) With many moving obstacles this'd get interesting. Unfortunately the only moving entities in the video are the attackers while the enemy is clustered at static positions (every officer of course would be very happy with such soldiers dying with so much discipline, moving not even one inch ;-)). Sad no dynamic obstacles are placed either. (in contrast to the video of @jeffnz, such debugging info would be helpful in the video of the giant storming of the static defender castle) Actually a good example of a pathfinder is already within pyrogenesis (despite maybe some bottlenecks at certain circumstances). Drunk on Nectar has a versatile pathfinding for flying units but even that one may require up to 3s (timeout) per unit per tic (of course a tic is way shorter, it's just a time out). Making that work with pyrogenesis' supported amount of moving entities is hardcore. (Yet the dynamic collision avoidance is fascinating and has many applications in the real world open source projects (where often there is just < two bot per computer whereas pyrogenesis' pathfinder deals with the thousands of entities within one part of one thread's time)!) Edited March 26, 2017 by Radagast. 3 Quote Link to comment Share on other sites More sharing options...
Palaxin Posted May 20, 2017 Report Share Posted May 20, 2017 On 2/20/2017 at 8:45 AM, wowgetoffyourcellphone said: Quote Link to comment Share on other sites More sharing options...
jonbaer Posted May 22, 2017 Report Share Posted May 22, 2017 While researching IDA* I came across this small project ... https://qiao.github.io/PathFinding.js/visual/ 1 Quote Link to comment Share on other sites More sharing options...
vladislavbelov Posted May 27, 2017 Report Share Posted May 27, 2017 (edited) On 22.05.2017 at 3:58 AM, jonbaer said: While researching IDA* I came across this small project ... https://qiao.github.io/PathFinding.js/visual/ If I'm not mistaken, it has wrong the breadth-first search implementation, it should be mostly equal to Dijkstra. Also Dijkstra is useless for this kind of graphs. About topic video, it could use some types of "formation fluid", when an algorithm doesn't control each unit, but all together, like sticky substance. Edited May 27, 2017 by vladislavbelov 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.