Scipio Posted April 20, 2008 Share Posted April 20, 2008 Seen today in a PBEM turn file. My infantry squad is in building A, I order it to quick move to building B. The building A has doors to the front and rear. Since I know an enemy platoon in a building in the upper right corner, I set a waypoint to the rear door (yellow line). What happend (red line): the squad leaves the house to the rear and splits up at the waypoint - the half squad runs to left as ordered, the other half runs to a building to the right and directly into enemy fire, causing 4 casualties and uncovered my movement. I made a test scenario with the map and units to figure out what went wrong! If I would have moved the waypoint a bit more to the road and to the left, that the movement would have been exectuted well, because parts of my squad wouldn't have moved that far to the right. But if I set the waypoint to the reardoor of Building B, my squad splits up again - one part leaves building A to the rear, the other half to the front door. More funny, the team that leaves to the frontdoor tries to get around Building B on it's LEFT side and regroup with the other team that waits at the rear door of building B. Since this is not possible within the timeframe of the 'regroup phase', they then went back and enter building B through the front door!!! The problem seems to be partely caused by the regroup an maintain phase at each waypoint(as written in the manual on page 54). My conclusion: it is very difficult to make a prediction about the execution of movement orders. Since I don't know where exactly I would have to set the waypoint(s) to achieve the result I want to see. My stupid question/idea of the topic: is the regrouping phase at the waypoints necessary? Would it help to remove it? 0 Quote Link to comment Share on other sites More sharing options...
handihoc Posted April 20, 2008 Share Posted April 20, 2008 I've seen that happen recently, too. Infantry given orders to go one way, but some or all taking another route where they already 'know' the enemy are waiting. 0 Quote Link to comment Share on other sites More sharing options...
SgtMuhammed Posted April 20, 2008 Share Posted April 20, 2008 This issue is being worked on. 0 Quote Link to comment Share on other sites More sharing options...
c3k Posted April 20, 2008 Share Posted April 20, 2008 That is good news. 0 Quote Link to comment Share on other sites More sharing options...
handihoc Posted April 20, 2008 Share Posted April 20, 2008 That is good news indeed! A similar issue, which hopefully is also being worked on, is that inf debarking from a vehicle STILL sometimes go for a long runaround in the open rather than high-tailing it pdq into the building they've been ordered to. 0 Quote Link to comment Share on other sites More sharing options...
Seanachai Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by SgtMuhammed: This issue is being worked on. Wow! And after only...how many months? 0 Quote Link to comment Share on other sites More sharing options...
Thomm Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by Seanachai: </font><blockquote>quote:</font><hr />Originally posted by SgtMuhammed: This issue is being worked on. Wow! And after only...how many months? </font> 0 Quote Link to comment Share on other sites More sharing options...
Scipio Posted April 21, 2008 Author Share Posted April 21, 2008 What I have seen so far makes me believe that pathfinding and unit placement is based on an underlying grid. IIRC Steve said something like that in a very old post. That would explain why so funny things happen frequently. Or why soldiers are not able to follow a trench without leaving it - as long as the squad or team is within the right grid tile, everything is okay for the program. Cover seems to be secondary. It's the same thing like ordering area fire - the target marker ussually jumps some meters, I assume to the center of the next map tile. With all respect to Charles, who makes a great job, especially if we keep in mind that he makes all the programming alone - but maybe the basic system has already a flaw. 0 Quote Link to comment Share on other sites More sharing options...
Thomm Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by Scipio: ... but maybe the basic system has already a flaw. No, no, no, I do not buy into that theory! The system needs more work, that is for sure, but I do not see any indication for it being flawed. Look at the trenchline problem, for example. Solving this is either a matter of adjusting weighting factors for pathfinding, or some special code that detects waypoints being plotted along a trench and concludes that the soldiers should stay inside of said trench. Sounds doable to me! Other misbehaviors are discussed internally. The Marine module, introducing 13(!) men squads will hopefully resolve some more of the issues. I feel that it has to, with squads as large as this! Best regards, Thomm 0 Quote Link to comment Share on other sites More sharing options...
SgtMuhammed Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by Seanachai: </font><blockquote>quote:</font><hr />Originally posted by SgtMuhammed: This issue is being worked on. Wow! And after only...how many months? </font> 0 Quote Link to comment Share on other sites More sharing options...
Huntarr Posted April 21, 2008 Share Posted April 21, 2008 Just a point. The only trench problem is with Zig-Zag trenches. Straight line trenches, troops move in perfect line formation staying in the trench. 0 Quote Link to comment Share on other sites More sharing options...
thewood Posted April 21, 2008 Share Posted April 21, 2008 Even in straight line trenches, I occasionally see one or two guys come up out of the trench. This seems to mostly happen at a waypoint when the milling about gene becomes dominant. 0 Quote Link to comment Share on other sites More sharing options...
ChrisND Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by Seanachai: </font><blockquote>quote:</font><hr />Originally posted by SgtMuhammed: This issue is being worked on. Wow! And after only...how many months? </font>Imagine that - difficult problems taking a long time to completely solve! Who'da thunk?! 0 Quote Link to comment Share on other sites More sharing options...
Scipio Posted April 21, 2008 Author Share Posted April 21, 2008 Originally posted by Huntarr: Just a point. The only trench problem is with Zig-Zag trenches. Straight line trenches, troops move in perfect line formation staying in the trench. This is unfortunatly not correct. I made a few tests with a simple trench. A six men engineer squad (organized in two teams of 3 men each) was ordered to enter a trench, follow it to it's end and then turn around an move to the other end. What happend: one half of the squad entered the trench and followed it, the other half runs to the trench, crosses it and moves on paralell to the trench in some distance. I assume that the path for both teams - of the unsplitted! squad - is calculated seperatly, and maybe both teams are not allowed to share the same tile, or take the same path at the same time, or something like that. 0 Quote Link to comment Share on other sites More sharing options...
Huntarr Posted April 21, 2008 Share Posted April 21, 2008 QUICK w/ARC & FACE at the end Two seperate units at two seperate trench locations. Everyone stayed in the trench unless given a waypoint out of the trench. One unit was given QUICK the other unit ASSAULT. Both were given ARCs for the move and FACE at the end. WEGO 0 Quote Link to comment Share on other sites More sharing options...
Huntarr Posted April 21, 2008 Share Posted April 21, 2008 These guys have problems with Zig-Zag trenchs Some of the problem can be reduced by tedious waypoints but not completely. I feel it is only a minor tweak in the Pathfinding AI. Charles could possibly give a higher priority to trench terrain thus drawing units into it not avoiding it. I feel that the AI is thinking of it as a hinderence to speed, not thinking of it as good cover. 0 Quote Link to comment Share on other sites More sharing options...
cool breeze Posted April 21, 2008 Share Posted April 21, 2008 Haven't checked but maybe a command slower than quick would keep them in the trench better? running on top is quicker. 0 Quote Link to comment Share on other sites More sharing options...
Scipio Posted April 21, 2008 Author Share Posted April 21, 2008 Huntarr, nice pics, but your approach seems different from mine. The teams on your pics start the movement in a trench, while mine start in open terraine and (shall) enter a trench. It seems to me that the starting point of the movement has some importance. On the last image it looks a little bit like the team leaves the trench because another team is blocking the way. But thats pure assumtion, I could be wrong with this. 0 Quote Link to comment Share on other sites More sharing options...
Thomm Posted April 21, 2008 Share Posted April 21, 2008 Originally posted by cool breeze: Haven't checked but maybe a command slower than quick would keep them in the trench better? running on top is quicker. No, the problem is the same for all movement types. Been there, tested that. Best regards, Thomm 0 Quote Link to comment Share on other sites More sharing options...
handihoc Posted April 21, 2008 Share Posted April 21, 2008 Also, I've had men leave a straight line trench when ordered to follow the line. It's worse when ordered to enter the trench from outside (some of them simply go prone or kneeling outside the trench), or when given different waypoints along the trench. It doesn't always happen that they leave or fail to enter the trench, but often enough to make me very nervous. 0 Quote Link to comment Share on other sites More sharing options...
Huntarr Posted April 22, 2008 Share Posted April 22, 2008 Scipio I run the tests as you said outside the trench and again the only problem was with Zig-Zag, sorry I couldn't reproduce your issue. I tried various movements on the zig and MOVE came up with some pretty good results. Agonizingly slow. Besides if your in a trench it's a pretty good chance you'll be in contact. So no MOVE there. 0 Quote Link to comment Share on other sites More sharing options...
Huntarr Posted April 22, 2008 Share Posted April 22, 2008 0 Quote Link to comment Share on other sites More sharing options...
Battlefront.com Posted April 22, 2008 Share Posted April 22, 2008 Yes, we are always looking into pathing issues as they are pop up. Charles is extremely busy with some other stuff at the moment, but I can assure you that he is going to look into these specific issues (and some others) very soon. Plenty of documentation awaits him, thanks to relentless testers Steve 0 Quote Link to comment Share on other sites More sharing options...
Guest Guest Posted April 22, 2008 Share Posted April 22, 2008 deleted per user request 0 Quote Link to comment Share on other sites More sharing options...
Battlefront.com Posted April 22, 2008 Share Posted April 22, 2008 That really wouldn't fix anything if the guys are going to the wrong spot in the first place Remember, those little green spots have to be determined by some process. The process we use is the TacAI in relation to terrain. For a game like FSW it was hardcoded into the map and had no relation to tactical situations. Steve 0 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.