Tweet
Edit: I believe the behavior of the path-finder as shown in the image below is explained by Jaybe's post below. I was able to find a save made just before I moved the workers in question, and in fact one of them had already expended a third of a move. Thus it could only move two road sqares thus the path-finder chose the more direct route (though it was still a poor route since it limited future options). I'm not completely positive this 2/3 movement unit was the one, but it seems plausible and in fact likely.
Original post: I have played four or five CivIII games now, and prior to the 1.29 patch I never had a failure of the path-finder. When I'd hit G or J the route that CivIII would pick was always an optimal one. Since the 1.29 patch I've seen three occasions where the path finder failed. The first two I was routinely doing a G goto move, and happened to notice the "2" sitting under my mouse cursor for what I knew was a "1" turn route. I cancelled the goto, and did it manually (actually used goto for part of the way, manually for part), and sure enough I got to my target destination in the current turn (and I'm sure I didn't pass any ZoC units along the way).
I chalked the first two errors up to a fluke, but this third and last occasion was too weird to ignore. I was using J on a stack of three workers. I was sending them to undeveloped terrain to build a road (destination is the square with the green outline). I hit J, move the mouse, click on my target, and the units move. The units are following the rail line coming in on the right side of the image. The first worker (I've labeled "A" in the image) takes the wrong path. Rather than getting off on the road, and following the red line, he moves into the square with no road as a step to get to the destination square. Since there's no road in this intermediate square he stops. His move will take two turns. But this is a J move, and the next two workers (labeled with "B" [I later moved a third worker into the square to complete the 1-turn road building trio]) come along. The next two workers in the same J move take the correct path. They follow the road, and get to the destination in one move.
What happened here? Did the first worker yell back, and say "go the other way"?
I'll admit my sample size before the 1.29 patch is very small (I could have just been lucky in not finding errors), so I was wondering if anyone else has noticed anything similar (either before or after 1.29).
PS: How do I put this image in the middle of my post?
Original post: I have played four or five CivIII games now, and prior to the 1.29 patch I never had a failure of the path-finder. When I'd hit G or J the route that CivIII would pick was always an optimal one. Since the 1.29 patch I've seen three occasions where the path finder failed. The first two I was routinely doing a G goto move, and happened to notice the "2" sitting under my mouse cursor for what I knew was a "1" turn route. I cancelled the goto, and did it manually (actually used goto for part of the way, manually for part), and sure enough I got to my target destination in the current turn (and I'm sure I didn't pass any ZoC units along the way).
I chalked the first two errors up to a fluke, but this third and last occasion was too weird to ignore. I was using J on a stack of three workers. I was sending them to undeveloped terrain to build a road (destination is the square with the green outline). I hit J, move the mouse, click on my target, and the units move. The units are following the rail line coming in on the right side of the image. The first worker (I've labeled "A" in the image) takes the wrong path. Rather than getting off on the road, and following the red line, he moves into the square with no road as a step to get to the destination square. Since there's no road in this intermediate square he stops. His move will take two turns. But this is a J move, and the next two workers (labeled with "B" [I later moved a third worker into the square to complete the 1-turn road building trio]) come along. The next two workers in the same J move take the correct path. They follow the road, and get to the destination in one move.
What happened here? Did the first worker yell back, and say "go the other way"?
I'll admit my sample size before the 1.29 patch is very small (I could have just been lucky in not finding errors), so I was wondering if anyone else has noticed anything similar (either before or after 1.29).
PS: How do I put this image in the middle of my post?
Attached Files
Comment