OpenTTD-patches
OpenTTD-patches copied to clipboard
JGRennison
Bug Report - Adaptive train speed sticks longer than it should
Version of OpenTTD
JGRpp 0.42.2
Expected result
As soon as the slower train pulls away onto a diffrent track and thus clearing the track for the faster train behind it. The fast train should immediatly return to its regular faster speed.
Actual result
The faster trains sticks to the adapted speed until the last signal where the slower train was on its path. (in case of a depot along a single line without any signals, this could be the entire single line until it reaches either a signal or the end of the line)
Steps to reproduce
Get a line, place some signals on it (long enough to allow adaptive speed to enable), then get a split off, and you put the slower train in front, the faster train behind it. Watch at the split off, the slower train will be on the branch, but the faster train is still going slower than usual until the last signal at the split off.
At present speed measurements and the changes to the adaptive speed limit only occur at signals. You would need to add signals on the far side of the divergence, you can make these reserve through to act as repeater signals. This may change at some point in the future. The adaptive train speed feature is mainly suited for homogenous layouts with regular signal spacing. At present it's probably not the thing for any sort of complex layout or single-line working.
https://drive.google.com/drive/folders/1n6SvOWJdOUjfQuYacjQczyCcV0TKSFXd?usp=sharing This happens in any similiar situation, which there are can be a lot of in big networks.
The tunnel/bridge issue seems to have been solved and for the remainder I suggest putting a timeout on the trains might work, so in case they haven't seen a signal in a long time, the max speed is reset? I think if the train hasn't seen a signal after 10 tiles it's probably safe to assume it's a single track or something similar. I'd use a number of tiles instead of ticks though, since it would have to be some installation on the ground updating the train.
As for the posted savegame, I don't see the issue with it? Only the junction signal knows that the train left in a different direction, so the train can only speed up after reaching the junction. Sure that's a small delay, but that's like one additional signal block before it can speed up again... I don't really see the issue with that. It's still way better than the behavior without speed adaptation.
