[11.35] Trains and trams reversing for no reason

[zook2]

(Win XP 32-bit)
In my old game (begun in 11.1x) I have noticed several spots where trains and trams stop and reverse for no apparent reason. That is, they stop, show "Reversing", wait until countdown, then continue in the same direction. When it happens in curves, as in the case of this tram, they run off the tracks to the end of the tile before "reversing".

And not all trains passing the spot do it, but the ones who do do it every time.

[Vladki]

Just a wild guess... Are you using waypoints in the schedules? I have seen similar graphical effect as in screenshot 2 in standard when there was a waypoint on the curve.

Sent using recycled electrons.

[jamespetts]

Can you upload a saved game in which thus can reliably be reproduced?

[DrSuperGood]

The only time I notice this happening is when you modify the rail such as removing a point, changing a curve etc. The train will try to keep the old route until it hits the modified section (where it cannot continue) and then teleports slightly to be aligned with the new track, reverses to take the new direction etc.

Is this a persistent problem? I suspect it could happen once if the save was freshly loaded with a new version as some internal stuff might have changed. However if it happens repeatedly (convoy does it every time it reaches that section on the route) then this is a bug. As mentioned, waypoints can do something similar to this so avoid them unless nescescary.

[zook2]

http://freakshare.com/files/32xrn22s/new-2.rar.html
(This is a different save than the one in the Weird convoy thread)

The green engine should display that behavior right after loading, but for some reason it doesn't (746,789). But if you wait a little, another train on the Wingington-Pewhead line will.

Another spot is the tram tracks at 308,1068. Hide the buildings to see it.

I just noticed that there are waypoints on both spots. I also saw other weird things connected to waypoints happen in the new version, like trains skipping them. What's suddenly wrong with waypoints?

[jamespetts]

Waypoints are not actually skipped: rather, they are not actually destinations, as it is necessary to prevent anomalies for vehicles (especially trains) to calculate routes through waypoints rather than only calculating the route from the waypoint on arrival.

Does this behaviour persist over time (i.e., over many iterations of a single convoy's timetable) if nothing has changed, may I ask?

[zook2]

It's persistent, i.e. immediately after loading the game (the one I uploaded) it didn't happen, but every time after that.

What has changed with waypoints (or where was that discussed)?

[jamespetts]

I will have to look into that when I get time, in that case.

As to what has changed with waypoints: for rail type vehicles, they no longer act as destinations, but as soon as the timetable reaches a waypoint and the convoy has found a route to it, it moves to the next timetable entry, and finds the route from the waypoint onwards, rather than waiting until the convoy has reached the waypoint first, unless the waypoint is a reversing waypoint (marked with [R] after it in the schedule), in which case the conventional behaviour is retained. May I ask whether [R] appears next to the waypoints in this case?

[zook2]

May I ask whether [R] appears next to the waypoints in this case?

Hmmm... it sure does. Which begs the question: where did the [R] come from? If I delete the waypoint and set it again, the R reappears.

I'm still going through the forum search results... is that feature explained in full somewhere?

[jamespetts]

I cannot remember now what discussion that there has been on the forum in respect of this topic: the [R] is marked automatically when a waypoint is added and then unmarked when a convoy passes a waypoint without having to reverse: it remembers the next time that it does not need to reverse, and de-sets the [R] designation. May I ask whether this is an old saved game that you have loaded or a new one that you have started in 11.35?

[zook2]

It goes back to 11.1x.

[jamespetts]

Interesting. Does the problem only occur with existing schedules, or does it also occur with any schedules created since the upgrade?