PBS line interference

[3iff]

See picture for a representation of the layout. The signals are really PBS signals but I don't have the required graphics on this computer...

Cardtown Transfer is a busy station. In the real circumstance, it serves a food processor in a desert scenario. About 6 food trains and 3-4 fruit trains regularly visit.

Another station delivers 5-6 maize trains to prevent deadlock at Cardtown transfer. Cardtown south has a single oil train and is not connected to any other part of the network.

Often, Cardtown transfer will fill with food trains and after a maize delivery, 2-3 food trains will fill up and try to leave. Often, the 3rd train will have to wait too long and 'give up'. This causes it to reverse and STOP...with the train window popping up.

I've discovered that the line to Cardtown south is interfering with the track detection system and the reversing train believes that this unconnected line is part of the track for Cardtown transfer.

I moved this track so that it didn't touch the western side of Cardtown transfer and waiting trains in that station never STOPPED or popped up a window (due to train unable to reverse safely).

I couldn't find any reference to this problem...but there's so many posts to look at.

[PikkaBird]

Often, Cardtown transfer will fill with food trains and after a maize delivery, 2-3 food trains will fill up and try to leave. Often, the 3rd train will have to wait too long and 'give up'. This causes it to reverse and STOP...with the train window popping up.

Personally, I think the "giving up" on a red signal and reversing is one of the least realistic, and least predictable, aspects of the original TTD signalling system. I've set both signal1waittime and signal2waittime to 255 (wait forever), and it makes life much easier.

I'm not sure if the PBS reversing problem is because of that extra track, or because when it tries to reverse it can't find another PBS signal to set a path to.

[3iff]

Ah, must have missed those signalwaittime settings... I could try that.

When I relocated that strip of train line, the stopping problem disappeared. With the track in the position as per the diagram, I would get a train stopped every game month or so. Certainly enough to annoy me.

I'll check out the signalwaittimes. I realise too that the problem might be due to TTD internals. I just wanted to draw attention to the problem...

[DaleStan]

If I understand you correctly, that's a known bug in PBS, triggered when a train occupies the entire platform. Make the platform one longer, or add a peice of rail to the end of the platform (preferably one the train can't use) and the bug should go away.

[3iff]

If I understand you correctly, that's a known bug in PBS, triggered when a train occupies the entire platform. Make the platform one longer, or add a peice of rail to the end of the platform (preferably one the train can't use) and the bug should go away.

Unfortunately, it's a bit messy to extend the station without a lot of reworking...but thanks for the tip. As long as the problem is known then I'm happy. Maybe I can redesign the oil route so it joins the main station as a dedicated platform...doing that will allow the main station to grow longer.

Thanks for the replies.

[Patchman]

Like I wrote on the PBS manual page, if trains are unable to reverse, that should be considered a bug and reported as such. So please post a savegame with the problem about to occur, along with the appropriate ttdpatch.cfg and newgrf(w).cfg.

[3iff]

Ok, will do...just have to wade through a collection of savegames...not a problem.

I did have a look at the signalwattimes and they're set at 70 and 20 (default??). I've now changed them to 70 and 70 and can keep an eye on train behaviour to see if that fixes things.

[3iff]

Got the savegame.

Now on TTD windows version. Save and 2 cfg files zipped up for you (and anyone who wants a look). I think only UStrainset and PBS signals are really relevant.

Lima Valley station, 11th Feb. Train 19 will stop in 4 days. It looks like it fails to exit the station, reverses and tries again and then halts with a permanent STOP. The only cfg change I have made is reducing signalwaittime2 down to 5 to ensure the effect happens (it was originally 20). Hopefully this won't affect the reason train 19 'gives up'.

If you remove the n/s track to the west and reposition it further west then the stop effect shouldn't happen. If fact, I'll try that now...

I did try it and if I reposition the track away from the rear of Lima Valley station, train 19 does not STOP...interesting, I think.