[11.21] End of choose signal ignored?

[Max-Max]

Train passing the choose signal can only choose platform P2 and P3 because an End of choose signal is placed before the platform P1.
Still the Train T1 (and all other trains arriving from the same choose signal)  has chosen Platform P1.

There is a block signal next to the choose signal blocking trains to enter P1 from that direction. Can it be that the choose signal isn't paying attention to the one-way block signal and chose P1 from the wrong direction. Then the train has been directed to P1 and ignores the End of choose signal?

[ӔO]

If T1 is scheduled to use P1, then that is all it will use if P1 is behind an end of choose.

[Junna]

As AEO says, that appears to be working correctly and as intended?

It's quite useful sometimes.

[whoami]

A bit late, but nevertheless: what I see in front of the platform P1 must be a choose signal, because it looks the same as the other one. An EOC is a different thing (not all paksets have it) and cannot be combined with a main signal on the same tile. Use a standard/simple main signal there instead.
The EOC descriptions in the Wiki are a little outdated, because the behaviour has been changed.

How is this related or confined to ST-Experimental? It seems like a generic problem, but there may be a difference.

[jamespetts]

Perhaps I should move this to "help requests"...?

[Ters]

My rule is to place choose signals so that the only thing behind them are the platforms the choose signal shall direct trains to. (Choose signals are in my opinion flawed in the sense that they are signals affecting all passing vehicles, while platform choice should be part of the schedule. In my style of play at least. Unfortunately, I don't see any easy way of making it part of schedules.)

[AP]

Choose signals are in my opinion flawed in the sense that they are signals affecting all passing vehicles

Agreed - if there are trains which aren't booked to stop, the presence of a choose signal can cause gridlock. I only use them at termini, unless the stations are very carefully designed.

[jamespetts]

In theory, a choose signal could be bound to a particular halt and work as a normal signal for any convoy whose next stop is other than that halt, but this would be a bit tricky to implement and a bit fiddly to use, I think.

[prissi]

Not to mention to find out which next stop is intended. But a choose singal works as normal signal if there is another choose signal on its route.

[Ters]

Much of the confusion comes from the fact that signals in Simutrans behave pretty much nothing like real signals. Perhaps the biggest difference is that Simutrans has no concept of a station area. It is easy to use choose signals as home signals, but it can't behave like one without at least a more full definition of a station. This would move Simutrans more in the direction of a model railroad simulator, although a more precise definition of areas could also benefit trucks, ships (only visually) and aircraft.

[jamespetts]

Much of the confusion comes from the fact that signals in Simutrans behave pretty much nothing like real signals. Perhaps the biggest difference is that Simutrans has no concept of a station area. It is easy to use choose signals as home signals, but it can't behave like one without at least a more full definition of a station. This would move Simutrans more in the direction of a model railroad simulator, although a more precise definition of areas could also benefit trucks, ships (only visually) and aircraft.

Interesting - the concept of a definition of a station area does not exist as such in UK signalling practice (although the concept of a yard area, not directly under control of the signalbox, does): the block and signal rules are the same for stations as for any other stretch of track, although sometimes route indicators on signals are based on platform numbers.

There is much to be said for making the signalling system more realistic, although note that the choose signal in Simutrans actually carries the function, not of a specific signal, but of routing (or "dispatching" in U. S. terminology), which, although closely associated with signalling, is actually distinct, which is probably the source of confusion. Simulating routing accurately in Simutrans is not easy (or, strictly, probably desirable, since it would involve repeated manual choices), but there might conceivably be a way of separating it from signalling at least, even if, when separated, it is then simplified. For the time being, the choose signal is a serviceable hack that gets the job done well enough.

[Ters]

Interesting - the concept of a definition of a station area does not exist as such in UK signalling practice

That's odd. When checking whether this concept was more or less unique to Norway, I found a description from India that seemed similar. I would have thought India used a British system. The difference between a station and a line seems to be that only one train may occupy a block on the line (often the entire line between stations is one block, most lines being single tracked), whereas a station is essentially one big block, where multiple trains are allowed (although a station can have some internal block like concepts). Passing loops are technically stations in Norway. There are also stops with platforms that are not technically stations, but part of the line. (This is where Simutrans needs the long-block signal.)

Simutrans signals would somewhat map for Norwegian signals in the following way:

Long block signal	Station exit signal (Can be placed on every track, or at the station boundary more or less opposite the entry signal. In the latter case, permission to approach it must be given by the station master in person, which is usually the procedure anyway.)
Normal signal	Block signal (doesn't work on single track)
Choose signal	Station entry signal
Pre-signal	Looks like a distant signal, but as we all (should) know, the pre-signal is a completely different beast

In most cases, such a correspondence would require signals to be one-sided, as opposed to one-way.

There is much to be said for making the signalling system more realistic, although note that the choose signal in Simutrans actually carries the function, not of a specific signal, but of routing (or "dispatching" in U. S. terminology), which, although closely associated with signalling, is actually distinct, which is probably the source of confusion. Simulating routing accurately in Simutrans is not easy (or, strictly, probably desirable, since it would involve repeated manual choices), but there might conceivably be a way of separating it from signalling at least, even if, when separated, it is then simplified. For the time being, the choose signal is a serviceable hack that gets the job done well enough.

Yes, the signalling in Simutrans does work, so my use of the word "flawed" was perhaps a bit harsh. One just have to be mindful of how Simutrans does things. In particular the lack of a unified schedule for all trains, which is what makes signalling work in real life.

[jamespetts]

Hmm, interesting. If a station is treated as one big block, how does it work when a fast train is passing through a station without stopping, and another train is in the station, albeit on a different platform? As I understand UK signalling practice, this is treated in exactly the same way as when a train is passing through any piece of track and another train is on a different line.

I should note that there are some specific rules about what drivers and guards must do when stopping at stations and starting from them and there are also things called "calling on signals", which are used mainly at stations, which allow one train to enter the same block as another train under caution (for a locomotive to couple to carriages or two multiple units to join or share the same platform, for example), but the signalling practices are based on the signal type, not a concept of a station area, as such. Station platforms and platform numbers have always been shown on signalling diagrams, and platform numbers are normally assigned as part of a train's schedule, so the concept of a station exists in signalling practice: just not the concept of a station area. UK signalling practice does not have a special type of signal for station entry and exit, although traditionally, a signal at the end of a platform was called a "starter signal".

[Ters]

Hmm, interesting. If a station is treated as one big block, how does it work when a fast train is passing through a station without stopping, and another train is in the station, albeit on a different platform? As I understand UK signalling practice, this is treated in exactly the same way as when a train is passing through any piece of track and another train is on a different line.

I guess interlocking ensures that the incoming train is routed through the station on an unobstructed path. The station (or certain parts) can also be opened for shunting, which allows trains to enter occupied tracks. One thing I've read is that only relatively few stations allow trains to enter from both sides simultaneously. The traditional procedure is for one train to wait at the entry signal until the other one has stopped at the proper place inside the station. The last train to enter the station may then pass straight through without stopping at all. (Bear in mind that there is likely only a single track at either side of the station.)

I should note that there are some specific rules about what drivers and guards must do when stopping at stations and starting from them and there are also things called "calling on signals", which are used mainly at stations, which allow one train to enter the same block as another train under caution (for a locomotive to couple to carriages or two multiple units to join or share the same platform, for example), but the signalling practices are based on the signal type, not a concept of a station area, as such. Station platforms and platform numbers have always been shown on signalling diagrams, and platform numbers are normally assigned as part of a train's schedule, so the concept of a station exists in signalling practice: just not the concept of a station area. UK signalling practice does not have a special type of signal for station entry and exit, although traditionally, a signal at the end of a platform was called a "starter signal".

There isn't much physical difference between a normal station signal and a line signal (station exit signals have a steady red light, the others flash), nor does there seem to be much difference in how drivers should read them. They are however described separately in the regulations. Old style station signals are somewhat different, and there are no corresponding line signals. With that kind of signaling, you have alternating blocks of line, station, line, station and so on. This requires the station to be staffed, and there is no exit signal. When there is no staff at the station, the signals are turned off, and the lines at either side appear to be merged into one block. There are also distinct prodedures for driving into a station, and out of a station (under different circumstances).

I'm not sure why Norway has this station concept. It might be that stations are the only place where trains may move in both directions at the same time and the only place with switches (apart from some industrial sidings, which are mostly closed now anyway). Traditionally, one therfore had stations, which were under full local control, separated by lines, which were controlled by the stations at either end in cooperation. Now more and more is centrally controlled, although some stations can apparently have some local control (shunting probably).

[jamespetts]

Ahh, that is very interesting - much of the Norwegian rules seem to rest on the assumption that traffic will be low density and on single track lines, whereas most of the UK rules seem to rest on the assumption of high density traffic and the use of double track or greater; perhaps that is the basis of the difference, ultimately caused by a differing population density in each country?

[AP]

Don't forget, before auto power signalling (LSWR main line was first, using pneumatics I think), the uk double-track network was limited to station and intermediate signal box posts, all of which were defined by the length of signal cable a signalman could physically pull (distant signal to distant signal).