Applying Andis's Signal Scripts

[PaulH]

Can you guide me to the reference manual AndiS?

[AndiS]

It is in the ZIP of the download, named "Reference Manual 1.0.pdf".
https://www.trainsimdev.com/forum/downl ... &df_id=141

[PaulH]

Thanks for that. I shall study it shortly.

Here's a odd one! I have (almost) successfully got my ground disc-home arm-ground disc arrangement working on my test line. This is a 3 -way split with diverging lines for the discs to left and right.

With a light engine they performed exactly as required so I tried a simple shunt move, making up a 3-coach train starting in advance if the new signal from one coach on each line, and returning to the start point. As expected, no signal would clear so I put a dummy distant in each line to allow me to proceed. I appreciate that you probably don't agree with this approach but I have tried ticking the right hand box (Limited Aspect?? I can't recall the name but not the "Approach Control" box) and found that the signal annoyingly went ON again as I was passing it.

On two of the lines it all worked exactly as required but on one route all was fine until I coupled the loco onto the coach when the signal that had gone ON at the right point immediately cleared OFF on coupling. Uncoupling the loco with out the movement at all and the signal went ON again, coupling signal OFF, uncoupling signal ON etc etc ad nauseam.

Could this be some instability caused by me unwisely applying distants at the throat of each siding?

I am now starting to juggle with the x, y and z co-ordinates of the elements as you described to get the discs and signal arms in the correct alignment.

[AndiS]

It has nothing to do with the distants.

When you couple and uncouple, the game sends funny messages to the signals around telling them that a train would have left the block on the other end when you uncouple and it tells them that a train entered the section in the wrong direction when you uncouple. The reasoning behind this is that the signals count passing trains to simulate track circuits. So when you uncouple the engine and drive it away, the game needs to tell the signals "there is one more train in the block" or else they let the next train crash into the uncoupled coaches.

How that aligns with your observations, I have no idea.

One thing is sure - when you change anything, reload for the main screen (where the "play again" button is). If you go from route editor or scenario editor directly into a test drive, you cannot expect anything to work. The reason for this is that only on scenario start proper, the game tells everyone to initialise and signals, switches and trains send a hundred messages to each other in order to get a halfway decent picture of what is going on.

[PaulH]

Right, thanks for that. I suppose it is a minor inconvenience that I will learn to live with. You have answered another item for me - whenever I changed anything I always actually logged right out of TS and started the game afresh before testing. Now I know that I only have to go back to "play again" to initialize everything.

I will check if the couple/uncouple glitch occurs on the other sidings.

Another anomaly has shown up whilst I was experimenting with the xyz coordinates of the elements. I tried to add 5m to the y coordinate for the main aspect. For some reason this lifted the signal head on the test signal I used on the test track, but not on other samples of the exact same disc-home-disc arrangement ii had put at another location. I them tried again (same signal arrangement) and this was to the original dimensions. I checked the main bin file and the additional 5m was still in the y coordinate for the main head but when I installed kit again it was still to the original dimensions.

Curious and curiouser!

[PaulH]

Don't waste your time on this AndiS.

The shunt anomaly does occur on all sidings so I'll just get used to it.

I have found out what I was doing wrong with the signal head y cords. I had simply made a copy of the original bin and sufficed the name with "orig" and placed it in a subdirectory called WIP. This was being read by TS so that when I placed it, it was to the original dimensions,

I have now taken the WIP, sub directory and moved I right out of TS directory. Problem solved.

Sorry for wasting your valuable time.

Sometimes taking a break works wonders!!

[PaulH]

Just a few pictures of what you have helped me to achieve, AndiS. I have tested these to death and they all work (almost) perfectly.

Note the bespoke disc-home-disc arrangement on the 8th image.

I know the scenery is a bit barren, having placed all of the p/way and the major structures including all platforms, cuttings, embankments, retaining walls, viaducts and major bridges I want to get all signalling functioning end to end before I embark on the rest of the scenery (including bespoke models of most of the buildings) between the boundary fences. The scenery outside of the boundary fence will follow.

The "almost" above is because whilst all signals respond exactly as required from my viewpoint as the driver, I have noted that once I have arrived at Bala town, the signals I passed OFF at Bala Junction which prototypically returned to ON after I passed have cleared again. When I return to Bala Junction, all of these signals at Bala town that returned to ON when I passed them have also cleared again.

As far as my viewpoint as the driver, I cannot see this, of course because I am 3/4 mile away so I am not too concerned as the errant signals have correctly retuned to ON before I get into sighting distance.

One thing I would like your advice on is distant signals. I am aware from previous posts from you that the distants are not simply passive as they are in real life where they simply indicate to the train that all stop signals through a block post are OFF along the main route into the next section, but in TrainSim they have an active role in the performance of home signals. Following your earlier advice I have placed dummy distants to simulate the distants at the next block posts along the main line shortly before the portals into hyperspace but what do I do about the distants I should be passing.

All of the actual distants on the real route were fixed as they were with many secondary single track routes and John Yelland has provided GWR fixed distants which are purely passive items of scenery with no track links. I have added dummy live distants to your scripts at these locations and sunk the arms underground.

One other thing that you may be able to help with is how do I get the points to default to the Normal position when I fire up the trainsim. I've tried everything I can think of but mine here seem to stubbornly default to sidings and cross-overs - very annoying.

Paul

[AndiS]

Regarding the fixed distants, there are two approaches.

1. You create a signal that has John's fixed distant for post and no child elements - no arms like all other signals have. In the script, you take out all the lines with gArmTable in them. Finally, you enter $E in the signal ID.
   I am not 100% sure that this will work because I envisioned a normal model with moving arm that is just immobilised by the $E in the ID.
2. You create a bog standard distant with one normal arm that happens to be 2 m below ground. This is a very safe road.

I have a dim memory of the problem with the switches but I would not say I remember something useful. There was one issue with manual switches acting different than automatic ones.

And there was the hint that you have to go to the scenario editor, set the switches there (but most likely just the manual ones, you cannot set the automatic ones, can you). Then you save the scenario and hope for the best. You have to repeat the exercise for every scenario, which makes sense as you may want a different path set in each of the scenarios.

[AndiS]

I forgot to drop my sincere congratulations of achieving what you wanted. This is not nothing here.

And never mind what the signals do when you are not there. I abandonned my programming effort three, if not five times because if you want a good impression by the observer, this is simply the wrong game. Once I focussed on giving the driver a good impression in complex situations, I got somewhere. I did try a fair bit to stop signals clear for no reason but no one knows what happened on the other end of the route. No one wants to read a ton of log files just to find out and without such a chore you have no chance to find it out.

[PaulH]

Agreed.

By the time I have returned there, they have reset to ON anyway.