Set junction using priority

[peterwhite]

I am looking for a way to set a facing junction by using the trains priority.
Basically I want the junction to set left for priority 2 and right for priority 1.
Is this possible?

Peter

 

[pware]

There is no session rule that I know of that will query the priority level of a consist (but that does not mean that such a rule does not exist).

 

[skittlekicks]

Is the invisible Enhanced Interlocking Tower smart enough to do something like this?

 

[MichelLmbrt]

Perhaps using two AI Routing Priority Markers ( kuid -3 10190 ). One on each track beyond the switch with priority 2 for one and 1 for the other ?

 

[peterwhite]

Is the invisible Enhanced Interlocking Tower smart enough to do something like this?

I use Interlocking Towers intensively but got to the stage where my session would not load. I assumed it was because of all the data involved so I am trying different ways to cut down the number of Towers and paths to see if that is the case.

 

[peterwhite]

Perhaps using two AI Routing Priority Markers ( kuid -3 10190 ). One on each track beyond the switch with priority 2 for one and 1 for the other ?

That would entail using Drive to/via and I only use autodrive. I don't like giving AI the option.

 

[justinroth]

I am looking for a way to set a facing junction by using the trains priority.
Basically I want the junction to set left for priority 2 and right for priority 1.
Is this possible?

Peter

I personally would use enhanced interlocking towers and then mission codes. The switch/route will be set depending on the mission code assigned to the driver.

 

[skittlekicks]

I personally would use enhanced interlocking towers and then mission codes. The switch/route will be set depending on the mission code assigned to the driver.

I use Interlocking Towers intensively but got to the stage where my session would not load. I assumed it was because of all the data involved so I am trying different ways to cut down the number of Towers and paths to see if that is the case.

This is something I had not considered when I originally suggested the EIT.

Are you dealing with single track mainlines or double track (or more mains) for what you want to do? You could use one way markers for certain priority trains and utilize track priority markers for said trains to follow exclusively follow. If there is too much data nerfing your session load with EITs, you might consider eliminating ones that might not be as necessary. I know that sucks because they are REALLY useful, but they probably need a complete code rewrite from scratch.

Until such time, the choices are limited in the automation world of the game. What works for some, doesn't work for everyone.

As I have found over the years of gaming and modding games, bad coding techniques that carry over from generation to generation leads to chaos with stuff failing to load.

I got to the point where I'll sporadically use the EIT, and sometimes not at all.

 

[justinroth]

I use Interlocking Towers intensively but got to the stage where my session would not load. I assumed it was because of all the data involved so I am trying different ways to cut down the number of Towers and paths to see if that is the case.

From what I understand and it's been a bit since I've messed with EIT's is that more "towers" with less paths is the way to go. Also I think they're supposed to go in the session layer. It's been a while though.

 

[peterwhite]

You are correct in saying EIT's should be in the session layer.
My route is not large but is very intense. It is around Carlisle in the late 1940's when it was in full swing.
It has one major station with 7 double or quadruple main lines radiating from it with another 13 stations on those lines.
It also has 10 large goods yards and 6 loco sheds, or 7 if I roll it back to the 1920's.
To control all of this I have 68 EIT's but can't remember the number of paths but it was well over 500.
I am running 22+ so I can pause the session and resave as the next session. I run the sessions for 48 minutes, pause, save and then run for the next 48 minutes etc.
I am working to a working timetable for passenger and freight for 1953 and traffic is fairly intense plus all the associated shunting in the goods yards where each yard has at least one shunter but most have two.
So, as you can see I have a lot of drivers and locos in each session. I got to about the tenth session before I started having problems, first paths started to disappear form an EIT and I got around this by resetting all rules when starting that session but then it came to the point that the session would not load.
So, here I am looking at alternatives to cut down on the amount of data to see if I can get any further.

Peter

 

[paulper]

I don’t know whether it’s the length of time that a session has been running (or subsequent multiple 48 minute loads as in Peter’s case) but for me it often gets to a point a few hours in where it’s like trains forget they have a mission code and it sits at a red. I guess in real life that’s a track/signal/junction failure so arguably authentic!

 

[JCitron]

I don’t know whether it’s the length of time that a session has been running (or subsequent multiple 48 minute loads as in Peter’s case) but for me it often gets to a point a few hours in where it’s like trains forget they have a mission code and it sits at a red. I guess in real life that’s a track/signal/junction failure so arguably authentic!

I have the same issue with my large route containing many AI trains in addition to portals adding more trains to the route. Like you I have my time set to about 48 minutes before the route contained the portal trains and after a few hours, the AI go on tea breaks and sit there at signals. Sometimes hitting pause and resuming will wake them, but sometimes that doesn't work. Other times, if I give them a poke by manually driving them, they'll come to and resume their schedules as if nothing is wrong.

I worked with N3V for quite a long time on this issue and the conclusion they came to is there are so many threads operating that they're clogging the pipelines causing long delays in responding to actions. At the time, they said that this is a fault with the thread manager and that the AI system needs to be completely rewritten from scratch. This was back in the early days of T: ANE and since then there have been some improvements but not what we'd like to see.

 

[Forester1]

I sometimes wonder if they have set up the database efficiently with stored procedures, etc. Understandably, a lot of us have humongous databases, but it really seems to work everything hard with even the slightest DB access.