I think your example is different because of the length and you have a portal. I've tried all the driver commands.
Line ahead is unsignalled...
- Thread starter dricketts
- Start date
I think your example is different because of the length and you have a portal. I've tried all the driver commands.
Looks like you have a portal there.
But the length is different I supose. Oh well. I tried
Just out of curiosity I was wondering if you built this route seperately and then merged the maps together? I ask because last year when I was using TRS2009 I had exactly the same problem occur with a route by PGibbons (I forget which one). I was able to run trains over the route for a few weeks with no problems, both AI and with myself driving.
While I was waiting for the next part of the map to be posted to the DLS I decided to add PGibbon's route to a map of my own that I had been working on so that I could practice blending terrain and textures between the two.
When I set up a session the AI train stopped at the end of the section and refused to enter it even though there were no other trains for many kilometres. When I took control I could run right through the section with no problems.
Interestingly, if I started the train past the starting signal and the put the AI driver back in charge it would carry on with no more interference from me.
I then placed another home signal onto the track just past the starting signal that controls entry to the section. The starting signal cleared but the new signal showed the track ahead being unsignalled. I then slid the new signal along the track and found that approaximately halfway through the section the new signal suddenly cleared. I slid it back a few millimetres and it again showed the track unsignalled. A level crossing was a short distance away and, thinking that maybe the culprit, broke the spline points and reattached them. Still the same problem.
I then tried deleting the track and relaying it. Same result. I tried putiing in extra spline points. Same problem. I tried building a passing loop. Still no good. I deleted that section and built new track to the west of it. Still the same problem.
In the end I just gave up, deleted the merged route and moved on with life. I've merged many maps over the years but that is the first and only time that has happened to me.
If anyone can explain it I like to know. The distance can't be the problem as I have much longer distances on my current map and trains run fine even when I was still using TRS2009.
Railwayz
While I was waiting for the next part of the map to be posted to the DLS I decided to add PGibbon's route to a map of my own that I had been working on so that I could practice blending terrain and textures between the two.
When I set up a session the AI train stopped at the end of the section and refused to enter it even though there were no other trains for many kilometres. When I took control I could run right through the section with no problems.
Interestingly, if I started the train past the starting signal and the put the AI driver back in charge it would carry on with no more interference from me.
I then placed another home signal onto the track just past the starting signal that controls entry to the section. The starting signal cleared but the new signal showed the track ahead being unsignalled. I then slid the new signal along the track and found that approaximately halfway through the section the new signal suddenly cleared. I slid it back a few millimetres and it again showed the track unsignalled. A level crossing was a short distance away and, thinking that maybe the culprit, broke the spline points and reattached them. Still the same problem.
I then tried deleting the track and relaying it. Same result. I tried putiing in extra spline points. Same problem. I tried building a passing loop. Still no good. I deleted that section and built new track to the west of it. Still the same problem.
In the end I just gave up, deleted the merged route and moved on with life. I've merged many maps over the years but that is the first and only time that has happened to me.
If anyone can explain it I like to know. The distance can't be the problem as I have much longer distances on my current map and trains run fine even when I was still using TRS2009.
Railwayz
No my map is not merged. It is a rather large DEM map though. About 80 miles.
Your Trainz behavior sounds identical to mine.
I should clarify that the passing loop I've tried was not signaled. Dermmy had suggested building an invisible loop with invisible levers but no signals and installing directional markers pointing away from each other on the loop. The idea was the signal/AI would see the loop and let a train pass the protecting signal into the long single track stretch. It didn't work. It does work perfect if signals are added to this loop but then this splits the block and makes cornfield meets inevitable. Which is what I'm trying to avoid to begin with.
It really is a simple setup and very common:
1. Left Diverging
2. Single aspect
3. Single aspect
4. Single aspect
5. Right Diverging
6. Single aspect
Trainz allows signals only to see so far. My single track stretch is about 17 miles long. Signals 2,3 and 4,6 can not see the diverging signals at the end of the block and thus read "line ahead unsignalled." AI trains will not pass these signals.
One could place additional signals half way between signals 1 and 5 and this would allow signals 2,3,4, & 6 to see the next signal and allow AI trains to pass. This solutions splits the single track into two blocks and creates cornfield meets.
I was hoping someone has come up with a unique solution. Maybe a signal that will see further down line or a creative work around. I've tried every signal in my collection of 25 gigs of content with no luck.
Last edited:
I've encountered a different variant of this issue when using long bridges. The train would fail to see the next track marker past the bridge, even if placed immediately after it. Signals also gave the track ahead is unsignaled message. I wound up disconnecting the track at both ends of the bridge, running a piece of invisible track across the bridge and lining up the spline points exactly with the bridge's own (but not connecting them) and suddenly the AI could see ahead again. I don't know how exactly once can create a bridge that the AI can't see past, but it fixed the problem so I have to assume that was it.
I also had this occur with a station on another occasion. In that particular case I replaced the station with another and it resolved the problem.
--Lamont
I also had this occur with a station on another occasion. In that particular case I replaced the station with another and it resolved the problem.
--Lamont
norfolksouthern37
what's a normal map?
the signals see as far as the tracksearch allows. after some testing, it appears that on a straight path the tracksearch only goes around 15 miles. i did not find any way to extend this. even if you could, what would be an upper limit? 15 miles is pretty high i think. move the protecting signals closer to the middle...
my smart signals let you place intermediate signals on sigle track without cornfield meets, but these are not released yet - that is the only solution i can give other than to move the protecting signals closer.
Man I can't wait for those new signals...
Hey that's a fresh idea about moving the signals closer to the middle! I wonder if the were moved say about a mile each if the turnouts would still work properly. I'll do some experimentation and post back.
norfolksouthern37
what's a normal map?
what didnt work exactly? you make it under 25400 m and they will 'see' them.
Moving the the diverging signals towards the center until the other signals stop reading "line ahead unsignalled" seem to work in surveyor at first. They would change to "no train approaching." But in driver the AI wouldn't pick up the oncoming train at the other end before it passed the diverging signaled that had been moved closer to the center.
Outcome: Cornfield meet.
Outcome: Cornfield meet.
This is how I set my signals at either end of a passing siding. My A.I. trains obey the signals and I've NEVER had a cornfield meet in three months of running one session.
If the siding is long enought to accommodate two trains, one in front of the other, I sometimes have a 04 on each opposing side of the tracks in the exact center.
Between passing sidings, I never have signals. This sets the tone for a confield meet.
Using this method, I will sometimes read a signal stating "Line Ahead is Unsignaled." However, even with that, I have found my A.I. trains still obey all commands.
If the siding is long enought to accommodate two trains, one in front of the other, I sometimes have a 04 on each opposing side of the tracks in the exact center.
Between passing sidings, I never have signals. This sets the tone for a confield meet.
Using this method, I will sometimes read a signal stating "Line Ahead is Unsignaled." However, even with that, I have found my A.I. trains still obey all commands.
norfolksouthern37
what's a normal map?
gotcha. well i cant seem to come up with any ideas utilizing the existing signal logic. i am afraid you are bound by the limit of the tracksearch graph. sorry 
this has made me think of some ways to counter the problem in my own signal logic that i wrote based on the original system, i just made it so it was a bit smarter.
this has made me think of some ways to counter the problem in my own signal logic that i wrote based on the original system, i just made it so it was a bit smarter.
gotcha. well i cant seem to come up with any ideas utilizing the existing signal logic. i am afraid you are bound by the limit of the tracksearch graph. sorry
this has made me think of some ways to counter the problem in my own signal logic that i wrote based on the original system, i just made it so it was a bit smarter.
Thank for trying though. I might just have to put in an additional passing loop with signals. It's not protypical to my route but at least AI traffic flows smoothly.
Are you the creator of the searchligh signals on Jointed Rail?
norfolksouthern37
what's a normal map?
yes i am. also a signal junkie, and i like writing signal logic for trainz. threads like yours usually get me writing new code for a few days.
I thought so. Those are awesome! There's nothing close out there.
I got another question or two. This should be much easier to answer. The route I'm working on is an old Frisco route that was dark territory. I want to use invisible signals for AI purposes and accurate modeling. I can't find an invisible right or left diverging signal. Am I missing something?
I got another question or two. This should be much easier to answer. The route I'm working on is an old Frisco route that was dark territory. I want to use invisible signals for AI purposes and accurate modeling. I can't find an invisible right or left diverging signal. Am I missing something?
norfolksouthern37
what's a normal map?
right or left diverging is simply for show. the AI can only see 3 signal states, STOP, PROCEED, and CAUTION. anything other than that is for us humans.
norfolksouthern37
what's a normal map?
it does. i am not sure how to read your question.
right or left diverging is simply for show. the AI can only see 3 signal states, STOP, PROCEED, and CAUTION. anything other than that is for us humans.
Thanks. I started thinking about that that possibility after I posted. Hey you learn something new everyday.