[Docmartin]

Here you are, Doug. And thanks in advance for any help you might be able to offer.

Martin

Attached File(s)

•  ChilternTDB.zip (2.58MB)
  Number of downloads: 26

[dajones]

It turns out dynamic track sections 53611 and 53608 are both missing and they are in vector section node 17161. Here is the output from my code that lists node 17161 and the junction nodes at both ends.

 bad section 53611 in node 17161 2
 bad section 53608 in node 17161 3
 node 17164 39401 -6080 14925 2843 0
  position -6080 14925 845.33759 28.557699 570.32452
  angles 0 1.9370679 0
 node 17161
  section 0 36366 39401 -6080 14925 2843 0 2 00
   position -6080 14925 845.33759 28.557699 570.32452
   angles 0 -1.2045245 0
   curve 500 3.92634
  section 1 38500 38500 -6080 14925 2633 1 0 00
   position -6080 14925 813.8302 28.557699 583.66315
   angles 0 5.1719713 0
   straight 1
  section 2 53611 43140 -6080 14925 7358 1 0 00
   position -6080 14925 813.07208 28.557699 584.04144
   angles 0 5.1631999 0
   unknown
  section 3 53608 43140 -6080 14925 7358 1 0 00
   position -6080 14925 788.12012 28.557699 596.74164
   angles 0 5.2031999 0
   unknown
  section 4 38711 38678 -6080 14925 2622 1 0 00
   position -6080 14925 772.68597 28.557699 604.99023
   angles 0 5.2031999 0
   straight 22.5
  calc 1 failed len1 0.000 radius 0.000 angle 0.008771 0.5 len2 0.847
  calc 2  len1 0.000 radius 699.989 angle -0.040000 -2.3 len2 0.000
  calc 3  len1 17.500 radius 0.000 angle 0.000000 0.0 len2 0.000
 node 17129 38678 -6080 14925 2622 0
  position -6080 14925 752.84216 28.557699 615.59583
  angles 0 5.2031999 0

The lines labeled calc 1 to 3 are what my code calculates for sections 1 to 3 (0 and 4 are ignored because they are part of the switches). The calculation for section 1 failed because the track direction lines don't intersect between the end points. The calculated results for sections 2 and 3 could be used to correct the route tsection.dat file, but it looks like the angle in the TDB for section 1 is wrong. The section 0 angle is the junction node angle minus PI, so that is correct. Adding 2 PI and 3.92634 degrees to that gives 5.147188. So I think the best way to fix the problem is to delete the existing track between the switches and replace it. A 370.769 meter radius turn of 3.21 degrees and a 25.541 meter straight should fit.

Here is a screen shot that shows where this track is:

The red square is between the two unknown dynamic track sections. It is near the southeast corner of the route.

Doug

[Docmartin]

Many thanks for your research, Doug. With your information, I will see if I can fix the problem.

Martin

[Docmartin]

This is not going to be straightforward.

This screenshot shows the offending track section. Note that there is no tdb line and about two-thirds of the way along there is a rogue yellow track section pole.


This is the same area with the offending track section and the adjoining track sections removed. There are two rogue blue poles, one at each end of where the offending track section was located. These are numbered 17965 for the nearer pole and 17968 for the farther pole. Neither of these nodes appear in the track database which only goes so far as node 17962. Thus to resolve this problem it is not a question of editing the tdb file as far as I can see.

So I am stumped for the moment. Does anyone have any ideas?

Martin

[Docmartin]

I have been experimenting. I managed to edit the rogue poles out of the tdb file but when it came to replacing the track sections I had deleted each time I placed a section TSRE crashed. I am not sure what to do now.

Martin

[Docmartin]

Well, through great patience and sheer persistence I have managed to correct the problem trackwork and eliminate the warnings in the log file. I am very grateful to Doug for having identified the location of the problem, otherwise I would not have known where to start. What I had not anticipated was the need to delete and replace much of the track in the area surrounding the section that was causing the problem.

This just leaves me with the original issue where OR and the tdb disagree on the location of one track node as illustrated here:

OR log file:
Warning: {TileX:-6080TileZ:14925} track node 2832 could not be found in TDB

TDB file:

TrackNode ( 2832
TrVectorNode (
TrVectorSections ( 1 38709 39508 -6086 14924 11315 0 1 00 -6086 14924 -726.68402 36.312996 113.368 0.000733177 1.3168 6.8259496e-09 )
)
TrPins ( 1 1
TrPin ( 2831 0 )
TrPin ( 2833 1 )
)
)
Any ideas as to why OR should be looking for a track node on the wrong WORLD tile?

Martin

[dajones]

That warning message is poorly worded, in this case 2832 is a world file UiD value and not a track node number. It is trying to say that there is a TrackObj in the indicated world file that has UiD 2832 and JNodePosn populated that was not found in the TDB. So, the points of the corresponding shape will not be animated.

Doug

[Docmartin]

Thanks for your reply, Doug. Now I know where to find the problem track section. I am working on the problem but so far without success. Unfortunately it does not seem to be a simple matter of deleting the offending section and replacing it. After the deletion I am left with rogue track database and tsection lines in the blank space and so far my attempts to edit the tdb to remove them have all failed with a TSRE crash. And of course since my last post several similar warnings are now appearing in the log.

I will persist!

Martin

[Docmartin]

I have managed to delete the offending track section and all the surrounding ones and finally got rid of any rogue blue or red poles that were left. I can replace some of the track but TSRE resolutely refuses to write the new sections to the tdb file. It simply crashes on me. It seems I have stretched TSRE to its limits or found a bug in it. As a result I don't think I can take things further and will probably just have to live with the track section warnings in the log file.

Martin