Shunting signal coding direction bug

Use this forum if you need help using TSRE.
Post Reply
Archimage
Posts: 115
Joined: 19 Feb 2019, 12:17
I have a problem with 2 shunting signals in one of the stations. They code the signal in the false direction in the game even in TSRE all looks good. RTS Dispatcher just reads the tdb file so i dont think there is a problem with it. I tried to replace the signals and still the newly placed signals on that track have the same bug. As far as i understand that signals just are not updated in TDB after they were flipped. But still in TSRE they look as they should be.
What can cause it? How to fix? Can it be fixed manually in TDB by hands?
You can see the attached screenshots. One from TSRE another from RTS Dispatcher (it just reads the tdb). The bugged signal in the blue circle, the green arrow is where it SHOULD code the signal and the red arrow is where it codes cause it is bugged somehow.
https://imgur.com/a/7v3JCRJ

User avatar
Goku
Site Admin
Posts: 357
Joined: 15 Jan 2019, 18:10
Location: Poland
Contact:
What happens if you flip it and it is facing other direction in TSRE? Does it faces different direction in RTS?

Archimage
Posts: 115
Joined: 19 Feb 2019, 12:17
Goku wrote:
22 Feb 2021, 13:47
What happens if you flip it and it is facing other direction in TSRE? Does it faces different direction in RTS?
nope. nothing happens. it always use the "red arrow" direction (see screenshot). And when i place it on that vector it uses that direction by default when placed (before flip). It seems to me that "flip" somehow is not written somewhere in tdb for the shunting signal on that track or tit even it looks flipped in TSRE after flip.

Archimage
Posts: 115
Joined: 19 Feb 2019, 12:17
I met that bug a few times. I can say that its more likely to happen when signal is placed on a short section between junctions (less than 10 metres, for example 5-metre section between 2 junctions as you see in screenshots), but a week ago a met that bug on a 15-metre section...
Replacing the signal, flipping lots of times or trying to place it anywhere else on that "oddy" track section doesnt help...
The only decision i found is a bad one - just rewrite (Z) the whole vector (the straignt section and 2 junctions connected with it) and trying to place again. But in this case its a very bad decision cause more than 20 other signals would be screwed after rewriting the track sections cause linked through that junctions....

User avatar
Goku
Site Admin
Posts: 357
Joined: 15 Jan 2019, 18:10
Location: Poland
Contact:
Maybe it's a RTS bug? There is something like minimum distance from junctions in MSTS ..
Check if TDB entry for this signal changes if you flip it.

Archimage
Posts: 115
Joined: 19 Feb 2019, 12:17
Goku wrote:
23 Feb 2021, 10:18
Maybe it's a RTS bug? There is something like minimum distance from junctions in MSTS ..
Check if TDB entry for this signal changes if you flip it.
TDB entry? how to check it? how to find that exact signal in tdb?
and i have an exit signal in other station that is 3 metres before the junction and it works ok and no bug...
as i know RTS just reads the tdb... and is there a way to fix the signal manually in tdb?

Archimage
Posts: 115
Joined: 19 Feb 2019, 12:17
I understood how to find a signal in tdb. And made a bug report. You can check the difference between two states (when signal is not flipped and when its flipped) and the state of its TrackNode in my text file: attached.
Attachments
shitty signal.rar
(465 Bytes) Downloaded 188 times

User avatar
Goku
Site Admin
Posts: 357
Joined: 15 Jan 2019, 18:10
Location: Poland
Contact:
TDB looks fine:
TrSignalType ( 00000000 0 4.2651339 T_Small_WB )
TrSignalType ( 00000000 1 1.123541 T_Small_WB )
As you can see, after flip, the second value changed from 0 to 1 indicating other direction and signal direction in radians was modified by 3.14 radians - 180 deg.
Try some hacks on these values, but it's probably RTS bug.

Post Reply