Hi James, I apologise to be a bother again. But, this time this is just a recent bug which has been discovered in-game. At Crewe, the choose signals clear trains (which don't call at Crewe) to the signal 2 blocks away (not shown on picture). The signal (which is meant to be 1 block away) at Crewe station (on the bypass line) has not cleared to a single yellow aspect (this is due to the signal off picture being at danger (red)). Could you look at this bug for me please?
Save: https://simutrans-germany.com/files/upload/UK2018-2_BACKUP.sve
Picture
(https://simutrans-germany.com/files/upload/Something_for_J._Petts.png)
I am afraid that I cannot reproduce this, since I cannot find any trains on the approach line that you show that do not stop at Crewe. May I ask - at exactly what in-game time can this be reproduced?
Erm look for Class 350 trains (in the old First Transpennine Express livery) and Class 221 voyagers which serve the Chartford Euston - Werningstone Central/Longchurch Waverley via Malborough West lines. They do not call at Crewe. Plus they go via the line where the Class 350 in the picture is.
At what in-game time does this occur next after loading the saved game that you provided?
At 3:00:49/6:24:00 in May 2015. This time from the WCML (the lines to the right of the picture).
I am afraid that I cannot see any trains not stopping at Crewe coming from the lines to the right of the picture at 3:00 in May 2015 (or thereabouts) in the saved game supplied.
Its properly because you have not played long enough
Quote from: SouthernTransport225 on August 18, 2018, 07:31:18 PM
Its properly because you have not played long enough
That makes no sense at all - I ran the game until the exact time that you gave, and a little while beyond. Which specific convoy number has the issue that you describe at 3:00:49 in May 2015 in that specific saved game?
Convoy: (24) BR Class 350/2 "Desiro" EMU. Time and date: 2:36:00-2:38:00/6:24:00 in May 2015.
At the times given, convoy 24 is nowhere near Crewe it is instead at Longchurch.
It is important to note that fast forwarding or altering time granularity may result in diverging game states due to it changing the simulation granularity. For this reason it may be important to have several saves captured around the time the instance occurs so that James can load up and literally a few seconds of normal simulation time later the bug happens.
Okay
Hi James, at Crewe (on the same map where a previous bug was not solved), the choose signal (in the Northbound direction) displays weird signal aspects, so for example. The signal would sometimes the choose signal would show a clear aspect (with/without the feather indicator on) to trains which are calling at Crewe and sometimes it would show a single yellow (as the case with the last bug (not solved)) to trains which are passing Crewe (the Glasgow Central fast services (named Train: Chartford E - Glasgow C 1 ingame)) and the signal at Crewe station (on the passing lines) is at danger and the signal protecting Polingpike Junction (south of Polingpike East station) which is at danger should be the only signal at danger (shown in the picture below) in this case, also the circles were added to show you what I mean by "the signal at Crewe station (on the passing lines) is at danger and the signal protecting Polingpike Junction (south of Polingpike East station) which is at danger should be the only signal at danger." Could you look at this bug for me please?
(https://simutrans-germany.com/files/upload/MDC.png)
https://simutrans-germany.com/files/upload/UK2018.sve (https://simutrans-germany.com/files/upload/UK2018.sve)
James? xD
There is somewhat of a queue of issues at present - this is noted, I will hopefully deal with it in due course.
I think the reason why this big occurs is because the block reservation does not route it to the tile where the signal is on. It routes the train to the tile before (the tile which does not have the signal on).
May I ask - at what specific in-game time/date does which specific signal show what specific aspect that you consider to be incorrect, and why do you believe that aspect to be incorrect at that time?
Sorry to waste time but this issue does not occur as often as before.
Does it occur at all?
Sometimes but not as often as before though.
Are you able to upload a saved game where this can be reproduced reliably at a specific (and specified) signal at a specific (and specified) in game time?
Sure. Map link: https://simutrans-germany.com/files/upload/UK2018_[OLD].sve
At what specific in-game time does what specific signal in that saved game show an incorrect aspect and in what way is that specific aspect incorrect?
I'll be honest with you, it can happen at anytime so a specific time isnt really worth giving as because as mentioned it can happen at anytime
A specific time is definitely worth giving, because without a specific time I have no way of being able to reproduce this reliably. I cannot fix a bug that I cannot reproduce reliably many times over from a saved game. I will therefore need a specific time and place, as indicated quite a few times now, in order to deal with this.
Okay, I'll play the map and then I'll see if I can get a time for you
Thank you.
Quote from: jamespetts on March 13, 2019, 10:11:21 PM
Thank you.
August 2016 at 2:24:18/6:24:00 (and open the game as I sent it (as in open the original map) as you might of missed the bug if you have saved it multiple times)
Thank you for that. May I ask at which specific signal on the map that the problem occurs at this time?
The choose signal facing trains on the down line at Crewe South Junction (towards the North and Scotland)
(http://oi65.tinypic.com/29zu139.jpg)
I am afraid that I cannot reproduce this: at 2:24:18 in August 2016 in the saved game to which you linked in this (https://forum.simutrans.com/index.php/topic,18435.msg178503.html#msg178503) message, the signal at 522,476 is correctly at danger. I get the same result at that time in this (https://forum.simutrans.com/index.php/topic,18435.msg175519.html#msg175519) saved game.
In the latter game, a few seconds before 2:24:18, a Voyager DMU is approaching Crewe and the signal is set to caution, which is also correct, as the next signal is at danger. The signal correctly does not display the feather aspect, as, in Simutrans-Extended, the feather aspect means that the train has been routed to a platform other than its default platform. In this case, the train has been routed to its default platform.
Can you explain what at that specific time and place you think is incorrect about that specific signal showing that specific aspect?
LIke a said it could happen at anytime and it only happens when a train is serving Crewe on a platform which it is not timetabled to serve
Quote from: thegamer7893 on March 19, 2019, 05:22:26 PM
LIke a said it could happen at anytime and it only happens when a train is serving Crewe on a platform which it is not timetabled to serve
Do you mean that the behaviour is indeterministic (i.e., it is different each time that the saved game is replayed from the same point without any intervention)?
However, I am not clear on what you think that the problem is exactly - the trains are not signalled through beyond the station if they are scheduled to stop at the station. The route beyond the station is only reserved once the train is ready to depart. Therefore, the signal at the end of the platform will be at danger and the signal immediately before that at the end of the platform will be at caution. This is the behaviour observed. May I ask how the behaviour that you observe differs from this?
Quote from: jamespetts on March 19, 2019, 05:50:27 PM
Do you mean that the behaviour is indeterministic (i.e., it is different each time that the saved game is replayed from the same point without any intervention)?
Yes
Without being able to reproduce this reliably at a specific time and place, it will not be realistically possible for me to fix this, or even confirm that this is an error.
Sorry.