Upcoming Games

5L32 on a spawning spree

You are here: Home > Forum > Simulations > Released > Oxford PSB > 5L32 on a spawning spree

Page 1 of 1

5L32 on a spawning spree 21/07/2017 at 12:26 #96521
AlexH
Avatar
85 posts
Running the 0445 timetable 2009.

5L32 arrive in platform 3 to divide into 2L32 and 2L02. It just goes insane and spawns 8 or so 2L32 and never 2L02.

Eventually I had to turn one of the 2L32s into a 2L02 and delete the spawning 5L32

Log in to reply
5L32 on a spawning spree 21/07/2017 at 12:29 #96522
Stephen Fulcher
Avatar
1352 posts
That's odd, what TT version are you using and do you have a save?

Edit: Just had a quick look at the timetable for 5L32 on the 0445 start v1.1 and can't see anything wrong that should cause that.

Last edited: 21/07/2017 at 12:35 by Stephen Fulcher
Reason: Extra info

Log in to reply
5L32 on a spawning spree 21/07/2017 at 18:19 #96529
VInce
Avatar
202 posts
Hi all,

I don't know if this adds anything to the issue but I've had a similar thing on the Exeter sim a number of times when trains have been dividing and/or joining.

A window pops up saying there has been an "access violation" and the trains list begins to show multiple examples of the trains that are dividing or joining.

For instance, I've recently revised a 1981 timetable for my own use (designed for an earlier version of Simsig) and implemented a detachment of vehicles at Exeter off 1B06 newspaper train. After the "access violation" window pops up and I dismiss it, I've seen as many as 10 1B06 entries and a similar number of 0S00 entries on the train list. If they are removed then then they just appear again.

I want to emphasise that this does not happen every time by any means. Indeed if I stop the sim, revert to a snapshot taken 10 minutes before and run it again then this doesn't happen. Until the next time! It will be very hard to pin down something like this which presumably depends on series of events happening a particular order!

I'll try and get a screen image of what exactly the "access violation" warning says next time it happens.

Vince

Last edited: 21/07/2017 at 18:23 by VInce
Reason: None given

Log in to reply
5L32 on a spawning spree 21/07/2017 at 18:38 #96530
mark265
Avatar
493 posts
Online
AlexH in post 96521 said:
Running the 0445 timetable 2009.

5L32 arrive in platform 3 to divide into 2L32 and 2L02. It just goes insane and spawns 8 or so 2L32 and never 2L02.

Eventually I had to turn one of the 2L32s into a 2L02 and delete the spawning 5L32
VInce in post 96529 said:
Hi all,

I don't know if this adds anything to the issue but I've had a similar thing on the Exeter sim a number of times when trains have been dividing and/or joining.

A window pops up saying there has been an "access violation" and the trains list begins to show multiple examples of the trains that are dividing or joining.

For instance, I've recently revised a 1981 timetable for my own use (designed for an earlier version of Simsig) and implemented a detachment of vehicles at Exeter off 1B06 newspaper train. After the "access violation" window pops up and I dismiss it, I've seen as many as 10 1B06 entries and a similar number of 0S00 entries on the train list. If they are removed then then they just appear again.

I want to emphasise that this does not happen every time by any means. Indeed if I stop the sim, revert to a snapshot taken 10 minutes before and run it again then this doesn't happen. Until the next time! It will be very hard to pin down something like this which presumably depends on series of events happening a particular order!

I'll try and get a screen image of what exactly the "access violation" warning says next time it happens.

Vince

It may help the developers if you are able to post as save or the "SimSig Loader0.log" (can be found in the Public folder - public documents - simsig, or C:\Users\Public\Documents [on Windows 7, cannot speak for other OS's])

The SimSig Loader0.log contains details of a session and if I remember correctly "access violations" are writern there (dev's please correct me if I am wrong).

Last edited: 21/07/2017 at 18:39 by mark265
Reason: edited

Log in to reply
The following user said thank you: VInce
5L32 on a spawning spree 21/07/2017 at 19:04 #96531
Meld
Avatar
303 posts
Having run the 0445 start 5 times this afternoon I have not had this happen at all.

5/5 successful splits and both services departing as booked

Log in to reply
5L32 on a spawning spree 21/07/2017 at 20:13 #96533
VInce
Avatar
202 posts
[/quote]

It may help the developers if you are able to post as save or the "SimSig Loader0.log" (can be found in the Public folder - public documents - simsig, or C:\Users\Public\Documents [on Windows 7, cannot speak for other OS's])

The SimSig Loader0.log contains details of a session and if I remember correctly "access violations" are writern there (dev's please correct me if I am wrong).[/quote]



Here's the whole Loader0.log.

To my untutored eye, there seems to be something amiss with the following entries :-

18/07/2017 at 11:43:52

18/07/2017 at 00:03:58

Hope this means something to someone..

Vince

Post has attachments. Log in to view them.
Last edited: 21/07/2017 at 20:14 by VInce
Reason: None given

Log in to reply
5L32 on a spawning spree 21/07/2017 at 20:49 #96534
clive
Avatar
1575 posts
Online
Raised as Mantis 17663. The log contains some information that I can look at. It appears you've had problems with Peterborough, Derby, and Exeter.

Access violation *always* indicates a bug in the core code, not the specific simulation.

Last edited: 21/07/2017 at 20:50 by clive
Reason: None given

Log in to reply
The following user said thank you: VInce
5L32 on a spawning spree 21/07/2017 at 21:36 #96535
clive
Avatar
1575 posts
Online
Unfortunately the information in the log isn't good enough to help. I can see that something happened in the train dividing process, but can't see what (and there's a lot of code).

I *suspect* something happened earlier and this is just showing the effects. I'm not sure how much help saved games will be, especially since the problem doesn't repeat from them. But please try and get as many saved games around the problem as possible, as they may help. Detailed notes as to what else happened with the train *before* the problem occurred would help as well.

Ideally would be something reproducable, even if the save is 30 or 40 minutes before the problem. Without it this is going to be a pain to investigate.

Oh, *EXACT* messages are important.

Last edited: 21/07/2017 at 21:37 by clive
Reason: None given

Log in to reply
5L32 on a spawning spree 21/07/2017 at 23:35 #96536
AndyG
Avatar
1718 posts
Online
Clive,
I also had this on York NS whilst testing a MGR TT at Gascoigne Wood (I assumed it was a activity error in the WTT), I'll see if I can get it to happen again.
FWIW, as it's only recently occurred, could it be a subtle change in the latest loader?

I can only help one person a day. Today's not your day. Tomorrow doesn't look too good either.
Log in to reply
5L32 on a spawning spree 22/07/2017 at 02:02 #96538
VInce
Avatar
202 posts
Clive,

Thank you - I'll keep a saved game next time, although as you say, there is no guarantee it will happen the next time it is run. I'll also take a screenshot of the "access violation" window as well....

Regards

Vince

Last edited: 22/07/2017 at 02:08 by VInce
Reason: None given

Log in to reply
5L32 on a spawning spree 22/07/2017 at 16:42 #96549
VInce
Avatar
202 posts
Clive,

Attached is a saved game where it happened once - I can't get it to happen again and I've run it three times since.

The sequence of events is as follows :-

1B04 enters at Castle Cary at 0216 and arrives at Taunton P2 at 0240.

The train divides and departs at 0245 non stop off the sim to Ivybridge.

0B06 leaves the Dock Siding at Taunton, reverses at E483 signal which is cleared with a subsidiary aspect on to P2 to attach the vehicles off 1B04 and form 1B06.

1B06 departs at 0256 to Exeter P4 where the train arrives at 0328 and a shunt engine 0S00 leaves Exeter Riverside onto the rear of the train to form 3B06, as a shunt with the one detached van to P2. The van is detached in P2 and left there. 0S00-1 then returns to Riverside Yard via Exeter New Yard, Exeter St Davids and Cowley Bridge.

1B06 then makes a new train 1B06-1 (this is done so as I could get the train lengths right in the analyser, as 1B06 then subsequently makes 1B10 at Newton Abbott)

1B06-1 departs at 0336 calling at Teignmouth before arriving at Newton Abbott P1 at 0403 and forming 1B10.

At 0340 a DMU 5B09 leaves Exeter West Yard at 0340 and runs via 677 sig (reverse) and Exeter North to P2, attaches the van as tail traffic and 1B09 departs at 0405 for Barnstaple.

The point at which the "access violation" warning happened was when 1B06 was in Exeter P4 with the shunt loco attaching to the detached van, and 1B06 was in the process of becoming 1B06-1. The trains list still showed 1B06 as "dividing trains" and 0S00 had not yet become 3B06.

I don't know whether this will help at all but we can but try.

Vince

Post has attachments. Log in to view them.
Last edited: 22/07/2017 at 20:04 by VInce
Reason: None given

Log in to reply
5L32 on a spawning spree 24/07/2017 at 02:43 #96577
BarryM
Avatar
1830 posts
Online
I did some testing of Exeter. I could not get an Access-Violation.

Barry

Barry, Sydney, New South Wales, Australia
Last edited: 24/07/2017 at 02:46 by BarryM
Reason: a

Log in to reply
5L32 on a spawning spree 24/07/2017 at 07:56 #96579
VInce
Avatar
202 posts
Barry,

As Clive said (above) it has nothing to do with which sim is used, there is very, very hard to find bug in the core code.

Its been documented by a number of users and is not repeatable.

See here for confirmation

https://www.simsig.co.uk/Forum/ThreadView/43823

Vince

Log in to reply
5L32 on a spawning spree 24/07/2017 at 10:57 #96580
kaiwhara
Avatar
439 posts
Strangely I had the same thing when testing Marylebone Yesterday, and I did get an Access Violation as well. Ended up with the same train on P1 at Banbury times 13...

I'll have a dig for saves in the morning.

Sorry guys, I am in the business of making people wait!
Log in to reply
5L32 on a spawning spree 24/07/2017 at 17:48 #96586
AlexH
Avatar
85 posts
Hi,

I've been travelling and just seen this.

Sim is v1, timetable is 2009 0445 start v1.

I played through from there and not sure I have a save anymore from that period. Sorry about that, but it seems like it's a problem which others have suffered with as well!

Thanks

A

Log in to reply
5L32 on a spawning spree 16/10/2017 at 08:31 #102302
clive
Avatar
1575 posts
Online
I hope people will be glad to hear that the root cause has been located and fixed.

My special thanks to the user who sent me a detailed dump file that allowed me to find out what was going wrong.

(For those interested, there are two objects within train timetables that are supposed to point to each other. When the timetable was being copied, *sometimes* the copy of one object pointed to the original of the other. As a result, if the original was deleted, there was a flailing pointer. I'm still not clear why dividing made it appear, but I think that's just luck.)

Log in to reply
The following users said thank you: kaiwhara, postal, VInce, Hawk777, Lyn-Greenwood, BarryM
5L32 on a spawning spree 16/10/2017 at 08:37 #102305
kaiwhara
Avatar
439 posts
Not going to complain!
Sorry guys, I am in the business of making people wait!
Log in to reply
5L32 on a spawning spree 16/10/2017 at 13:50 #102307
VInce
Avatar
202 posts
clive in post 102302 said:
I hope people will be glad to hear that the root cause has been located and fixed.

My special thanks to the user who sent me a detailed dump file that allowed me to find out what was going wrong.

(For those interested, there are two objects within train timetables that are supposed to point to each other. When the timetable was being copied, *sometimes* the copy of one object pointed to the original of the other. As a result, if the original was deleted, there was a flailing pointer. I'm still not clear why dividing made it appear, but I think that's just luck.)
Clive,

Thank you for this - not easy to find I bet..

Regards

Vince

Log in to reply
5L32 on a spawning spree 16/10/2017 at 19:21 #102326
AlexH
Avatar
85 posts
Wow, thanks for taking the time to figure this out. Glad it's resolved.
Log in to reply