The latest Trainz Plus patch 106618 dies every time editing BCSJ-507 route

horacefithers

New member
Wow! I got this via email this morning...



Trainz Plus 2020 Q1 Update - Official Release

The latest update for all our Trainz Plus and Gold Class members is now available on PC and Mac.
PC Build 105100 to 106618 (single patch)
Mac Build 105113 to 106755 (2 step patch)

This update includes Effect Layer tools and Bulk update/replace tools.To keep up to date on all the Trainz Plus updates, visit our Trainz Wiki here.

To update to the latest version, ensure your Patch Stream is set to "Trainz Plus".




Because I'm an idiot, I downloaded and installed the new version Trainz Plus version 106618.
Big mistake. Now Trainz dies every time when I try to Edit Route BCSJ-507.
Well, it does manage to enter Surveyor when editing Kickstarter County 2.
I rebooted my computer. Still dies before the Surveyor window appears.

I'm soo happy.... but I guess somehow I'm not surprised.

Horace (spittin' feathers) Fithers


 
I have no problem editing version 508 of the route in Build 106618. The only issue I get for now (as reported in the dedicated thread in Freeware section) is when resuming a driver session.
 
I had run routes with 106618 in my 19Plus beta install with no problems. Obviously so had many others as in beta testing there were no reports or a major issue such as this.

Interestingly, I had loaded your route in 105100 (previous plus release) with no issue.

This morning I patched to 106618 in my "main" install that contained your route. With your route and 106618 I too got a CTD. With 106618 I've had no problem with what few routes I've tried so far.

EDIT - UPDATE
The Route had crashed when I attempted to load a session I had setup previously in 106618. I just created a new session with no problem. So the crash came in opening an existing session.
 
Last edited:
It's not the processor, I don't think.

Try a database repair. You do this from the Launcher. If it doesn't do the trick, I recommend running an EDR. Extended Database Repair.

In both cases, close down Trainz.

At the Launcher click o the Developer menu item on the menu bar.
Press CTRL and click on Rebuild Database.

WARNING! This process can and will take a very long time depending upon how much data you have installed. This process maybe followed by a lengthy asset validation process as well. There may also be faulty assets listed in Content Manager. These maybe simple validation false errors, and the process of viewing errors and warnings usually removes the faulty status. If the faulty status does not go away, exit, wait a minute and start up again and check the faulty assets in Content Manager.

Opening up a route for the first time may take longer than normal. This is a normal process since the cached information has been deleted during the EDR. Once a route is opened in Surveyor, clicking on one of the side menus may take longer since there is further caching taking place. They menus will appear empty and caching will appear up on the top of the screen. Wait patiently for this to complete and the data will return to normal.
 
Multiple database repairs and an extended database repair did not solve my problem.

Neither did rebooting my computer and starting Trainz without any of the other junk running that I usually have.

The problem did not occur looking at a session. Trainz loses it mind and vanishes when used Edit Route at about the time the Surveyor interface should appear.

I was able to successfully edit an older version, BCSJ-502, route. 0f

I tried Edit session on BCSJ-507 and that also failed.

I have not yet checked compatibility mode.

The fact that some have reported being able to Edit Route the BCSJ-507 route while others are having troubles plus the heap-of-consists-in-the-paddoc makes me wonder if yet another database corruption has crept into the route.
Horace (overwhelmed in Oregon) Fithers
 
HF, maybe you should ask someone that has edited your route successfully to send their copy to you.
 
Last edited:
Previous version 501 from about a week earlier than BCSJ-507 loads into Surveyor and I can edit and save.

I was noticing there was considerable bloat in the config.txt file's String Table for assets of type Train. I had several trains configured in the Surveyor/F7/C area. So I tried removing them to see if that would reduce the String Table size.

It didn't.

But then 106618 came along and Surveyor started crashing before the Surveyor window appeared.

Now I'm remembering that the Portals at the east and west end of the route were generating trains using the pre-existing, named consists. As in the one's I'd removed.

This was not a problem in 105100. However, I'm curious as to whether this might be causing the problem in 106618 (I added this to the bug report ticket).

When I looked in BCSJ-501, I found that the consists were still present! Because versions after this don't load without Surveyor crashing I can't check this in 507.

Still, it seems a bit curious?

Also, there were a few trains in the session-layer in Salem that were built from these consists. I think though, that these were in the session file's session-layer.

Is it a violation to remove a consist definition after creating instances of a consist?

I was under the impression that custom consists listed in Surveyor/F7/C were not particular to a route, but were global for all of a user's custom routes. It seems this must be an incorrect assumption or why would they still be present in BCSJ-501?

Curioser and curioser,

I copied BCSJ-501 into a new route and have been removing the portals (and all Train type assets from the route file regardless of their layer). I wonder if that will behave better with respect to the derailed consist party that was occurring in the field north west of Lincoln for some people?

Horace Fithers
 
Something's munched in the route. It crashes on my Intel-based system so no it's not your AMD chip.

This initial crash is usually caused by a bad scripted object being initialized at start up. The problem is opening the route to find out the culprit and remove it.

Since the previous version works, what are the changes between 501 and the 507?

I tested your previous versions under the beta and they worked fine in that version so something's different here.
 
Thank you for confirming that it's not some strange aberration in the AMD threadripper instruction set causing random issues...

HF

It's also crashing on edit in an i7 6700K / 1080TI setup, so not related to Hardware, downloaded to have a look. Will let you know if I can find anything.
 
HF,

re consists:

In my experience, there are several types of consists.

One or more train cars coupled together on a layout become a consist. These cars are fully instantianed objects with a name and loading data, and I will refer to them as type consist_1. When the game is running, they are assigned a name automatically, such as consist1, consist2, etc. They persist through a save.

In the edit trains tab, you can select a type consist_1 with the consist selector and name it. This type consist_2 then becomes available for any route or session. I don't think the loads are preserved, I think the named consist is just a list of kuid's for the cars. I don't know if it is automatically uploaded to the DLS or is just local to the users installation. So when this consist_2 is placed on a track, it becomes a consist_1 without loading data.

Another type of consist arises when you define one in a portal. I did not write any portals, but I believe they are just a list of kuids. Call these type consist_3. When this consist succesfully emerges from the portal, it becomes a type consist_1, but without any loading data. That is why there are so many instant load type commands.
 
My non-crashing version is 508. What is the difference between 501, 506, 507 and this one?
 
I'm having the same problem. 506, 507, and 508 all crash 106618, as does the Levi & Appalachian. I can load any routes I had installed PRIOR to upgrading to 106618 with no problem.
 
Something's munched in the route. It crashes on my Intel-based system so no it's not your AMD chip.

This initial crash is usually caused by a bad scripted object being initialized at start up. The problem is opening the route to find out the culprit and remove it.

Since the previous version works, what are the changes between 501 and the 507?

I tested your previous versions under the beta and they worked fine in that version so something's different here.

iirc I started playing around with signalling and enhanced interlocking towers after version 501.

I wonder if the EITs are the problem?

HF
 
iirc I started playing around with signalling and enhanced interlocking towers after version 501.

I wonder if the EITs are the problem?

HF

I was thinking that too. Maybe there's too many scripts and things got stuffed. Just guessing. I'm thinking how this can be undone. I might try merging your route into a blank one of mine for kicks.
 
I'm having the same problem. 506, 507, and 508 all crash 106618

Strange as I am running Build 106618 too and can edit the route (version 508) without any issue (my 106618 isn't the one-step patch from 105100 but the last of several successive beta patches).
 
Back
Top