[rrfan96]

Hello all. So lately I've been running into an issue, a really weird one. To sum it up, usually about between 10-20 minutes into a run all of the sudden the train would stop and the speed would read (on all displays) "NaN". That was with the "Version 1" release. Curious I updated to the latest build today. Didn't have an issue...but then upon running a grain train everything broke...this time horribly. The cabview went black, the entire train disappeared, and the gradient and speed readings read "NaN". The route is the Version 1 of the BNSF Scenic Sub. The locomotives...well it was happening with pretty much everything. A deksoft unit, an SLI unit, and a Lou Volland loco. All the locos are stable, and the freight cars are stable as well. I haven't touched the settings, and it has happened with and without the advanced adhesion model.

 OpenRailsLog.txt (24.82K)
Number of downloads: 168






P.S. I've had no issues with the Allegheny Rails V2.0 and FBL Commuter Edition routes. Which were the only ones I've used after downloading OR V1.0. Is this just a route issue?

[Lindsayts]

In computing terms "NaN", (Not a Number), is a particular error flagged by a floating point processor which usually results from it (the FPU) being sent nonsense data. In OR it would usually indicate some kind of corrupt data is around.

Lindsay

[rrfan96]

Lindsayts, on 02 July 2015 - 03:00 PM, said:

In computing terms "NaN", (Not a Number), is a particular error flagged by a floating point processor which usually results from it (the FPU) being sent nonsense data. In OR it would usually indicate some kind of corrupt data is around.

Lindsay

That makes some sense actually...just after I posted this I went again trying to grab a screenshot when the game crashed and told me about a shape error of some kind.

[R H Steele]

Learn to use your OR log to find solutions to problems when they arise....it's a wonderfully rich resource.

Just a quick scan of the log shows many missing sound files...that should be corrected. Once corrected the sound you experience in OR will be improved

The sound stream errors are not so important...just a mistake in adding up the total number of streams within a specific .sms file.

The "Cannot parse the constant number 27.2.2 in C:\1MSTS\trains\trainset\PGWR_Grain\PGWR_LO_100312_LD.wag:line 58" is easily fixed....goto line 58 in that wag file and change the 27.2.2 to read 27.2 >>> this is a common error in SLI equipment.

[rrfan96]

R H Steele, on 02 July 2015 - 03:31 PM, said:

Learn to use your OR log to find solutions to problems when they arise....it's a wonderfully rich resource.

Just a quick scan of the log shows many missing sound files...that should be corrected. Once corrected the sound you experience in OR will be improved

The sound stream errors are not so important...just a mistake in adding up the total number of streams within a specific .sms file.

The "Cannot parse the constant number 27.2.2 in C:\1MSTS\trains\trainset\PGWR_Grain\PGWR_LO_100312_LD.wag:line 58" is easily fixed....goto line 58 in that wag file and change the 27.2.2 to read 27.2 >>> this is a common error in SLI equipment.

Alright then :sign_thanks: . Thank you sir. :furious: I have some work to do obviously. :)

[R H Steele]

rrfan96, on 02 July 2015 - 03:59 PM, said:

Alright then :sign_thanks: ...I have some work to do obviously. :)

You are very welcome. I chip away at the errors and warnings, each time I load a route. There is plenty of work to do. I cannot say for sure that all the missing sound files would eventually cause a failure in OR - like you experienced. It would take someone with knowledge of the code to determine that.

[James Ross]

rrfan96, on 02 July 2015 - 02:35 PM, said:

P.S. I've had no issues with the Allegheny Rails V2.0 and FBL Commuter Edition routes. Which were the only ones I've used after downloading OR V1.0. Is this just a route issue?

Different routes and rolling stock may trigger the problem, but any NaNs showing up (and the joint weird visual behaviours) are always OR bugs. They can be caused by calculations which are unstable or have very unusual inputs but we should be ensuring stability wherever we can and at least logging very unusual inputs as warnings, if not outright ignoring them.

[disc]

I've seen some NaN-s too, only restart of the activity solved it. Full restart, as save-load brings back the NaN.

[Hamza97]

Yes, I encounter NaN on following occasions :

1) When I try to add ORTSEngine() defination to an .eng file....
The loco disappears alongwith its consists, although pressing F7 shows the PLAYER tag at top of now invisible loco.

2) When I time accelerate greater than 4x.....
The train doesn't disappear, although no Tractive or braking forces appear to work and train gradually loses speed.

The only solution is restarting the activity or explorer mode.

[Csantucci]

Hamza97, on 09 July 2015 - 01:36 AM, said:

Yes, I encounter NaN on following occasions :

1) When I try to add ORTSEngine() defination to an .eng file....
The loco disappears alongwith its consists, although pressing F7 shows the PLAYER tag at top of now invisible loco.

2) When I time accelerate greater than 4x.....
The train doesn't disappear, although no Tractive or braking forces appear to work and train gradually loses speed.

The only solution is restarting the activity or explorer mode.

I also got a NaN time accelerating very much. The NaN referred to the speed value: in the cabview speed digital indicator the speed value disappeared, and a big, big NaN strign appeared. Obviously I wasn't able to reproduce this.