P3D 4.1 CTD near Labrador

[av8r172 2]

Is anybody getting this crash running Pointsoft's Pro ATC X?

[av8r172 2]

Okay I find this very interesting. I run the aero.sors.fr update, but before I do, I use the "check conflicts" function. It spits out a long txt file containing existing conflicts between the scenery.cfg and addon bgl files that contain navdata aka FTX vector. There are a whole lot of American airport on the list, but the last few dozen entries are Canadian. Several of which are on the eastern coast.

 

ZJT  (NDB)     Freq: 340.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYJT.bgl]
ZMN  (NDB)     Freq: 366.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYQM.bgl]
ZNF  (NDB)     Freq: 270.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYYT.bgl]
ZQM  (NDB)     Freq: 304.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYQM.bgl]
ZSM  (NDB)     Freq: 286.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYAM.bgl]
ZST  (NDB)     Freq: 397.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYSJ.bgl]
ZXJ  (NDB)     Freq: 246.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYXJ.bgl]
ZXY  (NDB)     Freq: 353.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYXY.bgl]
ZYB  (NDB)     Freq: 404.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYYB.bgl]
ZYT  (NDB)     Freq: 246.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYYT.bgl]
ZZD  (NDB)     Freq: 308.0  [Path: C:\Program Files\Lockheed Martin\Prepar3D v4\ORBX\FTX_VECTOR\FTX_VECTOR_APT\scenery\APT_CYEG.bgl]

Edited July 13, 2018 by av8r172

[G-YMML1 582]

To my knowledge PMDG does not use default Navaids, rather those from their own database. Correct me I'm wrong

[G-YMML1 582]

18 minutes ago, av8r172 said:

Is anybody getting this crash running Pointsoft's Pro ATC X?

I do, but not on the long-hauls.  Zero issues

[av8r172 2]

2 minutes ago, G-YMML1 said:

I do, but not on the long-hauls.  Zero issues

Are you using FTX Vector?

[G-YMML1 582]

3 minutes ago, av8r172 said:

Are you using FTX Vector?

Yes, the whole sandwich - FTX Base, Vector, OLC NA and Europe. 

[av8r172 2]

Just now, G-YMML1 said:

Yes, the whole sandwich - FTX Base, Vector, OLC NA and Europe. 

I am very interested in these four .bgl's that were identified as duplicates between vector and default.

CYJT
CYQM
CYYT
CYSJ

I think I will try to disable the vector ones (use default bgls for those airports) and try to recreate this.

[G-YMML1 582]

Vector itself (and Regions especially) have made a LOT of changes to default *bgls by adding new bgls, not modifying the defaults and if this would have been a culprit for CTD, it would be uproar all over the flighsimming boards already. 

[Bob Scott 7,718]

Years ago, I wrote an FSUIPC-based utility called QuantumLeap that should still be in the AVSIM library.  It allows you to reposition the jet to any named navaid/waypoint in the Navigraph PMDG waypoint database, as well as resetting the time.  That's how I fly the long legs anyway--takeoff heavy, climb to cruise, coast out, then use QL to reposition myself ~200 nm short of destination for the descent/approach/landing.  So for my EU-to-CONUS legs, QuantumLeap lets me avoid trolling for CTDs in the Newfie Triangle.

Thinking back many years about similar vexing problems, I recall a few pernicious issues in previous ESP platforms:

1. Multiple ATC facilities visible on the same frequency from high altitude (ATC settings, and where are your radios set?)

2. A bad texture on a region-specific AI acft that spawned or came into view in the problem area.

3. Multiple AFCAD/ADE files or a bad AFCAD/ADE file in the problem area

I think the ORBX FTXGlobal NA Freeware airports include a few buffed-up airfields up in Newfoundland, and also MyTraffic includes a bunch in its scenery folder.  I had to do a lot of work to deconflict those two ADE sets.  If using a utility like ScruffyDuck's Simple Airport Scanner with P3Dv4, you'll want to use Lorby-SI's Addon Manager to temporarily consolidate all your sceneries from the add-ons.cfg files together with those in the scenery.cfg while you scan for duplicates.  It's an option on the backups tab, IIRC.

Regards

 

[G-YMML1 582]

1 hour ago, w6kd said:

Years ago, I wrote an FSUIPC-based utility called QuantumLeap that should still be in the AVSIM library.  It allows you to reposition the jet to any named navaid/waypoint in the Navigraph PMDG waypoint database, as well as resetting the time.  That's how I fly the long legs anyway--takeoff heavy, climb to cruise, coast out, then use QL to reposition myself ~200 nm short of destination for the descent/approach/landing.  So for my EU-to-CONUS legs, QuantumLeap lets me avoid trolling for CTDs in the Newfie Triangle.

 

 

Bob,

Sounds like that I have been wanting all these years. Do it  work with P3D v4.3? I'd like to give it a try.

Thanks

PS. Searching by your name returns zero results 😞

Edited July 13, 2018 by G-YMML1

[G-YMML1 582]

Found it!

quantumleapv1.1.zip

I will test it later tonight under P3Dv4.3 conditions

 

[Bardia 11]

19 hours ago, johnnycrockett said:

Hands down, you've just made the most constructive comment in here.

Truthfully though, regardless of whether GSX is or isn't the problem, a fix has been offered so just enjoy the good customer service and keep flying.....

As a paying customer, I don’t remotely consider this to be good customer service. 

We went through vector, navaids, etc. etc. many pages ago and they all got affirmatively eliminated through verifiable means. GSX remains the only plausible offender.

Devoper must be held accountable by the community. Lousy excuses dished out by the dev make very little sense. 

[Bob Scott 7,718]

1 hour ago, Bardia said:

As a paying customer, I don’t remotely consider this to be good customer service. 

We went through vector, navaids, etc. etc. many pages ago and they all got affirmatively eliminated through verifiable means. GSX remains the only plausible offender.

Devoper must be held accountable by the community. Lousy excuses dished out by the dev make very little sense. 

First, nothing has been verified until somebody does a fairly sophisticated audit of the processes and files in use at the time of these crashes.  People are operating over this area with GSX and without experiencing these crashes...how does that fit into "verifying" that GSX is the only plausible source of the problem?  The data is too inconsistent for the logic to track convincingly, much less conclusively...NOTHING has been proven here. 

I remember a story once told by a former VietNam POW...when he was brought into the office (hut) for interrogation, the single lamp on the desk would flicker, and the guards would scurry over to straighten out the power cord, thinking that electricity wouldn't travel through a bent/kinked wire as water might not travel through a kinked hose.  In the process of moving the power wire around to straighten it, it would jiggle the bare wires where they were partially inserted into the outlet, they would again make contact, and the lamp would work once again.  Those guards repeatedly reinforced a completely false belief on the physics of electricity based on a false assumption of causation, but had you asked one of them, I'm sure they'd tell you they could prove beyond a doubt that bending that wire would stop the flow of electricity.  Similarly, no smoking gun has been described here that clearly and without question establishes causation between GSX and these CTDs. 

If a CTD occurs because, in the course of routine ops, GSX asks the sim to read scenery bgls in a standard way--a normal process request--and in so doing while flying in this area the sim then finds a corrupt file that crashes it, the problem is with the corrupt data (and how the sim badly handles the error) and not with GSX per se.  The attempted corrective action Umberto has provided is a change in the way GSX requests data that tries to minimize the possibility that the sim will encounter a bad scenery bgl in the problem area, if (and it's a big if) that's the problem at all.

I, for one, appreciate Umberto's effort.  And lousy accusations are at least as bad as lousy excuses.

Regards

 

[KL Oo 698]

2 hours ago, Bardia said:

As a paying customer, I don’t remotely consider this to be good customer service. 

We went through vector, navaids, etc. etc. many pages ago and they all got affirmatively eliminated through verifiable means. GSX remains the only plausible offender.

Devoper must be held accountable by the community. Lousy excuses dished out by the dev make very little sense. 

How do you/we know that GSX is definitely the problem? I know there are a number of users that claim they can avoid the crash by restarting/closing COUTL or GSX but there are plenty for whom that does not fix the issue.

 

Umberto took the time to alter his software, despite not being sure that it was causing the problems (from what I can tell of his responses). I for one am grateful that he did so - Not sure we'd see the same level of support from other developers.

 

[ZKOKQ 984]

3 hours ago, Bardia said:

As a paying customer, I don’t remotely consider this to be good customer service. 

We went through vector, navaids, etc. etc. many pages ago and they all got affirmatively eliminated through verifiable means. GSX remains the only plausible offender.

Devoper must be held accountable by the community. Lousy excuses dished out by the dev make very little sense. 

Some have CTD without having GSX installed or disabled. Umberto from FSDT is one of the devs who I have a lot of respect for, and if you read some of the rude people posting on his forum, I am surprised he still is interested in developing.

I think what you just posted fits into that category.