SimConnect problem - no 1sec system event

[kiek 1,215]

35 minutes ago, cknipe said:

@kiek - I believe I can reproduce the issue on demand.  I need to do a few more tests, but I've just deliberately crashed PSXT (no position update received).

It's not a problem in the data feeder, it's not a problem in PSXT, it's a general, very rare, issue that DOES happen and occur in real life as far as MODE-S transponder codes go.  Hint: Whilst they are SUPPOSED to be unique, there IS a chance, that a duplicate can occur.  - that should already point you towards where I am heading with this.

I suspect you are relying on the MODE-S codes to be unique, and I suspect the errors occur when the data feed injects two (or more) aircraft with the same MODE-S.  The AI Object is removed (seems to be generated based on the MODE-S code), but PSXT still tries to update it after it has been removed from the sim (the update occurs from the other aircraft, not the object deleted, but I think SimConnect removes -all- objects with the same ID).

FYI: https://en.wikipedia.org/wiki/Aviation_transponder_interrogation_modes  Mode-S Transponders and how they can be reprogrammed,

FYI: https://www.pprune.org/atc-issues/550950-duplicate-icao-24-bit-aircraft-addresses-ssr.html an interesting forum read about duplicate Mode-S codes in real life at Heathrow.

In a nutshell, inject two duplicate aircraft (MODE-S codes), remove one (i.e. PSXT executes the DeleteObject), then go to another airport which puts you out of range and PSXT will (fail) in attempting to execute the second DeleteObject because it doesn't exist in the sim anymore.

Follow AAY1791* in test.txt (what I inject into PSXT), and watch what happens in PSXT after AAY1791 is removed at KSDF, and after I move to KJFK.  She's dead in the water.  Data feed no longer receiving position updates from PSXT (connects, but receives no data).

Logs at https://filebin.net/tcnw9wic5ldzv0ex - your latest 64bit build with trace logs.

 

This will explain why this is so rare, and it IS this rare in real life as well - but even in real life, it does cause absolute havoc from what I am reading / understanding.

Should be fixable in PSXT, or in the data feeder fairly easily, if I am right about this.

I have not yet looked into your files but have serious doubt you are dealing with the SimConnect issue.

If two aircraft have the same id only the first encountered in the data stream will be detected as new aircraft in PSXT. The data from the second one will be taken as next sample for the first. This may lead to very strange behaviour of that aircraft but there will be only one.

Will look into that in more detail tomorrow.

[cknipe 13]

Just now, kiek said:

I have not yet looked into your files but have serious doubt you are dealing with the SimConnect issue.

If two aircraft have the same id only the first encountered in the data stream will be detected as new aircraft in PSXT. The data from the second one will be taken as next sample for the first. This may lead to very strange behaviour of that aircraft but there will be only one.

Will look into that in more detail tomorrow.

Only one aircraft was visible in the SIM (correct), and the aircraft was positioned at the first (not second) position update received.  If what you are saying was true, the aircraft should have been at the last, or second position (manual injected position by me),  which was not the case.  It also doesn't explain why as soon as I went out of range of the aircraft, PSXT stopped with position updates (after it has been running stable without issues - literally for hours).

This is exactly my point though.  Only one of the two planes makes it to the sim (yes, I agree).  Once PSXT removes that first aircraft however, the "second one" is *updated*, and the SimObject obviously doesn't exist because it was removed.

I will write something up that we can all test to confirm my theory as well.  Will post it here as soon as it's done.

[kiek 1,215]

9 hours ago, cknipe said:

 If what you are saying was true, the aircraft should have been at the last, or second position (manual injected position by me),  which was not the case.

If a position update arrives within 2 seconds of the previous one it will be skipped, that's why?

Edited August 15, 2020 by kiek

[kiek 1,215]

11 hours ago, cknipe said:

This is exactly my point though.  Only one of the two planes makes it to the sim (yes, I agree).  Once PSXT removes that first aircraft however, the "second one" is *updated*, and the SimObject obviously doesn't exist because it was removed.

As long as PSXT receives position updates for that "id" that are within range it will not be removed from the sim. 

Edited August 15, 2020 by kiek

[kiek 1,215]

9 hours ago, cknipe said:

@kiek but I think SimConnect removes -all- objects with the same id

No.

Objects in SimConnect have a unique id given to them at creation time.

That id is different from the mode s id in your data stream.

SimConnect can only delete one object at a time.

[kiek 1,215]

19 hours ago, cknipe said:

sure, of course.

Hi Chris,

Try new version 18.5 first, I think I have fixed a lot of these 'is hanging' problems.

Nicoi

[fluffyflops 1,856]

6 hours ago, kiek said:

Hi Chris,

Try new version 18.5 first, I think I have fixed a lot of these 'is hanging' problems.

Nicoi

can you give me a link to the new 18.5 file please so i can just overwrite 18.4 without uninstalling and reinstalling ?

 

thanks 

[kiek 1,215]

There is no need to uninstall. Just run the setup file. Piece of cake.

[mike stryczek 23]

On 8/14/2020 at 4:31 AM, mike stryczek said:

 

This is inconclusive yet but I eliminated from my DLL.XML  "traffictoolbox.dll"   which is STK FSX addon, and I never suspected it but when I try without it I had 2 uninterrupted successful flights with 18.4 standard from the website.

Maybe going through one by one and eliminating temporarily addons from  DLL.XML   trace which conflicts with PSXT

 

mike

unfortunately I’m still getting disconnects 10 sec error with 18.5

It lasts longer, sometimes I don’t get it but it still happened

 

mike

[kiek 1,215]

12 minutes ago, mike stryczek said:

unfortunately I’m still getting disconnects 10 sec error with 18.5

It lasts longer, sometimes I don’t get it but it still happened

 

mike

Could you also provide a SimConnect.log to see what is wrong in SimConnect.

[mike stryczek 23]

2 hours ago, kiek said:

Could you also provide a SimConnect.log to see what is wrong in SimConnect.

it could be tricky because it takes a while for the error to happen

last time I tried it was more than 1gb (!#$^)

I well try

 

[mike stryczek 23]

7 hours ago, kiek said:

Could you also provide a SimConnect.log to see what is wrong in SimConnect.

https://www.dropbox.com/s/spp2f4h1cme9z6u/SimConnect0.zip?raw=1

I hope that will give some information , very large file.

I didn’t stop the flight right away because I was on the Vatsim so I continued for about 20 minutes after the error

 

mike

 

[kiek 1,215]

5 hours ago, mike stryczek said:

https://www.dropbox.com/s/spp2f4h1cme9z6u/SimConnect0.zip?raw=1

I hope that will give some information , very large file.

I didn’t stop the flight right away because I was on the Vatsim so I continued for about 20 minutes after the error

 

mike

 

Thx for the file.

6211.36242 [21] I/O Error! (2, 80004004)

Looks like the old threading problem, reading / writing at the same time, or maybe MaxReceive = 16384 fixes this?

BTW I see lots of SimConnect exceptions in your Log.txt  file, there is still work to do for me... :-[  

EDIT: From SimConnect0.log I see that FSUIPC AI manager deletes some of the PSXT objects... You should not use the AI trafic limiter function of PSUIPC!

> 10.11046 [ 5, 1]Open: Version=0x00000004  Name="FSUIPC4_AItraffic"
> 162.38761 [21, 1]Open: Version=0x00000004  Name="PSXseeconTraffic"

> 2204.85457 [21, 191073]SetDataOnSimObject:DefineID=401, ObjectID=436, Flags=0, ArrayCount=1, cbUnitSize=32, pDataSet=1138349180
< 2204.85501 [5] ObjectAddRemove: EventID=14  Type=2  ObjectID=436
 

Edited August 16, 2020 by kiek

[mike stryczek 23]

8 hours ago, kiek said:

EDIT: From SimConnect0.log I see that FSUIPC AI manager deletes some of the PSXT objects... You should not use the AI trafic limiter function of PSUIPC!

 

 

 

The error  happens no matter what I set in FSUIPC, if I have it at 0 it happens and if I put limit there.

I was just trying to limit and try seeing if it helps

mike

[kiek 1,215]

2 hours ago, mike stryczek said:

 

The error  happens no matter what I set in FSUIPC, if I have it at 0 it happens and if I put limit there.

I was just trying to limit and try seeing if it helps

mike

Cannot you get rid of FSUIPC all together? I don't use it anymore, is old school. 🙂

Edited August 16, 2020 by kiek