FSUIPC5.DLL APPCRASH [using LINDA]

[Pete Dowson]

5 hours ago, Nervures said:

Can you tell me if it seems to you correct ?

New entries:

ConsoleWindow=25,25,702,384

38=777COM - Copy
39=777NGX - Copy
40=Demauto - Copy
41=GSX - Copy
42=GSX_Arr - Copy
43=Autobrake - Copy
44=choix - Copy
45=Followme - Copy
46=RAAS - Copy
47=RWstart - Copy

ConsoleWindow will be in the [General] section and it is where FSUIPC remembers the console window size and position for the Log. You must have checked the Console option in the Logging tab at some time.

I assume those other entries are in the [LuaFiles] or [Macros] section? They just list the files you have in your Modules folder. Looks like you made copies of those listed there!

5 hours ago, Nervures said:

[Window.Message Window]
Docked=389, 4130, 15927, 234

[Window.SimConnect Message Window]
Docked=1438, 1105, 14183, 690
Undocked=3028, 616, 553, 275

[Window.SimConnectWindow]
Docked=67, 414, 16248, 15859
Undocked=2634, 552, 391, 417

These are where FSUIPC records the size and position of the different types of SimConnect window, so it can restore them in the next sesion. It's part of the enhancements which will be documented for this release when I have heard from yo that it fixes your Lua problems.

Pete

 

[Nervures]

Hi Pete,

9 hours ago, Pete Dowson said:

I assume those other entries are in the [LuaFiles] or [Macros] section? They just list the files you have in your Modules folder. Looks like you made copies of those listed there!

I understand now. Before entries were manual.

Yes backup copies.

Yesterday I made the same flight with the crashes => no crash during flight (6h) and no crash when the sim closes.

I continue the flights this weekend.

Francis.

[Pete Dowson]

37 minutes ago, Nervures said:

Yesterday I made the same flight with the crashes => no crash during flight (6h) and no crash when the sim closes.

Excellent.

37 minutes ago, Nervures said:

I continue the flights this weekend.

Okay. Let me know. It looks like I can make a new release in the week, Monday or Tuesday.

Thanks,
Pete

 

[Nervures]

Hi Pete,

12 hours ago, Pete Dowson said:

These are where FSUIPC records the size and position of the different types of SimConnect window, so it can restore them in the next sesion.

I have some difficulties to understand how the parameters of windows are recorded.

1 I deleted all the parameters of windows:

[Window.Message Window]

[Window.SimConnect Message Window]

[Window.SimConnectWindow]

2 I launched the simu:

[Window.Message Window]
Docked=194, 925, 15994, 234
[Window.SimConnect Message Window]
Docked=1057, 1450, 14268, 690
[Window.SimConnectWindow]

Red parameters are perfect.

3 I opened GSX menu ( CTRL+F11 ):

[Window.Message Window]
Docked=194, 925, 15994, 234

[Window.SimConnect Message Window]
Docked=1057, 1450, 14268, 690

[Window.SimConnectWindow]
Docked=6609, 6962, 3173, 2749

Still perfect for the three windows

4 Resized [Window.SimConnectWindow]

Docked=6609, 6962, 3173, 4365

Still perfect.

5 Undocked the [Window.SimConnect Message Window] and [Window.SimConnectWindow]

[Window.Message Window]
Docked=194, 925, 15994, 234
[Window.SimConnect Message Window]
Docked=1117, 1851, 14259, 676
Undocked=3028, 616, 554, 276
[Window.SimConnectWindow]
Docked=67, 414, 16248, 15859
Undocked=2634, 552, 391, 417

Undocked, position and size ( from linda.gui.ini ), are perfect.

Why the [Window.SimConnect Message Window] and [Window.SimConnectWindow] Docked parameters changed ?

I undock these two windows from my main screen to my second screen by using mouse.move mouse click.

ipc.display("Tests",1,5)
    ipc.sleep(1000)   
    mouse.move(329,109)
    mouse.click(2)
    ipc.sleep(200)  
    mouse.move(379,121)
    mouse.click()

When I restart the simulator, as positions are not good, I cannot any more undock these windows. It's as if another program modified these parameters.

Francis.

 

 

 

 

 

 

 

[Nervures]

Pete,

Same for the window [Window.Message Window] when the simulator restarts.

[Window.Message Window]
Docked=194, 925, 15994, 234

after restart:

[Window.Message Window]
Docked=313, 1726, 15977, 234

Francis.
 

 

[Pete Dowson]

3 hours ago, Nervures said:

Why the [Window.SimConnect Message Window] and [Window.SimConnectWindow] Docked parameters changed ?

FSUIPC is only reading the positions when it sees that the windows are created, moved or resized (whether visible or not). The undocked ones will probably be screen coordinates. Not sure about the docked ones -- probably not. I'd have to check both.

The facility you are worried about is not finalised yet. I'm still testing it on different PCs. I've not yet tested the undocked windows. The docked ones are behaving perfectly here on two systems so far.

Don't forget you are using an as-yet unreleased version of FSUIPC5, for the purpose of checking that the LINDA stuff (plus additions) don't crash the sim.

Pete

 

 

[Pete Dowson]

To help me with further testing of the SimConnect window saving facilities, can you tell me a few things please?

1. Are you running P3D4 in full screen?
2. If not full screen mode, if P3D4 maximized?
3. Do you have other screens where you move undocked windows to?

So far I cannot make any docked or undocked saved window parameters go wrong.  They all seem to restore correctly on reload -- not necessarily in docked or undocked state, as I don't control or save that. Just positions and initial sizes (the vertical sizes of the Windows can be lenghened when needed for additional test or lines -- i'm not in any way involved in that).

Pete

 

 

[Pete Dowson]

I've been checking the Window saving, and there are some inconsistencies between use in a maximise / whole screen and normal windowed P3D4. 

The "docked" values saved are not Window coordinates but proportions of the P3D4 window size, as a portion of 16384 (so 16384 = 100% of height or width). I'll check that this works correctly. The idea was to keep the SimConnect window in proportion and relative position in P3D4's window no matter what happened to it. Maybe I should only do this with the position not size -- though both would need adjusting if they don't fit the new P3D4 window size.

But the undocked values should be screen coordinates, with negative values for screens to the left of the default, as so on. That seems to work pretty well here but needs more testing.

I'm going to be very busy with other things in the next 3-4 weeks, so I might disable this new facility for now so I can make a release with the Lua fix which you are testing so ably for me!

Pete

 

[Nervures]

1 hour ago, Pete Dowson said:

1. Are you running P3D4 in full screen?

No.

1 hour ago, Pete Dowson said:

2. If not full screen mode, if P3D4 maximized?

Yes.

1 hour ago, Pete Dowson said:

3. Do you have other screens where you move undocked windows to?

No.

1 hour ago, Pete Dowson said:

So far I cannot make any docked or undocked saved window parameters go wrong.

While waiting for a solution I delete all saved parameters from the three windows before starting the simulator. It works very well.

20 hours ago, Pete Dowson said:

I assume those other entries are in the [LuaFiles] or [Macros] section? They just list the files you have in your Modules folder. Looks like you made copies of those listed there!

I found a lua file in the Modules folder that was not added => Sauvegarde_Auto_777.lua. Why ?

I added it manually.

Francis.

[Nervures]

3 minutes ago, Pete Dowson said:

But the undocked values should be screen coordinates, with negative values for screens to the left of the default, as so on. That seems to work pretty well here but needs more testing.

I confirm. Work pretty well here too.

9 minutes ago, Pete Dowson said:

I'm going to be very busy with other things in the next 3-4 weeks, so I might disable this new facility for now so I can make a release with the Lua fix which you are testing so ably for me!

No problem Pete.

I can fly and this is the most important.

Francis.

[Pete Dowson]

2 hours ago, Nervures said:

I found a lua file in the Modules folder that was not added => Sauvegarde_Auto_777.lua. Why ?

I added it manually.

The name is too long. it breaks the rules (max 16 characters).

The only purpose of that list is to enable button or keys to be assigned to those Lua files. If that file is never used in such a way it doesn't matter. But adding it manually won't do any good. it will make no difference.

Pete

 

[Nervures]

Hello Pete,

5 hours ago, Pete Dowson said:

The only purpose of that list is to enable button or keys to be assigned to those Lua files

No keys no button. I removed it.

Another long-haul from YPPH to FMEE ( 9h34 ) without crash. The simulator closes perfectly :-)).

Francis.

 

FSUIPC5.log

IVAO_YPPHFMEE.log

[Nervures]

Hi Pete,

After a flight of more than 13h without any problem, leaving the runway, all the flight software closed. LINDA P3D4 AS16 without any crash message (nothing on the desktop or in the event viewer)
Do you think FSUIPC5.111 can be the culprit ? I do not think so but I would like to have your confirmation.

Francis.

[Pete Dowson]

23 minutes ago, Nervures said:

After a flight of more than 13h without any problem, leaving the runway, all the flight software closed. LINDA P3D4 AS16 without any crash message (nothing on the desktop or in the event viewer)
Do you think FSUIPC5.111 can be the culprit ? I do not think so but I would like to have your confirmation.

After 13h? No, I don't think so!

I've never heard of anyone having a session lasting that long in any case! That's increadible! Try doing than in FSX or P3D3!

How did the FSUIPC5.LOG end?

Pete

 

[Nervures]

1 hour ago, Pete Dowson said:

How did the FSUIPC5.LOG end

It was late and I restarted manually my backup which has overwritten the file . My backup is corrupted (first time it does not work).
I'm flying at the moment on an identical flight and I scheduled a backup of the FSUIPC5.log if a crash occurs. I know you save it but if I am not in front of the screen at the time of the crash the restore of the backup is done automatically and overwrites the file.
I keep you informed.

Francis.

[Nervures]

1 hour ago, Pete Dowson said:

I've never heard of anyone having a session lasting that long in any case! That's increadible!

Twice a month I do OTHH-NZAA ( 15hor 16h ).

Francis.

[Pete Dowson]

1 hour ago, Nervures said:

It was late and I restarted manually my backup which has overwritten the file

Odd, as in the FSUIPC you have I now save the previous Log as FSUIPC5_Prev.Log.

Pete

 

[Nervures]

 

Good morning Pete,

Unfortunately the simulator crashed 5 hours after the take-off. Flight YPPH-FACT (12h45 ).

You said: " Previous crashes reported in that DLL have been eventually attributed to other problems in P3D."

I know you can not do anything.

It's just to keep you informed.

Event viewer report:

Faulting application name: Prepar3D.exe, version: 4.0.28.21686, time stamp: 0x594a7255
Faulting module name: ucrtbase.DLL, version: 10.0.10586.788, time stamp: 0x5879ab76
Exception code: 0xc0000005
Fault offset: 0x000000000000443a
Faulting process id: 0x1608
Faulting application start time: 0x01d3150293632fce
Faulting application path: C:\P3D4\Prepar3D.exe
Faulting module path: C:\Windows\system32\ucrtbase.DLL
Report Id: d5acde7b-8128-11e7-b7ec-e03f491af142

Launch the simulator from backup.

After landing when I closed the simulator it crashed.

Event viewer report:

Faulting application name: Prepar3D.exe, version: 4.0.28.21686, time stamp: 0x594a7255
Faulting module name: ntdll.dll, version: 6.1.7601.23807, time stamp: 0x5915fdce
Exception code: 0xc0000374
Fault offset: 0x00000000000bf3e2
Faulting process id: 0x22f4
Faulting application start time: 0x01d31535ff1812b8
Faulting application path: C:\P3D4\Prepar3D.exe
Faulting module path: C:\Windows\SYSTEM32\ntdll.dll
Report Id: aee83b42-8169-11e7-b7ec-e03f491af142

Francis.

 

 

FSUIPC5 crash.log

FSUIPC5 close.log

[Pete Dowson]

As you said, I can't really do anything with the crash log. I'm pleased that you've had so much success in any case, and as I said, I'm surprised that you can manage such long flights in any case. My impression from reading the Forums is that it is quite rare to get such flights done without a crash at some time. that's why autosave is so popular.

I don't get crashes at all, but then I enjoy takeoff and landing, not cruising, so I fly 737 on short routes of max 2 hrs and usually 1 hr or less (cruise). It often takes longer to plan the flight and prepare the cockpit, and taxi about at large airports and queue (realistically) behind a few AI aircraft for takeoff! If I want to reach an airport a long way from base I plan a series of hops via interim airports, and re-boot the system between each leg for safety.

The close log gets a long way through closing everything down. My best guess from the log is that one of the Lua threads didn't manage to properly close its dealings with a device before FSUIPC Terminated.  The last thing FSUIPC did was to close the "Global" Lua thread, which is the one supporting Global variables. I know LINDA uses these a lot to share data. With the Global thread closed it would get "nil" returned to any query, so possibly there's a place in their code where this isn't checked. Though that shouldn't cause a crash -- unless it was used as a parameter for some Windows API call, or a call to one of its other DLLs dealing with devices.

I've now got a couple of other facilities to sort out in FSUIPC5 (and 4) before I release an update. It'll probably be 5.112.

Pete