IMFD IMFD 5.4 CTD with Orbiter 2010

[johan]

Hi all

I'm facing a problem with IMFD v5.4 running on Orbiter 2010. Other modules I have running are:

• OrbiterSound 3.5
• ScnEditor

Other stuff installed includes:

• All 3 XR vessels including soundpack
• UMMU & UCGO
• DGIV including Soundpack
• Earth L11 textures
• LaunchMFD (disabled)
• TransX (disabled)

The symptoms are:

• Start with the stock "DG docked at ISS" scenario... so none of XR ships etc should be loading, right?
• Switch to Shuttle-A on Brighton Beach so I can use AlignPlanesMFD to find a good window to launch from the ISS... in the process, I fast-forward about 600k seconds.
• Switch back to DG, undock, fire up IMFD and set up off-plane intercept burn using the Delta-V program. (delta-V on the left, map on the right... map is set to share info with the delta-v instance)
• Hit Autoburn and fast-forward... watch as it does the eject burn.
• Fast-forward more as we coast away from Earth
• At the border of the moon's SOI, slow down to realtime, now start poking around the delta-V program so I can do a course correction.
• As I poke about on the MFD, Orbiter has a CTD.

I've done this once or twice, and tried somebody else's recording... it crashes consistently. It doesn't seem to have anything to do with exactly where you try to do the course correction, I've seen it crash long before you get to the moon's SOI, as well as on the border (as per map MFD).

I haven't tried to rigorously isolate which button in IMFD causes this, I could do this if anyone thought it would help. I tried looking at some logs but it didn't make me any wiser, they don't seem to record any errors, except for this one in orbiter.log:

Module ScnEditor.dll ......... [Build 100606, API 100606]
Module OrbiterSound.dll ...... [Build ******, API 060425]
---------------------------------------------------------------
>>> WARNING: Obsolete API function used: oapiRegisterMFDMode
At least one active module is accessing an obsolete interface function.
Addons which rely on obsolete functions may not be compatible with
future versions of Orbiter.
---------------------------------------------------------------
Module InterMFD54.dll ........ [Build 100616, API 100603]

And that strikes me as more likely related to OrbiterSound.

Any help would be greatly appreciated... :tiphat:

 

[jarmonik]

Can you reproduce this if you save the scenario before the CTD and reload it. If you can reproduce it then could you attach the scenario into your post. Also you could try it with a clean installation of orbiter. There are some addons those will replace Orbiter DLLs like planet modules with their own replacements.

Based on your report it's the Orbiter that will CTD not the IMFD. Is that right ?

I made a quick run with the IMFD 5.4 and I couldn't reproduce the CTD. If I can reproduce it with your scenario then I should able to fix the cause pretty quickly. If not then it will become more complicated. Do you know how to run debugger. I could upload a debug version of IMFD.

 

[johan]

Can you reproduce this if you save the scenario before the CTD and reload it. If you can reproduce it then could you attach the scenario into your post. Also you could try it with a clean installation of orbiter. There are some addons those will replace Orbiter DLLs like planet modules with their own replacements.

Okay, I'll get a chance to look at this, probably tomorrow night. I'll try those things and post here how it goes and/or attach the scenario.

Based on your report it's the Orbiter that will CTD not the IMFD. Is that right ?

Doesn't that work out to be the same thing? The simulation window disappears, and the launchpad disappears, and I get a message from Windows saying "Sorry for the inconvenience". I'll get a screenshot of the message so you can see what I mean.

Do you know how to run debugger. I could upload a debug version of IMFD.

Sure, I made my living as a dev for a few years recently. I mainly used Java, but I've used C++ before (Orbiter addons are written in C++, right?), so if you can give me some lightweight instructions I should be able to do some debugging for you.

Lets first eliminate the possibility that other addons are causing this? I'll get back here with results in a day or two.

 

[orb]

Doesn't that work out to be the same thing? The simulation window disappears, and the launchpad disappears, and I get a message from Windows saying "Sorry for the inconvenience". I'll get a screenshot of the message so you can see what I mean.

There should be something like: "To see what data this error report contains, click here." or "What data does this error report contain?" on that window. If you click on it, there should be detailed report displayed (or you' d need to click on another "Details" button - I don't remember now), with type of exception, module name which caused that exception, and the address of the instruction that caused that exception. Posting that would help (more or less) to identify the problem, even without debug version of the module.

 

[johan]

There should be something like: "To see what data this error report contains, click here." or "What data does this error report contain?" on that window. ... detailed report displayed ...

Yeah that's the one it displays; I must say, I never even considered that there may be useful info behind that dialog somewhere :embarrassed:

I'll make that the first thing I check out... thanks for pointing it out!

 

[johan]

Hi all

Okay, so I have more information now. I tried two different things last night:

1. Do the run to the moon manually again, this time taking care to record some stuff from the resulting windows crash dialog. See the contents of crash_1.zip (attached).
2. Rename my orbiter folder to some backup name, and re-extract a clean install of Orbiter from the zip file. Then extract my IMFD5.4 zipfile into the new folder, and run the scenario again. It crashed again... see the contents of crash_2.zip (attached). Unfortunately I managed to not get the *appcompat.txt file this time around... hope it doesn't matter? If it does I'll just go back and get it.

A strange thing I noticed, is that transX seems to be already installed on my "clean" install... I didn't think transX came with the default orbiter install?

This raises the question: Do I maybe need to re-download it and start from there?

So I used md5sum from GNU textutils (UnxUtils download) to determine whether or not my download had gotten interfered with somehow.
Just for good measure, I did a checksum of my IMFD download too, as well as the .dll that got mentioned in the error report.

ede8a82ab323d75ae90d58c7372a6a38 *orbiter100606.zip
570400a133ffcbe39ff91af1e8616308 *IMFD54.zip
8079998f9bb7246889c3fe3b15a8dbf8 *Cowell2.dll

If someone could please check their versions of these files against my md5's, the answer may jump out at us...

Cheers,
Johan

 

[johan]

I found this page on the wiki that has MD5's, but they look like they're for Orbiter 2006:

http://www.orbiterwiki.org/wiki/Orbiter#How_to_install

 

[jarmonik]

TransX belongs in the default orbiter installation so it's ok to have it in a clean install.

I think I found one bug. Let's hope it was the one giving trouble.

 

[johan]

TransX belongs in the default orbiter installation so it's ok to have it in a clean install.

I think I found one bug. Let's hope it was the one giving trouble.

Ok, will try it out tonight. Thanks for the quick response!

 

[johan]

It got boring at work, so I gave it a whirl, it looks like IMFD refuses to load (it isn't available on the MFD menu when I hit Shift-F1)...
I installed it by renaming the stock Cowell dll file to have some other extension, then unzipping your new version into its place. Then, I unzipped your new InterMFD55.dll into the Modules\Plugin folder, next to its *54.dll counterpart.

I found the following in Orbiter.log:

Error loading module Modules\Plugin\InterMFD55.dll (code 14001)

Logfile attached.

 

[orb]

I found the following in Orbiter.log:

Error loading module Modules\Plugin\InterMFD55.dll (code 14001)

Strange, the new module is being loaded fine for me.

You could be missing one of the dependent assemblies from MSVC++ runtimes:

<assemblyIdentity type="win32" name="Microsoft.VC90.DebugCRT" version="9.0.21022.8" processorArchitecture="x86" publicKeyToken="1fc8b3b9a1e18e3b"></assemblyIdentity>
...
<assemblyIdentity type="win32" name="Microsoft.VC80.CRT" version="8.0.50727.4053" processorArchitecture="x86" publicKeyToken="1fc8b3b9a1e18e3b"></assemblyIdentity>
...
<assemblyIdentity type="win32" name="Microsoft.VC90.CRT" version="9.0.21022.8" processorArchitecture="x86" publicKeyToken="1fc8b3b9a1e18e3b"></assemblyIdentity>

 

[johan]

You could be missing one of the dependent assemblies from MSVC++ runtimes:

Sounds like a possibility, unfortunately I've never worked with MSVC++ before, or anything that even remotely smells like .NET for that matter... how do I check if this is the problem?

 

[jarmonik]

Looks like I forget to make a static build. I made a new build that doesn't need those runtime libraries anymore.

 

[johan]

Thank you jarmonik! That did it. The recording runs through 100% now.

:tiphat:

 

[jarmonik]

It's good to hear that. I'll try to upload a new distribution tomorrow.

 

[Tommy]

Thanks for solving this, Jormo. The playback he's trying to run is an advance "beta" of a part of the IMFD Full Manual Version 2, and I was worried that it might be an issue with the recording, or with 2010.