Hi Steve,
first place to check is the tm.log of the Aerofly executable which is in your documents\aerofly folder when it loads the aircraft it will write out what is wrong.
If you have copied the tmd file from another aircraft then you have to basically clean out the graphics section except one rigidbodygraphics object which is attached to the fuselage. You can then add all objects in the geometry list for testing.
I'm actually not sure what the minimum tmd looks like nowadays. We usually start with a similar aircraft and than modify the physics (e.g. add wheels, gear, more wings, actuators...) and redo the entire graphics section and also adjust the sounds.
if you check out the wiki, I've recently added a few code snippets:
https://www.aerofly.com/dokuwiki/doku.php/aircraft:tmd
Essentially:
<[file][][]
<[modelmanager][][]
<[pointer_list_tmuniverse][DynamicObjects][]
// leave this as it is from a similar aircraft
>
<[pointer_list_tmuniverse][GraphicObjects][]
<[rigidbodygraphics][Fuselage][]
<[uint32][PositionID][Fuselage.R]>
<[uint32][OrientationID][Fuselage.Q]>
<[string8][GeometryList][ ]> // all your model parts go in here
>
>
<[pointer_list_tmuniverse][SoundObjects][]
// keep or delete
>
>
>Hi Steve,
ummm it says loading dynamics end: 0 dynamics, 0 graphics, 0 sound so it didn't create any objects, and it says mass 0 (no rigidbodies have been created).
Are you using the new syntax for the tmd structure? The one with <[pointer_list_tmuniverse][DynamicObjects][] and with "modelmanager" and not the old <[int32][dynamics][0] ...
You can ignore the copilot.tmc file. I mean you could just create a blank file to get rid of this message but it shouldn't affect the loading at all.
if you check out the wiki, I've recently added a few code snippets:
Thanks for progress on the Wiki Jan. I'd love to do an aircraft project one day, but lack of documentation puts me off at the moment. This is already a step forward. (I will fly a Cessna 208 in Aerofly even if I have to make it myself 
)
Is the XML-ish format that IPACS uses something developed in house? I've been tempted to whip-up a quick app that converts it to XML and back again
Something like XML Notepad could then be used to edit this stuff hierarchically, which I think would be a lot easier for the large TMD files.
https://github.com/Microsoft/XmlNotepad/wiki
Example:
<[string8][object][jet_engine]
<[string8][Name][JetEngine]>
<[string8][Body][Fuselage]>
<[string8][ThrottleControl][ThrottleInput.Output]>
<[float64][MaximumThrust][10000.0]>
>Converts to
<object type="string8" value="jet_engine">
<Name type="string8" value"JetEngine" />
<Body type="string8" value"Fuselage" />
<ThrottleControl type="string8" value"ThrottleInput.Output" />
<MaximumThrust type="string8" value"10000.0" />
</object>Hi Jan
Having a bit of a mare this morning as aircraft converter refuses to work, tried re-install and registry clean but still no work so loading fs2 onto my development system to see if it will work there.
I am using the f18 as a working copy. In an effort to make things easier I attached all my geometry into 2 objects to make working with a base TMD file easier, ie only 2 entries, but I have not been able to check this out due to the above problem.
Hi Nickhod
Any tools to make things easier will be much appreciated, I believe the only thing stopping aircraft development is the TMD file. To programming types it will look ok but to us mere modellers it's going to take some getting used too. Perhaps an export tool that writes a basic TMD file might help. I have noticed the tm.log file produced by the exporter contains much of the information required, ie names, location, pivot axis etc
The way you whip up files you should have it done by, err what time is it now 11.21, ok I'll give you to 11.24 
Steve
To programming types it will look ok
It looks horrendous to this programming type too!
I get why they've done this, you can parse this a lot faster than you can parse XML, but XML as the intermediate format would have been preferable.
For example, if Jan's bare bones TMD file was XML, you could see the hierarchy in a free XML editor and also edit the values in the text box in the bottom left of the screen. Any syntax errors would be validated.
I'll have a look at writing a two-way converter as I think it would be useful for the scenery stuff too.
I vote nickhod for the New Years Honours List.
A CBE at least.
Hi Steve,
yes you need a copy of the tmd and tmc files in the intermediate directory, together with the exported model in form of a tgi file and all of the textures as bitmaps. On that folder you can launch the converter which will copy the tmd, tmc over to the aircraft directory, convert all textures, 3d model and sounds.
The tmd that you have in the intermediate folder has to be a tmd that could be loaded by aerofly, at least for the preview image rendering.
DId you find the aircraft tutorial in the wiki and does the converter work for that aircraft? That would be a good check if your tools are installed correctly, the rest is probably just caused by some mistakes in your tmd file and not actually an issue of installation. Do you have skype? Write my a PM if you do, I'd like to explain the tmd in a screen-sharing session, maybe together we can find the mistake(s).
Maybe you could upload your tmd file here (or via pm) and I'll go over it, maybe I can find an error.
[off topic]
nickhod Please don't spend too much time with trying to make aircraft tools, I already have a decent aircraft editor which works in 3D and has become totally OP in the last couple of months. This tool is not yet available to the public, there are things I need to iron out... needs an undo/redo function for example.
It can load a tgi.txt file (from the exporter) to show the geometry, select parts and extract pivots from the export or manually pick vertices, It shows the physics geometry (e.g. wing and fuselage, position of masses, etc.), and it has a very handy debugger, instant refresh when a file is saved externally, functions like move, rotate, mirror or physics objects and control clickspots (with multi-select) and a lot more. And last but not least: it preserves the comments and indentation when you edit the file and then save it.
This is the perfect tool to assist in making an aircraft or will be when I eventually get to publish it. Personally for editing the files I use Visual Studio 17. Its text editor is pretty good, allows multi-row block edit and offers other useful tools and allows me to split the text editor view into pretty much any configuration I can think of (left/right, all four corners, 3 columns, 4 columns, top/bottom,...)
Please don't spend too much time with trying to make aircraft tools, I already have a decent aircraft editor which works in 3D
Looks amazing Jan and just what AFS needs to allow more people to have a go at aircraft development. Looking forward to its release.
Looks amazing Jan and just what AFS needs to allow more people to have a go at aircraft development. Looking forward to its release.
Six years+ in the making 
Six years+ in the making
Yeah, I can tell that's not something you throw together in a weekend.
Don't be afraid to give us an alpha version when you can. I could manage without undo and redo.
Its not helped by the Starfighters complicated gear.
Hi Steve,
I take it you are building a Lockheed F-104 Starfighter, that's a very nice aircraft, my old boss had a hundreds hours on them
Looking forward to seeing your screenshots
Cheers,
Paul
Hi Jan
Could you possible explain how you work out the inertia length
<[tmvector3d][InertiaLength][ 0.404930 -0.355130 0.765910 ]>
Also
How do we input an axis that is rotated. The Starfighters gear rotates at approx 45 degrees.
I can see why people give up when they reach the TMD file, good thing about being old is we get stubborn so I'll keep bashing away at it. Yes I know I could have gone for a simple aircraft but where's the fun in that 
Thanks
Steve (with a brain that hurts)
The rigidbodies in Aerofly are box-shaped with a mass and a certain dimension, the InertiaLength. That length is used to calculate the inertia of the rigidbody. The higher the values the more torque required to rotate the object (principle of angular momentum).
We use a solid cuboid to easily modify the inertia around the each of the three main axes of inertia:
https://en.wikipedia.org/wiki/List_of_moments_of_inertia
We get these values either from bounding box dimensions from the modelling tools or by estimation. The precise value is not that important, I wouldn't be able to tell a difference in behavior even if its 5% increased or decreased.
For a landing gear: just estimate the physical size (e.g. 0.4m long, 1.1m wide and only 0.2m thick... that would give you: 0.4 1.1 0.2
Make sure all values are positive though... I see you set a negative value (or found one, where??? :D) that would be like dealing with negative masses.
Note: the inertia length represents the dimensions of the object in the local coordinate system. So if you rotate the rigidbody with the B0 matrix the inertia axes rotate with it, so you need to work out at which angle the gear is attached to the fuselage, say 30 degrees, then just plug that into my handy rotation matrix generator: https://www.aerofly-sim.de/download/downl…atrix-generator
and out you get the B0 matrix for your gear: (I hand drew it in)
<[rigidbody][LeftGear][]
<[float64][Mass][100.0]>
<[tmvector3d][InertiaLength][ 0.4 1.1 0.2 ]>
<[tmvector3d][R0][ -0.5 1.5 -1.0 ]>
<[tmmatrix3d][B0][ 1.0 0.0 0.0 0.0 0.866025 0.5 0.0 -0.5 0.866025 ]>
>
<[rigidbody][RightGear][]
<[float64][Mass][100.0]>
<[tmvector3d][InertiaLength][ 0.4 1.1 0.2 ]>
<[tmvector3d][R0][ -0.5 -1.5 -1.0 ]>
<[tmmatrix3d][B0][ 1.0 0.0 0.0 0.0 0.866025 -0.5 0.0 0.5 0.866025 ]>
>I can offer to set up the landing gear for you. That is the most tricky thing you need to do in Aerofly.
If you send me your converted aircraft files (your starfighter aircraft folder) together with the tgi.txt file (needs to be enabled in the plugin before exporting I think...) then I can view your aircraft in my editor and set up all the annoying physic objects.
We know that it is a bit overwhelming to set this up without any help and we don't want the user projects to fail just because of this, so we're here to help you with your first aircraft(s).
Hi,
I hope it's okay when I hijack the thread. I was batteling with the tmd file since yesterday and didn't get anywhere.
I have been reading in the wiki and converting the DR400, so I understand the basic idea and my setup works.
So far I have only exported the fuselage and wings of my own project. I used the tdm file of the DR400 and cutted it down to what I need
That's how it looks like:
<[file][][]
<[modelmanager][][]
<[pointer_list_tmuniverse][DynamicObjects][]
// telemetry
<[telemetry][Telemetry][]
<[telemetry][Telemetry][]
<[string8][Body][Fuselage]>
>
<[event_value][DEV0][]
>
// cameras start
<[camera_head][CameraFollow][]
<[string8][Body][Fuselage]>
<[tmvector3d][R0][-9.0 0.0 2.0]>
<[tmvector3d][Direction][1.0 0.0 0.0]>
>
<[camera_head][CameraPilot][]
<[string8][Body][Fuselage]>
<[tmvector3d][R0][-0.104 0.27 0.95]>
<[tmvector3d][Direction][1.0 0.0 -0.1]>
<[bool][InCockpit][true]>
<[string8][Tags][cockpit pilot left]>
>
<[camera][CameraCoPilot][]
<[string8][Body][Fuselage]>
<[tmvector3d][R0][-0.104 -0.27 0.95]>
<[tmvector3d][Direction][1.0 0.0 -0.1]>
<[bool][InCockpit][true]>
<[string8][Tags][cockpit copilot right]>
>
<[camera][CameraLeftWing][]
<[string8][Body][LeftWing]>
<[tmvector3d][R0][ -0.2 6.0 0.7 ]>
<[tmvector3d][Direction][ 0.0 -0.999 0.0 ]>
<[bool][InCockpit][false]>
<[string8][Tags][left wing]>
>
// cameras end
// rigidbody systems start
// airframe
<[rigidbody][Fuselage][]
<[float64][Mass][387.0]>
<[tmvector3d][InertiaLength][4.0 1.116 1.08]>
<[tmvector3d][R0][0.612 0.0 0.396]>
<[tmmatrix3d][B0][1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0]>
>
<[rigidbody][LeftWing][]
<[float64][Mass][80.0]>
<[tmvector3d][InertiaLength][1.53 3.5 0.3132]>
<[tmvector3d][R0][-0.1 2.1 0.1116]>
<[tmmatrix3d][B0][1.0 -0.0 0.0 -0.0 0.999800 0.019000 -0.0 -0.019000 0.999800]>
>
<[multibody_joint][JointFuselageLeftWing][]
<[string8][Body0][Fuselage]>
<[string8][Body1][LeftWing]>
<[tmvector3d][X0][1.0 0.0 0.0]>
<[tmvector3d][R0][-0.27 0.558 0.018]>
<[float64][Kp][1000000.0]>
<[float64][Kd][10000.0]>
<[uint32][Type][0]>
>
<[rigidbody][RightWing][]
<[float64][Mass][80.0]>
<[tmvector3d][InertiaLength][1.53 3.5 0.3132]>
<[tmvector3d][R0][-0.1 -2.1 0.1116]>
<[tmmatrix3d][B0][1.0 -0.0 0.0 -0.0 0.999800 0.019000 -0.0 -0.019000 0.999800]>
>
<[multibody_joint][JointFuselageRightWing][]
<[string8][Body0][Fuselage]>
<[string8][Body1][RightWing]>
<[tmvector3d][X0][1.0 0.0 0.0]>
<[tmvector3d][R0][-0.27 -0.558 0.018]>
<[float64][Kp][1000000.0]>
<[float64][Kd][10000.0]>
<[uint32][Type][0]>
>
// rigidbody systems end
>
<[pointer_list_tmuniverse][GraphicObjects][]
<[rigidbodygraphics][Fuselage][]
<[uint32][PositionID][Fuselage.R]>
<[uint32][OrientationID][Fuselage.Q]>
<[string8][GeometryList][ Fuselage Leftwing Rightwing
]>
>
<[pointer_list_tmuniverse][SoundObjects][]
>
>
>It compiles, but the preview is empty and the report tells me for all 3 geometries "unreferenced geometry"
I see in the wiki
Referenced object does not exist (e.g. trying to call .Output on an object that is not defined in the scope of the same section)
To be honest, it doesn't tell me anything. Could someone please bring a bit light into it?
Thanks a lot!
Cheers,
Kai
Hi Kai
I did the same as you and achieved the same results. The preview is nothing more than an image of the compiled aircraft, strangely enough its the 3d model that's in the modelling program so something is going right. If its been produced you can look at this image in your intermediate folder. In my case its preview_f104g.tiff and shows the model perfectly
I believe things are a lot more involved than just a bare minimum approach which is what we are trying to achieve, ie a model we can just look at in the sim. In my view the exporter would be better to produce a sample TMD file with all the relevant entries as a starting point.
By the time we're done we'll all be working for NASA as rocket scientists
Steve