Military AI Works

All three of these packages went to AVSIM just a few minutes ago, so give it a couple of hours for the next pass of the library posting. And just in case you haven't figured out what I knew all along, I'm quite mad, lol. This is 44 models and 210 paints total, and it's not done yet. The last part will be the countries which did not take-up orders, plus paint kits. This will be a combination of fact and a little make-believe, as the following countries will be represented:

Norway
Greece
Turkey
Switzerland
USAF
US Army (I think this would have made a great FAC aircraft).


And lastly, Euro-Scooter is a generic term, there will be another new aircraft on this thread shortly, a big departure for what I normally design, but it's a favor for a friend (you know who you are....).

Just picked up the German and Italian packages....nice!

Pretty cool Mike. Embarrassed to admit I am completely unfamiliar with this aircraft but it looks fast and cool like a Mirage F-1. What was this plane used for, is this a trainer?

It was conceived as an entry for a NATO standard lightweight fighter. It won the competition, beating aircraft such as those that that became the F-5A and Etendard IV.
The idea was the winner would be ordered by all NATO members, but the French decided to go with the Etendard and the Brits ignored it completely and that kind of killed the concept and all future NATO design competitions.

You can read about on wikipedia.

Simply amazing, the speed at which your models and paints are released...

Thank you so much for these !!

More than happy

Giorgio

This one was a winner from the very first primitive onwards, Mike

Hi MikeS, both MP and MM , I was testing AMI's G-91 and noticed some problems, some probably due to FDE (just my opinion), some strange problems I can 't verify at every airport or during all tests.
This is absolutely not a criticism, I'd just like to understand if it is just my problem or if also other users have experienced the same. (...Martin...are you there? )
Tests has been done with Traffic view board and Traffic Toolbox
Here what I noticed:
G-91 Y : the nose wheel is not on the ground.
Some of them seem to tremble a little bit too much during taxi (at x4 speed start to jump and crash).

After 20-30 minutes from the start of the flight, FS9 gives the error:
Runtime error!
R6025
-pure virtual funnction call

I had this problem in the past and I'm quite sure it is due to some effect used by flyable aircraft used as AI, may be some problem with effects used

The other problems I was referring, don't happen every time.
Some aircraft are not at ground when parked, but the same aircraft do not have the same behaviour at every airport where I tested. I'll do other tests to be more precise.

All test have been done with only the G-91 installed.

Thanks, Giorgio

Addressing Gorgio's input....

Aircraft stuttering and failing, they run fine at normal speed, if you speed up the sim that's the risks sometimes. I make every effort to make them stable through all speeds of the sim (4x is the fastest that AI aircraft will function, go to 8x and they re-spawn), but it's not a perfect process. I won't be a name-dropper, but I have seen established AI aircraft from experienced designers do the same thing in the past (outside of MAIW, I'm not poking at anybody). The AI traffic engine is designed for normal speed, and stuff that happens outside of that is kind of at your own risk.

Aircraft not touching at different airports. This one's kind of on me, but still a quirk of the sim. I can blame everything on the AI traffic engine, but this one I can work around a little bit. The aircraft will perform differently through flight at different altititudes, and I did the FDE at a sea level airport. This is something Mike Mac mentioned during my FDE education, but it's the first time I have experienced it. What I can do is do an FDE update, adjusting the contact points for the low-to-medium-altitude airfields of Europe, rather than sea level. It still won't be perfect, and the aircraft may be sunken or elevated elswhere, but they will sit appropriately at the airfields they are supposed to. Ultimately I'm blaming it on the sim, but I can fix it a little bit.

The crash-to-desktop runtime virtual cell error, I can't help you with that one, these aircraft never crashed any one of the three computers I tested them on. Read the last part of my documentation.

bismarck wrote:(...Martin...are you there? )

Sure my friend.

I happen to had a closer look at the German variants, and naturally in FSX. I did not notice what you have pointed out above. I agree with Mike - I just speed up time when I want something to happen soon and when I do so I don't care about erratic behavior given the aircraft stays in the sim.

I will try to replicate your findings this evening and report back.

bismarck wrote:After 20-30 minutes from the start of the flight, FS9 gives the error:
Runtime error!
R6025
-pure virtual funnction call

I had this problem in the past and I'm quite sure it is due to some effect used by flyable aircraft used as AI, may be some problem with effects used

You are quite right about the normal cause of this crash. The most common cause is that it is a flyable effect used by an AI aircraft, like burner or vortex etc, but it can also caused by ground scenery using effects as well. I once had ship scenery with duff effects.

You say that you only have the G-91 installed but I am guessing that you only have a traffic file with G-91 aircraft in the usual scenery folder. It maybe though that you have some traffic files in addon scenery folders. We used to use this, e.g. original Coningsby and Topgun packages, We are not the only ones that did/do this. I have several others that add traffic files and if you are not careful they will load without you realising it.

bismarck wrote:Hi MikeS, both MP and MM , I was testing AMI's G-91 and noticed some problems, some probably due to FDE (just my opinion), some strange problems I can 't verify at every airport or during all tests.
This is absolutely not a criticism, I'd just like to understand if it is just my problem or if also other users have experienced the same. (...Martin...are you there? )
Tests has been done with Traffic view board and Traffic Toolbox
Here what I noticed:
G-91 Y : the nose wheel is not on the ground.
Some of them seem to tremble a little bit too much during taxi (at x4 speed start to jump and crash).

The other problems I was referring, don't happen every time.
Some aircraft are not at ground when parked, but the same aircraft do not have the same behaviour at every airport where I tested. I'll do other tests to be more precise.

All test have been done with only the G-91 installed.

Thanks, Giorgio

Hi Giorgio,

Mike P. soloed on his AI "Pinocchio" models and created both the models and the AI FDE's himself. I haven't looked at it at all, so I can't give you any definitive answers, but I can make some general comments.

The trembling during taxiing at normal acceleration and the bouncing disappearance when taxiing at 4X acceleration can be caused by an imbalance between the gear compression built into the model itself and the contact point gear compression settings, the contact point location settings and/or the CG location setting in the aircraft.cfg file. I always test my AI FDE's at both the normal and 4X acceleration before I release them in order to catch issues such as this. If there is any issue with an AI FDE it will usually show up much more readily and dramatically at 4X acceleration.

When you observed the nose wheel off the ground, were the main gear wheels in proper contact with the ground

With regards to Steve's comment about some effects causing a CTD, I used all standard effects which are well-tested, except for the G91Y AB effect, which was modified from the original HTAI F16 AB. All I did was delete the long flame emitters and change the core emitter color to a darker orange-yellow, to achieve that austere little early AB glow in the pipe that doesn't shoot out fifteen feet behind the aircraft. I did this same sort of modification with the blue recognition lights for my Mirage series, and had no issues at all.

If these aircraft are being used in FSX, I am not responsible for any weirdness in their behaivor; they were designed primarily for FS9. I do know that a lot of evaluatory comments in the past have come from use in FSX, you can tell from the screenshots.

There will be a small model and FDE update, to be sure, but it's going to be easy and small, since I only did three packages, and it will only be models and dynamics, no paints; so it will be a drop-in-and-overwrite mod. I did flight-test this pretty aggressively, and had at two other beta-testers involved, but will make it right, for sure.

First of all, thanks to all for your answers and suggestions.

mikewmac wrote:Mike P. soloed on his AI "Pinocchio" models and created both the models and the AI FDE's himself. I haven't looked at it at all

...sorry Mike, reading the comments in the CFG, once I read Michael, I didn't waste time in reading also the surname...

I found the cause for a couple of problems.
The "Runtime error" is caused from fx_g91_ab effect. Removing it from cfg, the problem doesn't appear anymore. Very strange, it uses only the default texture fx_2.bmp, used in other effects without problems.
I'll install some Mirages to verify if I have the same problem.

I've changed a value in the contact point for the nose wheel of the Y, the wheel is ok on the ground, on the other hand, the yellow support for the tail is no more on the ground....

Mike P., if ok for you, I'd try some other changes and let you know, also for the T1

I know that test at 4x speed can cause problems, but I do not yet understand why some models at an airport stay on ground, at another float. Doesn't matter for now

Thanks, Giorgio

I don't have that pure virtual function call error in FSX, Giorgio.

The front wheel is a tad sunken in, but not much (I just realize that you did not say that it affects all models, so I tested the wrong one most likely ).

Patience, patience, everybody, I will be home from work tomorrow, and this will get sorted out......

Awesome !!!

Many thanks,
Paul

The fix is in the works. I anticipate finishing all aircraft with in-sim checks by later tonight my time. The T, PAN, and R series were just contact point adjustments, but the Y required a modeling adjustment to the rear tripod ground stand, and nosegear CP adjustment as well. I normally don't have the aircraft shadows enabled, and this threw off my perception of wheel-to-ground contact, so sorry about that one. The AB effect will be fixed as well, a big thanks to Steve H. for the insight about the mandatory end line I accidentally deleted when I was stripping long flame emitters out of the effect. This was more of a learning curve for me, so I don't feel too bad about it, lol.

I will post when I publish the fix to AVSIM.

Okay, the G91 fixes are off to my two faithful beta-testers, so the Gina project should be wrapped up except for the last package of users-might-have-been and paint kits. I am going to keep this thread very much alive with a new aircraft in advanced state, pretty much finished in alpha state except for texture mapping and part cutout/animation.

This is the Fouga CM.170 Magister, a French-built jet trainer/light attack aircraft which was a highly visible aircraft all over Europe, Africa, and the Middle East. It's another great companion aircraft for the Cold War theme that I have been doing most of my modeling on, very much at home with Mirages, Starfighters, and Ginas. As you can see, I already have the flight dynamics in play, and I am trying to get this one done by the end of October or early November.

Ah, the Magister.. Will it come with the mandatory ear plugs to cover up the infernal noise that thing creates?

There's one (OY-FGA) based at my old local airport, which used to do rounds at the biannual airshow. For a hefty fee (~€400), one could get a 15 minute ride in it