Flying is MAGIC !

[bomber]

I'm sure you've read the differing opinions as to what's the principle of flight....

I want to thrown my hat into the ring, just as Dumbo the Elephant flys so do a hell of a lot of the planes within Flightgear.. by magic.

There's no other explaination becaue if it weren't so there wouldn't be so many 'magic numbers' within the flight models..

When we see a magic number we need to seriously question it's validity..... afterall it's magic.

Simon

[IAHM-COL]

I restate my opinion, be that be a hat in a ring.
I think the numbers behind the JSBsim engine are second in importance to an accurate simulation.
Sometimes, I feel, the most real values underperform your classical -hare in a hat- numbers, for a more realistic JSBsim simulation.



I blame JSBsim. It could very well be a maze of mistakes all interacting within each other on a faulty FDM

But I am far to being the expert here. JSBsim still has an aura of magic, and mystery to me

[jwocky]

Well, it is not a JSB specific problem. There are some quite wild Yasims as well. So lets not go in this my FDM your FDM thing. The point is, for some planes, we don't have the accurate data, for some planes, it was a quick shot to make the whole flying in the specs, for some, and I admit to it, it was just for the fun (the Koenigsadler with cheat mode to enable her to fly with a flock of big birds in 39,000ft for example, well, it's not actually magic but more a tinkerer doing something he knows it's not "authentic").

And sometimes (now we talk about the psychology of plane tinkering), you start out with a plane that has a model and a very basic, usually Yasim conf (a certain French developer does it all the time. Any plane has exactly two tanks, almost always of the same size, the same flap settings and the same whatnot and flies like a wet sack). So you want to make something better with it and you start but there is a lack of data, so, you guess, you take values out of the model and such stuff. A lot of stuff, that would be in rl world engineering simply bull****. And while you do the first commits, just to get work out of the dev folder, the "authenticity discussion" starts. Someone bothers you about one number, again and again. Maybe even a war of the buttons begins. Newsflash, you know, not all the numbers are real, that wasn't the goal to begin with. I had about the Koenigsadler an endless discussion. The real one had 32 hours of air endurance, flown right. And someone who gave it about 40 minutes of air endurance tried to start a discussion with me, some parts of the engine wouldn't be "authentic" ... yeah, right! Now, in lieu of the actual data, what can you do? Give up on the project? Have some fun? Or even add, with a malignant grin on your face a "magic" mode? You all know me, you know what I do in such cases.
So, while I enjoy the level of detail work, Bomber does, I make planes for the fun of people. Users shall have fun. I want them to have fun. If they want to fly a route that would be entirely possible with the real plane and drop out of the sky because someone made one part "authentic" and didn't do the other parts ... that's no fun.

I think, maybe our rating system needs under that consideration a rework. Not only that rating should be out of the hands of a developer (I usually fill it all with zeros if I remember it at all) because honestly, no dev is objective about his little flying babies. And it needs an overall authenticity rating and an overall fun rating. Just my thinking. A lower authenticity rating and a higher fun rating tells a user what he has to expect. Maybe the plane is faster than it should be, climbs better, but it is just for fun to fly. Vice versa higher authenticity rating, less fun rating means, this is serious machinery, it will torture you, but it will also give you a sense of achievement. Or take the Lake of Constance 707. A beginner has to be happy if he makes it an hour without burning engine or iced windows or a little decompression. But she is fun and she is authentic in many aspects (of course someone will again play the critic and say, it could be even more authentic and torture the amateur in front of a monitor more, and yes, that is always a point. Theorethically between FDM, Nasal and some clever XML, we can simulate it down to the level in the pilot's coffee cup and the PMS status of every flight attendant).

[bomber]

I'm not against people creating the 3d exterior model and posting it up as a means in which to generate interest about developing the plane further..... I'm dead against them being called the author of the plane and having to go to them to ask if it's possible to improve it and it's one of the reasons I prefer fgmembers to fgaddon.

I also do flight models so as people can have fun flying, it's not some sort of hair shirt fetish I have here.

My aim is to raise the bottom bar of flight modelling... which shouldn't be too hard considering how low it is.. To show that it's not that difficult to create authentic engines, props and airfoils, to understand which bit of generic code snipet should be used for which type of plane... To make it an easier more intuitive task to create a starter FDM.

Non of my posts are critisism of individuals....I understand that what we have in flight modeling has developed over time, with people using others code, and using other code and adapting it is exactly the way to go

My critisism if any is that there's no foundations, there's nothing solid on which this copying other work is based upon... no supporting documentation of data or calculations.

If I create an authentic flight model of a plane that is a pig.... I want to understand why it's a pig and then if someone wants a fun flight model for it, removing it's faults as we should have understood why they exists should be an easy task as opposed to just flidling a magic number here and there and hoping like hell it makes the plane more fun.

Simon

[IAHM-COL]

I'm dead against them being called the author of the plane and having to go to them to ask if it's possible to improve it and it's one of the reasons I prefer fgmembers to fgaddon.

I completely relate. And it's bottomline of why I rejected the proposal of stop FGMEMBERS development as an step to find some calm.

[LesterBoffo]

There's a couple of fiddly details with YASim that are a bit annoying, concerning rotary engined pre-WWI and WWI aircraft.

One: that they are not having an authentic gyroscopic feel from the rotating engine mass, known as procession, and two: 'windage' power loss from all the rotary engine's cylinders whirling around against the incoming airflow. The first we can fudge some, but the underlying code for YAsim propeller mass momentum seems so weak for this effect, that it's really not worth bothering with. The second would have to be implemented with a bit of crafty Nasal coding. Actually in YASim, ( and I've had Jason from Fighter Squadron-WWI check this out..) our engines, even when set up exactly to spec for power/rpm and displacement/compression-ratio, over perform.

The procession issue is particularly annoying because it was a big factor in how all rotary engined fighters performed when the controls where thrown about, like in a dogfight, annoying because the gyro effect programming does exist, it's there in the propeller section of the prop engine part of the YASim FDM, it's just that someone didn't take the time to scale it to reality.

[jwocky]

@Bomber
I didn't say, you accused me of anything. But over the thought of who gets accused of what, the point I tried to went a little bit under. I think, what we could do is set parameters to give planes several "difficulty settings" to adapt to the skills and styles of users better. Some days, I like flying a 707, having all those buttons and switches, all the thongs to do, on other days, it's just give me a few turns over Grand Canyon and I am good.
But I suggested to mark somehow in the rating system, whether a plane is built more with authenticity in mind or rather has some magic to just have fun.

Having said that, it doesn't mean, an easier to fly and less "iffy" plane has to be based on entirely surreal FDMs. I don't even think, the planes with "magical" numbers have those numbers from bad intentions. It is just a lot of work to do research in the necessary depth and even more work to sit down and do the math. You can do that kind of math but not everyone can. And even you, with your skill set, spend already a lot of time on one plane. So I am wondering, we still have to do more to organize users as test pilots, but maybe we can do the same thing with research and recruit users as researchers? You idea of better FDMs, given we talk about hundreds of planes, is not only a question of doing the math right but before that, of getting data. I admit, at the moment, getting a research team together to supply FDM devs with more and better data is just a blurry idea on the horizon, but maybe one worth thinking about it.

@Lester
Yasim ... hey, I can find the engines in a Yasim plane. Sometimes also the pilot in the cockpit ... so I know technically what you are talking about but how to word it for Yasim ...

[bomber]

Not acussing anyone of saying anything, just treading carefully.

[jwocky]

But what do you think about the idea of getting some help with research?
And then, maybe better tools, so we don't need to do everything with paper and calculator or in spreadsheets anymore?

[bomber]

I love the idea...