Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[HiFlyer]

Honestly I just assumed that if anyone was the logical choice for being the ones to codify and organize the accumulating knowledge, it would be the developers......

I strongly believe there should be simple example models of the right way to do things, so that newcomers can take a look and orient themselves from there. Right now, its very much a hodge-podge of competing techniques.

Its especially bad I think with the cars, where there is nothing even resembling a standard. At least right now with planes, Levi is the standard!

Peace and long life to Levi!!! 

[bomber]

I was struggling to explain, and then the first detractor (don't take it personally) as to why there shouldn't be a Wright Flyer has explained just why there should be one....

Because it's a simple example and something that can be used to progress from...

Flight Modelling is my bag.... but don't be confused that it's my passion, because it isn't it's Heavy bombers of WWII, of which I have a large selection of 3d models, 11 Lancasters , 3 Manchester, B17's, B24's...

Why spend the last 3 years on a single glider flight model then ?

Because JSBsim flight models are not what WE want, neither in quality, functionality or transferability to another plane... it's too poor, vanilla and requires far to much customisation to 'hit the numbers'

To attract hobby developers WE need a cross section of planes, that are the required level of quality and functionality that demonstrate what can be done, what can't and how to use this plane to modify and create another of the same era.

A flight model for a glider, bi-plane, heavy bomber, helicopter, fast jet are not the same...

My worry is that jumping straight into 5th generation combat planes with eventually just get us nice looking models but everything else, avionics and flight models will be poor or identical to other planes..

ie Another Flightgear.

Don't take this the wrong way, because I know we're al putting in the hours, but lets not adapt a culture here, like flightgear where we're all working alown.

Simon

[josem75]

The concern I have is that at this development stage everything we do flight sim, combat flight sim should be to lay foundations... In skills, knowledge and techniques so as to build a large community of mod developers and not a large community of people waiting to fly the next plane made by one or two people...

One of the reasons why I like the MSFS series is that the did make thing relatively simple for developers.  Between all the information they included in the SDK and the sources from www.fsdeveloper.com, it was real easy to learn.  The way Outerra is right now I can do static models, but everything else I am at a total loss on.

This is totally my case.

I was thinking those days on what wriah said here. It would be nice to generate a knoweledge so many people can begin to work, and not a few people actually working with all the models.
For me the answer, more than wiki, is videotutorial in Hd or full hd with every step. There its almost imposible you loose. Wiki can also have all the technical details but today for today a tutorial is what all the people will und and fast.

I saw a video tutorial for put a car in outerra from max, but it was not hd and for me was dificult to read the things unfortunatelly in my screen.
We have Wriah, Levi and more, maybe they can document in a video some of their current jobs. Even LEvi can do in his other language, Spanish, so more comunity can come.

[bomber]

Communicating flight modelling is the next major challenge I'm facing....

My flight models are nothing like what you'll find anywhere else, so the jsbsim and flightgear community are of no help there.

I've spent the last 2 months on ensuring standardisation and naming convention consistency throughout the flight model.

I'd like to do another glider, detailing the process that needs to be gone through.

[M7]

so Bomber if i'm getting you right, you start your flight model from scratch, then use somekind of virtual wind tunnel to find how plane surfaces react, all in software?

if so it sure sounds like the best approach. i can see why you would need to use simple model first then move on the evolution path. idont know crap about flight sim but i were, i'd be really excited by your approach.

so you first glider was done using that method? starting from scratch then finding flight data from the model itself?

[bomber]

M7, I started looking at flight modelling when I was mod leader for Target4Today at Targetware and what I saw was that a Hurricane, a Be109, a Spit, a Fw190 and all their variants should have a relationship to each other.. they were not designed in isolation but in response to a particular flight characteristic of another. But this relationship didn't exist within Targetware as each plane was created by an individual for the game in apparent isolation and brought into the sim with the declaration 'this plane hits all the numbers'......

So simply put if all planes brought into a sim 'hit all the numbers'.... the relationships don't matter a jot as all planes are perfect and behave as they should.... But here's the rub.

Flight Modelling is often described as being a bit of a 'black art', what this translates too is that it's an exercise in compromise. Most jsbsim flight models created only have data for inbetween the plus minus angle of attack stall points with no attempt to understand what's happening once a wing stalls, so no spins or effects associated with climb stalls, high speed stalls  http://flighttraining.aopa.org/students/maneuvers/skills/stalls.html.

Then there's the issue of what's important to a flight model, and you'll hear alot about determining a planes co-efficients of lift and drag and there are programs out there that'll do that and yet here again you'll find only for certain configuations (bi-planes. no way) and only within the safe operating envelope... and only for a wing made up of a single airfoil and incidence... There must be a plane whose wing uses a single airfoil and incidence but apart from a paper one, I don't know what it is. So from a flight modelling co-efficients stand point you're compromised from the get go, before you even start adding your own compromises into the mix.

Then there's a Mass Ballance, moments of inertia and CoG position of a plane and getting the data for this, and there are genuinely very few sources for this, so it's invariably a made up set of of numbers with no relationship to the materials and construction techniques used in the design and build of the plane.. These are important values and JSBsim gives you the opportunity to use them if you have them (you can't omit them) but determining them for say a 1915 bi-plane or for finding the actual values for an f18 or f22 , good luck.

And now I come into the fray..... I simply refuse to compromise, and I never use a 'magic number' to make my planes 'hit the numbers'.. So I'm totally at odds with the whole flight model building community as exists in Flightgear and JSBsim. And as such have had the tendency to ruffle a few feathers to say the least.

I've come here to Outerra because you guys are at the beginning of a flight sim creation. You have no baggage, no thousands of hours invested and all the politics that go with it... of "that's the way it's done here"... You have no Guru's who's opinions are above questioning and you're as equally open to combat as non-combat aircraft. There are simply no walls that need tearing down before we can start doing something correctly, we simply start doing it correctly.

M7, the answer to you question is so complicated that we don't just need a new thread for it.....but a new flight modelling forum....

So Outerra developers, could we please have a new child flight modelling forum ?

Regards

Simon

[Acetone]

This debate is really interesting, and is demonstrating the modding paradox. It's true that it's always better to make things step after step and not try to build a car before having invented the wheel; but in the other hand, it's not possible to expect that a bunch of enthusiast modders will follow a normal developement pace, because they are by definition creating content they would like to have.
As a result, modding is nothing else than a field of anarchic projects gravitating around a well more (hopefully) organized game/sim. There is also a delay in the global learning curve, since more users will come later, with their own projects.

In the end, I guess there is only one response to this paradox, and it was already mentionned many times in this thread :

I think what we really should be doing is documenting everything officially, from modeling, scripting to building advanced JSBSim FDMs in a public wiki that all of us devs work on together to consolidate our knowledge in one location. There is a lot of great knowledge in the forum, but it is unstructured and fragmented through thousands of threads. If it were consolidated, structured and refined by all of the devs on a wiki that would make it much easier for new devs to find the information they are looking for.

If we manage to expand the wiki, it will probably help a lot to raise the quality of the community content and attract people ready to try and learn with a proper documentation.
I will do my best to write something about the few things I've learned in scenery creation.

I consider myself as a moderate experienced flight simmer and I would like to hear a bit more about these flight model questions, it's really an interesting subject.

[HiFlyer]

This debate is really interesting, and is demonstrating the modding paradox. It's true that it's always better to make things step after step and not try to build a car before having invented the wheel; but in the other hand, it's not possible to expect that a bunch of enthusiast modders will follow a normal development pace, because they are by definition creating content they would like to have.
As a result, modding is nothing else than a field of anarchic projects gravitating around a well more (hopefully) organized game/sim. There is also a delay in the global learning curve, since more users will come later, with their own projects.

Honestly, I was just gonna leave it there. I would say though, that one can risk tilting against the grain of a very individualistic endeavor, practically begging for frustration in trying to tightly organize a process that as you say, Acetone, is anarchic by nature.

I'm perfectly willing to watch the experiment, but i've found myself in the hotseat, having been voted into leadership of several volunteer team efforts before (Yikes! I should have run away!) and the concept of herding cats comes to mind. Very temperamental, egotistical and feisty cats, especially as the group gets larger and larger.

I wish everyone the best of luck, but you will have a much better chance if the Dev's step in proactively as the ultimate authority when the nearly inevitable disputes, and teste-fests arrive.

Be careful, guys!

[PytonPago]

He has a point. Many will want to start to do stuff anew like newbies and having them "boxed" can be quite impossible. Doe, having them take on an more "standardized" approach by an non-abiding way may prove be better ... simply by putting a ton of how-to´s converging on the actual "standard" in mind. No need to say people they should do this or that beyond giving a helpful point of view. If they get a help package, that doesnt need them to stick to it, whyle giving them a voluntary inside of the pros of that showed approach ... most will find something to get them to use them on their own.

I kinda like a thought of having more how-to vids about importer, that will show them some nice magic, that can be done already, and are potential in the future. And ill certainly hope to finally get to the moment to show off the stuff i done at least in some vehicle stuff (hopefully i can resolve the most stuff that stands in my way for that this year).

The few here making a good job so far (actually, for ya Levi, Uriah and some more people here, it should be a little more than just good job to give the proper credits ), but new ones would need a little helpful hand ... in show-offs for all stuff that is already possible to do and some, that each in the community envisions.

 ... so, ill have to start some serious work on some props and vids. But basics are ... as i can see it in chemistry, there are too "institutions" and "organization" and "standards" ... but all suck at something, so in the end, its mainly the persons bringing imaginations to life for whoever it is, walking that road, rather than those "box-ers", opening up the future for them ...   

[HiFlyer]

I have a question, though.

It's already been noted by others, that Simulation-wise, Outerra is already getting a bit "top heavy" on the aviation front, with the amount of detailed work on aircraft far exceeding any equivalent in ground and watercraft.

Will any similar efforts be devoted to those areas, and by whom?

And how is the whole JSBSIM/BULLET interaction problem working out at the moment?

[bomber]

I'm not advocating a straight jacket approach to directing peoples efforts but something more subtle...

If left to their own devices a person interested in flight sims will come up with a plane he/she likes and 'have a go'... nothing wrong with this but equally a person if presented with a choice of planes from which to work on, and even join in the development with others, will chose this.

Offering the latter isn't a bad thing and gives the flight simming community here an avenue for continued growth and presenting our works as a group, hopefully pulling in others.

So a shopping list so to speak of 2 planes that have a relationship to each other...

Focker I and Camel
Spitfire II and Be109
Sabre and.......

Roman Galley and Greek Galley
Spanish Galley and RN privateer

Tiger Tank and a Ronson.

I'm sure you get the drift...

Simon
 

[jskoker]

It's already been noted by others, that Simulation-wise, Outerra is already getting a bit "top heavy" on the aviation front, with the amount of detailed work on aircraft far exceeding any equivalent in ground and watercraft.

There are tons of combat sims, driving and racing sims, and general open world sims.  Flying we have FSX/Prepar3d and X-Plane basically.  FSX/Prepar3d has an outdated game engine and X-Plane is lacking in a bunch of features that has become expected.

Us pilots have waited too long for this...   

[Uriah]

I think what HiFlyer means is that there are some extreme bugs with Bullet, sometimes hitting the ground can send the vehicle into space or through the ground.

Also, Pyton said he couldn't get anywhere with the gear system on the Ural, and there is not only limitations to the vehicle configuration script, but no standard or examples of more complex vehicles and watercraft.

Although my interests lie in aerospace, it is a valid point that Outerra needs to be a full scale seemless experience, and for that we need a fairly balanced level of quality for both aircraft and vehicles. I will try to put in some time to make a different type of watercraft, and also it would be great to have an amphibious vehicle!

We seem to have completely hi-hacked the original topic of the Boeing 737. If a model can be found it is a very straight forward process to import and configure it for Outerra.

I think the most agreeable solution at the present time is to import and develop as many mods as possible and if a better method for building flight models is developed an effort can be mounted to convert all of the old FDMs to the new system. As devs we should all document our findings in the OT wiki which anyone with an OT account can edit/contribute to by the way.

Regards,
Uriah

[PytonPago]

While its about 737s ...

http://interestingengineering.com/boeing-737-flight-simulator/

 ... he certainly needs one for pure perfection reasons ... i would want too. 

[ZeosPantera]

The ease of the flight sim aspect is why it is "taking off" once you are off the ground all outerra really is is a scenery generator. When you do vehicles on the ground collisions and physics are a HUGE part of making the experience enjoyable. The trees all being non-collide sprites also lessens the fun of being on the ground. If the physics were fixed and those 3D procedural trees implemented the other aspects of the sim would fall right in line.