Print

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

[DarkBlueBoy]

How goes the head scratching Stainless? 

[Stainless]

Gone from head scratching to experimenting.

Give me some more time to play.

[DarkBlueBoy]

Of course Stainless! You take all the time in the world!

[VF111Sundowner]

Hey Stainless,

First I think it should be defined what the overall scope will be, IE: a improved IL-2 1946 clone, covering many fronts, or battle specific Cliffs, and Stalingrad. Open world, where what if scenarios rule. Tomcats vs Zeros al la Final Countdown, or a Flight of A-10s vs Romels Panzers in North Africa. ( i Personally feel the current plane list in the current mod community is overwhelming)

Then once that is decided the next step would be to publish a formal list of the intended abilities, features and specs. Dynamic campaigns, multi-player, and flight modeling

Then all the niceties, effects, graphics damage modeling(which should tie into flight modeling).

[Pursuivant]

To my mind, the core of an excellent flight sim is flight models, damage models and human models (i.e., AI behavior).

Pretty terrain and photorealistic graphics are secondary to getting the planes to fly right, fight right and take damage right. The third aspect is creating realistic aerial "terrain" in the form of weather, visibility effects, and so forth.

Otherwise, you're just creating an airborne FPS like War Thunder.

While IL2 has its faults in all four areas, right now it's the best we've got for WW2.

What would get me excited about a new flight sim would be improvements in the areas I've mentioned as compared to IL2.

Flight Models: Dynamic Center of Gravity, Realistic engine, fuel and oxygen management, Realistic radio management, fogging of windows at altitude, Possibility of systems failure, Truly amphibious planes, Realistic water landing and ditching, Skidding on mud and ice, bogging or ground looping due to mud or deep snow.

Damage Models: Custom ammunition belting, Better consistency of damage models (e.g., some planes are much better at surviving damage than others, even though they're of about the same weight and construction, the same engine in different planes survives damage very differently), Better fire starting and fire damage modeling (self-sealing tanks are seriously nerfed), Realistic damage texture modeling (i.e., damage textures based on type of bullet and hit location), Better critical hit modeling (to crew, guns, fuel and oil systems, oxygen systems, control surfaces and control lines), Realistic injury to crew (i.e., hits by large caliber bullets should be lethal, as should nearby hits by canon shells), Realistic smoke based on exact damage, Realistic coolant/oil/fuel leaks leak, Fuel shut off and diversion, Manual fuel pumping, Manual fuel dumping (for the planes that had it), Realistic oil/coolant/blood splattering.

Human models: Human reactions to hypoxia, frostbite, fatigue & combat stress, Ability for AI planes to get lost, Ability for AI to "recognize" landmarks on the ground, Ability for AI to accidentally attack friendly targets based on familiarity and type of target, Ability to command bomber crews fully, Ability to perform first aid on wounded crew, More intelligent info from crew and squadron mates (like calling out locations of enemy fighters, ground targets, parachutes, etc.), More intelligent AI crisis management behavior (like deciding when and where to bail out, or whether to bail out, ditch, or crash land), Realistic flight, squadron and group tactics based on year and air force, Realistic ground control/tower interaction.

Environment: Having a map for the entire world, or at least a big portion of it (like was done in FSX), AI for ground units, ships and AAA, Submarines that submerge, travel underwater and surface, player controls AAA guns, Ability to place real trenches, craters, etc. which make holes in the map, Dynamic weather with the ability customize cloud types, altitudes and percentage, icing in cold weather, Dust storms, Different levels of dust kicked up depending on terrain type, Partially destroying buildings, more realistic building damage models, Larger fires, Fires that spread on the ground, Wave/water action and appearance based on wind and weather, simulation of boggy/muddy ground, simulation of deep snow, simulation of icy/slippery conditions, contrails for prop tips in humid conditions, realistic radio, radar and navigation systems.

Perfection would be the the Air Traffic Control, Dynamic Weather and open-source of FSX, damage models of something like Rise of Flight or Cliffs of Dover, Flight models of CloD, RoF or BoS (or maybe X-Plane) and the content and focus of IL2.

Finally, I don't think it's that hard to bolt a new graphics engine onto IL2. It's been done before with IL2: Wings of Prey. It would certainly make the game look better, but as I said, it's not how the game looks, it's how it plays that matters.

[Stainless]

However a lot of the things you are talking about ARE graphics based.

"fogging of windows at altitude"
"Realistic damage texture modeling"
"Realistic smoke based on exact damage"
"Having a map for the entire world"
"Ability to place real trenches, craters, etc. which make holes in the map"
"customize cloud types, altitudes and percentage, icing in cold weather, Dust storms"
"more realistic building damage models, Larger fires"
"Wave/water action and appearance based on wind and weather"
"contrails for prop tips in humid conditions"

These are ALL features of a graphic engine.

I can tell you now that the combination of all of these means there are NO FOS engines that can do all you are asking for.

I've always assumed that if I go ahead, I will be extending the graphical capabilities of whatever engine I choose to adopt. No engine does everything you want, but I want to minimise the work I have to do in the graphics side of the code to allow me to spend more time on the rest of the game.

The choice of an engine for me is the most critical decision. If I get it wrong, the game will NEVER come out. Simple as that.

[Herra Tohtori]

These are ALL features of a graphic engine.

Mmmhh... actually in my opinion most of those are things that just happen to have a graphical representation of some kind.

For example, icing conditions first require simulating the weather conditions (temperature, precipitation, dew point etc.) that would tell when the aircraft would encounter icing and how rapid it would be.

Secondly you would need to make the flight modeling take into account the accumulation of ice, both by increasing weight and degrading aerodynamic performance - and finally, just as an indication to the pilot as to what's going on, you'd want to have visual icing on the wings, fuselage, and canopy.

This kind of feature requires some pretty profound software engineering, not just a graphics feature that you turn on or off.


Similarly the cloud types, dust storms, water animation based on wind conditions, and contrails for prop and wing tips as well as engines are all features that are more tied to how accurate you would want to make your weather system in the game, than the graphical representation for them. And I'm pretty sure that's something that would need to be a custom addition to whatever game engine you wish to use.


I would absolutely love to see a flight simulator with complex weather system based on simplified physical reality, but the sheer scope of the concept is somewhat disturbing. If it could be done, it would be absolutely magnificent, and basically you wouldn't need to ever worry about making custom cloud coverage because the weather would always produce unique conditions.

But I have no illusions that, for gameplay purposes it comes pretty late on the list of priorities after good flight modeling, well-modeled aircraft with good damage model, accurate weapon ballistics, and good looking environments to fight in. Static weather conditions can provide a perfectly good illusion of reality - the main advantage of dynamic/physical weather is that you would get a lot of phenomena that are a part of real flight, and it would have a nature of unpredictability to it that I feel would add quite a bit to the simulation side of the experience. Things like thermal currents, turbulence, wind, gusting, crosswinds... these would be things that are quite difficult to emulate with static weather conditions, but would emerge naturally from a well-designed physical weather system.

[VF111Sundowner]

Or we could wait until Holodeck technology is available...

[SAS~Tom2]

Hahaha! Spot on, always wished for that since this(these) episode(s). Go, Chief and Doc Bashir!

[Fresco23]

I am following this thread excitedly, and all the ideas in this thread are exciting and awesome... but if you guys make something with Stalingrad like graphics(as mentioned earlier in the thread...) then the only way I can run it is to get a new job, and buy a whole new PC with the greater income... because my PC would die if I just threatened to run Stalingrad on it.. IL'2 is fine, even Lock-On FC2(graphics set low), but beyond that it gets terribly hit or miss whether games run. Best of luck though, wish I could do something terribly useful and be of help in some way.

[Stainless]

For example, icing conditions first require simulating the weather conditions (temperature, precipitation, dew point etc.) that would tell when the aircraft would encounter icing and how rapid it would be.

Secondly you would need to make the flight modeling take into account the accumulation of ice, both by increasing weight and degrading aerodynamic performance - and finally, just as an indication to the pilot as to what's going on, you'd want to have visual icing on the wings, fuselage, and canopy.

This kind of feature requires some pretty profound software engineering, not just a graphics feature that you turn on or off.

Rubbish,

Simulating a physical system is easy compared to the technical difficulties of displaying it.

The conditions for icing to occur are easy to calculate. It's simply a case of working out a few equations and applying them.

Doing a good looking shader so the player can see it, then doing another one for when the ice starts to break up, that's bloody complex.

[Herra Tohtori]

I both agree and disagree with you.

The difficulty of simulating a physical system depends on how complicated you want to make it. The system I was thinking about was a fully dynamic weather system, using air volumes and tracking the relevant meteorological data for each "cell" of air - and those parametres would determine whether that cell of air moves up or down, north, south, west or east - whether that air cell should have cloud or precipitation in it, etc. etc. This would also include conditions that were conducive to icing.

Now, if you just make a simpler system which would feature a more static atmosphere, with pre-positioned clouds and you would just encounter icing conditions if the temperature and precipitation was right - then yeah, I agree, the difficulty of coding that system is probably quite simple.

But in my view, the visual representation of this kind of thing should be the last piece of the puzzle, complementing the physical part of it: Weather system that provides icing conditions, the physical effect of ice accumulation on the weight and aerodynamics of the aircraft, their effect on the aircraft's systems (carburettor, pitot tube, radiator, coolant freezing, jammed leading edge slats, limited control surface range of movement...).

Much like the physical part of the simulation, the visual part can be made simple or complex. At its simplest, it would just be a texture that gradually expands to cover the surface of the aircraft and canopy glass - much like damage decals. You could make it much more complex, like using refraction and diffusion shader to create the effect of gradually degrading visibility through the canopy. You could differentiate between clear ice and "white" (snowy) ice accumulation. Should the ice slowly thaw or break away in pieces - I can definitely see how that could get really complicated as well.

What I'm saying is, it all depends on how complex you want to make a feature like that. It's possible that the simulation side is more complicated to make, or it's possible that the visual side is more complicated to make. Depends on the implementation.