What Sparfell said ^^
There is a reason we have a boost mechanic so that we can rapidly change vector when needed. The regular thrusters simply can’t handle the momentum of the ships in high velocity combat / racing.
See the video I posted above on that exact topic. A wing surface generates a huge amount of force relative to the engines so either INS can create artificial wing surfaces, or they make the thrusters overpowered.
As I said earlier, if thrusters are powerful enough to rotate on the spot, either while hovering or while in motion, why wouldn’t a pilot use that capability? In order to force dog-fighting rather than turreting/jousting, you have to take away that capability.
There are two ways to do that:
Undesirable because we want to be able to hover and rotate at slow speeds for takeoff etc.
ie add huge ‘virtual’ fins way behind the ships centre of mass so that the more the nose is pointed off the velocity vector, the more it is forced back onto the velocity vector.
This would force swoopy aeroplane-like manouvres but I think this would feel really artificial.
If you point your nose too far into the airstream, it’ll flip you around like if you try throwing a dart backwards.
Flight Assist would control the thrusters to prevent the pilot from doing this meaning the tightness of turns would be limited depending on your speed.
With Flight Assist off, pilots could try to fly right at the limit, risking getting flipped out of control while trying to pull off slightly tighter turns/crazy aerobatic manoeuvres.
This is kinda similar to modern jets like the Eurofighter which are deliberately unstable and the software controls the control surfaces to keep it from flipping out.
In case it’s not obvious, number 3 is my preferred method of achieving dog-fighting.
Rotating is not a replacement for making rapid turns and it doesn’t make them useless, there is no need to nerf it to make turreting an undesirable play style.
How are you going to dodge a missile for example by only rotating? Or how are you flying fast through a canyon? What use is rotating your ship into a direction you are not traveling? Why not just mount your weapons on a gimbal If you want to fire in all directions? Just because not moving and just shooting is better in the current prototype because dogfighting is impossible, doesn’t mean it is the superior tactic in general.
Good points, thanks.
However, rotating allows you to point your main thrusters in any direction and then thrust at many g.
A Harrier can hover. Hovering doesn’t require that much in the way of thrust. Everything that modern jets do to hover can be applied to spaceships. At high speed, they put the majority of their power into shields that give them an aerodynamic shape. At low speed, they can rely on thrusters. The thrusters are nerfed, so even putting all your power into them doesn’t give you the control that the shields would give, but it would give you a different style of control.
To a degree, spaceships would have the choice of helicopter or jet controls. That might make for an interesting toggle on ships.
As for the observations about stability, the shields can be as magical as we like. If the shields provide the aerodynamic shape, then they can also subtly warp and twist as needed to provide aerodynamic stability.
The shields could also do things like form a razor-sharp disc around the spaceship, letting it fly “backwards” or “sideways” at high speed. There are all sorts of control modes possible when magic technology gets involved.
So the real question is one of what people want. Flavien presented his poll in terms of “realism” which I think is the wrong question to ask. I think the actual question is whether people want to keep flying the ships as if they were in space, as helicopters, as bullets, as jets, as dirigibles, or as something else altogether. Perhaps atmospheric flight should be a microcosm of space travel, where you fly like an airliner from point to point, then fight as if you were in a helicopter. Many things are possible.
There is a difference between rotation of your ship, witch means changing where the nose of your ship is pointed and turning your flight vector, witch means changing where you ship is heading to. You normally have to use the flight assist to turn your flight vector to where your nose is pointing.
Realistically you need much less trust for rotating a fast flying (winged) brick around his own axis then turning his flight path.
I was under the impression, that Flavien suggested only that he add a little aerodynamic to help when pulling up in a dive, or let you fly a tighter turn, if you roll and pitch instead of yaw, without totally loosing the unique nature of the vector trust flight model.
I am hoping he didn’t mean to artificially restrict the rotation of the ships nose, like in these other typical biplane/submarines in space sci-fi games.
But if Flavien wants to make the position of the nose matter more when you are trying to turn your ship, because of aerodynamic shields and different strengths in directional thrusters, it would make atmospheric flight much more interesting.
(If dogfights in the atmosphere became boring because of turreting/jousting, I would wish for the addition of some special designed missile or tractor beams etc. and not a dumbed down flight model. But hopefully this is a different topic.)
I wonder if gimbaling the main thrusters would help with changing the ship’s vector faster in an atmosphere.
I think that would just achieve faster rotation, but I’m happy to be corrected.
I fly with FA off except when just cruising from A to B, and its simple to use the main thruster and the lateral (and vertical) thrusters to point your velocity vector wherever you want it, all while pointing the ship in any unrelated direction.
This is the assumption I was operating under, too. I’m perfectly happy with the realistic approach with lower speeds and being able to use the manoeuvring thrusters to their fullest extent; but when getting up into 1000m/s+ velocities (where the terrain becomes the most fun to fly around), being restricted to what feels like making minor tweaks to a straight line seems way too limiting.
So what became of this discussion? I’m sorry to necro the post but as we edge towards the alpha I’m keen to keep it a hot topic. Looking at the competition (Starfighter Inc being one such game) it is clear that the key selling point of Infinity is the atmospheric / space crossover aspect of the game. I feel like this needs to be the focus as there are numerous space arena combat games out there and this is what makes Infinity shine. It’s becoming more important now due to the announcement that another key selling point of the game, solar system wide battles are being shunted towards sequential planet / moon scaled battles. This makes Infinity more similar in gameplay to the competition. I imagine the core gameplay to be more of an arena battle than a long drawn out war of attrition at this stage.
None of these points are criticisms as game design is indeed an iterative process which evolves throughout its production.
Anyway, my point is that now we have arrived at this juncture with the gameplay evolving along the path it finds itself, the distinction between space and atmosphere (including different density atmospheres with different amounts of gravity) should become a key strategy. I feel like there should be a choice between ships which perform better in space and those which perform better in atmosphere, thus making the decision to plunge into the atmosphere of a planet a tactical one if being pursued by ships more suited to space or visa versa.
What does everyone think? Flavien? Do you think this is an area to think about or is it going to be a situation where all ships are equally good in space and in atmosphere?
I welcome debate on this. 
We’ll know more on this once we see objectives implemented. That will help shape what causes players to transition between locations.
We already have some interesting effects to do with the size of capital ships - they aren’t very good in atmosphere at all!
There are already different atmospheric densities and gravity on the planet/moons in the game right now. As a consequence different ships handle differently and have different speed caps in atmospheres.
Iterating on the atmospheric flight model will happen at some point, but right now for the alpha we’re focusing on implementing the missing gameplay features.
In general, alpha should have most features implemented even if they’re in a rough shape. Iterating on these features will happen between alpha and beta. Polishing, bug fixing, optimizing, adding missing fx and balancing will come after beta.
Brilliant! You’ve thought of everything
Is it a possible future addition to have two variants of each ship class? One atmospheric and one space? Both able to fly in either but each being more suited to their favoured location? Possibly related to spawn location. If you spawn in space you get the space variants but if you spawn on a planet you get the atmospheric version. It may add an interesting twist to the gameplay.
That’s an interesting idea but I’d limit this to just the interceptor as that would reduce the amount of work needed and the interceptor would benefit the most from it in the end.