[Taristin]

Seriously... because FS is so big on newtonian physics...

[FireCrack]

*cough*modifications*cough*

[StratComm]

But you don't change basic behavior to support a mod.  It should be built on top of the engine, not in spite of it.

[FireCrack]

*cough*flag*cough*

[StratComm]

It's unfortunately nowhere near that simple.  The physics controls a lot of stuff, and to incorporate true Newtonian would involve negating some of what's already there.  At absolute best, you'd have to find someone to reprogram the physics engine, and then have it only active via a command line argument.

[FireCrack]

yeah, i know, i'm jsut being anal. last few posts i posted obnly for the sake of posting them, cause i was bored.

[StratComm]

Fair enough  I just see so many requests for true Newtonian from people who don't understand what a big undertaking it would be for a feature that would not normally be used.

[Flaser]

-Translating animation support.

Either as a true implementation or as hack of the current rotational code - though I think the first will be better in the long run.

-A dive into the turret code: It would be the next wonder after multiship docking - I don't request anything new, just the a dive itself into that portion of the code so it can be understood/optimized.

[WMCoolmon]

Originally posted by Flaser
-A dive into the turret code: It would be the next wonder after multiship docking - I don't request anything new, just the a dive itself into that portion of the code so it can be understood/optimized.

It's really not all that monstrous. A turret basically:

- Finds something to shoot at
- Aims at where it will be
- Shoots at it

Most of the code involves figuring out what to shoot at who.

[Flaser]

Than why wouldn't anyone touch the code with a 10 foot long pole?

BTW some more efficient targeting algorythms may help the AI immensly.

What do I mean?

Instead the current, pick a target and fire at its lead designator, the AI could merly shoot in the general direction of a fighter/bomber wing.
Turrets would immensly benefit from this - they will be able to truely unleash hell in a curtain of lead as point defense weapons are meant to be.

On another issue: is it the AI that's so horrendous then?

[WMCoolmon]

Because it's still big/complex, especially when there are missions that depend on the AI being a certain way. Ex: escort missions through an asteroid field. If capship turrets went all out, there wouldn't be an asteroid field.

'Fire in the general direction of a wing' wouldn't work too well, because shots would be continually ending up behind ships or you'd have to get the average predicted position for all the ships in the wing, and if one flew away, the turret would end up firing at nothing.

Anyway, there's also a bunch of stuff to make turrets less accurate. Flak is always slightly randomized, if the weapons system is below 70% turret performance is degraded, in a nebula range is degraded (unless the ship is tagged), and of course the AI class plays a part in things as well.

And actually base newtonian physics wouldn't be that bad either. You just need someone good at physics, but most of the code is contained in a couple of functions that are applied globally. What FS2 does is set a desired_velocity variable, and then has the object try to match that to the best of its ability. All the throttle really seems to do is set that variable.

[aldo_14]

I can't help but wonder if you could rewrite some of the dogfighting AI to use a set of 'rulebooks' that can provide simple escape solutions... i.e.  if enemy on 6 and enough space clear, then do an Immelman turn.  Simple, deterministically called (but random enough to be predictable) strategies that complement existing behaviour.

 (this is something used for chess playing AIs, for example, but i'm not sure how useful or feasible it is for real-time games.  I am curious about it, though)

Unfortunately, I've heard the current code for AI is a complete and utter bastard, so I'm scared to even contemplate looking at it.  And, to be honest, I dont even have the first clue how it's stuctured or the behaviour is setup on the top level anyways.

[Lynx]

3d brifings would be teh ****. I'm not talking about 3d icons, but a real 3d briefing room. The ability to insert custom pictures or animations  in the briefing, like a pilot asking something, or a target profile being magnified. Kind of like the briefings for Star Trek: Elite force. Immersion, baby!

3d mainhalls. You could have multiple rooms, look at it from different perspectives, and so on.

[Flaser]

Originally posted by Lynx
3d brifings would be teh ****. I'm not talking about 3d icons, but a real 3d briefing room. The ability to insert custom pictures or animations  in the briefing, like a pilot asking something, or a target profile being magnified. Kind of like the briefings for Star Trek: Elite force. Immersion, baby!

3d mainhalls. You could have multiple rooms, look at it from different perspectives, and so on.

Custom mainhalls and in effect, several custom places to go around are already possible AFAIK.

[FireCrack]

so can anyone say somthing about non-flak muzzle flashes?

[Turnsky]

Originally posted by FireCrack
so can anyone say somthing about non-flak muzzle flashes?

i reckon that's quite doable, since it's already quite possibly present in the original code..

it's nice to see people discuss, brainstorm, and whatnot.. i generally mean these kinds of threads to be just a brainstorming platform for either the SCP or Ferrium..

and Lynx, My point exactly, if you could have character modeling in the game (may or may not be possible with the SCP...i won't speak for Ferrium, however) but that's the kind of thing i was talking about, that, and the possible use of more immersive cutscenery as a game progresses.. i mean, who wants to see the bosch monologues all kitted out in-engine, with all the SCP goodness to boot at high res without the video graininess..

or better yet... Hallfight.

[MovGP0]

Ok, many of the guy's here would like to have a more Newtonian Flight Model. But you don't see that in the most space flight simulating computer games. [1]

But why is that?

The Answer is quite easy: The way of controlling a ship within a Newtonian space-environment would be to different from the effects you have in Air[2]; You would fly straight ahead with current speed if you turn of your engines[3].

This means that if you want to make an ideal 90° turn you have to turn by 180° trottle your engines to the max until you stop, make a -90° turn and trottle your engines until you have reached the speed you want - Nobody can really control that in 3D.

A possible solution might be the NASA one: use some kind of flight computer - you say you want to turn by 90° and the flight computer makes the needed engine stuff.

I think the current flight model shoul'd be kept in the most part from the view of the user[4] - therefore the arcarde like input need to be preserved as much as possible

--------

The effects of an newtonian model would be:
(1) destroyed ships would fly straight ahead and the wreck hits other ships or falls into planets, stars and black holes
(2) "nozzle-engine-exhaust-stream"[5] effects like you know them from the film Apollo 13 or from Babylon 5
(2.1) engine stream only at accelerating; stream goes to front at decellerating
(3) turning of the ship around axis with an change of the flight direction will be mostly slow
(3.1) a possible solution would be to let ships only flight slow so that there is not to much decellerating work to do
(3.2) to let ships flight also fast there could be a kind of booster - but than the ship could only turn to engage enemys - not to change the flight direction
(4) there needs to be an indicator for the distance of the point-of-no-return when being in the near of black holes[3] - such high-gravity objects will change the way of flight dramatically because flight to them in will lead to faster speed; flight from them to slower speed
(5) last but not least - the whole physics engine need a complete redesign - this is a very heavy stuff!!!

--------

[1] As Exceptions see: Asteroids, Moon Lander, etc.
[2] The same problem with sound - in space there is none, but that don't matches your all day experience (turn of your speakers - you won't like that)
[3] not so with warp engines
[4] an exception might be that accelerating is faster than decellerating
[5] not sure about the translation - hopefully the meaning is clear

[Flaser]

MovGP0, some of what you write of is completly irrevelant, while for others there are already tried and prooven methods to counter.

First off: Play some I-War. Than play some more I-War. Then play I-War 2: Edge of Chaos and then play it some more.

They have a perfectly working newtonian physics for all intentions and limits of a game engine.

////////////////////////////////////////

As for your suggestions:

What you didn't write about is the tendecy of true bodies in space is to start uncontrollable spinning.

The next true problem is that there is no speed limit in a true newtonian engine. The common mistake of a beginner space pilot is overshooting his target - even I-War hadn't had good enough readings / help to counter this problem.

Moreover in such a "realistic" environment, the only likely form of engagement is BVR - beyond visual range - with missiles as your primary arnament. At that point the game will be more of a game of Tycondera or Harpoon, a chess play of radar hide and seek. That can be fun, but most hardlighters don't come to us to do so.

So if you want a true newtonian engine in a space shooter you have to:

1) Make sure the engagement takes place at visual range (30-15 km max).
2) Provide the pilot with a control system that automatically stabilizes the ship's spin.
3) Provide a control system or accurate readings that help the pilot do a good merge on his target without unintentionally overshooting.

////////////////////////////////////////

The things that is irrevelant for a game engine or as you put it is Gravity. Forget the Buck Rogers physics.
Take a physics book look at the equation of gravity:

F = g * m * M / r^2

F = Force in [N]
g = gravitic constant a very small number
m = mass of one object in [kg]
M = mass of other object in [kg]
r = radius or distance of objects in [m]

There is no distance limit to the effect of gravity! I cold be half a galaxy away and the sun will still pull me. The effect will be minimul, but there will be. All objects pull me in this manner.

Therefore forget about falling into blackholes, suns, planets meteors. There isn't a thin line or border where they suddenly start to pull me.

In mission therefore such an effect will be already present from T- minus zero and throughout the mission.
In actual space battle the only gravity I will take into account is a stellar body I'm orbiting.
Nothing else will have a considerable effect within the timefram of a mission! Stellar travel takes months without going at relativistic speeds.

////////////////////////////////////////

To clear things up I will do my best to put togheter the 101 of convenional space-travel in a different thread.
(I will also use a lot of stuff from the Wikipedia since they're better at giving simple and clear representations).

[MovGP0]

I'm missing the way of docking from X - Beyond the Frontier or Wing Commander - you had to flight into a given region of an ship/station to land and therefora accomplish the mission.

When you look at Star Trek all ships have the ability to warp-out. but to always warp-out has 2 drawbacks:
(1) warping out to land on a ship 100 meters away looks strange
(2) warping of ships witch normally can't warp looks strange, witch is not a problem in Star Trek but in Babylon 5 and some others.

In the Babylon 5 example jump-gates (or big ships) opens a jump-hole and then all the ships have to flight into them. There is no way for the ships to jump without jumpgate -- A behavior that is more docking than jumping.

This would also lead to more tactical deep - because if the jumpgate is destroyed or closed there is no way to escape

[MovGP0]

I don't want to say any different than yours - and right, I've didn't played I-War.
But give me a change to explain:

Stability Problem:
not so if you have perfect balanced ships (simplified physics model) and a good auto-pilot that helps you - I call it "flight-computer", you call it "control system" - the meaning is the same

BVR-Problem:
therefore I suggest to limit the speed or at least make it overproportional harder to accelerate for high speeds.
For big distances there is the afterburner-idea
-- I think this is a problem where we should avoid to much realism...

Infinite Gravity:
right, but when gravity is low you don't model it; just wanted to say that in the extreme near of very heavy objects like black-holes you need to respect a special behavior[1] - and right, during the mission the gravital effect doesn't change

[1] there is a thin line where the player doesn't has a change to fly away from the gravital object - you model it that the player will fall into the object and die when he is to near
But it will be interesting to be in the near of black holes when shoots doesn't go straight ahead but makes an bow