There is nothing wrong with that. It just works.
Elite Dangerous plans for 2024
- Thread starter Arubeto
- Start date
I'm more worried that now n-body discrete Newtonian physics has been determined "fake" an awful lot of real-world planes and such are going to fall out of the sky and NASA gonna lose a whole lot of assets.
Are you saying it's not <insert aircraft manufacturer name of choice here> that their planes are crashing?
Well, that's not very typical at all, I'd like to make that point. I was thinking of the other ones. Some of them are built so the side doesn't fall off at all!
Pretty sure once you get to NASA level you need to go beyond the Newbtonian fake stuff and use the relativistic real stuff instead.
For games it's all fake and the question of implementing something becomes how many layers of code hacks does it take to make it seem real enough at the level of detail the game is going for. Having your own engine could be an actual advantage here as long as you still have people proficient in it.
It’s more of a YMMV thing IMO.
I can definitely see how walking from cockpit, through the ship, and to the top of the boarding ramp can be an immersive experience. I do it occasionally myself, when I’m in the right mood. Mostly, if there’s no gameplay to be had traveling from point A to point B, then I find it an annoying experience, which breaks immersion. Starfield simply reinforced what I’ve experienced in other games.
Sorry, no, NASA does not use GR on the regular, for three reasons:
The constant G itself is only known to four sig figs, so if you're making any calculation that relies on G, there is absolutely no point worrying about millionths of a metre, because we don't actually know how strong gravity IS to that sort of accuracy. The uncertainty in the estimates for the mass of the Earth are there because we can measure how the satellite moves to a greater precision than we can measure the strength of gravity.
It makes no difference in-system anyway for trajectory planning, excepting Mercury or closer. (Ooo, I might go and look up whether the Lucy team had to worry about it)
Rockets, gyros, actuators, gimbals, whatever, just ain't that accurate. If Einstein says you need 2ns more burn or an extra milliradian... tough, you can't control the burn that well anyway!
You do hear about GPS as an example because that relies on decades of accurate station-keeping and continual recalibration, so there's long enough for the tiny deviation from Newton to grow into something significant. Even then the noise and jitter is greater than the correction factors so you still need 5 or 6 sats rather than 3 to get down a sub-metre accuracy on the ground.
As for "games are fake" well OK, but so are digital twins and they are a very solid part of deployment of precision complex physical engineering these days, so I'm not quite sure what you mean here. If you mean games don't need to be 8 sig figs accurate, true, but that's what we're all saying anyway.
The constant G itself is only known to four sig figs, so if you're making any calculation that relies on G, there is absolutely no point worrying about millionths of a metre, because we don't actually know how strong gravity IS to that sort of accuracy. The uncertainty in the estimates for the mass of the Earth are there because we can measure how the satellite moves to a greater precision than we can measure the strength of gravity.
It makes no difference in-system anyway for trajectory planning, excepting Mercury or closer. (Ooo, I might go and look up whether the Lucy team had to worry about it)
Rockets, gyros, actuators, gimbals, whatever, just ain't that accurate. If Einstein says you need 2ns more burn or an extra milliradian... tough, you can't control the burn that well anyway!
You do hear about GPS as an example because that relies on decades of accurate station-keeping and continual recalibration, so there's long enough for the tiny deviation from Newton to grow into something significant. Even then the noise and jitter is greater than the correction factors so you still need 5 or 6 sats rather than 3 to get down a sub-metre accuracy on the ground.
As for "games are fake" well OK, but so are digital twins and they are a very solid part of deployment of precision complex physical engineering these days, so I'm not quite sure what you mean here. If you mean games don't need to be 8 sig figs accurate, true, but that's what we're all saying anyway.
There is no such thing for FD. It's a business and the mileage shouldn't vary. Now how much money does a business put at risk to develop a niche feature that maybe a fraction of players use somewhat regularly, a bigger fraction once or twice and another fraction doesn't use at all?
I wouldn't call it an RPG because I have a very specific definition of that. I'd call it an Aventure game though, and combat just happens to be part of it.
Open world sandbox RPG, maybe? I think there is a strong aspect of Role Playing going on but for sure not wrapped up in an overarching quest framework that normal RPGs have. Though I'd definitely say that First Person Shooter is one of the most inaccurate primary descriptions of Odyssey that can be made.
Even if there’s no ”point and click this obscure pixel” gameplay to be found.
Having to type GO NORTH, GET MANUFACTURING INSTRUCTIONS, SHOOT GUARD WITH L6 ROCKET LAUNCHER came to my mind when the term Adventure game is used. lol
It's one of the reasons why I think Frontier should, through the Guardian narrative, introduce limited artificial gravity to the game. It can be handwaved to be unrealistic to apply to a Coriolis station due to exponential scaling of the technology required compared to the size of the 'gravity' field generated. There can be gravity zones in outposts and megaships etc.. Would love to see some Super Mario Galaxy shenanigans going on in Elite.
Unless calculating the motion of four player avatars bunny hopping like loons on a ship additionally trying to escape an interdiction whilst traveling at 55c flying through a multi-moon binary Jovian system?
That really shouldn't be a problem still.
So long as anything connected to the ship has its position defined relative to the ship rather than relative to the primary star - this is the obvious way to do it and since your hardpoints don't fall off in a sharp turn or your holo-displays go out the window, almost certainly how Frontier has implemented it - it doesn't matter how either the ship or its interior contents move. You then just translate those ship-centric coordinates to instance-centric ones (or even world-centric ones, possibly) as needed, and possibly to camera-centric ones for display purposes depending on how you prefer to implement that bit.
All you need then is for both separate bits to work: the ship can already be interdicted through a multi-moon system without obvious problems [1]; players can already jump around each other in a shared instance ... that the shared instance is itself changing position and orientation quickly isn't any more of a problem in this case than it is if you land on Mitterand Hollow for your exercise routine.
(Coordinate transformations in a game like Elite Dangerous should be absolutely routine operations carried out flawlessly thousands of times a second. I can think of a lot of potential technical and implementation challenges for ship interiors, but there is no circumstance under which this should be one of them)
[1] Leaving aside the bug reimplemented from FE2 where if you move between two reference frames which are moving relative to each other, your speed relative to the current reference frame stays constant rather than suddenly jumping. It's not normally noticeable because you need to have a speed which is in the same order of magnitude as the relative speeds of the two reference frames but it's fairly obvious if you're supercruising towards a fast-orbiting planet from "behind" it in its orbit. But that's not really relevant in this case.
So the thing that gets me, and still does, even after some (highly appreciated) thorough and detailed replies, is how the game works out the huge disparity of simulated velocities/distances relative to all the objects interacting with each other in the example I gave. When dealing with celestial bodies, they are large enough that any potential rounding errors/lag effects can be multiple kilometers out or more before they are visibly perceptible, but millimeters? I think of the few times I've driven over a planet surface and have found myself clipping through the surface terrain, and I'm wondering if that may be an example of a desync between the SRV & the planet? I've also seen rubber banding between ships in supercruise when in mulitplayer, which looks like network lag, and seeing a ship jump a quarter of an inch across the screen isn't terrible but in supercruise that would represent many light seconds distance. I get locking the coordinates from the ship to the avatars means that both objects are moving as if as one through supercruise, but it's the disparity of working out the offset when one starts to move around relative to the other that starts to unravel that.
I get that game development has an element of smoke and mirrors and I wonder if there isn't some distance scaling workarounds that make some of these simulated distances more manageable, but again, that doesn't necessarily jump out at me as solving the differential equations that need to be made to account for freeroaming avatars on ships in freeroaming supercruise.
Stephen Jay Gould would be proud.
Mainly the key is that the game doesn't need to solve those equations because the answers are irrelevant.
If you're on ship A in supercruise, and a party of CMDRs is running around ship B also in supercruise ten Ls away, then exactly where those CMDRs are on board the ship is entirely irrelevant to you - you can't see them or interact with them in any way where their position is irrelevant. From the point of view of the CMDRs on ship B, they're all sharing a common reference frame for their interactions with each other, so that reference frame being in motion with respect to the star system is also irrelevant.
You can store the coordinates as a nested set [1] (illustrative, not necessarily what Frontier does)
- planet is at these system coordinates
- instance is at these planetary coordinates
- ship is at these instance coordinates
- player is at these ship coordinates
Operations within a coordinate system are straightforward (and generally all on the same scale). If the game can support CMDRs on a planet surface, or ships flying around an instance, it can support CMDRs on a ship. Most of the actual interactions take place between objects defined in the same coordinate system - a player inside a ship can't shoot one on a planet surface - which makes things simpler too.
Operations between coordinate systems are mostly straightforward too - you just "add" [2] the difference between the coordinate systems after doing the calculations for each object in its own coordinate system. You can lose precision at this stage if the difference between the coordinate systems is large ... but if you do, it doesn't matter, because the distance between the two objects should be so much larger than the lost precision that it's still "good enough" for whatever calculation you wanted to do. Most inter-coordinate translations are going to be done for display purposes, where the final imprecision is going to be thousandths of a pixel if that.
Essentially the hard parts of the problem had to be solved to do things like a Coriolis station with a ship landing at it and a pilot looking out the ship's window, all orbiting a planet - which worked fine well before 1.0! - and everything else should just be implementations of that.
[1] If you try to do everything in a unified worldspace coordinate system, then yes, you'd get huge problems with imprecision on the scale of an ED system. Even just moving or rotating a ship might start to fail once you get too far from the star.
[2] Not quite the right term - once you get into the coordinate systems themselves likely being rotated relative to each other then "multiply" might be more analogous - but good enough.
Shouldn't be in this sense - more likely a problem with collision detection or the terrain generation.
You only talk about yourself in every single sentence, did you notice that? You have so many discoveries. You don't care about the number of players as long as you can play well. You are not participating in CG. You play only with a few friends at most. Man, this is an MMO. They won't maintain a game just because of you, which is only profitable with thousands of users. Your narrow-mindedness infuriates me, i'm sorry.
This game only works with players.
In my opinion, I've never really considered Elite an MMO ? BGS and PP could be NPC only for the majority of the players. I've played open and come across a few players but generally they disappear when a 07 is sent. And I started in 2016.
To change this game to single player wouldn't affect any gameplay .
But that's been my take on the game . Players don't make the game never really have .
If I remember correctly, in the early KS days it was some backers who came up with labelling ED as an MMO - which FDev eventually adopted.
The FDev stated reason as to why we don’t have single player (offline single player was announced part-way through the KickStarter and dropped less than one month before release (yes, I’m still disappointed with that
)) is that players are required for the BGS - the Galaxy “would be unacceptably limited and static compared to the dynamic, ever unfolding experience we are delivering" if the game was a fully offline experience.But I think the point being made by @terasz is that without thousands of other players the game will not be profitable and therefore facing the chopping block, not that they are required to enjoy playing the game. If you dismiss someone else’s comments with variations on “well it doesn’t bother me” then it strikes me as a failure to address the bigger picture.
No need to be sorry... You hold your own interpretation on what a game should be, for sure.
Naturally I only talk about me playing a game - and my in-game friends too (did you miss that?) as this is the only relevant factor, for me.
If you, personally, base your gaming on other player's participation, that is fine...