Star Citizen - Chris Roberts, lots of spaceship porn, lots of promises

Meanwhile, over there…

Assume that I am and provide an answer. For example, in which other games can I pilot a space ship with other people, stop piloting it while in space, get off my chair, open cargo hold, jump into another space ship in zero G and go pilot that, land on a planet and go shoot some people, all seamlessly without loading? If there are other games like this I want to know! Afaik No Man’s Sky doesn’t quite allow this.

The engine-side changes were mostly double-precision maths code and a brand-new networking library.

Does Star Citizen allow this, other than theoretically? I honestly don’t know.

Yeah, you actually can do all that stuff currently.
Some of that stuff you’ve been able to do for a long while. I posted some videos of it months ago.

Videos are for suckers.

I mean Unreal Tournament, Halo, and many others had multi-user vehicles since forever. Zero-G isn’t a technical difference that makes it more complicated (in fact it probably lowers the technical load). Even MMOs have had multi-user vehicles for some time so the scale isn’t even new.

Edit: The hard part is tying it all together into a coherent and fun game (which as previous examples show isn’t new) and yet Star citizen hasn’t yet shown it can do that.

Edit2: And before this turns into a “yeah but those are much smaller vehicles” that’s not a valid argument. The other games had smaller vehicles because they didn’t want to spend months on a single vehicle of art time (which SC does) because that’s a huge time and money waste. The technical and gameplay aspects have definitely been done in previous games though and the added scale SC is trying to do doesn’t add complexity to that.

Heh, it was a video of me playing it.

Or maybe Allegiance by Microsoft (great space game/RTS).

Its still a tech demo, not a fully playable game.

It’s not about the zero G, it’s about the seamless transition into different physic frames of reference, each with its independent gravity, in an open (networked) world. It is quite a technical achievement, not absurdly so, but I do know of no game that integrates such a system (most of what comes to mind it’s obviously scripted and not systemic).

I mean, it is fully playable game. It is not finished, it lacks lot of content, but it is fully playable. I played it. Tech demo implies something much more limited than what it already there.

Plus the unified first person and third person, also not very common just yet, and the combination of scale and fidelity.

Still, Squadron 42 is going to be the proof in the pudding for me…

I will grant you that the scope is ambitious, but let me give you another ambitious scope: a pizza tycoon management game, where you build pizza stores. Except that the game simulates you getting out of bed every day, making your coffee, getting in the car and driving in a fully simulated city, and then getting to your store and managing it in first person. Each keystroke on your in-game keyboard is meticulously simulated and requires skill.

This is fully doable with today’s tech, but the question is: where is the gameplay loop? How do you put this thing together in a way that doesn’t suck? What are the design decisions? And, in case I want to turn it into a bustling MMO with hundreds of players, how the heck do I do that, because at that point, I need super tech.

That’s what SC is. It’s a slice of life simulator that is completely doable with the tech we have in the single player case, and impossible in the MMO case. The reason it wasn’t done before is both because space games didn’t have a sufficient market until recently, and because figuring out how to turn this particular slice of simulated life into something interesting and worthwhile is difficult. Most coherent space game designs don’t really need you to get out of your ship – you’re going to be walking across familiar, boring terrain repeatedly, and any FPS bolted on will suck in comparison to just about any other FPS. Most will cut out the boring parts and not make you sit on a static 5 minute train ride every time you land on a planet. Good designs will let you skip the boring parts of space, and there are very, very many boring parts in space. Good designs also have to wrestle with providing some sense of scale in terms of number of planets, since randomly generated planets are boring (see NMS). Good designs also need to figure out the possible strengths and weaknesses of each ship – something impossible to do when you’ve sold ships to backers at insane prices before you’ve figured out how they fit into your design. And that’s what SC completely and utterly lacks – a design. And that’s why it’s a tech demo rather than a game.

Uhhh, again those have all been done in vehicle combat games. Do you really believe that in Battlefield 2 (the really old one) players had the same physics if they were in an airplane, helicopter, or running through the terrain? Or all the different phsycis that GTA IV vehicles have been shown to have?

Most games don’t try to have a unified physics model for everything, they have different physics models based on what system you are attached to at any given time, which is fundamentally the same as what SC is doing.

Edit: Here is anothe way to think about it. SC is just a game with Gravity set to zero always, and the player has a velocity modifier that changes depending on if they are in a vehicle or not or if they are in proximity to a planet’s surface. That’s no different than Just Cause games where you jump out of a plane and have a parachute, in which case you are affected by a base gravity and the parachute gives you a velocity modifier that points away from the ground instead of towards the ground…

I’m assuming this isn’t meant to be in response to me, because I agree with you. Nothing they have done or showed off so far is actually new on either an individual level or even on a lot of aggregate levels. Their ambition is stupid not because they are trying so many new things but because they wont’ focus on “a thing” to create and have scope creeped the hell out of the game that they don’t even know the fine details of the gameplay loop to put everything they’ve created together.

Sorry I was looking for the point in the chain to respond to and messed up. Cheers!

Yo dawg, we heard you liked space ships.

Attaching an object to a physics body so that it inherits a translation rotation and speed is fundamentally different than having parallel physics systems running within distinct frames of reference and transitioning from one to another. These are different problems and the games you give an example of are not examples of what SC is doing (in the case of modern Battlefields and Just Cause, that you quote, they are doing precisely the opposite).

Edit: to be specific, having a collection of physics object A, that are enabled and interacting with each other, moving freely within physics object B, that is moving, receiving accelerations and rotating, would cause the solver of pretty much any third party physics middleware to become unstable at certain (low) thresholds of complexity. You can hack at it a little, but eventually even the hacks won’t be able to deal with the instability. In the case of spaces as big as those in SC, you are also dealing with floating point precision. The solution to the problem is to run several instances of the physics middleware in isolated “rooms”, each room using a single frame of reference, and defining windows of transition between rooms so that you can convert momentum, velocity, etc, between frames. It’s not rocket science, but it’s a significant code engineering problem, specially because the physics engine is middleware. And it’s nothing the games you mention attempt, at least agaik, because they don’t need to and it would be weird for the to spend resources that way.

Could you explain what you mean by this? The way it works in SC currently is, gravity starts X km above a planet. On one side of that boundary, you will be pulled towards the planet. On the other, you’ll float in space. They have no atmospheric modeling of drag, lift or anything like that because their ship designs are not aerodynamic.

EDIT: Also no gradual onset of gravity, which would be the realistic approach.

See above. I don’t know how they deal with planets but the really effective things is how they deal with walking within ships engaged in combat.

I see so you’re talking about having gravity inside the ship vs no gravity outside of it, or just transitioning from one to the other? Or is it about the networking needing to carry over all the state that’s in a ship so it can display what’s inside the ship correctly? Or just the fact that there are containers inside ships that behave differently from outside of it? It wasn’t obvious which of these you consider to be the real challenges.

I explained it as well as I can muster in the post just above the first of yours.