The Star Citizen Thread v8

Status
Thread Closed: Not open for further replies.
Atrocious physics, acceptable FPS, nice atmospheric shader.

Very atmospheric - some monster has taken a bite out of the shade in this shot and that event horizon is back again. It must be a horror movie ;)

2kgU8Dq.png
 
Tough crowd in tonight! :D

I tend to agree. With all of its shortcomings, it looks quite fun. And they have the basic physics there. Sure the ground acts like glue, there's LOD lines, frame rate issues (understatement), but as a first approximation to a drive model, it's ok. But if CIG says it's anything but a first approximation (pre-pre-alpha?), then they need to think again.

Especially as 'our' gravity ranges from 0.05g to 9.9g, whereas SC has only 1g. :)

Yes, ED spoils us in that regard. I'd like to see the CIG rover in that range of gravity. It'd be fun! (in perhaps very unexpected ways)
 
Do they look awesome?

I can clearly see the LOD line where the pebble texture layer gets streamed onto the ground below the vehicle and seems to be tied to the stuttering. Later in the video that LOD transition gets more prominent and starts to become an incorrectly shaded border around the texture/LOD line. I wonder if it is due to that texture being too high res or the engine failing?

Yeah, it does look awesome. They have a long way to go in terms of frame rate but I remember playing Crysis back in the day when no one could run it over 30fps.

I hope they can soon stop expanding the scope of the game and focus on delivering what they have in a playable state. I wouldn't be surprised if this is in a playable state in a year. Not the full promised game, but something good enough to play like a sandbox game.
 
Yeah, it does look awesome. They have a long way to go in terms of frame rate but I remember playing Crysis back in the day when no one could run it over 30fps.

Does it look "awesome"?

The answer to that depends on what you are looking at. There are aspects of the "game" that look great.....and there are parts of the game that look awful. That's the way it has been for ages now.

And I ain't talking FPS but actual graphical quality.

It must also be borne in mind that Crysis was designed more as a "look what I can do" advertisement for CryEngine than a game. It was supposed to struggle at 30fps on most systems.

SCs end game is that it is supposed to provide acceptable performance with 1000 player instances and high quality graphics in VR.

SCs more reasonable performance goal will be acceptable performance with noone else in draw distance. It might achieve that in a year.

I hope they can soon stop expanding the scope of the game and focus on delivering what they have in a playable state. I wouldn't be surprised if this is in a playable state in a year. Not the full promised game, but something good enough to play like a sandbox game.

My....you ARE optimistic, aren't you?

Unfortunately, from what I can see, when CIGs devs say they are at the limits of what the engine can do...I believe them. They are already talking in terms of diminishing returns from their optimisation efforts and while we do know that the netcode they have is woefully inadequate for what they have now, never mind what they plan to have, replacing it will only be a band aid...CIG need to get their server tech up and running as well, need to optimise local routines and the simple truth is that CIGs focus on fidelity and detail comes with a high price, and that price is performance.

Even if SC had no other issues, having better, higher quality graphics simply requires processing power. Which is why other developers work to strike a balance. Looks are important, but looks can be enhanced through using a distinctive style rather than a high polygon count....such as NMS...but performance is one aspect which can't really be cheated. The game needs so many free processor cycles available that aren't dedicated to graphics so it can work.

Fixing the netcode will be an important part of that. But at best, it will only reduce the times when such performance issues matter. Unfortunately, so far nothing CIG have announced seems likely to offer performance improvement when it is needed most. When multiple players and ships get together.

Data culling et al will help...should help...performance when you are all by your lonesome.

Server integration and instances will also help when talking about an entire server.

But these don't really work when you get players together in one spot. Indeed, they can make things worse because of the extra processing overhead.

So being in a playable state next year MIGHT be doable if...big IF...CIG get their netcode fixed AND they give up on the idea of "1000 player instances" (or rather, the players give up on it since what CIG means
by "instance" isn't what players are expecting) and get used to working on their own.

Acceptable performance...in other words...will depend on how many other players you want in the game with you.
 
Last edited:
Yeah, it does look awesome. They have a long way to go in terms of frame rate but I remember playing Crysis back in the day when no one could run it over 30fps.

That's quite a different scenario though. Crysis was pushing more polygons than most PCs could handle, it was designed to be taxxing in that manner.
SC is struggling because they can't get things working as intended.
 
That's quite a different scenario though. Crysis was pushing more polygons than most PCs could handle, it was designed to be taxxing in that manner.
SC is struggling because they can't get things working as intended.

I'd love to see how it'd run if every single asset was just simple a simple untextured cube...
 
I'd love to see how it'd run if every single asset was just simple a simple untextured cube...

Yeah. It's a funny thing about game development, that part…

At this stage, that's what they should be showing in their videos because that's what you need in order to nail down the core essentials of making an actual game. The fact that they started with all these high-detail art assets is — more than anything — a clear indication that not only are there nowhere near any kind of alpha stage, but that they are also quite simply doing it wrong. Until they stop fiddling around with these absolutely irrelevant trinkets, they can never move forward towards actually producing a game because they are constantly working on stuff that provides no useful testing data; that steals immense amounts of resources with no way of knowing if they actually generate anything useful; and that will by far age the quickest out of anything they produce, so they have to keep working on it and thus never get any of that absolutely critical data or measure of usefulness — they'll just be stuck in an infinite loop of irrelevance.

The backwards process they've chosen can only ever result in one (or both) of two outcomes: the game ends up not working as a game, and the game ends up looking very outdated and ugly. Everything they've shown so far points to their achieving both of those.

There's a reason why actual game development — and indeed all software development — has this prolonged grey-box/placeholder/concept design phase, long before any real work on art even begins. CIG skipped straight to the “real art” part, and then didn't even particularly pay attention to it either, but skipped straight to the art polish phase. And they have no game to put any of that art in.
 
Last edited:
The backwards process they've chosen can only ever result in one (or both) of two outcomes: the game ends up not working as a game, and the game ends up looking very outdated and ugly. Everything they've shown so far points to their achieving both of those.

Yep they designed the shoe without measuring the foot, and are shoehorning it inside all the while carving the shoehorn with a spoon. Never done before!

And they can't help themselves constantly putting new socks on the foot.
 
Last edited:
Status
Thread Closed: Not open for further replies.
Back
Top Bottom