Star Citizen Discussions v7

Dare I say Starsiege: Tribes (1998) and its 64vs64 matches?

Oh yeah i remember that one :) Although the quality and power of multiplayer netcode has improved quite a bit since then. That's why i wonder how CiG managed to completely fail on that, i believe there are enough coders on the market who have the knowledge to do a proper netcode. My hypothesis is they just used the cryengine code as is and didnt touch it, but even that wouldnt explain the ties with the framerate. I cannot imagine what kind of spaghetti code would include network updates in between render cycles, and i really hope that's not the case there for CiG sake...
 
Just to clarify: it's not the "looks and plays like crap" part that is the worry. Rather, the question is why are they polishing at all at this stage? They're in some kind of featureless pre-alpha. They are three entire development stages away from when it's time to start polishing anything at all.

One top of that is the question of how on earth they've managed to be so ridiculously inefficient and abysmally slow that they're not even in a proper alpha phase after six years.

Of course, the lack of any kind of forward movement is handily explained by the fact that they're developing backwards and therefore everything they do is done wrong and a waste of time, but neither of those should even be possible in any kind of half-remotely pseudo-professional setting. The utter and complete incompetence on display here is nothing short of staggering, and anyone who's involved with this project at this time is stained by it. If they ever displayed an ounce of competence in a previous instance, SC only ever indicates that it was pure chance and by accident. They simply got lucky last time and their fumbling was left unexposed. The argument "...but he worked on X" had lost all value and the only way for anyone associated with this disaster to regain any credibility at this point is to either quit or for SC to exceed every last expectation, very very soon. The latter will not happen.

A clean game is a healthy game. That's why they polish. Also: Croberts heard that that the Witcher was made by polish(ed - he omits the ED where he can, though) company and he wishes SC to shine just like that.
 
They can't do another free-fly weekend until 3.0 is available to everyone, as it would draw attention to the reality, that 2.6 is the released Alpha version. Highly unlikely to be this year, I'd have thought. Perhaps that was what you were thinking!

Lol Yeah, I wouldn’t waste my hard drive space installing 2.6 again. Sometime next year though, I want to see this planet tech they’ve been crowing over. I too want to experience having cargo boxes stuck to my fingers as I salute other commandos. I want to ride in someone’s Idris. Maybe it’ll stick to my fingers too, who knows?

I’m looking forward to checking it out. I don’t expect to be floored or have a vision of Jesus Croberts, but I hope, hope, hope it will be fun.

Most of all, I can’t wait to haul some bio-waste!
 
Once they start adding gameplay its going to be very interesting to compare the fantasy with the reality. I mean, if they can pull it off, then wow... but you know, i'm not sure its possible to do even half of what they have talked about, at least not within a timescale of less than 10 - 20 years.

Even if they do implement everything they've talked about, that won't account for all the crazy dreams that all the hardcore fans have accumulated during the long, long wait. When you really look at quotes like that one about the missions, he didn't describe hands-on mechanics at all, just the idea of chaining hand-crafted pieces. They've been so vague with the actual details that they could conceivably implement things like the golf-swing while being true to the letter of what they've said. But it won't come close to satisfying their more enthusiastic backers. Those backers aren't imagining just another game where you press some buttons and some things happen, they expect Roberts to deliver something utterly revolutionary that works like no other game before it. That's why CIG is petrified by the idea of locking anything down, making it "real", and having to deal with the inevitable backlash. Which is why they just ignore it and work on shopping instead.

In a normal multiplayer game that doesnt work like that. The GPU render thread is not correlated to the network thread (unless someone made a huge mistake in the code..), and the volume of data to transmit is also not correlated to all of this.

For an engine designed with those goals in mind, sure. But SC continues to be a glorified CryEngine mod, c(r)obbled together on top of an engine that was never designed for most of what they're trying to achieve. The independence of systems you described is something that needs to be designed and engineered, it's not just a law of nature that applies automatically. And it looks very much like, in the case of SC, it wasn't.
 
Just to clarify: it's not the "looks and plays like crap" part that is the worry. Rather, the question is why are they polishing at all at this stage? They're in some kind of featureless pre-alpha. They are three entire development stages away from when even it's time to start polishing anything to begin with.

On top of that is the question of how on earth they've managed to be so ridiculously inefficient and abysmally slow that they're not even in a proper alpha phase after six years.

Indeed. When somebody says "polishing" I understand the term as in "fixing bugs, making it run smooth, optimizing etc" NOT GLOSSING EACH INDIVIDUAL PIXEL for maximum visual fidelity. Graphics have never been a guarantee for a good game but many core backers seem to believe exactly that which is a bit worrying. We have been wondering with CiGs priority years ago but that was at best confusion. Apart from DS who touted that this will never work lots of people were giving CiG the benefit of the doubt and waited to see if they managed to get it work. After all this time with THIS result I can safely say "Derek Smart was right" and chuckle at the bit as well.
 
That's why CIG is petrified by the idea of locking anything down, making it "real", and having to deal with the inevitable backlash. Which is why they just ignore it and work on shopping instead.

I love how that make JPEGs expensive preorders for a giant lootbox as no one can say what SC will be. The two "plagues" of modern AAA games that infuriate gamers, though CIG seems to be immunized to gamers' ire.

Though I'm sure the most hardcore will contort their mind to convince themselves CIG achieved the new jesus, whatever they release.
 
Last edited:
On top of that is the question of how on earth they've managed to be so ridiculously inefficient and abysmally slow that they're not even in a proper alpha phase after six years.

Isn't it simply that what they're trying to do is crazily ambitious? And probably impossibly, undebuggably ambitious with finite time and money, regardless of which engine you're using or how talented your team are.
 
Isn't it simply that what they're trying to do is crazily ambitious? And probably impossibly, undebuggably ambitious with finite time and money, regardless of which engine you're using or how talented your team are.

Well that’s the weird part. What they’ve talked about, dreamed about, hypothesized about is incredibly ambitious. What they have so far is pretty run of the mill. I could knock together all of their current gameplay in unity in a week or less. 3.0 is more ambitious, but certainly not more ambitious than Elite Horizons which took less than half the time SC has burned through to get us on the surface of planets.

The only standout in SC is that it looks so damn pretty. That’s the polishing part. It makes for a good presentation as long as you only pay attention to the superficial elements.
 
I think it actually is. Not because of the performance per se, but because this 6yo mess is revelatory of their bad habits since the beginning. Roberts looks like he's living in a marvellous 'verse™ where time, money, bandwith, RAM, CPU cores and speed, polycount, backer's trust are infinite.

As raised some posts before, even if the netcode was near perfect there's many other bottlenecks they have to overcome, starting with the loony "no-limit to fidelitay!" that laden any asset. As always, they should have built things from the ground up: lowpoly to testbed then ramp-up towards what they can be realistically achieved and balanced between every sides. Now they're just hopelessly to try to shoehorn things in an already bug-crippled and choke-full leaking cardboardbox. And they didn't even started to implement AIs.

The point was that performance isn't a major concern given that the game is in a preAlpha state.

The problems are why are we seeing it like this? Why are CIG wasting time and money doing work they will likely have to throw away? CIG have too much work to do on the games foundation to be worried about stress tests

The game is in a preAlpha state and...for SC at least...the game engine isn't complete. There should be no polishing or balance passes or even ships beyond the bare minimum and the fact that CIG are polishing the game and holding stress tests that prove nothing says a lot about their priorities and management.

CIG squandered at least three years of development because of CRs change in direction from space sim to MMO life simulator. They've wasted more with their refusal to follow standard development practises.

There is a reason most companies get the basic game engine working first. CIG...OTOH...seem to think developing an MMO without the netcode needed to support it is a good idea.

And preAlpha is not the time to be polishing the game, conducting stress tests or creating game assets. The game is likely to change so much between now and release, any such work done now will simply need to be redone.

For a preAlpha, what is on show is impressive.
But for six years of development, it's an embarrassment.
Nor is it supposed to be impressive. It's supposed to be a work in progress. CIG should be digging the foundations, not choosing the furniture.
 
Last edited:
Isn't it simply that what they're trying to do is crazily ambitious?
No.

Being ambitious would explain things if it took long to finish, and that you might at some point have to surrender to the fact that you have to split things up and deliver it in chunks. That's not the problem CIG has demonstrated. They're not having problems reaching the end goal — they've tripped over their socks trying to cross the starting line. Hell, they've tripped over deciding what floral pattern is supposed to be on the socks.

Also, what they're trying to do is not particularly crazy or ambitious — just poorly described, the the point of being almost completely undefined. They have mumbled incoherently about lots of fanciful ideas, yes, but those are just dreams. Having dreams are not the same thing as having ambition, much less having any kind of ability to actually realise those ambitions.
 
Last edited:
Isn't it simply that what they're trying to do is crazily ambitious? And probably impossibly, undebuggably ambitious with finite time and money, regardless of which engine you're using or how talented your team are.

There is a difference in being ambitious and saying you are going to do something, when you know full well you can't.
 
Isn't it simply that what they're trying to do is crazily ambitious? And probably impossibly, undebuggably ambitious with finite time and money, regardless of which engine you're using or how talented your team are.

What's ambitious about Star Citizen?

Other than Chris seeing how much money he can rake on?

Star Citizen is very doable. If you think about it, they already have much of the game working. All they need do is merge their tech demos and use their planet generation tech to create a system...

That they haven't done so yet implies they can't.

More importantly, even if it were ambitious, ambition doesn't excuse the wasteful way they've developed this game or the managerial incompetence that seems to be behind much of the problems affecting development.
 
Last edited:
What's ambitious about Star Citizen?

Other than Chris seeing how much money he can rake on?

Star Citizen is very doable. If you think about it, they already have much of the game working. All they need do is merge their tech demos and use their planet generation tech to create a system...

That they haven't done so yet implies they can't.

More importantly, even if it were ambitious, ambition doesn't excuse the wasteful way they've debeloped this game or the managerial incompetence that seems to be behind much of the problems affecting development.

Too much truth here, brain overload :D
 
What's ambitious about Star Citizen?

Ooh, ooh! I know the answer to this. The answer is wrong, granted, but I know what it is.

“It has never been done before at this scale and with this level of detail.”

That's the very heart of the problem: the idea that detail and scale somehow makes it 1) new, and 2) ambitious. Of course, in reality, it does neither. First of all, everything CIG is doing has been done before, and when asked to specify, even the backers will eventually reduce that claim to talking about visual fidelity. The problem with that is that visual fidelity is just standard evolution and having better graphics is not new. It has happened every week, month, and year for the last half-century. The rest is just mechanics, methods, and gameplay we've seen dozens of times before (and CIG hasn't even gotten around to actually doing the vast majority of it yet).

Secondly, the problem is that detail is not ambitious. In fact, in most instances, it's just stupid. Simulating the oxygenation of blood and its effects on stamina is ambitious if you're programming a body-physology simulator as a medical teaching tool or as a kind of experimental environment — it's a complex system but getting it all right will improve the end result proportionally to how much effort went into it. For anything else, ambition would be to make a really clever abstraction of all of that with the least possible complexity, because while it's neat if it's there, it is so wholly irrelevant that it needs to have zero impact on performance. Complexity serves no purpose in this case — complexity for the sake of complexity means you've just lost the plot altogether.

I'll quote the Jennison letter at this point, because the underlying confusion with this not-actual-ambition is endemic to the whole project:
David Jennison said:
Chris Roberts might have a vision but he can’t communicate it. And therefore, no one on the team knows what it is. This is known to every team member, certainly of the art team. Roberts is not an artist and it is clear he is not a visual communicator. The basic understanding of macro vs micro, what is essential to the piece and what is not, completely escapes him. Everything is of equal importance- the laces on the boot are just as important as the overall value pallet and silhouette, in many cases more. This is indicative of Robert’s extreme lack of understanding of the most basic of artistic principals.
Substitute… well, anything, really for “artistic principals” and suddenly, a lot of the nonsensical and outright false claims about ambition become clear. The inability to distinguish what's important from what's irrelevant; what's worth spending effort and computing power on and what's not; and what makes it a game and what actively works against that goal is apparent in everything they discuss and show off to try to wow the fans.

When I look at SC these days, I just see the Monty Python sketch of a man trying to eat a cathedral. It's not ambitious. It's not impossible. It's just very dumb. Perhaps SC is slightly less bad for their teeth… perhaps.

More importantly, even if it were ambitious, ambition doesn't excuse the wasteful way they've debeloped this game or the managerial incompetence that seems to be behind much of the problems affecting development.
Also, this, many times over. Ambition does not magic away waste or the incompetence that time and time again creates that waste.
 
Last edited:
Indeed. When somebody says "polishing" I understand the term as in "fixing bugs, making it run smooth, optimizing etc" NOT GLOSSING EACH INDIVIDUAL PIXEL for maximum visual fidelity. Graphics have never been a guarantee for a good game but many core backers seem to believe exactly that which is a bit worrying. We have been wondering with CiGs priority years ago but that was at best confusion. Apart from DS who touted that this will never work lots of people were giving CiG the benefit of the doubt and waited to see if they managed to get it work. After all this time with THIS result I can safely say "Derek Smart was right" and chuckle at the bit as well.

"He who negleteth the Pixel shall not be worthy of the Big Picture!"

A deafening voice sounds in your ear as a reminder of your unworthiness in matters of that what was never done before!

*That showed him! The Man of Destiny reclines satisfied in his Chair of Lecturing.*
 
The basic understanding of macro vs micro, what is essential to the piece and what is not, completely escapes him. Everything is of equal importance- the laces on the boot are just as important as the overall value pallet and silhouette, in many cases more. This is indicative of Robert’s extreme lack of understanding of the most basic of artistic principals.
I’ve always been confused about this... is he saying Chris overvalues the shoelaces, or is he saying Chris doesn’t understand that the shoelaces are so important? It kinda reads like the second, but what part of this project screams “we don’t care about trivial minutiae “?


Edit: and then I read it again and it seems like he’s saying just that: he cares too much about shoelaces, he doesn’t understand macro.
 
Last edited:
I’ve always been confused about this... is he saying Chris overvalues the shoelaces, or is he saying Chris doesn’t understand that the shoelaces are so important? It kinda reads like the second, but what part of this project screams “we don’t care about trivial minutiae “?

The shoelaces are unimportant. Most people only see them from a significant distance compared to the size of the object. The important bits are the basic shape/colours - the stitching and laces being perfect can come later.
 
Wrong choice of engine, tools and inexperienced management and partially rookie team.

Classic AAA story, but this time people actually paid for that, not corporation.



Framerates are kept in sync of networking, if networking sync struggles, it impacts graphics.

It *should* in theory be different at later date, but I doubt current code does this async (which would just lag other players, but give constant frame rate).


Networking might put a general cap on FPS is there's a sync issue - which would explain why it seems to struggle at 70 fps while staring at a blank wall: You still have to keep track of stuff even when you can't see it. But the massive drop when looking at commandos must be a rendering issue. Too many fidels, I suspect.

Having said that, they did manage that 'Schrödinger's door' effect a while back, so there might some level of zen going on that I simply can't comprehend. If a commando T-poses on Olisar and no one is watching, does he still glitch through the floor?
 
Last edited:
Back
Top Bottom