Vapourware Scam Citizen - Only people with too much money can become StarCitizens! WOULD YOU LIKE TO KNOW MORE?

[fantadomat]

AMAZING,CLOUDS!!!

 

[fantadomat]

 

[Myobi]

https://archive.fo/Yhnx8#selection-797.0-797.136

After 3 years of silence, we got another release window to miss!

When will we see Persistent Streaming and Server Meshing in the PU?

Our current aim is to release Persistent Streaming and the first version of the Replication layer, ideally, between Q1 and Q2 next year. We’ll then follow up with the first version of a static server mesh, barring any unforeseen technical complications, between Q3 and Q4, of next year.

and... mofos ain't even sure the damn thing will be able to host more than 50 players x,D

Q: When you have a party moving (quantum travelling or other) from one object to another, and another DGS node, object, or instance is full , will T0 / Static Meshing create another DGS node pre-emptively? Or how will this be handled?

With Static Server Meshing, everything is fixed in advance, including the number of server nodes per shard and which game server is responsible for simulating what locations. This does mean that if everyone in the shard decides to head to the same location, they will all end up being simulated by the same server node.

Actually, the worst case is if all the players decide to spread themselves out between all the locations assigned to a single server node. That way, the poor server will be trying to deal not only with all of the players but it will also need to have streamed in all of its locations. The obvious answer is to allow more servers per shard, so each server node has fewer locations it may need to stream in.
However, because this is a static mesh and everything is fixed in advance, having more server nodes per shard also increases running costs. But we need to start somewhere, so the plan for the first version of Static Server Meshing is to start with as few server nodes per shard as we can while still testing that the tech actually works. Clearly that is going to be a problem if we allow shards to have many more players than the 50 we have right now in our single-server “shards”.

So, don’t expect player counts to increase much with the first version. That avoids the issue of a single server node becoming full before players get there since we’ll limit the maximum player count per shard based on the worst case. Once we’ve got this working, we’ll look at how the performance and economics work out and see how far we can push it. But to make further expansion economically viable, we’ll need to look at making Server Meshing more dynamic as soon as possible.

 

[GhostCow]

Just cancel this trash and make S42 already

 

[RobotSquirrel]

Just cancel this trash and make S42 already

If freelancer taught you anything its that Chris Roberts has to be physically removed from a project in order for it to release by a more competent project lead.

 

[ADL]

and... mofos ain't even sure the damn thing will be able to host more than 50 players x,D

Q: When you have a party moving (quantum travelling or other) from one object to another, and another DGS node, object, or instance is full , will T0 / Static Meshing create another DGS node pre-emptively? Or how will this be handled?

With Static Server Meshing, everything is fixed in advance, including the number of server nodes per shard and which game server is responsible for simulating what locations. This does mean that if everyone in the shard decides to head to the same location, they will all end up being simulated by the same server node.

Actually, the worst case is if all the players decide to spread themselves out between all the locations assigned to a single server node. That way, the poor server will be trying to deal not only with all of the players but it will also need to have streamed in all of its locations. The obvious answer is to allow more servers per shard, so each server node has fewer locations it may need to stream in.
However, because this is a static mesh and everything is fixed in advance, having more server nodes per shard also increases running costs. But we need to start somewhere, so the plan for the first version of Static Server Meshing is to start with as few server nodes per shard as we can while still testing that the tech actually works. Clearly that is going to be a problem if we allow shards to have many more players than the 50 we have right now in our single-server “shards”.

So, don’t expect player counts to increase much with the first version. That avoids the issue of a single server node becoming full before players get there since we’ll limit the maximum player count per shard based on the worst case. Once we’ve got this working, we’ll look at how the performance and economics work out and see how far we can push it. But to make further expansion economically viable, we’ll need to look at making Server Meshing more dynamic as soon as possible.

50 player count servers are pretty standard for ARMA unless you're willing to deal with the game constantly shitting itself so that's what I was expecting for their first iteration since Star Citizen is pretty much ARMA in space and frankly I don't think they need any more from a gameplay perspective in the short-term. Imagine a grid-based map and each coordinate on that grid having its own instance and you can manipulate the bounds of that instance. That's how server meshing is meant to work for a highly populated area. Theoretically you could have each district of a landing zone (city) be its own instance and while you're hopping on a train to another district, it moves you to a separate point on the grid containing a whole new 50 player instance. A landing zone like Area 18 could contain 150-200 people within city limits between 2-3 districts and a spaceport and that seems right around what I'd see in the top MMO's cities during prime time and those are much smaller than Lorville or Area 18.

For the outskirts of a planet that don't see much activity? That instance would be physically larger and if there's no current activity in that coordinate, it sounds like their intent is to be able to turn that portion of the mesh off until someone enters it and they can spin it up as needed. Factor in the connections-of-connections system that basically works as the algorithm that recommends you friends on Facebook to fill up your instances so you're always interacting with people you know or have reason to interact with such as organization members or enemies of your organization, people you've accepted contracts from, the server mesh concept makes a lotta sense and it's the only way you could have this sort of high fidelity systemic gameplay in a massively multiplayer online game like this.

 

[Myobi]

What a load of apologetic bullshit. Disgusting.

 

[RobotSquirrel]

50 player count servers are pretty standard for ARMA

city life has higher capacity and is also sharded meaning you can move your account to other servers in the network, your data is persistent regardless of what server you join. Some of the Arma servers I've seen running city life have had 100-200 pop with 1-4 servers.
If they can't pull that kind of pop then its pretty pathetic because Arma already does its just that it doesn't use AI for those servers. I feel as if Roberts made a big mistake with the size of the game and with the necessity of having AI.

 

[Perkel]

If they can't pull that kind of pop then its pretty pathetic because Arma already does its just that it doesn't use AI for those servers. I feel as if Roberts made a big mistake with the size of the game and with the necessity of having AI.

AI won't be run on player servers in near future. They have quanta in works that will offload all AI out of player servers creating backend service that communicates with player servers.

Static server meshing role is to replace one server per system with several/dozens which will improve both performance and player count and be test bed for their dynamic server meshing at the same time. Currently with 50 players their servers operate at around 20-30fps tick rate and in some really spread out population even lower than that which causes huge problems.

In theory static server meshing is enough for great player count sizes as well as huge world with good performance. The problem is scalability as for it to provide those numbers you would have to put shitload of servers in places where are no people thus wasting $$$. The point of dynamic server meshing is to allocate and divide load on the go creating way in which server A can split into A and B and leave B load to new server all automatically adjusting on the fly.

 

[Myobi]

The project left pre-production 12 years ago, and the 1st version of the technology that the entire project is riding on MIGHT be ready just next year, a friendly MIGHT BE, because just like everything else in this project when the time comes, it will get delayed…. WTF is the 1st version? Well, I can tell you what is not, is not enough to have 2 Idris, 2,000$ worth JPGs, fighting each other.

Plus, even ignoring the company incompetence, just on pure limitations on hardware and internet connections, I highly doubt CiG will ever get that shit working properly in our life-time.

 

[RobotSquirrel]

They have quanta in works that will offload all AI out of player servers creating backend service that communicates with player servers.

I mean logically it makes sense to offload that processing to dedicated servers in the cluster. There's one big problem however, the system falls apart if PU becomes popular. They won't have the power necessary to deal with that kind of demand I don't think there's a data centre on earth that could. The AI will be expensive to run especially without much abstraction. I have no doubt they will be forced to decouple the systems at some point and wait until demand lowers that's where I see this going.

They could fix the issue if they just made AI exclusive to PvPvE specific instances and have exclusive PvP on a separate instance making up the majority. There's no reason to roll it out on every instance the problem is that the core of the game seems to be revolving around that AI.

 

[Myobi]

There is also this:

Also, one thing that should be mentioned: even if the magic server meshing tech works as promised and you can fit like 100 people and their millions of polygons in the same room - that doesn't mean your GPU will actually be able to render those 100 high fidelity character models and all of the physics calculations that are going on. When Skyrim released in 2011, people were thinking "wow, a decade from now when GPUs become advanced enough and affordable enough, I could play Skyrim ultra modded at 4K 60 FPS!". A a decade later that's still not really possible, partly because GPUs aren't good enough and affordable enough (also prices just tripled and are unlikely to return to pre-pandemic values now that the manufacturers have realized that they can continue selling at these hiked up prices), and also because the Skyrim engine was really old and just doesn't scale well). By the time dynamic server meshing becomes a thing in 2023 or 2024, the Star Citizen game engine will also be rather old, so who knows if it will be possible. Right now, unless you're out in the middle of nowhere in deep space, Star Citizen chugs at 30 FPS even with a high end GPU and the game set to 1920 x 1080 and low graphics settings.

… and this is a fucking backer of the game by the way, jokes on him though, by the time the game is out, everyone will have access to quantum computing.

I'm also curious how jesus tech will work out with people with shitty internet connections...

 

[Perkel]

I mean logically it makes sense to offload that processing to dedicated servers in the cluster. There's one big problem however, the system falls apart if PU becomes popular. They won't have the power necessary to deal with that kind of demand I don't think there's a data centre on earth that could. The AI will be expensive to run especially without much abstraction. I have no doubt they will be forced to decouple the systems at some point and wait until demand lowers that's where I see this going.

Quanta will be divided into 3 "tiers"

1st tier is pure statistic. Meaning that most of quanta don't really exist they only exist as probability cloud and they are very simple in how they work. You can run 100s of thousands of them on normal non server rig.

2nd tier are npcs which come into existence when player interacts with 1st tier probability cloud. Meaning for example that player was moving through space and in that space there was high chance of pirate attack. Pirates spawn and start to fight with player. Player either kills them or not. In any case most of them despawn and go back to being probability cloud.

3rd tier. This is when player decides for example to follow one of pirates. So that NPC now is getting fully simulated. IT has its own goals, vices etc. Maybe he will go to bar on grim hex to drink or hunt someone else. It will stay that way until it gets killed or player loses interest and stops tracking it.

Basically scalable system that won't burn their servers and still provide ample simulation on any scale.

I'm also curious how jesus tech will work out with people with shitty internet connections...

It will be like any other game in that aspect. Meaning people with high ping simply either will deal with it or will not play.

 

[RobotSquirrel]

everyone will have access to quantum computing.

Not gonna happen. The average consumer doesn't need Q-Computers this is why everything has gone cloud. There's no reason grandma needs a Q-Computer.

 

[Perkel]

speaking of which. CIG is on track to beat previous record year in funding. At thing point they have same amount of money gathered as last record year.

 

[Myobi]

There's no reason grandma needs a Q-Computer.

There will be if she wants to play this shit if it ever releases.

It will be like any other game in that aspect. Meaning people with high ping simply either will deal with it or will not play.

Yeah, for some reason I highly doubt that.

GTA Online is probably the most recent nightmare of a mesh connection shitshow I’ve seen recently, I can’t wait to see how it will be with a server mesh connection in a game like this.

Correct me if I’m wrong, I mean, I’m not fucking expert here, but my data will have to go from my client, to the server I’m playing in, from that server to multiple other servers, them from all those servers to the players that I’m interacting with… constantly, back and forth, right? Or am I missing something here?

 

[Perkel]

Correct me if I’m wrong, I mean, I’m not fucking expert here, but my data will have to go from my client, to the server I’m playing in, from that server to multiple other servers, them from all those servers to the players that I’m interacting with… constantly, back and forth, right? Or am I missing something here?

You didn't miss anything. Current (instancing) and static server meshing doesn't really care that much about latency because at all times you are tracked by one server. The real problem comes from dynamic server meshing because in case of static server meshing we are talking about say station A = server A but in dynamic server meshing that station could be didived into 20 servers and someone shooting from server 3 with bullet trying to hit someone at server 8 and flies through server 4,5,6,7 all has to be synchronized.

As far as i know there isn't anyone outside of CIG that even tries to make attempt at dynamic server meshing. Usually MMO games use what current SC uses. Instancing. And this is why they talk about huge ass amount of players playing together in future unlike any other game or mmo.

This is basically how CIG want to resolve this problem. By introducing cache like layer which has huge ass memory pool to cache everything server nodes have in memory and work as communication layer between player and server node on which he plays. So the bullet position is kept on replication layer instead of server node and it is only sent to server node to display that bullet position. That layer doesn't do any calculations it is strictly for cashing latency sensitive data and sending non latency sensitive data for cold storage (like your money amount in bank etc).

Thanks to this you don't have situation in which 9 servers tries to trade information which each other about that dude who shot someone from server 3.

 

[RobotSquirrel]

I like the ... some actual honesty from a change from CIG, yes it can handle 16 clients lol. Beyond that is a bonus.

At least who ever is designing the network actually has some clue of what they're doing because having a separate shard for persistent databases is the best way to do it the issue is going to be the connections to that database and how it gets loaded in and out of the simulation part.

 

[Perkel]

I mean they have been working on it for years and they have pretty much unlimited budget. IT is only natural to expect that they come up with something. If CIG finishes it and it works i have no doubt there will be plenty of developers who will want to license their tech for other MMO games much like HERO engine is everywhere in mmo these days.

CIG is pretty much only studio that could achieve it as they don't have any milestones and they have essentially infinite money. In any normal development situation owners wouldn't put single dollar on new tech if you didn't have designed from a to z system on paper and no vision for ROI.

 

[zapotec]

That reminds me Einstein famous quote.
Two things are infinite: CIG funds and human stupidity.

 

[Grotesque]

If CIG finishes it and it works i have no doubt there will be plenty of developers who will want to license their tech for other MMO games

So, when they will start to license their engine, that would be the sign that they ran out of money?

they have essentially infinite money.

infinite now, 0 tomorrow.

 

[Perkel]

So, when they will start to license their engine, that would be the sign that they ran out of money?

They can't license engine because they are using heavily moified cryengine on lumyard license. But whole multiplayer netcode tools associated with it and their whole setup.

 

[Corbin Dallas Multipass]

There is also this:

Also, one thing that should be mentioned: even if the magic server meshing tech works as promised and you can fit like 100 people and their millions of polygons in the same room - that doesn't mean your GPU will actually be able to render those 100 high fidelity character models and all of the physics calculations that are going on. When Skyrim released in 2011, people were thinking "wow, a decade from now when GPUs become advanced enough and affordable enough, I could play Skyrim ultra modded at 4K 60 FPS!". A a decade later that's still not really possible, partly because GPUs aren't good enough and affordable enough (also prices just tripled and are unlikely to return to pre-pandemic values now that the manufacturers have realized that they can continue selling at these hiked up prices), and also because the Skyrim engine was really old and just doesn't scale well). By the time dynamic server meshing becomes a thing in 2023 or 2024, the Star Citizen game engine will also be rather old, so who knows if it will be possible. Right now, unless you're out in the middle of nowhere in deep space, Star Citizen chugs at 30 FPS even with a high end GPU and the game set to 1920 x 1080 and low graphics settings.

… and this is a fucking backer of the game by the way, jokes on him though, by the time the game is out, everyone will have access to quantum computing.

I'm also curious how jesus tech will work out with people with shitty internet connections...

I am willing to bet that by the time quantum computers exist, star citizen will be fully playable and fun. That day will arrive in approximately never, of course. Quantum computers aren't real, and never will be, and neither will Star Citizen.

But a magic computer with more logic gates than there are atoms in the universe playing a space game with more fidelity than every stereo ever sold probably would be a blast.

 

[Peachcurl]

Quantum computers already exist.

 

[Corbin Dallas Multipass]

Quantum computers already exist.

Not really. What currently exist are simulations of a quantum computer. It's comp-sci cold fusion, people can describe what it is, but nobody can demonstrate it.

I'm well aware intel and some other companies claim to have quantum computers you can rent time on. But read more about it and you'll find they're not actually quantum computers at all.

I'd be so happy to be wrong on this though, because super duper ultra fast computers would be wonderful for gaming. And, I guess, science shit and what not. But mostly games.