And what do you thing that means? lol
Game Discussions Star Citizen Discussion Thread v12
- Thread starter Yaffle
- Start date
-
- Tags
- star citizen
And what do you thing that means? lol
It means more gallons of netcode and a cleaning cycle of the coffee machine.
The database is always invariably reset, it happens every patch...The thing you need to note is whether long term persistence is enabled or not, and it's certainly still enabled in the latest 3.13.0a patch having played it this evening, nothing changes with your inventory, bought in game ships or credits as long as persistence is enabled...barring mishaps of course
Last edited:
Am I wrong then by thinking that server meshing is just a glorified term for a extreme case of server clustering? cause that’s exactly what I thought back then when CIG announced the damn thing... and that’s why I said some MMOs habe been doing this for decades... the differences, for me, lies on the scope of “space they each server is handling”
CIG's (fantasy) description is of an extremely dynamic system where servers multiply as needed, even getting to a point where one server is managing a single room. The mind boggles at the AWS bill.
More likely it will be like playing ED in Open.
You join "the" server and if there are slots free on the server where your friend is, you might get to join up with them. Or maybe not.
In ED its not usually a problem since space is big, and unless you are in a really busy system you'll probably be instanced with your friend anyway. And of course, if you wing, the game does its best to make sure you instance together. That's how the mass meetups in ED are often done, by people winging with others to pull them into the same instance even after the soft cap has been hit.
In SC, there will likely be a hard cap, not a soft cap per server and due to the small size of the game the chances of meeting up (assuming the game is really popular) is going to be slim. Of course, it all depends on how many servers they throw at the problem. If its one server for a whole system, then its basically like what we have now. If they were to add 1 server per station/planet, that would make things better, as long as you don't get more than 50 people at Port O at once (assuming the limit per server remains at 50).
Problem is, the more granular they are with servers, the higher their server costs are going to be.
I guess that would depend on how long you have to wait. Some seconds, ok. Some minutes? Probably not.
What they should do is have multiple servers per area, so you don't have to queue. New servers should be spun up on need. (which will cost of course).
I guess the current queue is because CIG are trying to keep their server costs down.
Yeah, i think its technically possible. I just don't think CIG are capable. And the dream of hundreds or thousands of players all together in the same area is beyond them. Even games that have been built from the ground up to handle lots of players in the same area (Planetside, Dual Universe) have their limits.
Now imagine SC with its "fidelity" with the servers having to track a ton of players, components, the argon in the air in each compartment of each ship, the state of cargo in the hold and ships. All that being communicated to every client in the same area. I think no way that's going to work, especially with CR at the helm.
The dream they have been sold, of massive fleet battles, with capital ships fully crewed and lots of smaller ships flying around, hundreds of people and ships (well, CIG said thousands at one point) is just that, a dream.
Pretty much, but CIG don't like to acknowledge something has already been done before, often many times, often better.
from this brilliant thread, the OP is hilarious
So how long do I have to wait? - Star Citizen Spectrum
I know, I know... it'll be done when it's done and don't rush them.... Buuut I remember buying the lti version of the aurora for 40 euros and sq42 when I was 16-17 YO with pocket money I earned...robertsspaceindustries.com
A lot of people agreeing and asking why CIG focus on trivial stuff while core stuff languishes.
Of course, there are defenders of the faith present.
Completely missing the point because its only trivial stuff we see actually being added most of the time with CIG offering up excuses of blockers with core tech stopping more important stuff being added.
Funny how its always the essential items that are blocked by this, never things like the ability to push trolleys.
EDIT: Ooooh, and i knew someone would come up with this old tripe
In short they are saying in 10 years, to date, CIG have only come up with trash. Also completely not willing to acknowledge that the gaming industry has put out some great games over the last 10 years. And "earthshattering"? Never been done before? Some of these guys live in their own reality bubbles.
Last edited:
Yep that server provisioning dynamic is, to some extend, handled by the underlying virtualization technology... in other words AWS.
But that feature (dynamic servers) are more or less given for free, the real money is in compute time, bandwidth and storage.
so yeah, that’s why I said in other post that due to complexity CIG might need to go with a subscription or another way of monetization after launch (if the game gets launched )...
Well I can’t say I disagree with you on the fidelity front... specially when people forget that that at some point point their pc will need to render the hundred of big dogs like Idris.. (and the swarm of small fighters) that will appear on their character’s viewport!!!
that graphic fidelity he promised is going to plummet quickly.
Last edited:
You're hitting another point here: even if possible by a structured design handled on a sane foundation and managed by coherent people, they constantly look like they're going through a naive approach for every dream they have, forgetting half of the problem by never taking into consideration the other side of the double edged swords they try to use as shovel to dig their hole deeper and deeper.
And let's take another example of thing they boasted about but ended exactly like that: bartenders.
And let's take another example of thing they boasted about but ended exactly like that: bartenders.
Viajero
Volunteer Moderator
IANAD, but wouldn´t that be possible by splitting Stanton in multiple areas each handled by a different server? I mean, ok, more servers means a higher cost for CIG and there would have to be some kind of new instance loading when moving from one area to another and the overall star system would not be the seamless unique instance it is today but I thought that was precisely the point of the first phase of the meshing thingie, the "static" server meshing no? Pretty sure CIG could hide those instance transitions fairly well with shiny quantum animations or what not.
Edit: Ninjaed by Golgot, need to read all before posting!
To be honest, the more I scratch my head with how SC is doing things the more I appreciate Elite´s stellar forge procedural approach (to limit the impact of assets in server or client memory) and the p2p local instancing system. Also from what I can see when, for example, the galaxy server allows us to travel in a supercruise instance and then we drop into a another player´s normal flight instance, that is in fact the same exact logic behind the goals in "server meshing". Elite meshing is even dynamic as it can occur anywhere where players are and it follows them anywhere they go afterwards. The only thing I dont think Elite´s meshing can do is to transition shot projectiles or render objects from one instance to another. There is though at least one thing that I am aware of that manges to communicate accross instances, the wing beacon signals.
Last edited:
Various MMOs cluster differently and depending on the topology of what they do (consensus protocols used, request paths, distribution of common state layers next to computation parts etc.), it can look vastly different. This is why "server meshing" is such a cool marketing term - it is so broad it is almost as meaningful as "distributed systems" or "cluster computing" (which may mean really whatever). So on one hand it is easy and has been done before, on the other, once you start adding up all the promises CIG have made, it is becoming a fools' errand. Honestly, if they could deliver on their promises, they would be working on a super secret project at Google or DARPA, not on the next iteration of Freelancer.
Viajero
Volunteer Moderator
Not sure DU is referring to the same scale of bandwidth requirements discussed for SC or ED for that matter though. Havent seen large scale seamless ship battles or encounters in it yet that would suggest DU is indeed "doing it now". We have discussed this before a bit but as far as I can tell what DU is doing is "simply" adding in a range based culling system. Highly doubt projectiles shot outside the culling range of a player will transition to that player in range asset list for example.
In some aspects that actually sounds like Elite galaxy server instancing does it, and as we know that has its limitations. Either way, I have not yet seen large player numbers transitioning through the described DU culling system though, would be very interesting to see how that culling performs when hundreds, never mind thousands, of actual players or ships (all with hugely diverse latencies) decide to congregate within a few hundred square meters. What is the actual practical cap of players in the same culling range that can all see and interact with each other in real time? What happens when that cap is exceeded? It will also be interesting to see how culling conflicts will be managed, i.e. when a player in culling range of another shoots at a 3rd player that the other player can not see because it is out of range for him, etc. Or is the system rather based on arbitrary dynamic space partitions? I mean I am happy to be shown evidence if there is any videos etc that show how all this works. But so far it sounds a lot like the most wild SC claims.
Remember that the heart of the DU system, if it ends up working, is that magical, dynamic and range based culling tech.
Last edited:
IF they could deliver on their promises - they would, overnight, become the richest company on the planet and utterly revolutionize the networking, banking, insurance, transport and security industries.
But Genuine Robert's is incapable.
Why do they need to be in the same instance in the first place?
Sure it has its lot of limitations and problems, but it's clever in many ways, starting with technical and financial costs side. It's almost as if there was these petty inferior so 'old world' terms like "forethinking" and "scalability" at work in project management...