Server Meshing for Space Enginner 2
I saw the Server Meshing in Star Citizen, and it's ****** amazing if something like that were implemented in Space Engineers 2.
Quick Summary of Server Meshing: Server Meshing is the technology that connects multiple servers together, allowing for the creation of a seamless, scalable world without lag. This enables more players to participate and improves overall performance.
How Space Engineers 2 Can Benefit from Server MeshingServer Meshing could revolutionize Space Engineers 2, enabling a larger and more dynamic universe. With distributed servers, the game could support a greater number of players simultaneously without sacrificing performance. This would reduce lag, improve data persistence, and allow for smoother interactions across vast regions of space. Large-scale battles, station construction, and exploration would be more stable and scalable. Factions and alliances could operate seamlessly across multiple servers without the limitations of traditional server structures. Overall, this technology would make multiplayer in SE2 more immersive, efficient, and expansive.
Technical ChallengesServer Meshing is still a relatively new and underexplored technology, particularly in games like Space Engineers 2. One of the biggest challenges is the lack of large-scale examples; Star Citizen is one of the few games pushing the boundaries of its implementation. Additionally, it's worth mentioning that Rockstar have a patent related to server meshing, which (I hope) provides some credibility to the technology and this topic.
Concept of Server Meshing with Limitations on EVE Online
Server Meshing in Ashes Of Creation (Game still in alpha but is worthy mentioning)
The complexity of integrating multiple servers in a way that feels seamless is extremely high. Managing real-time synchronization, player transitions, physics calculations, and data persistence across different servers requires sophisticated networking and data management solutions.
Given these challenges, implementing Server Meshing in SE2 would require substantial research, development, and optimization. However, its implementation could be a game-changing feature for the series, potentially offering an experience unlike any other in the genre.
Implementing Modding and Custom ServersIntegrating modding with Server Meshing is a complex task, as it involves deep integration with networking and persistence systems. However, if done properly, it could vastly expand the game, adding new mechanics and possibilities.
A potential approach could be mirror how VRChat handles custom content or Garry’s Mod with infinite custom mods. In this scenario, mods with custom code be loaded dynamically (Maybe it's worth discussing this more). This would require Keen Software House to validate mods for stability and compatibility for official servers if they are planned to be added.
With Server Meshing, players could buy specific game areas for mods, automatically downloading them when new players enter and disabling them when they leave. This would allow for smooth integration while also providing Keen Software House with a potential avenue for monetization through rent a server (or multiple servers, or some new way that I haven't thought of, with monetization in server meshing).
(Note: This would not limit the ability for players to create fully private custom servers. but additionally, in private servers, players who are alone in a certain region will be able to play the game almost entirely locally and sync when other player are close).
The Possibilities Are Game-ChangingIf successfully implemented, Server Meshing could elevate Space Engineers 2 to one of the greatest multiplayer experiences of all time—potentially even surpassing Star Citizen’s vision.
Imagine players building entire galactic civilizations, complete with economies, politics, wars, and trade, all driven by the community. Massive space battles, fully player-driven economies, and large-scale engineering projects could become a reality.
This technology could push SE2 beyond a sandbox experience, turning it into a persistent, evolving universe where players shape history together. The possibilities are limitless.
I like this feedback
It sounds like you want an MMO - and given the relative failure of Starbase and Star Citizen (at the moment) - I can see the potential here. The problem besides those mentioned would be service federation - ensuring that all servers remain updated at all times and ensuring data privacy is respected. Here are a few thoughts of mine (assuming third party services are involved to run these servers):
# Syncing Behaviour: Player's Perspective
Allowing syncing across private and modded instances could cause larger scale issues as well if ships are constructed with those in mind. Present SE1 behaviour would have modded blocks disappear, leaving a gap. To promote a combined non-modded and modded server federation would require ships to be built with vanilla redundancy for the sections of the 'universe' which don't have advanced reactors for instance; or vanilla guns. It would make ship design difficult and cumbersome. Thus, I'd recommend ensuring uniformity across the cluster instead - all vanilla or all similarly modded.
# Syncing Behaviour: Data Ownership and Privacy - GDPR (EU) / APP (AU) / Other
Something which doesn't seem to be considered is the fact that if this syncing is also permitted across third-party services (players or other hosting providers besides Keen's), one needs to be careful about data ownership and international sovereignty. The players would need to be made aware of everywhere their information is being processed and stored (as requirements of GDPR and APP) and on third party services there is little enforcement Keen can do to ensure that providers act in good faith.
# Syncing Behaviour: Security
Lastly, this kind of service federation would require a careful pass regarding security. A network such as this could enable effective DDOS mitigation (via stateless firewalls and other cloudfront solutions), though the more likely issue here is the potential for its misuse. All elements of the federated network would need assurance that all severs therein are acting by the same rules; no third-party server management solutions - this is to prevent malicious insider risk. Secondarily, one needs to consider the wider ramifications of this network: if someone would discover a way to tunnel malware through this kind of federated network, the consequences would be disastrous.
Finally, this kind of mesh is somewhat susceptible in and of itself as there is usually no root of trust. This means the network cannot effectively take authoritative action to take malicious members out of the network. These risks would need to be mitigated carefully.
Understand me: I would love such a solution - but the practicality needs to be considered first.
It sounds like you want an MMO - and given the relative failure of Starbase and Star Citizen (at the moment) - I can see the potential here. The problem besides those mentioned would be service federation - ensuring that all servers remain updated at all times and ensuring data privacy is respected. Here are a few thoughts of mine (assuming third party services are involved to run these servers):
# Syncing Behaviour: Player's Perspective
Allowing syncing across private and modded instances could cause larger scale issues as well if ships are constructed with those in mind. Present SE1 behaviour would have modded blocks disappear, leaving a gap. To promote a combined non-modded and modded server federation would require ships to be built with vanilla redundancy for the sections of the 'universe' which don't have advanced reactors for instance; or vanilla guns. It would make ship design difficult and cumbersome. Thus, I'd recommend ensuring uniformity across the cluster instead - all vanilla or all similarly modded.
# Syncing Behaviour: Data Ownership and Privacy - GDPR (EU) / APP (AU) / Other
Something which doesn't seem to be considered is the fact that if this syncing is also permitted across third-party services (players or other hosting providers besides Keen's), one needs to be careful about data ownership and international sovereignty. The players would need to be made aware of everywhere their information is being processed and stored (as requirements of GDPR and APP) and on third party services there is little enforcement Keen can do to ensure that providers act in good faith.
# Syncing Behaviour: Security
Lastly, this kind of service federation would require a careful pass regarding security. A network such as this could enable effective DDOS mitigation (via stateless firewalls and other cloudfront solutions), though the more likely issue here is the potential for its misuse. All elements of the federated network would need assurance that all severs therein are acting by the same rules; no third-party server management solutions - this is to prevent malicious insider risk. Secondarily, one needs to consider the wider ramifications of this network: if someone would discover a way to tunnel malware through this kind of federated network, the consequences would be disastrous.
Finally, this kind of mesh is somewhat susceptible in and of itself as there is usually no root of trust. This means the network cannot effectively take authoritative action to take malicious members out of the network. These risks would need to be mitigated carefully.
Understand me: I would love such a solution - but the practicality needs to be considered first.
Hm, yeah, Server Meshing is great and all, for a MMO. What you totally miss here is, that a MMO is something complete other. With such a big player base you need much more features and mechanics in the game so that it didn't end in a total chaos. It is not that easy to make out of a normal solo, multiplayer concept one in a MMO scale. And a MMO didn't need 1000+ players in the game world to deal with typical MMO issues, that starts already with 100+ or even lower numbers of players.
Also a general rule: the more people you get together, the more problematic people are under them. People who love more to destroy others the fun at the game then following game rules and having fun together. People who are not able to follow rules. People who like to cheat. People who like to cause trouble in general. Simply egoistic people and we have a lot selfish people today.
And this problems rise at least with factor 10 if it is a PvP game. PvP creates with more players always a toxic surrounding. PvP on a large scale always attracts a certain type of PvP ‘fans’ in larger numbers, who have their very own idea of PvP, as described above. PvP sounds in theory nice, in practice it struggles on the human factor. Look how toxic bigger online ego shooters are today. And even Star Citizen has already problems with such players.
With 10 players it can work, with 20 players it already gets a lot more difficult, with 50 players you will have for sure trouble with them. And no one want to deal with such players, every other player simply want to have fun together with others.
So like said, Server Meshing is a awesome technology (I'm myself a StarCitizen fan since day 1), especially the dynamic one that is still missing, but I doubt Keen has the resources to deal with such problems and the plan to scale the multiplayer so much up, it would also need changes in the gameplay. I guess they have other plans with SE2.
Dual Universe totally failed and showed the difficulty of a building game in MMO scale very good and some problems had even SE1, for example that you have a big cruiser, a big space station with only 1-5 player on board. That feels totally bad and in MMO scale it feels like a dead world, because from our reality we would expect that on such places are a minimum amount of people walking around. In Star Citizen is for example the plan that players are only 10% of all characters in the universe, 90% are NPCs to give this universe a feeling of a living and breathing world. Many NPCs are still missing (they need dynamic Server Meshing for that) and so even StarCitizen feels still too lifeless. You would have the same problem with Space Engineers 2, btw. it is already a huge problem Keen should working on because it is already a SE1 problem.
-------
But to think smaller, maybe you can give the players something like server "meshing", server software that makes it easier to bind servers together, where one server is a star system and you are able to jump with your ship to a other system on a other server. In that way some server owners can work together to create a little mesh of servers with same mods and rules with still a little manageable player count. They could start with one star system and add another later.
I think that would be a better approach. Better more smaller communities than bigger ones which risk to fail more easily.
Hm, yeah, Server Meshing is great and all, for a MMO. What you totally miss here is, that a MMO is something complete other. With such a big player base you need much more features and mechanics in the game so that it didn't end in a total chaos. It is not that easy to make out of a normal solo, multiplayer concept one in a MMO scale. And a MMO didn't need 1000+ players in the game world to deal with typical MMO issues, that starts already with 100+ or even lower numbers of players.
Also a general rule: the more people you get together, the more problematic people are under them. People who love more to destroy others the fun at the game then following game rules and having fun together. People who are not able to follow rules. People who like to cheat. People who like to cause trouble in general. Simply egoistic people and we have a lot selfish people today.
And this problems rise at least with factor 10 if it is a PvP game. PvP creates with more players always a toxic surrounding. PvP on a large scale always attracts a certain type of PvP ‘fans’ in larger numbers, who have their very own idea of PvP, as described above. PvP sounds in theory nice, in practice it struggles on the human factor. Look how toxic bigger online ego shooters are today. And even Star Citizen has already problems with such players.
With 10 players it can work, with 20 players it already gets a lot more difficult, with 50 players you will have for sure trouble with them. And no one want to deal with such players, every other player simply want to have fun together with others.
So like said, Server Meshing is a awesome technology (I'm myself a StarCitizen fan since day 1), especially the dynamic one that is still missing, but I doubt Keen has the resources to deal with such problems and the plan to scale the multiplayer so much up, it would also need changes in the gameplay. I guess they have other plans with SE2.
Dual Universe totally failed and showed the difficulty of a building game in MMO scale very good and some problems had even SE1, for example that you have a big cruiser, a big space station with only 1-5 player on board. That feels totally bad and in MMO scale it feels like a dead world, because from our reality we would expect that on such places are a minimum amount of people walking around. In Star Citizen is for example the plan that players are only 10% of all characters in the universe, 90% are NPCs to give this universe a feeling of a living and breathing world. Many NPCs are still missing (they need dynamic Server Meshing for that) and so even StarCitizen feels still too lifeless. You would have the same problem with Space Engineers 2, btw. it is already a huge problem Keen should working on because it is already a SE1 problem.
-------
But to think smaller, maybe you can give the players something like server "meshing", server software that makes it easier to bind servers together, where one server is a star system and you are able to jump with your ship to a other system on a other server. In that way some server owners can work together to create a little mesh of servers with same mods and rules with still a little manageable player count. They could start with one star system and add another later.
I think that would be a better approach. Better more smaller communities than bigger ones which risk to fail more easily.
I would be happy to host a dedicated server on Linux. Then we can think of mesh 😅
I would be happy to host a dedicated server on Linux. Then we can think of mesh 😅
A Torch plugin called Nexus can make a set of SE1 servers all host parts of a shared world, specifically each server hosting a defined (possibly non-contiguous) volume of space.
I want to have something similar integrated into SE2, where multiple servers under the control of a single administrative entity can all host a shared world.
I'm unsure how Server Meshing differs from what Nexus does (synchronized shared state between SE1 servers, so players and grids can move between servers).
A Torch plugin called Nexus can make a set of SE1 servers all host parts of a shared world, specifically each server hosting a defined (possibly non-contiguous) volume of space.
I want to have something similar integrated into SE2, where multiple servers under the control of a single administrative entity can all host a shared world.
I'm unsure how Server Meshing differs from what Nexus does (synchronized shared state between SE1 servers, so players and grids can move between servers).
I would like to be able to host servers with thousands of players.
I would like to be able to host servers with thousands of players.
I'm surprised at the amount of dissent in this thread considering the sheer number of third party servers working to share a unified world across multiple shards. The nexus seamless plugin allows players to change from server to server so a single instance only has to worry about a planet or a region of space instead of the entire map at once. This is absolutely something i'd love to see integrated in SE2. Spread the workload across multiple servers for 100 player communities.
I'm surprised at the amount of dissent in this thread considering the sheer number of third party servers working to share a unified world across multiple shards. The nexus seamless plugin allows players to change from server to server so a single instance only has to worry about a planet or a region of space instead of the entire map at once. This is absolutely something i'd love to see integrated in SE2. Spread the workload across multiple servers for 100 player communities.
Replies have been locked on this page!