So Intrepids version is called dynamic gritting, while the tech CIG is developing is server meshing (first static) and later dynamic server meshing using an entity graph system. The point is that CIG is currently testing static servermeshing internally and some public tests, where as Itrepid wants to test it qith alpha testers now. I am curious about what works how good and when. Not that it would be compareable. Intrepid has relatively "low" polygons with limited actions and CIG needs a totally different system that aids action combat and huge moving spaceships with multiple fully physicalised planets. Lets see how Intrepid does it.
Also server meshing is nothing new planetside 2 had it in 2011, its hex based system on 64 square km map with 1200 slot per continent, and it need sync bullets and stuff and fps is much more dynamic game especialy in 100v100 situation than an mmo. I remmber like 10 years ago they said star citized relase in one year i was studying then, now i am 34 and star citizen didint relased yet lol.
Network engineer, MMO enjoyer and occasional SC player here. As one commenter already pointed out, this is not fair comparsion. SC deserves tons of criticisms, do not get me wrong please. But this is not accurate. You are basically comparing motorcycle engine and truck engine, saying "this one is better because" but that in itself can not make sense because while both are pushing piston by expanding combustion gases, they are working on completely different scale and serving completely different purpose. Again, do not get me wrong, Interpid is making beautiful achievemtn which may push MMO genre forward a lot. I love what I see here. But it has absolutely no place in comparsion with SC because their tech is simply being developed for different purpose. if you would like more technical details and tech-blabla, I am happy to share buit this is already too long post which nobody will read anyway.
I am super excited for AOC and not a Star Citizen fan boy at all so I hope this taken objectively, but to say what AoC is planning is the same as what Star Citizen is working on, is not fair at all. While I don't have a degree specifically in this area, I do in servers and databases, and from my understanding I would say the Dynamic Server Meshing of Star citizen is still a tier above what AoC has created and is planning for. BUT that is not a problem, because you don't need the fidelity needed like they are trying for in Star Citizen for an MMO, you don't need bullet tracking and high speed space sim elements tracking across multiple servers interacting with each other Or entire ships being servers themselves interacting with other people in servers around them. I do still have some questions on how AoC is going to handle abilities and things interacting across these mico servers without there being major bugs because I can't see them being able to get there minimum server size down to a point where that isn't needed. Async programing is HARD, extremely hard, and it will be exciting to see how they handle these issues.
To be fair, we don't know what AOC has planned for the future. What they showed are just things that will be present at the start of A2. Like what one of the developers said during the live stream, they will continue to improve the technology over the next few years.
Its more like star citizens Static Server Meshing that they have implemented right now. Static Server Meshing is all a traditional mmo would need. Dynamic Server Meshing would be overkill I think.
A few things on what Narc said about both server meshing solutions is a bit incorrect. Microservices is not how the servers communicate together, they are the services like chat, mail and others that need to be distributed for management. It basically takes that job from the server so that the server doesn't need to worry about it and players can use these services no matter how broken the server they are on is. The servers are linked together not through microservers but through the grid communication. The servers only speak to what is adjacent and orthogonal to them. The replication layer would indeed be an issue for SC if it was like what the graph Intrepid showed, but it's more close to what Intrepid are doing. The rep layer in SC is just a bunch of services that communicate the client data and the server data with the database. We have yet to really see a bottleneck since the replication layer scales horizontally just like the server mesh. Server meshing isn't live in SC or AoC so it will be anyone's guess on how well either work when you have tons of players doing whatever they want. Exciting times are ahead for games. Something like Greyzone Warfare would be pretty cool with something like Server Meshing in a vast large world.
Correction, you are right that Narc is wrong, but you're slightly wrong in that the Microservices were primarily designed so that there could even BE things like communication across the server, since the server mesh technology inherently prevents servers sharing data past the immediate ones surrounding it. It's just that these microservices come with a few perks with it working that way(perks like you stated). Just like the centralized replication layer was done away with primarily due to the bottleneck, yet it came with perks like realm stability
While true that Server meshing is not yet "live" in SC, the replication layer is and has show to work. In the tech tests with static meshing things also held up pretty well.
SC just did a test with 800 players, that is the most any fps online game had ever done. In the last video they were saying they probably would have 800 players, but they needed to stress test it. They admitted to bugs, but it is overall working. Already our basic servers went from 50 player servers up to 100 player servers and that has become the new standard. Also, we can actually play SC and 4.0 is supposed to launch this fall/winter, which will bring Pyro, and server meshing will be live. It might get delayed, but by years end we are supposed to see SC server meshing live in action.
i will sincerely admit that ASHES OF Creation is my most waited thing in this year. i am from indonesia.. from SEA server.. ive played lots of western subs-based mmorpg but i got super terrible ping . exceeds above 240ms.. that killed all my hope of enjoying a mmorpg.. and i hope the subs fee is no way above 15bugs.
there is one name that echoes through the ages that i can never forget. one name that never had more than a spot on a leaderboard. one name all my friends wanted to be as cool as... and that one name was... Zezima.
I'm a fan of both projects but It's pretty crazy to say that AoC team has made a better Server Meshing solution than Star Citizen when you compare the real tests of both. - AoC Alpha 1 test with replication layer: 72km² static map + 1000 players + very limited dynamic objects and npc's = Completely bottleneck and need to change the entire SM system to work. - Star Citizen Alpha 3.22 tests with Replication Layer and Persistence Entity streaming (Entity Graph) : 750 millions km across of simulated space (full volume is way way bigger) + 17 full planets and moons with cities rotating in real time + thousands of npc's + 800 players with a minimum of 800 ships with physic grids + Millions of dynamic entities = Test worked better than expected. Maybe is way more complicated comparing both SM solution because the games are vastly different. Star Citizen combine PES(Entity Graph) with Micro-services + a decoupled replication layer + Object container streaming both client and server side etc... A lot of tech that sound similar but are way more complicated for a game like this :)
The tool that allows the devs to see all of the NPCs and players on the map at once "Shol's Eye" is named after the God of Truth in Ashes of Creation lore Cheeky detail
Comparing this so intensely (competitively) to star citizen's server meshing tech is not really helping your argument. It looks a little disingenuous and isn't something you can compare cleanly; the data and precision needed for these two completely different games require different technology and specs in the meshing technology, so it's not like ashes 'figured it out where star citizen couldn't. I wanted to know more about Ashes' tech, but this just felt like propaganda lol
All the hype about tech... What about actual gameplay? Will crafting be interesting or just a grind? Will resource harvesting have some depth or will it be clicking on a tier 1,2,3,4 node? etc etc...
What SC is doing is the Transformers of sever meshing, what Intrepid is doing is the Go-bots of server meshing. In the end, it's the one that works that is the best. If they both work we all win. SC will have the better version for SC and intrepid will have their version for AoC. I don't like either game, I just want to see the sever meshing make it in one form or another.
Long ago, Steven mentioned that scientific nodes will be able to teleport between other scientific nodes. How cool would it be (talking about no loading screens) if you open a portal and you can see the other side in real time. Just walk through, and you're able to look back and see your friends waving from the other side of the portal.
I think this probably would’ve been a better response to the stream if every aspect of Intrepid’s solution to server tech limitations was covered in a separate video. It’s just too hard to condense everything they talked about into a bite sized summary like this. But good attempt, this is largely accurate with a decently strung together narrative arc.
I hope this works well but I don't think dynamically partitioning the world will be enough once players really try to gather in the same location if the servers are fully authoritative. I did prototype a similar system once and what I noticed is that at a certain occupancy level the split granularity made players change servers fast enough that it will actually add overhead. Of course this will only matter if ashes is popular enough and at that point I'd just be happy to have a new mmo to play.
Can litterally not compare the 2 games... Ashes is playing 2D and Star Citizen is playing 3D chess... They have to have this tech working on 360 degrees in ever direction possible while Ashes is a single world that is mostly 1 plane. That being said tho Both are impressive! And im really hyped for both games.
This might not have had the draw, or wow factor of some of their previous dev diary updates, but it was by all means extremely important. Anyone who's ever taken part in large pvp battles in an MMO before, should know how significant a problem it poses. They seem like they're on the verge of taking a step that all modern MMO's will follow in one way or another, at least if they're serious about providing players with a more reliable in game experience.
With data moving between servers and microservices I worry about dupping bugs. It takes time to confirm transactions and I wonder how Intrepid does that smoothly without loss of data and without duplication of items. I'm thinking of all the test cases I'm going to try in A2 (trading across, dropping items, picking up items, splitting stacks and mixing in logging off etc)
It is quite funny for old-timer IT people because this is an old architecture model that was used on fault tolerant hardware in the 1980/90s like Tandem and System/88, except in those days we couldn't connect multiple servers together faster than LAN speeds. And of course it was horrendously expensive.
Indeed, it is interesting to see people get all excited about tech which has been around for decades. At the same time though, the requirements of this tech in the context of these games is vastly more unique and demanding than it was back then.
@@blazemonger1 Yeah, back in the 90s it was used for processing real time transactions like for point-of-sale, or stock markets. Which was pretty demanding, because you had to deal with authorization transmission times, or real-time pricing, given the WAN speeds. Remember no high speed Internet at the time. Think about racing cars back in the Ford vs Ferrari days. Just because the cars were going 30+mph slower, and the technology was older, doesn't mean that they were not pushing what they had to the limit, in terms of design or process. It is more that we haven't had to resort to complex software architecture, because processing times for most requirements are easily met more reliably these days with hardware upgrades. In fact the main difficulty (from what I can tell), is making server meshing work with Unreal Engine. If Unreal Engine was written with this architecture in mind then it wouldn't be as much of an issue
Overall great review Narc, I really felt like you took that heavy preview and thinned it out well for the rest of the community who couldn't be bothered to sit through it. One small issue about the microservices statement you made, but someone else here already covered that. Well 2 issue really, you also talked like Intrepid doesn't do layering because they won't lower their standards like that, but in reality they literally can't. The game is not designed in a way that accommodate layering or sharding WHATSOEVER. Due to the dynamic nature of the world like citizenship status, guild status, node status, quest arc status, ect ect. They have no other choice but to do this. Steven even mentioned that this was all a part of development, they didn't choose to do this after the fact, they knew from the outset that this was getting done. The game just flat out wouldn't work if they didn't do it, the worlds too big and too dependent on lots of large groups. Either they get this right or the game fails completely. I'd say that the same could be said about Star Citizen but lets not fool ourselves, the games going to fail completely either way. P2W dev welfare economy sim trash.
I didn't realise others had the same feelings as me about the last video until I saw this video and the comments under it.. I saw asmongold watch the last one and I felt like there was a lot of specific details missed out which sets this apart from the others, it was underwhelming. You've absolutely smashed this one out the park & is a testament to your dedication to us & the game. Really appreciate the fact you've taken comments of the last one on board...
The "gridding" tech AoC is showing is not new either.. It's been around for a long time and was successfully applied in the otherwise failed MMO Dual Universe as an example. And you do not need a tick rate of 60 or even 30. Tick rate is only part of the solution and can be much lower than that, in fact a tick rate of 15 would probably be fine for a MMO like AoC. And indeed micro services ar not new either as they and you mentioned, it is a core part of the EVE server tech for instance and has been from the start. In the end, none of what Intrepid has shown is new except for maybe how they had to rework the way UE5 handles large players the way they need it to, as that engine by default can't. What is unique is how they apply these separate server technologies and on paper make it look like a great solution for the type of game AoC is. I hope they will be able to show their tech works, absolutely. But for now it is just talk and unproven. We'll find out during A2 if it all holds up, provided they get enough people to play.
oh they will def have enough people to play. But I'm sure they will be shitting up a storm if the servers keep crashing. Let's hope that doesn't happen.
As a long-time software developer I believe that success hasn't been achieved despite the core team only consisting of eight people, but because of it. Sometimes, less is more. A small team of highly talented and experienced people who "can smell" each other is often vastly more effective than a big one which usually gets mashed together by managers who only know how to throw money at problems and who think that nine women can get you a baby in a single month.
The multi-treading isn't done, not because it cannot be done, rather the game they are making has to run on the minimum spec machines. Now that AoC did this, I could only imagine that in order to play this game, I have to spend hundreds of dollars to upgrade my machine -- not gonna happen. If the minimum spec to play this game is too damn high, I'll play something else. If the can find a way to apply this multi-threading at low spec, that would be very impressive. So I'm still waiting and see what other IRL magic Intrepid can pull off.
My main concern now are.. bugs. When you are doing everythign from scratch, you can expect lots and lots of unexpected bugs. Just look at new world, for example. Had some simple dumb bugs, that ruined economy of the game.
Server meshing was never meant to be exclusive to Star Citizen. When Chris Robert announce SC he thought the tech would be rdy.. and it's not been rdy since yearsss.. Every big MMO, like pax dei, gonna be server meshed ;)
Pepperidge Farms remembers when Rift was going to change the future of MMOs with a teleporting sniper and EQNext was, too. After FFXIV, I went back to where I started, consoles. MMOs just weren't the same anymore, the community got far worse and console RPGs are a lot less 'update' or mod-reliant to raid. If this was ready when I quit PC MMOs 3 years ago, I would have been a Day 1 player. Hope this does well and the tech somehow eventually works with the limitation of console controllers.
Ashes of Creation server mesh that was never attempted before, meanwhile Star Citezen looking at AoC "i was first kid"... Then comes GTA6 Online with fully working server mesh, released game and tech laughing at both.
Not concerned with a race. Don't care that one is harder than the other. Just excited that it seems like a reality for both and that's good for everyone.
I mean, I wouldn't even call this copium here. They've already done it, and all it shows is how capable their team is, with solid proof. Or ya know, I'm just coping with copium addiction 🤷♂
Ashes is the best what could happened to the mmo genre. I dont want this to be a WoW Killer, but this will spark a good competition from wich we player will profit a lot. At least i hope so.
I "retired" from gaming after finishing classic and TBC remakes but my bros got me to buy an alpha key and re-enter the dungeon. Gonna stock up on adult diapers and take 2 weeks off work for alpha release and binge this shit till my eyes bleed 🤣, see you in there nerds!!
Question is there gonna be voice chat any idea? i feel like it would be so fun setting up a raid and stuff and rp with it and any thing aboutn housing so far ? Love the good explaining in the video keep it up
Mortal online 2 does 1 server and for the longest time, 1 character. When you shit where you eat, you tarnish your reputation, and people will be more hostile towards you over time. If we can have that on an even larger scale! Holy, that'd cool.
Dont worry people , they ll run alpha for many years ,so there is enough time to fix any server problem. Sleep-eat good so we are healthy and sometime in the future we ll play aoc
Does this game have a PvE component or not? Do you level up via traditional grinding, is there a cap on leveling? Is it all PvP like Black Desert or is it kind of hybrid like Archeage (which has dungeons and such)?
having 500 people in one area should be possible if the server doesnt load the stash of each person every second like D4 for real though, there should be a way to optimize the amount of data , you dont need new magic server tech if you combine optimization with new server tech though, it should be possible to have 1000 in one area, dont see the need for more. When questing I hope you dont see 40000 people in one area
Great video (obviously). Interesting topic. Technically, very impressive. However, I fear it will have the opposite of the desired effect. People often think they want things, that they, if fact, do not. A smaller server size is vastly superior, across the board, for the vast majority of the players, the vast majority of the time. A bigger pond, means fewer, and smaller, 'big fish' at the end of the day. You don't need an MMO to feel like a forgotten nobody, who knows nobody, and whom nobody knows, that is what IRL is for. It also is effected by how human brains work, but that gets kinda technical. So, I gotta say the tech is great, but the result will be an inferior experience for +/- 99% of the players.
If we think massive open worlds are MMOs biggest problem, we're in even deeper than I thought. Can we please talk about _real_ problems plaguing the genre?
Theres definitely many problems, but seamless servers has to be at the top of what is expected when building a massive open world game. I think its one of the main reasons why WoW felt like a real world.
@@missionpupa WoW's mostly open world is a boon, but when an open world comes at the cost of discussion about things like gameplay we're screwed. EQ's original design immersed me immediately even though it has loading zones. Mind you I grew up with WoW first for over a decade. Seamless worlds are def not at the top, but near. Not required, but really important.
@@poisonated7467 Its just one of the many conversations we can have, in fact AoC barely discusses these things in their videos. Im saying that technically, this definitely has to be the most challenging part of creating the game, the game loops, skill set, theyre different problems but theyre not technical problems.
@@missionpupa Technically, everything is a has a technical side to it, but I understand what you're saying and agree that it is a technically challenging problem to solve. Just that it's not MMO's biggest problem right now as the video suggested. One of the largest problems is trying to solve community and player choice problems with tech. It seems unsolvable.
@@poisonated7467 The MMO can have many problems its not just 1 thing, and im definitely not saying once you solve the seamless servers it means you solved everything, im just highlighting the importance of it. Not really sure what you mean by player choice problems.
I don't know why we're trying to compare Ashes of Creation and Star Citizen's server meshing. It's like comparing apples to oranges, they're both server meshing but they're inherently different in their own right. Star Citizen has a lot to be criticized about but the random digs throughout the video make it tough to watch. I come to watch Narc videos for AOC content, if I wanted to listen to SC criticism I would go to any of the other more informed SC content creators. We should be hoping both succeed in making that kind of tech work instead of "hurr durr my meshing is better than yours" when they're not even a real comparison. Either of them succeeding means a brighter future for MMO games
You can actually play star citizen. Ive actually seen normal people testing Server Meshing. Aoc hasn't done shit till the actual public gets to test it... Believe it when i see it
So whats the solve for like a busy starting area? Even server based MMO's that are popping have you sitting around for stuff because of over crowding. Unless i'm understanding it wrong? Is the map and content big enough to handle every last player sitting on the map?
Good question. We'll have to see how AoC and Star Citizen handle massive amounts of players in single locations. One solutions is to cap the player numbers at a number that supports such massive congregations, another is layering. Another problem besides the gameplay/map supporting vast amounts of players is the optimization of the game itself so that your PC can handle all the things it has to render on screen.
@@herobrine024 Yeah I guess that's my worry. If they layer it kind of defeats the goal. I'm not concerned with processing as much as I'm concerned with overcrowding causing the game to be boring. Lol. I'm thinking about lines of people in wow classic for a quest Mob, or never being able to get a crafting node. I guess the tests will go a long way toward figuring that out I hope
@@johnnynometer in Star Citizen when they raised player counts from 50 to 100 and had a server wipe, on day one, there were literal lines at the ship retrieval terminals. Because they are now adding physicalized cargo and persistent personal hangars, meaning players will spend much more time inside hangars, theyve relocated these terminals inside of the hangars. This indirectly solves this issue and prepares the gameplay for many more players playing on the same server, because these hangars are layered. But thats only possible because theyre rooms that only specific people have access to (think GTA Online Apartments with seamless doors to the outside). AoC is primarily in open environments so this wouldn:t be applicable to the game.
The entire framework they're developing is the solution to that issue. As more players would arrive in a single area, their dynamic server meshing would dynamically split and assign resources as needed, without the need for layering. It's a wait and see for how they'll implement starting areas but the zones themselves are vast. Quick Edit: I could see them having a few similar starting points per race to divvy players out and help with those issues. Could have players congregate a bit later on. Maybe you're assigned one of those locations and could override it if you wanted to start with your friend, etc. All speculation on that of course but only someone like Asmongold would be ruining that for most people I would think lol.
No, its the other way around. A tick rate represents the amount of times the engine performs a simulation step per second. The higher, the more accurate.
I would hope it's a case of getting this server mesh tech to work first; then you'd be working on the end of testing how to keep people from cheating. I feel like this type of mesh network will make some weird duping issue of items etc be a lot easier to pull off; but of course no way to know then to go into Alpha 2 and try and break the game etc and report it.
Tech like this isn't usually visible in gameplay because the goal is to not show you it is there. If you notice it there must be lag or something wrong in that grid server location.
we're live tonight with SaltEmike discussing all things Server Meshing, Ashes of Creation and Star Citizen so come join us: www.twitch.tv/Narciverse
Holy fucking shit, I cant wait, what time?
@@AnTicXz look at teitch it says, i think 2 hours from now
That's gonna be exciting! Prolly gonna miss it due to a weird post 4th of July shift but I'm looking forward to the vod
So Intrepids version is called dynamic gritting, while the tech CIG is developing is server meshing (first static) and later dynamic server meshing using an entity graph system.
The point is that CIG is currently testing static servermeshing internally and some public tests, where as Itrepid wants to test it qith alpha testers now.
I am curious about what works how good and when. Not that it would be compareable. Intrepid has relatively "low" polygons with limited actions and CIG needs a totally different system that aids action combat and huge moving spaceships with multiple fully physicalised planets.
Lets see how Intrepid does it.
Also server meshing is nothing new planetside 2 had it in 2011, its hex based system on 64 square km map with 1200 slot per continent, and it need sync bullets and stuff and fps is much more dynamic game especialy in 100v100 situation than an mmo. I remmber like 10 years ago they said star citized relase in one year i was studying then, now i am 34 and star citizen didint relased yet lol.
Network engineer, MMO enjoyer and occasional SC player here. As one commenter already pointed out, this is not fair comparsion. SC deserves tons of criticisms, do not get me wrong please. But this is not accurate. You are basically comparing motorcycle engine and truck engine, saying "this one is better because" but that in itself can not make sense because while both are pushing piston by expanding combustion gases, they are working on completely different scale and serving completely different purpose. Again, do not get me wrong, Interpid is making beautiful achievemtn which may push MMO genre forward a lot. I love what I see here. But it has absolutely no place in comparsion with SC because their tech is simply being developed for different purpose. if you would like more technical details and tech-blabla, I am happy to share buit this is already too long post which nobody will read anyway.
Yep, software engineer here; I've worked on net code for games. What SC is doing is in another class of complexity...
I read it ❤
you guys like comparing starcraft I haven't played that since I was a kid
@@HorribleCoopGamers lmfaaoo
@@HorribleCoopGamersStar Citizen young padawan
I am super excited for AOC and not a Star Citizen fan boy at all so I hope this taken objectively, but to say what AoC is planning is the same as what Star Citizen is working on, is not fair at all. While I don't have a degree specifically in this area, I do in servers and databases, and from my understanding I would say the Dynamic Server Meshing of Star citizen is still a tier above what AoC has created and is planning for. BUT that is not a problem, because you don't need the fidelity needed like they are trying for in Star Citizen for an MMO, you don't need bullet tracking and high speed space sim elements tracking across multiple servers interacting with each other Or entire ships being servers themselves interacting with other people in servers around them. I do still have some questions on how AoC is going to handle abilities and things interacting across these mico servers without there being major bugs because I can't see them being able to get there minimum server size down to a point where that isn't needed. Async programing is HARD, extremely hard, and it will be exciting to see how they handle these issues.
To be fair, we don't know what AOC has planned for the future. What they showed are just things that will be present at the start of A2. Like what one of the developers said during the live stream, they will continue to improve the technology over the next few years.
Tell my why everybody talking server meshing like it is somehting new? planetside 2 had it since 2011.
They address this in the video they use proxy actors across borders. It very good and looks great in the footage they showed.
Its more like star citizens Static Server Meshing that they have implemented right now. Static Server Meshing is all a traditional mmo would need. Dynamic Server Meshing would be overkill I think.
@@MrBalrogos PS2 doesn't use server meshing and it hasn't been done on a big scale like AoC has
A few things on what Narc said about both server meshing solutions is a bit incorrect.
Microservices is not how the servers communicate together, they are the services like chat, mail and others that need to be distributed for management. It basically takes that job from the server so that the server doesn't need to worry about it and players can use these services no matter how broken the server they are on is. The servers are linked together not through microservers but through the grid communication. The servers only speak to what is adjacent and orthogonal to them.
The replication layer would indeed be an issue for SC if it was like what the graph Intrepid showed, but it's more close to what Intrepid are doing. The rep layer in SC is just a bunch of services that communicate the client data and the server data with the database. We have yet to really see a bottleneck since the replication layer scales horizontally just like the server mesh.
Server meshing isn't live in SC or AoC so it will be anyone's guess on how well either work when you have tons of players doing whatever they want. Exciting times are ahead for games. Something like Greyzone Warfare would be pretty cool with something like Server Meshing in a vast large world.
Correction, you are right that Narc is wrong, but you're slightly wrong in that the Microservices were primarily designed so that there could even BE things like communication across the server, since the server mesh technology inherently prevents servers sharing data past the immediate ones surrounding it.
It's just that these microservices come with a few perks with it working that way(perks like you stated). Just like the centralized replication layer was done away with primarily due to the bottleneck, yet it came with perks like realm stability
While true that Server meshing is not yet "live" in SC, the replication layer is and has show to work. In the tech tests with static meshing things also held up pretty well.
SC just did a test with 800 players, that is the most any fps online game had ever done. In the last video they were saying they probably would have 800 players, but they needed to stress test it. They admitted to bugs, but it is overall working. Already our basic servers went from 50 player servers up to 100 player servers and that has become the new standard. Also, we can actually play SC and 4.0 is supposed to launch this fall/winter, which will bring Pyro, and server meshing will be live. It might get delayed, but by years end we are supposed to see SC server meshing live in action.
Server meshing is live in AoC, the thing that isn't live yet is the dynamic part.
@@AshU-ug2fy 800? that isn't the most fps has done lol planetside 2 has 3k on a single server
i will sincerely admit that ASHES OF Creation is my most waited thing in this year. i am from indonesia.. from SEA server.. ive played lots of western subs-based mmorpg but i got super terrible ping . exceeds above 240ms.. that killed all my hope of enjoying a mmorpg.. and i hope the subs fee is no way above 15bugs.
A winner take all game with only one server is a disaster waiting to happen.
Im not a copium addict, I swear!
why i alredy smell copium around u!?
I could validate you are not if you validate I am not 😂
Sir, you blew a 25.83 on the copalyzer test. You're under arrest!
But copium feels so good!
there is one name that echoes through the ages that i can never forget. one name that never had more than a spot on a leaderboard. one name all my friends wanted to be as cool as... and that one name was... Zezima.
I'm a fan of both projects but It's pretty crazy to say that AoC team has made a better Server Meshing solution than Star Citizen when you compare the real tests of both.
- AoC Alpha 1 test with replication layer: 72km² static map + 1000 players + very limited dynamic objects and npc's = Completely bottleneck and need to change the entire SM system to work.
- Star Citizen Alpha 3.22 tests with Replication Layer and Persistence Entity streaming (Entity Graph) : 750 millions km across of simulated space (full volume is way way bigger) + 17 full planets and moons with cities rotating in real time + thousands of npc's + 800 players with a minimum of 800 ships with physic grids + Millions of dynamic entities = Test worked better than expected.
Maybe is way more complicated comparing both SM solution because the games are vastly different. Star Citizen combine PES(Entity Graph) with Micro-services + a decoupled replication layer + Object container streaming both client and server side etc... A lot of tech that sound similar but are way more complicated for a game like this :)
Maybe wait until it is released and proved to work before saying it is solved :-)
The tool that allows the devs to see all of the NPCs and players on the map at once "Shol's Eye" is named after the God of Truth in Ashes of Creation lore
Cheeky detail
multithread utilization on UE is just bonkers, that's probably one of the best inventions in online gaming so far..
Is that not just the replication graph portion and not the entire engine itself?
Comparing this so intensely (competitively) to star citizen's server meshing tech is not really helping your argument. It looks a little disingenuous and isn't something you can compare cleanly; the data and precision needed for these two completely different games require different technology and specs in the meshing technology, so it's not like ashes 'figured it out where star citizen couldn't. I wanted to know more about Ashes' tech, but this just felt like propaganda lol
Thanks for adapting the Mic., its very pleasant now! I can tell u really deep dived to explain as much as possible, thx mate.
honestly one of the best videos you've made yet narc!
All the hype about tech... What about actual gameplay? Will crafting be interesting or just a grind? Will resource harvesting have some depth or will it be clicking on a tier 1,2,3,4 node? etc etc...
What SC is doing is the Transformers of sever meshing, what Intrepid is doing is the Go-bots of server meshing. In the end, it's the one that works that is the best. If they both work we all win. SC will have the better version for SC and intrepid will have their version for AoC. I don't like either game, I just want to see the sever meshing make it in one form or another.
Exactly, this tech will push games to new levels, and also will be picked up and utilized outside of gaming too furthering it's own acceleration
I m so glad to have back up this game in 2018. Intrepid puting so much work for us to have a great mmo🎉
Long ago, Steven mentioned that scientific nodes will be able to teleport between other scientific nodes. How cool would it be (talking about no loading screens) if you open a portal and you can see the other side in real time. Just walk through, and you're able to look back and see your friends waving from the other side of the portal.
I think this probably would’ve been a better response to the stream if every aspect of Intrepid’s solution to server tech limitations was covered in a separate video. It’s just too hard to condense everything they talked about into a bite sized summary like this. But good attempt, this is largely accurate with a decently strung together narrative arc.
I hope this works well but I don't think dynamically partitioning the world will be enough once players really try to gather in the same location if the servers are fully authoritative.
I did prototype a similar system once and what I noticed is that at a certain occupancy level the split granularity made players change servers fast enough that it will actually add overhead.
Of course this will only matter if ashes is popular enough and at that point I'd just be happy to have a new mmo to play.
TLDR: "All your replicators are now belong to us" Intrepid 😃
Can litterally not compare the 2 games... Ashes is playing 2D and Star Citizen is playing 3D chess... They have to have this tech working on 360 degrees in ever direction possible while Ashes is a single world that is mostly 1 plane. That being said tho Both are impressive! And im really hyped for both games.
This might not have had the draw, or wow factor of some of their previous dev diary updates, but it was by all means extremely important. Anyone who's ever taken part in large pvp battles in an MMO before, should know how significant a problem it poses. They seem like they're on the verge of taking a step that all modern MMO's will follow in one way or another, at least if they're serious about providing players with a more reliable in game experience.
With data moving between servers and microservices I worry about dupping bugs. It takes time to confirm transactions and I wonder how Intrepid does that smoothly without loss of data and without duplication of items. I'm thinking of all the test cases I'm going to try in A2 (trading across, dropping items, picking up items, splitting stacks and mixing in logging off etc)
I think they talked about that in the presentation. The part about the proxies. I am no expert, but you might find the answer if you watch the video.
Was thinking the same thing
I'm getting more n more excited with every passing day mates!! My RC hobby is the only thing keeping me sane as I wait for a good MMO!
It is quite funny for old-timer IT people because this is an old architecture model that was used on fault tolerant hardware in the 1980/90s like Tandem and System/88, except in those days we couldn't connect multiple servers together faster than LAN speeds. And of course it was horrendously expensive.
Indeed, it is interesting to see people get all excited about tech which has been around for decades. At the same time though, the requirements of this tech in the context of these games is vastly more unique and demanding than it was back then.
@@blazemonger1 Yeah, back in the 90s it was used for processing real time transactions like for point-of-sale, or stock markets. Which was pretty demanding, because you had to deal with authorization transmission times, or real-time pricing, given the WAN speeds. Remember no high speed Internet at the time.
Think about racing cars back in the Ford vs Ferrari days. Just because the cars were going 30+mph slower, and the technology was older, doesn't mean that they were not pushing what they had to the limit, in terms of design or process.
It is more that we haven't had to resort to complex software architecture, because processing times for most requirements are easily met more reliably these days with hardware upgrades.
In fact the main difficulty (from what I can tell), is making server meshing work with Unreal Engine. If Unreal Engine was written with this architecture in mind then it wouldn't be as much of an issue
Overall great review Narc, I really felt like you took that heavy preview and thinned it out well for the rest of the community who couldn't be bothered to sit through it.
One small issue about the microservices statement you made, but someone else here already covered that. Well 2 issue really, you also talked like Intrepid doesn't do layering because they won't lower their standards like that, but in reality they literally can't. The game is not designed in a way that accommodate layering or sharding WHATSOEVER. Due to the dynamic nature of the world like citizenship status, guild status, node status, quest arc status, ect ect.
They have no other choice but to do this. Steven even mentioned that this was all a part of development, they didn't choose to do this after the fact, they knew from the outset that this was getting done. The game just flat out wouldn't work if they didn't do it, the worlds too big and too dependent on lots of large groups. Either they get this right or the game fails completely.
I'd say that the same could be said about Star Citizen but lets not fool ourselves, the games going to fail completely either way. P2W dev welfare economy sim trash.
Planetside 2 won the Guinness world record for largest MMOFPS on same map (1500). Wiki that
Love his SpaceTomato style intro lol
I didn't realise others had the same feelings as me about the last video until I saw this video and the comments under it.. I saw asmongold watch the last one and I felt like there was a lot of specific details missed out which sets this apart from the others, it was underwhelming. You've absolutely smashed this one out the park & is a testament to your dedication to us & the game. Really appreciate the fact you've taken comments of the last one on board...
Great Vid Narc, keep up the good work.
Everyone, server meshing is not a novel idea.
Pantheon MMO implemented the same thing like 5 years ago and did the exact same video about it AOC did.
The "gridding" tech AoC is showing is not new either.. It's been around for a long time and was successfully applied in the otherwise failed MMO Dual Universe as an example. And you do not need a tick rate of 60 or even 30. Tick rate is only part of the solution and can be much lower than that, in fact a tick rate of 15 would probably be fine for a MMO like AoC.
And indeed micro services ar not new either as they and you mentioned, it is a core part of the EVE server tech for instance and has been from the start.
In the end, none of what Intrepid has shown is new except for maybe how they had to rework the way UE5 handles large players the way they need it to, as that engine by default can't. What is unique is how they apply these separate server technologies and on paper make it look like a great solution for the type of game AoC is.
I hope they will be able to show their tech works, absolutely. But for now it is just talk and unproven. We'll find out during A2 if it all holds up, provided they get enough people to play.
oh they will def have enough people to play. But I'm sure they will be shitting up a storm if the servers keep crashing. Let's hope that doesn't happen.
Thanks for dumbing it down for us Narc!
As a long-time software developer I believe that success hasn't been achieved despite the core team only consisting of eight people, but because of it. Sometimes, less is more. A small team of highly talented and experienced people who "can smell" each other is often vastly more effective than a big one which usually gets mashed together by managers who only know how to throw money at problems and who think that nine women can get you a baby in a single month.
Here to say Narc will be the first AOC CC to reach 100k subs when Alpha 2 comes out.
Can’t wait to see how this tech performs under a massive seige war 💛
Great recap and summary of the tech. Good Job.
The multi-treading isn't done, not because it cannot be done, rather the game they are making has to run on the minimum spec machines. Now that AoC did this, I could only imagine that in order to play this game, I have to spend hundreds of dollars to upgrade my machine -- not gonna happen. If the minimum spec to play this game is too damn high, I'll play something else. If the can find a way to apply this multi-threading at low spec, that would be very impressive. So I'm still waiting and see what other IRL magic Intrepid can pull off.
My main concern now are.. bugs.
When you are doing everythign from scratch, you can expect lots and lots of unexpected bugs. Just look at new world, for example. Had some simple dumb bugs, that ruined economy of the game.
Yo narc this style of intro is better feel more emotionaly touching for some reason .
Server meshing was never meant to be exclusive to Star Citizen. When Chris Robert announce SC he thought the tech would be rdy.. and it's not been rdy since yearsss..
Every big MMO, like pax dei, gonna be server meshed ;)
Pepperidge Farms remembers when Rift was going to change the future of MMOs with a teleporting sniper and EQNext was, too. After FFXIV, I went back to where I started, consoles. MMOs just weren't the same anymore, the community got far worse and console RPGs are a lot less 'update' or mod-reliant to raid. If this was ready when I quit PC MMOs 3 years ago, I would have been a Day 1 player. Hope this does well and the tech somehow eventually works with the limitation of console controllers.
Nice one, Narc. It's not an easy topic to digest but you did a great job. Very excited to see how it works in game.
Ashes of Creation server mesh that was never attempted before, meanwhile Star Citezen looking at AoC "i was first kid"... Then comes GTA6 Online with fully working server mesh, released game and tech laughing at both.
Not concerned with a race. Don't care that one is harder than the other. Just excited that it seems like a reality for both and that's good for everyone.
Same here I'm not really getting on the whole dunking on the other train. I'm just glad it's getting further along.
I mean, I wouldn't even call this copium here. They've already done it, and all it shows is how capable their team is, with solid proof. Or ya know, I'm just coping with copium addiction 🤷♂
Star Citizen tried the entity graph method already, it failed then and it will fail now.
Ashes is the best what could happened to the mmo genre. I dont want this to be a WoW Killer, but this will spark a good competition from wich we player will profit a lot. At least i hope so.
I "retired" from gaming after finishing classic and TBC remakes but my bros got me to buy an alpha key and re-enter the dungeon.
Gonna stock up on adult diapers and take 2 weeks off work for alpha release and binge this shit till my eyes bleed 🤣, see you in there nerds!!
Now it makes sense why their development studio is called "Intrepid"
ashes of creation = dynamic server griding
star citizen = dynamic server mashing
Question is there gonna be voice chat any idea? i feel like it would be so fun setting up a raid and stuff and rp with it and any thing aboutn housing so far ? Love the good explaining in the video keep it up
Game that's not out is competing with other game that's not out's technology that no one has access to test. Amazing.
Mortal online 2 does 1 server and for the longest time, 1 character. When you shit where you eat, you tarnish your reputation, and people will be more hostile towards you over time. If we can have that on an even larger scale! Holy, that'd cool.
It is exciting tech that has huge potenital moving forward from gaming to the real world
Dont worry people , they ll run alpha for many years ,so there is enough time to fix any server problem. Sleep-eat good so we are healthy and sometime in the future we ll play aoc
the replication layer used by star citizen is also multple servers. not jsut one
Does this game have a PvE component or not? Do you level up via traditional grinding, is there a cap on leveling? Is it all PvP like Black Desert or is it kind of hybrid like Archeage (which has dungeons and such)?
Hopefully they've solved MMOs biggest problem. Let's see how well it does during spot testing before we declare it solved
having 500 people in one area should be possible if the server doesnt load the stash of each person every second like D4
for real though, there should be a way to optimize the amount of data , you dont need new magic server tech
if you combine optimization with new server tech though, it should be possible to have 1000 in one area, dont see the need for more.
When questing I hope you dont see 40000 people in one area
Great video (obviously). Interesting topic. Technically, very impressive.
However, I fear it will have the opposite of the desired effect.
People often think they want things, that they, if fact, do not. A smaller server size is vastly superior, across the board, for the vast majority of the players, the vast majority of the time.
A bigger pond, means fewer, and smaller, 'big fish' at the end of the day. You don't need an MMO to feel like a forgotten nobody, who knows nobody, and whom nobody knows, that is what IRL is for.
It also is effected by how human brains work, but that gets kinda technical.
So, I gotta say the tech is great, but the result will be an inferior experience for +/- 99% of the players.
If we think massive open worlds are MMOs biggest problem, we're in even deeper than I thought. Can we please talk about _real_ problems plaguing the genre?
Theres definitely many problems, but seamless servers has to be at the top of what is expected when building a massive open world game. I think its one of the main reasons why WoW felt like a real world.
@@missionpupa WoW's mostly open world is a boon, but when an open world comes at the cost of discussion about things like gameplay we're screwed. EQ's original design immersed me immediately even though it has loading zones. Mind you I grew up with WoW first for over a decade. Seamless worlds are def not at the top, but near. Not required, but really important.
@@poisonated7467 Its just one of the many conversations we can have, in fact AoC barely discusses these things in their videos. Im saying that technically, this definitely has to be the most challenging part of creating the game, the game loops, skill set, theyre different problems but theyre not technical problems.
@@missionpupa Technically, everything is a has a technical side to it, but I understand what you're saying and agree that it is a technically challenging problem to solve. Just that it's not MMO's biggest problem right now as the video suggested.
One of the largest problems is trying to solve community and player choice problems with tech. It seems unsolvable.
@@poisonated7467 The MMO can have many problems its not just 1 thing, and im definitely not saying once you solve the seamless servers it means you solved everything, im just highlighting the importance of it.
Not really sure what you mean by player choice problems.
Im a History major. So yes to all.
I just need one more copa- cola . So glad to see the game isn’t a scam 😊
Great Vid Narc!
I need me a documentary style, copium filled 2 hour movie about Ashes' journey. If that is something you consider doing at some point.
SC is pushing the boundaries with this tech, AoC is only doing the light version
I like Guild Wars 2, FFXIV, and modern WoW.
I don't know why we're trying to compare Ashes of Creation and Star Citizen's server meshing. It's like comparing apples to oranges, they're both server meshing but they're inherently different in their own right. Star Citizen has a lot to be criticized about but the random digs throughout the video make it tough to watch. I come to watch Narc videos for AOC content, if I wanted to listen to SC criticism I would go to any of the other more informed SC content creators. We should be hoping both succeed in making that kind of tech work instead of "hurr durr my meshing is better than yours" when they're not even a real comparison. Either of them succeeding means a brighter future for MMO games
throne and liberty has it too, (about server layout) and before ashes. GGz
absolutely insane video, loved it.
God, I hate when my load gets split.
Epic News, thanks Narc!
i will be excited in 5 years, when i manage to play it !
cannot wait to embarress myself to everyone and not being able to run off to a different server, because there will only be one
Publisher - You make game, we pay for servers, GOT IT ?
Developers - Uh sure, just pay us.
intrepid studios - Hold my beer.
You can actually play star citizen. Ive actually seen normal people testing Server Meshing. Aoc hasn't done shit till the actual public gets to test it... Believe it when i see it
I'm sad that there was no Server Meshi anime opening
so there will be no griding spots for everyone in the game cuz we will not have chanels?
I'll believe it when I see it working.
So whats the solve for like a busy starting area? Even server based MMO's that are popping have you sitting around for stuff because of over crowding. Unless i'm understanding it wrong? Is the map and content big enough to handle every last player sitting on the map?
Good question. We'll have to see how AoC and Star Citizen handle massive amounts of players in single locations. One solutions is to cap the player numbers at a number that supports such massive congregations, another is layering.
Another problem besides the gameplay/map supporting vast amounts of players is the optimization of the game itself so that your PC can handle all the things it has to render on screen.
@@herobrine024 Yeah I guess that's my worry. If they layer it kind of defeats the goal. I'm not concerned with processing as much as I'm concerned with overcrowding causing the game to be boring. Lol. I'm thinking about lines of people in wow classic for a quest Mob, or never being able to get a crafting node. I guess the tests will go a long way toward figuring that out I hope
@@johnnynometer in Star Citizen when they raised player counts from 50 to 100 and had a server wipe, on day one, there were literal lines at the ship retrieval terminals.
Because they are now adding physicalized cargo and persistent personal hangars, meaning players will spend much more time inside hangars, theyve relocated these terminals inside of the hangars.
This indirectly solves this issue and prepares the gameplay for many more players playing on the same server, because these hangars are layered. But thats only possible because theyre rooms that only specific people have access to (think GTA Online Apartments with seamless doors to the outside). AoC is primarily in open environments so this wouldn:t be applicable to the game.
The entire framework they're developing is the solution to that issue. As more players would arrive in a single area, their dynamic server meshing would dynamically split and assign resources as needed, without the need for layering.
It's a wait and see for how they'll implement starting areas but the zones themselves are vast.
Quick Edit: I could see them having a few similar starting points per race to divvy players out and help with those issues. Could have players congregate a bit later on. Maybe you're assigned one of those locations and could override it if you wanted to start with your friend, etc. All speculation on that of course but only someone like Asmongold would be ruining that for most people I would think lol.
Great video Narc! Ignore the haters!
I love how GW2 did so well in aspects that WoW was lacking, then wow stole everything that made gw2 good, turning gw2 irrelevant
both games have become pretty lame now however...
WTF Narc taught me something. This is timeline meshing
im on the copium waiting to see how this turns out when testing happens fingers crossed no dumpster fire pls
I feel like the first server meshing work was done with Ultima Online. It was much smaller, but the concept was the same.
in the game software, High tick rate is bad, low tick rate is good
No, its the other way around. A tick rate represents the amount of times the engine performs a simulation step per second. The higher, the more accurate.
@@herobrine024 I could have swore they said the opposite, but ill take your word for it
I would say Ultima online did it before WoW
Great summary!
well AoC shows its will be mediocore game not actual mmo everybody waiting for so i am afraid its gonna end like new world
Haven’t you titled another one of your videos this exact thing before?
tech has always been there, its just expensive to run.
Narc i love your videos but i am kinda getting tired of corporate is trash talk being constant
There's one question these mmos haven't answered. How do they stop cheating. It's still a big problem that's just kinda ignored
I would hope it's a case of getting this server mesh tech to work first; then you'd be working on the end of testing how to keep people from cheating. I feel like this type of mesh network will make some weird duping issue of items etc be a lot easier to pull off; but of course no way to know then to go into Alpha 2 and try and break the game etc and report it.
Beep beep, here comes the copium truck 🛻
Next-gen server tech, gameplay from 2010. Overhyped vaporware.
Show, dont tell. i'll belive it when they show it in some massive server test
Tech like this isn't usually visible in gameplay because the goal is to not show you it is there. If you notice it there must be lag or something wrong in that grid server location.
@@perilousintent2936 well wow lags with 30+ players, this should be able to support alot more or did i missunderstand it?
Actually a well explained video.