It's NOT really highly instanced, and lets use the obvious example. Eve only has one server. Fights in Eve basically turn into clip-fests. No thanks.
Now for the obvious answer, because there are queues, there need to be more servers. (this is the second time I've had a queue, the first outside of prime weekend time)
But it is highly instanced. If you look up at the top left of the screen it says the zone and some number like 100. That's the number of people in the zone. Do you honestly believe there's only a few copies of that zone on launch day? There could be 30 copies of that zone or more.
Why are there even cues? Just make more instances.
If you have several hundred million dollars to make a game wouldn't it be possible to have a computer that could generate thousands of instanced zones...or as many as were needed?
It's not a single computer, it's actually a cluster of servers working together.
Exactly, thank you. So if you have 100's of millions of dollars, it's just a matter of buying the right amount of clusters then?
Yes, you could probably create thousands of layers (instancing or phasing) for the same zone, that is not the problem.
The problem is that you can only reliably have ONE DATABASE supporting all those layers. And if there are too many simultaneous transactions being fired at the database, things go very horribly wrong.
Ahhh...this is the first thing that is starting to make sense to me. Even though there's lots of copies, one database for all. So each server has it's own database.
Indeed, you're getting the picture
You could (in theory) also duplicate the database along with the instances and then sync the copies at regular intervals, but it is so incredibly risky that no sane systems guy would contemplate it. Think along the lines of dissappearing inventory items, lost XP, quests resetting, etc.
And besides, the system load that will occur during the synchronization will cancel any advantage of the instancing in the first place.
Actually you just preempted my next question. If the problem is the database, why not do the same as in the virtual world. As soon as demand gets to heavy, create instances within the database. I could imagine how that could be disasterous though.
Look.. i'm sure if there were a way to do this with the current batch of tech, it would have been done. I mean really, these companies hate opening a ton of servers, as when and if you have to merge them, it's choas with the players. However, I would much rather have a smooth running game then have them try to find a way to jam more people on the servers, impacting my gaming experience. I mean, how many people does a TOR server hold? You honestly need more?
How many does a TOR server hold?
I wish it could hold 1 more right now
lol, I hope you get in soon - I'd give you my spot if I could.
Techinically some places would be instanced in the game just for you kind of how guild wars has it in your own copy of the map. But for zones that they made it purposefully to see other people its not possible to make a incalculable amount of copies of the same zone becuase it takes up memory in the server they use. It really isn't I see where your coming from about having multiple servers for pve pvp and rp. I think the reason they want to keep it all together is because in the end its one whole experience they don't want people to just pvp in a server because it would ruin their work in making the stories because they want people to see the hard work they've done. Vice versa they dont want just pve players not realizing the fun aspects of pvp, competitive in that hand, that many people enjoy.
The game needs only three servers for NA: pve, pvp, and rp. If everyone was on the same server, we wouldn't have to server hop to be with other groups of friends on other servers.
First, design a server capable of hosting a million players plus. Then we can talk.
So you will only talk to me if I am a computer engineer?
Well, it first must be physically possible to do what you suggest. While it isn't, there's not much point in discussing the idea, is there?
it is possible, but normally they limit a server to a certain number of people online because of maps and people competing in certain quest mobs and such, also most of game maps are small so would be a pain to compete a mob agaisnt other 10/20 peeps daily basis.
by hardware you don't have a limit per see, the limit is a dev thing
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
If you're really interseted, PM me and I'll go through the documentation provided to me when I got my Hero Engine license. They clearly state that you could have over 1 million, there truly is no limit if you really wanted to take advantage of phasing.
The Hero Engine very possibly CAN control a million instances. In theory.
The game engine runs in the middle layer, top layer is the client (you) and at the bottom is the database engine, which will be Oracle, Microsoft Sequel Server, DB2 or something similar.
So even though Hero can host the million instances, it has to send and receive data to the SINGLE database engine continuously to manage the goings-on in each of those instances. THAT's where the wheels come off.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Im glad its designed to do that...however it is not currently doing so.
There are a lot of startups in the technology industry that make wild claims and when you actually get down to brass tax its not quite what you thought it was.
I remember a year ago there was talks of a new engine that could host billions and billions of on screen pixels then it came out they couldnt be animated correctly. So while gorgeous it wasnt really applicable.
I have never heard of the hero engine, and its quite possible that when its setup it can host good numbers. Hell eve's server can run 60k - 70k people. The question is what is the draw back?
Eve's system works well because of how the game is designed, you never have more than say 2000 people in the same system / on the same node. However, with theme park games thats not the case. If EvE were a theme park game their setup wouldnt work.
Hero's engine will definitely have limitation. It just depends on what they are and if a game is acctually acceptable based on said limitations...or...even used for that matter.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Im glad its designed to do that...however it is not currently doing so.
There are a lot of startups in the technology industry that make wild claims and when you actually get down to brass tax its not quite what you thought it was.
I remember a year ago there was talks of a new engine that could host billions and billions of on screen pixels then it came out they couldnt be animated correctly. So while gorgeous it wasnt really applicable.
I have never heard of the hero engine, and its quite possible that when its setup it can host good numbers. Hell eve's server can run 60k - 70k people. The question is what is the draw back?
Eve's system works well because of how the game is designed, you never have more than say 2000 people in the same system / on the same node. However, with theme park games thats not the case. If EvE were a theme park game their setup wouldnt work.
Hero's engine will definitely have limitation. It just depends on what they are and if a game is acctually acceptable based on said limitations...or...even used for that matter.
The fact is, you don't know what it is currently doing or not doing.
The Hero Engine manages things in much the same way as Eve's, by limitting the number of people you are actualy interacting with. Their documentation focuses on designing the game to handle as many players as you need, but reminds you that if you have, for example, 100 players all standing together, every single movement or action of each player has to be sent to 99 other players. So, everyone jumps to the left and there are 100x100 updatings being sent out.
One of the devs talks about a test he did with 5000 AI characters active in a single node, all completely aware of each other, essentially testing this out.
lol you link an article that is over 2 1/2 years old to try to contradict me? Seriously, please try harder. I'm correct, you are incorrect. If you truly want to know, the information is available.
Nope. Seriously interested in documentation, if that has changed.
"HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Like the op, I'd love to see that kind of capability in the live world.
That's pretty exciting news, actually. Hope we see someone attempting it soon.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
It's NOT really highly instanced, and lets use the obvious example. Eve only has one server. Fights in Eve basically turn into clip-fests. No thanks.
Now for the obvious answer, because there are queues, there need to be more servers. (this is the second time I've had a queue, the first outside of prime weekend time)
But it is highly instanced. If you look up at the top left of the screen it says the zone and some number like 100. That's the number of people in the zone. Do you honestly believe there's only a few copies of that zone on launch day? There could be 30 copies of that zone or more.
Why are there even cues? Just make more instances.
If you have several hundred million dollars to make a game wouldn't it be possible to have a computer that could generate thousands of instanced zones...or as many as were needed?
They are not doing this due to added overhead. At the moment, all those instances are running on one server. To do what you want, we would need clusters of servers. Now you need to have software that tracking where on what server each instance is running. Other functions would need to be made cluster aware. Guilds? Need to track what servers members are on. What do you do if server X dies, Now you need active - active failover. How do you scale the databases to be 100 times bigger, You would have to do a ton of infrastructure design and work to do this, making the product much more complex and from their point of veiw for little return. I expect if they had tried this, it would have cost them twice as much to do and probably took twice as long to release.
another way of viewing this is if you put 2 graphic cards into you computer, you do not get twice the FPS, cycles are lost due to the cards having to coordinate. Same thing with clustering 2 servers, you would not be able to put twice as many people on the cluster.
lol you link an article that is over 2 1/2 years old to try to contradict me? Seriously, please try harder. I'm correct, you are incorrect. If you truly want to know, the information is available.
Nope. Seriously interested in documentation, if that has changed. Nothing on the HE site indicates otherwise.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
They've been updating their site quite a bit recently. There's a lot of really great documentation there. They even talk about why, even though the engine can handle it, it may not be a good idea for your game depending on the design. There's a lot of focus on proper world design to maximize performance. It's a really powerful engine, with a lot of amazing potential. I've only had my access for a short time and I'm just scratchign the surface.
But it is highly instanced. If you look up at the top left of the screen it says the zone and some number like 100. That's the number of people in the zone. Do you honestly believe there's only a few copies of that zone on launch day? There could be 30 copies of that zone or more.
Why are there even cues? Just make more instances.
If you have several hundred million dollars to make a game wouldn't it be possible to have a computer that could generate thousands of instanced zones...or as many as were needed?
That's not the instance number, that's the number of players in the zone.
And to add to the number of good reasons already offered... Imagine trying to think of a name that's unique after a million other players have created several charcters each. Then think about doing it after they've released a trial...
lol you link an article that is over 2 1/2 years old to try to contradict me? Seriously, please try harder. I'm correct, you are incorrect. If you truly want to know, the information is available.
Nope. Seriously interested in documentation, if that has changed. Nothing on the HE site indicates otherwise.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
They've been updating their site quite a bit recently. There's a lot of really great documentation there. They even talk about why, even though the engine can handle it, it may not be a good idea for your game depending on the design. There's a lot of focus on proper world design to maximize performance. It's a really powerful engine, with a lot of amazing potential. I've only had my access for a short time and I'm just scratchign the surface.
Yep. The proof is in the pudding, feasible vs. plausible.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
lol you link an article that is over 2 1/2 years old to try to contradict me? Seriously, please try harder. I'm correct, you are incorrect. If you truly want to know, the information is available.
Nope. Seriously interested in documentation, if that has changed.
"HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Like the op, I'd love to see that kind of capability in the live world.
That's pretty exciting news, actually. Hope we see someone attempting it soon.
It would really come down to game design... like in WoW, for example, what would be the point?
It wouldn't make sense for the game I'm working on, but I hope to take advantage of these capabilities on a larger scale. For example, it will primarily take place in Rhode Island, and we want to pack everyone in to make it a living, breathing world. The key is to have enough things for people to do that they're not all waiting in line for the same things.
lol you link an article that is over 2 1/2 years old to try to contradict me? Seriously, please try harder. I'm correct, you are incorrect. If you truly want to know, the information is available.
Nope. Seriously interested in documentation, if that has changed. Nothing on the HE site indicates otherwise.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
They've been updating their site quite a bit recently. There's a lot of really great documentation there. They even talk about why, even though the engine can handle it, it may not be a good idea for your game depending on the design. There's a lot of focus on proper world design to maximize performance. It's a really powerful engine, with a lot of amazing potential. I've only had my access for a short time and I'm just scratchign the surface.
Yep. The proof is in the pudding, feasible vs. plausible.
Feasable vs plausible vs workable
It actually might be interesting designing a game from the ground up with this idea in mind.... we want 1 million people on a server.... and then go from there. It would certainly have it's "gimmick" to make it stand out and get media attention, which isn't necessarily a bad thing. But where do we go from there?
I'd really be interested in hearing ideas on this.
Yep. The proof is in the pudding, feasible vs. plausible.
Feasable vs plausible vs workable
It actually might be interesting designing a game from the ground up with this idea in mind.... we want 1 million people on a server.... and then go from there. It would certainly have it's "gimmick" to make it stand out and get media attention, which isn't necessarily a bad thing. But where do we go from there?
I'd really be interested in hearing ideas on this.
As I was saying earlier, if names have to be unique, it would be a bad, bad, BAD idea to have a million on one server. I'm surprised Eve isn't more horrible than it is for names, and it's still pretty horrible.
But single server for games like STO and CO? Works fairly well, with less than 100k subscribers...
Yep. The proof is in the pudding, feasible vs. plausible.
Feasable vs plausible vs workable
It actually might be interesting designing a game from the ground up with this idea in mind.... we want 1 million people on a server.... and then go from there. It would certainly have it's "gimmick" to make it stand out and get media attention, which isn't necessarily a bad thing. But where do we go from there?
I'd really be interested in hearing ideas on this.
As I was saying earlier, if names have to be unique, it would be a bad, bad, BAD idea to have a million on one server. I'm surprised Eve isn't more horrible than it is for names, and it's still pretty horrible.
But single server for games like STO and CO? Works fairly well, with less than 100k subscribers...
Allow first name, last name, and middle initial
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Im glad its designed to do that...however it is not currently doing so.
There are a lot of startups in the technology industry that make wild claims and when you actually get down to brass tax its not quite what you thought it was.
I remember a year ago there was talks of a new engine that could host billions and billions of on screen pixels then it came out they couldnt be animated correctly. So while gorgeous it wasnt really applicable.
I have never heard of the hero engine, and its quite possible that when its setup it can host good numbers. Hell eve's server can run 60k - 70k people. The question is what is the draw back?
Eve's system works well because of how the game is designed, you never have more than say 2000 people in the same system / on the same node. However, with theme park games thats not the case. If EvE were a theme park game their setup wouldnt work.
Hero's engine will definitely have limitation. It just depends on what they are and if a game is acctually acceptable based on said limitations...or...even used for that matter.
The fact is, you don't know what it is currently doing or not doing.
The Hero Engine manages things in much the same way as Eve's, by limitting the number of people you are actualy interacting with. Their documentation focuses on designing the game to handle as many players as you need, but reminds you that if you have, for example, 100 players all standing together, every single movement or action of each player has to be sent to 99 other players. So, everyone jumps to the left and there are 100x100 updatings being sent out.
One of the devs talks about a test he did with 5000 AI characters active in a single node, all completely aware of each other, essentially testing this out.
I am aware it is not running a single game live atm.
I am also aware that AI characters do not = actual players. Darkfall pretty much proved that to us.
My point is, until its picked up and used its all just hype.
I've only read 3 pages, so someone may have answered already, but I think it has to do with bandwidth. There is only so much you can pipe into a server without spending silly money, so it is much more economical to use smaller pipes on more servers.
Before you give the obvious answer, because there are so many people, my question stems from the fact that the game is instanced. For example, if two many people are in a zone, a copy of that zone is made. As many copies as needed could theoretically be generated, or is this technically impossible and why?
The game needs only three servers for NA: pve, pvp, and rp. If everyone was on the same server, we wouldn't have to server hop to be with other groups of friends on other servers.
To my knowledge the instancing system is supposed to be just a backup system for launch load, propably expansions etc, not as a feature that divides every zone into a thousand instances, I recall someone from BW saying the idea is to have "one world" but the instancing is there to help high loads to lessen pain on especially starter zones etc. Could be wrong.
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
You'd almost need the ability to expand the map dynamically as your population increases... which Hero Engine could manage pretty well, actually.
Traffic direction would be a major issue... Guards at city gates to hold people away? heh Of course, if 10,000 players storm the gates, we're screwed.
I suppose in a mob of 10,000 people things would be very chaotic in real life, so a little lag isn't the end of the world!
The issue is not going to be your engine, its going to be your infrastructure. Google can move a whole datacenter to a new site in real time. The only company in the world that can do that. the initial hold up was not the servers but the connection when servers are disconnected and are unable to verify data during transactions. A problem that still shows up in large virtual clusters. The back end routers and connects need to resolve this issue start in the high 6 figures and quickly move north. It is not plausible to spend a 50 million on hardware so that you can say AHHH in a game that is more story driven than world driven. Price any high end server cluster and the related database SAN and you will see what i am saying.
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
There's six billion or so people on planet Earth....doesn't feel empty to me.
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
There's six billion or so people on planet Earth....doesn't feel empty to me.
...the hell?
I might get banned for this. - Rizel Star.
I'm not afraid to tell trolls what they [need] to hear, even if that means for me to have an forced absence afterwards.
P2P LOGIC = If it's P2P it means longevity, overall better game, and THE BEST SUPPORT EVER!!!!!(Which has been rinsed and repeated about a thousand times)
Common Sense Logic = P2P logic is no better than F2P Logic.
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
There's six billion or so people on planet Earth....doesn't feel empty to me.
Comments
lol...thanks brother
Techinically some places would be instanced in the game just for you kind of how guild wars has it in your own copy of the map. But for zones that they made it purposefully to see other people its not possible to make a incalculable amount of copies of the same zone becuase it takes up memory in the server they use. It really isn't I see where your coming from about having multiple servers for pve pvp and rp. I think the reason they want to keep it all together is because in the end its one whole experience they don't want people to just pvp in a server because it would ruin their work in making the stories because they want people to see the hard work they've done. Vice versa they dont want just pve players not realizing the fun aspects of pvp, competitive in that hand, that many people enjoy.
it is possible, but normally they limit a server to a certain number of people online because of maps and people competing in certain quest mobs and such, also most of game maps are small so would be a pain to compete a mob agaisnt other 10/20 peeps daily basis.
by hardware you don't have a limit per see, the limit is a dev thing
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
The Hero Engine very possibly CAN control a million instances. In theory.
The game engine runs in the middle layer, top layer is the client (you) and at the bottom is the database engine, which will be Oracle, Microsoft Sequel Server, DB2 or something similar.
So even though Hero can host the million instances, it has to send and receive data to the SINGLE database engine continuously to manage the goings-on in each of those instances. THAT's where the wheels come off.
Im glad its designed to do that...however it is not currently doing so.
There are a lot of startups in the technology industry that make wild claims and when you actually get down to brass tax its not quite what you thought it was.
I remember a year ago there was talks of a new engine that could host billions and billions of on screen pixels then it came out they couldnt be animated correctly. So while gorgeous it wasnt really applicable.
I have never heard of the hero engine, and its quite possible that when its setup it can host good numbers. Hell eve's server can run 60k - 70k people. The question is what is the draw back?
Eve's system works well because of how the game is designed, you never have more than say 2000 people in the same system / on the same node. However, with theme park games thats not the case. If EvE were a theme park game their setup wouldnt work.
Hero's engine will definitely have limitation. It just depends on what they are and if a game is acctually acceptable based on said limitations...or...even used for that matter.
The fact is, you don't know what it is currently doing or not doing.
The Hero Engine manages things in much the same way as Eve's, by limitting the number of people you are actualy interacting with. Their documentation focuses on designing the game to handle as many players as you need, but reminds you that if you have, for example, 100 players all standing together, every single movement or action of each player has to be sent to 99 other players. So, everyone jumps to the left and there are 100x100 updatings being sent out.
One of the devs talks about a test he did with 5000 AI characters active in a single node, all completely aware of each other, essentially testing this out.
Nope. Seriously interested in documentation, if that has changed.
"HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
Like the op, I'd love to see that kind of capability in the live world.
That's pretty exciting news, actually. Hope we see someone attempting it soon.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
They are not doing this due to added overhead. At the moment, all those instances are running on one server. To do what you want, we would need clusters of servers. Now you need to have software that tracking where on what server each instance is running. Other functions would need to be made cluster aware. Guilds? Need to track what servers members are on. What do you do if server X dies, Now you need active - active failover. How do you scale the databases to be 100 times bigger, You would have to do a ton of infrastructure design and work to do this, making the product much more complex and from their point of veiw for little return. I expect if they had tried this, it would have cost them twice as much to do and probably took twice as long to release.
another way of viewing this is if you put 2 graphic cards into you computer, you do not get twice the FPS, cycles are lost due to the cards having to coordinate. Same thing with clustering 2 servers, you would not be able to put twice as many people on the cluster.
From their site, which is NOT 2 1/2 years old: "HeroEngine is designed for Massive Multiplayer games, easily accommodating tens of thousands, hundreds of thousands or even millions of players in a cluster."
They've been updating their site quite a bit recently. There's a lot of really great documentation there. They even talk about why, even though the engine can handle it, it may not be a good idea for your game depending on the design. There's a lot of focus on proper world design to maximize performance. It's a really powerful engine, with a lot of amazing potential. I've only had my access for a short time and I'm just scratchign the surface.
That's not the instance number, that's the number of players in the zone.
And to add to the number of good reasons already offered... Imagine trying to think of a name that's unique after a million other players have created several charcters each. Then think about doing it after they've released a trial...
Yep. The proof is in the pudding, feasible vs. plausible.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
It would really come down to game design... like in WoW, for example, what would be the point?
It wouldn't make sense for the game I'm working on, but I hope to take advantage of these capabilities on a larger scale. For example, it will primarily take place in Rhode Island, and we want to pack everyone in to make it a living, breathing world. The key is to have enough things for people to do that they're not all waiting in line for the same things.
Feasable vs plausible vs workable
It actually might be interesting designing a game from the ground up with this idea in mind.... we want 1 million people on a server.... and then go from there. It would certainly have it's "gimmick" to make it stand out and get media attention, which isn't necessarily a bad thing. But where do we go from there?
I'd really be interested in hearing ideas on this.
As I was saying earlier, if names have to be unique, it would be a bad, bad, BAD idea to have a million on one server. I'm surprised Eve isn't more horrible than it is for names, and it's still pretty horrible.
But single server for games like STO and CO? Works fairly well, with less than 100k subscribers...
Allow first name, last name, and middle initial
And I'm thinking one million players in a virtual world, not just sharded servers. So the world itself would have to be massive (bring the massive back to MMORPG, right? hehe)... and probably very sandboxed....
I am aware it is not running a single game live atm.
I am also aware that AI characters do not = actual players. Darkfall pretty much proved that to us.
My point is, until its picked up and used its all just hype.
I've only read 3 pages, so someone may have answered already, but I think it has to do with bandwidth. There is only so much you can pipe into a server without spending silly money, so it is much more economical to use smaller pipes on more servers.
Probably.
To my knowledge the instancing system is supposed to be just a backup system for launch load, propably expansions etc, not as a feature that divides every zone into a thousand instances, I recall someone from BW saying the idea is to have "one world" but the instancing is there to help high loads to lessen pain on especially starter zones etc. Could be wrong.
Massive also implies (indirectly) "feels empty", doesn't it?
Maybe not, if you could cram enough players on a server. But I can see a "congregate in one spot" (Lagforge) problem developing, if you're not a very careful traffic director.
Self-pity imprisons us in the walls of our own self-absorption. The whole world shrinks down to the size of our problem, and the more we dwell on it, the smaller we are and the larger the problem seems to grow.
You'd almost need the ability to expand the map dynamically as your population increases... which Hero Engine could manage pretty well, actually.
Traffic direction would be a major issue... Guards at city gates to hold people away? heh Of course, if 10,000 players storm the gates, we're screwed.
I suppose in a mob of 10,000 people things would be very chaotic in real life, so a little lag isn't the end of the world!
The issue is not going to be your engine, its going to be your infrastructure. Google can move a whole datacenter to a new site in real time. The only company in the world that can do that. the initial hold up was not the servers but the connection when servers are disconnected and are unable to verify data during transactions. A problem that still shows up in large virtual clusters. The back end routers and connects need to resolve this issue start in the high 6 figures and quickly move north. It is not plausible to spend a 50 million on hardware so that you can say AHHH in a game that is more story driven than world driven. Price any high end server cluster and the related database SAN and you will see what i am saying.
There's six billion or so people on planet Earth....doesn't feel empty to me.
I might get banned for this. - Rizel Star.
I'm not afraid to tell trolls what they [need] to hear, even if that means for me to have an forced absence afterwards.
P2P LOGIC = If it's P2P it means longevity, overall better game, and THE BEST SUPPORT EVER!!!!!(Which has been rinsed and repeated about a thousand times)
Common Sense Logic = P2P logic is no better than F2P Logic.
The Earth.