KhajitFurTrader wrote: »Waaaay before launch, there was talk in developer interviews about different zone phases, and a questionnaire for new players at the beginning of the game, according to which player characters would be put in appropriate phases, sorted by preferences: solo, group, RP, etc.
At the time of the stress test betas, this system didn't make an appearance, and questions were asked about it on the forum. I can't remember whether there had been an official statement, or there were just educated guesses, but I do remember that the word on the street was that plans for zone phasing by preference had been discontinued due to severe performance restraints put on the server architecture.
Given how server performance is a hotly discussed topic to this day, zone preference phases failing to materialize was probably for the best.
So, to answer the question: in theory, yes, they are possible, but may just not be practical. Each of the original zones has, at the minimum, three phases already: one for the Alliance it belongs to, and two VR copies for the others. I don't know if, and how often, this is done, but there may also be multiple copies of the same zone to funnel and distribute population, i.e. once a zone reaches its pop limit, another copy is opened to alleviate system load by overcrowding.
Well, that depends.I could see the performance issues being a problem. In hindsight, the Megaserver technology seems to be more a hindrance to ESO than it is a boon.
KhajitFurTrader wrote: »So, to answer the question: in theory, yes, they are possible, but may just not be practical. Each of the original zones has, at the minimum, three phases already: one for the Alliance it belongs to, and two VR copies for the others. I don't know if, and how often, this is done, but there may also be multiple copies of the same zone to funnel and distribute population, i.e. once a zone reaches its pop limit, another copy is opened to alleviate system load by overcrowding.
lordrichter wrote: »It is the use of channels to enforce the separation of players that are in the same location that may contribute to people periodically making population-based claims that the game is dead, dying, or going to be. They destroy our perception on how busy the server is, to the point where we cannot tell the difference between low server population and high server population. Channels might be only a couple hundred players in size, and we will not be seeing all of them during the course of play. We don't know if there is one channel or 50 channels, and we have no idea how balanced the channel populations might be.
KhajitFurTrader wrote: »I find the channel concept intriguing, and I don't remember reading about it before. The old battle.net chat rooms come to mind: if one chat room filled up to its cap (100 per channel in the days of D2, iirc), another one was spawned and joining players were placed in it. The reason for handling things this way is simple mathematics: each message from any member to the chat room has to be sent back to all participants of said chat room; the number of messages to be handled by the chat room thread grows linearly to the number of participants (complexity of O(n)). With a graphical MMO client, there's another catch: player characters are usually the objects with the highest polycount and the most diverse armor textures in a scene (I'm not even speaking about particle effects caused by those players). If there were 600 concurrent players in same place, and the client had to display all of them, performance would suffer for everyone. Split them up into neat slices of, say, 300, and you spread the load on two threads.