Does a person need to buy a bunch of computers to increase size for online games?

#1 Ehhhmmm.. This sounds like you have (sorry my wording) no clue about it. You should really read up on it a lot before even considering buying a server a farm. Start small... Are you aiming to do something like a MMO?
Either way you should take a look at this thread it has a lot of valuable information about servers and the like.

#2 I was aiming to MMO and all by myself. I know it seems crazy, but I would like a GOOD challenge even though I do I understand the hurdles of doing this very extreme. If there is a seemingly impossible obstacle to face I will try to figure it out. I try this challenge even if it ultimately ends up with my game stolen hacked and myself having to surrender the challenge, but I would like to try. I do have some game design experience behind me already, but I am just less familiar with Torque and certain aspects of game design.

#3 If you are aiming at a MMO you should definetly have a look at the beforementioned thread, but you should be aware that ootb Torque supports about 64 players. You can expand this support since you have access to the source (this has been done many times before).

One way you could do it, which would be the simplest way is to have support for approx. 100 players per server, and then let each server run a zone. When players then step from one zone to another they will simply transfer from one server to another. To expand to more computers simply execute your server on those computers, and let a masterserver point your clients to the right machine.

Konrad Kiss made an excellent blog about how he had planned to do it for his game Xenocell this blog


Shameless self-advertising
WinterLeaf Entertainment is currently working on a MMO kit you should have a look out for future updates on that.

#4 I get quite irritated by the fact that every time somebody mentions MMO they get to sent to that thread and people dont answer the questions asked.

For a start that thread is barely applicable to MMO development, it makes unreasonable and unfair assumptions on that score, and most of the rest of it should be read by everybody who wants to make any kind of game, theres maybe half a dozen comments that are worth reading, many of the rest are condescending imo.


anyway to get to the question in hand, the original question is similar to the now infamous 'how long is a piece of string' question.

If we take an MMO game server as a concept this consists of both hardware and software servers. you have or can have several software servers running per hardware server, how many exactly is impossible to know without first benchmarking the software so that you know how much each software server costs as far as memory, CPU and bandwidth, i'm sure you can look up some examples on the net somewhere, as this question will have been asked and answered elsewhere, as well as studies done on specific hardware and software configurations of servers.

So having explained all that, the simple answer to the question is no, initially you wont need to add hardware servers for testing and development, however you should design the server components in a modular way so that they can be distributed across multiple hardware servers based on what functionality you want, for example you might want a large central city where many players will converge at one time, it would make logical sense for this to have its own hardware server, while if you have a 7 or 8 dungeons say that only hold 7-8 people then these could all be hosted on one hardware server. n top of that you will want chat servers, login servers database servers. ll could be run on one PC initially (decent CPU and a good stock of ram will help you here).

there is one gotcha i will point out if you are using T3D and that is most of the master servers available at the minute tend not to play nice with the game and server being run on the same PC, the solution being to modify the master server of virtualise the master server, or use one of the online master servers tho i do recommend one of the first two solutions really.

#5 @Bloodknight, you know that thread is the most valuable resource for MMO development, a lot of it applies to general game development as well because.. You know it's still a game even tho it is a MMO.

Also I did post a link to Konrad Kiss' server solution as well which is a short and very good example of how to set up a server which can host MMO's so I didn't "just" link the MMO thread, I actually answered the question.

#6 Oh heres another one, more in depth.

#7 i wasn't just specifically picking on you, its the continuing condescending nature of these forums and this community towards anybody who is interested in making that particular game type.

Ive read the thread, it should be read by everyone who comes here regardless of what game they are making, perhaps as an induction trial or something :p

#8 A calculator, or a watch can be a computer. Doesnt mean people can play games on them en mass. In Your mind, Your thinking computer ie lap-top, and desktop. Technically, yes you can hook up a few together bare bones, and accomplish something on a smaller scale for development or if your audience is going to be just a few friends. If your looking to go Huge MMO Your looking at a server farm, that would handle clusters. Within development aspect or small scale, pretty much depends on Your computer. For instance back in the day I ran a live private server for a game, and I was able to only get 50 players on per computer comfortably as my load increased up to 100 players things would come to a stand still.

Id like to add some other things that I didnt see mentioned above, that I did come across from reading different Net articles/topics.

Running a huge MMO is expensive, when dealing with server farms. Im not saying its not possible. Runescape started out by one person If I remember right with 3 zones. Players were charged a small fee of $3.00 per month. The player base was probably like 500-1k players for the first few months. The owner Andrew, was a one man team to begin with. He did his own programing, networking, and art. But as his game grew he had to upgrade, and hire more staff to compete with the huge player base. I really personally idolize this guy for his accomplishments.


The other thing is just not hardware, its connection type, ISP provider, the type of line's you will be using. You can only get so much information through one line, and you also have bandwidth up, and down to consider. ISP Providers dont like massive bandwidths on low level lines for example, they want their cut of the pie so to speak. Theyll throttle your connection after some time if You try. So you have to purchase an ISP business account, with probably a T1 line, and theyll add/increase more bandwidth for extra $ amounts as you need.

Another thing to keep in mind is theres a build difference in mind from regular computers, and server farms, routers, ISP connection types while dealing with internal software. Basically what Im saying is they are made differently apart from each other with different intentions.Some are made just for home use. Take a home windows O/S computer vs a window server O/S computer for example. A home computer which is what I use is not made for large server type activity, but with some tweaking it can be used as such, but not on a large scale. The way sent, and received packet information, handling is handled is different. The companies know this, and this is once again so they can get their cut of the pie. Tweaking this stuff is a huge headache, Ive done it trust me.

So honestly alot Depends on How big You want the game to be, how many players,how much you are willing to spend on servers, bandwidth, and maintenance. A couple computers (maybe 3) can get you started in development, but will not be the final end result when it comes to open beta or release.

#9

Quote:... its the continuing condescending nature of these forums and this community towards anybody who is interested in making that particular game type.

I actually find that many people come here (any game engine forum, really) wanting to make games with no idea what a game is. Instead of picking up an introductory game development book or reading through forums to gain some understanding of what is involved, most people just jump in and start asking the same question that we've all answered a thousand times over - and that they would have found if they'd have just looked. I am not "condescending," I am tired. Oddly enough, the best way to learn about something is to research it - Google is far more patient than any human when it comes to answering the same question time and time again. I know, I'm grumpy, arrogant, crotchety and just plain rude (and I've been called much worse over the years). I got that way from being used as Google far too often.

Need input on solving a problem? Sure. Need reference material that is available via a library or a web search? Go there and do that first; if you still have questions I'm game. If you don't put in the effort why should I?

#10

Quote:I actually find that many people come here (any game engine forum, really) wanting to make games with no idea what a game is.

I actually have designed my own game before it was not like a small one but a pretty decent one. The thing is thought that it was a non multiplayer 2D free game that I used making a game engine that makes 2D game design easy. So I was trying to take next big step up. I resisted urge to make an MMo back then. I do not even know for sure if I going to go exactly that far as to MMo, but I would like to try. The question I asked here might seem really dumb, but I asked this question not just for question it self but to see what else I might learn.

#11 Sorry - sort of derailed the thread. I wasn't griping about this thread directly, but sort of indirectly. There are two books that address the basic sketch outline of building MMOs (and they're probably not really that good) -

http://www.amazon.com/Massively-Multiplayer-Development-Charles-River/dp/1584502436

http://www.amazon.com/Massively-Multiplayer-Game-Development-Charles/dp/1584503904

I own and have read them both, even though I don't have any plans to make an MMO.

My point is, there are a million ways to get this information - almost all of them are better than asking on any particular game engine's forums (unless that game engine is a turnkey MMO solution).

For the record, I made a Lunar Lander clone on the Apple ][+ about 30 years ago - doesn't mean I know much about large scale game production, especially to the tune of making something like a MMO. That's why I read extensively....