Is Torque scaleable?
by Mike Stoddart · in Torque Game Engine · 10/12/2002 (9:33 pm) · 4 replies
I know that there is a lot negative press on these forums regarding MMPOGs, but hopefully I can get some honest opinions on the subject at hand.
Can the Torque engine (obviously with a lot of work) scale to be used for a massively multi-player online game? I'm talking of a persistent world, with a database backend that can contain thousands of players in a seamless world, with no pauses for loading data?
I know the obvious answer is that yes you could probably make Torque do anything, but do you think it's really possible?
Thanks
Can the Torque engine (obviously with a lot of work) scale to be used for a massively multi-player online game? I'm talking of a persistent world, with a database backend that can contain thousands of players in a seamless world, with no pauses for loading data?
I know the obvious answer is that yes you could probably make Torque do anything, but do you think it's really possible?
Thanks
#2
In this case, are you really gaining anything by using Torque? Are you modifying and/or removing so much code that you may as well not use Torque?
Thanks
10/14/2002 (8:59 pm)
Quote:I think that with a dedicated team, you could get an MMOPG up with Torque. You'd probably have to modify most heavily the networking layer, and do some revamping of the object management code.
In this case, are you really gaining anything by using Torque? Are you modifying and/or removing so much code that you may as well not use Torque?
Thanks
#3
Torque has some of the most effcient network cdoe for transfering object state over the wire that I have ever seenm, and I have been doing high volume transactional systems for more an a few years now.
The network layer is the MAIN reason I selected Torque for my development.
Once I get the ShapeBase stuff cleaned up and rework the player controlled classes, I plan on the "easy" part which is enabling federated clustering of the Torque engine.
Which means multiple servers will be able to participate in a federated cluster to support massiving battleground / play areas accross machines.
Imagine an entire city environment and as you moved thru the city you were seemlessly transfered from server to server.
Or imagine you were in the middle of an long running clan match and the server crashed and you never knew it, because you were automatically re-routed to another machine in a cluster that seemless picked up the game play .
I have written more than a few application servers that do just these things, but with business transactions instead of game transactions, the principals are EXACTLY the same, the contents of the data are only different, e-commerce or b2b transactions or game events it does not matter.
I am planning on useing the Torque network code as a stepping stone for this kind of robustness.
10/14/2002 (11:51 pm)
In almost all cases, starting from a blank page is harder than not.Torque has some of the most effcient network cdoe for transfering object state over the wire that I have ever seenm, and I have been doing high volume transactional systems for more an a few years now.
The network layer is the MAIN reason I selected Torque for my development.
Once I get the ShapeBase stuff cleaned up and rework the player controlled classes, I plan on the "easy" part which is enabling federated clustering of the Torque engine.
Which means multiple servers will be able to participate in a federated cluster to support massiving battleground / play areas accross machines.
Imagine an entire city environment and as you moved thru the city you were seemlessly transfered from server to server.
Or imagine you were in the middle of an long running clan match and the server crashed and you never knew it, because you were automatically re-routed to another machine in a cluster that seemless picked up the game play .
I have written more than a few application servers that do just these things, but with business transactions instead of game transactions, the principals are EXACTLY the same, the contents of the data are only different, e-commerce or b2b transactions or game events it does not matter.
I am planning on useing the Torque network code as a stepping stone for this kind of robustness.
#4
I don;t know that it's so much a factor of whther Torque itself is scalable, it's the architecture you choose to deploy to support your MMO. I think picking the right db system will have a greater affect on your scalability question than TGE itself, in it's current state.
Lots of threads from about a year ago that start poking at possible architecture, that thread is just one, seach the same forum around the same time for more. I think Jarrod and I may have talked briefly about this already, but I agree, the principals applied to business problems are the same as those required here.
10/15/2002 (6:06 am)
http://www.garagegames.com/index.php?sec=mg&mod=forums&page=result.thread&qt=539I don;t know that it's so much a factor of whther Torque itself is scalable, it's the architecture you choose to deploy to support your MMO. I think picking the right db system will have a greater affect on your scalability question than TGE itself, in it's current state.
Lots of threads from about a year ago that start poking at possible architecture, that thread is just one, seach the same forum around the same time for more. I think Jarrod and I may have talked briefly about this already, but I agree, the principals applied to business problems are the same as those required here.
Associate Kyle Carter
The source of all the "bad press" here is because most of the people who want to do MMOPGs don't really understand what they're getting into - they just want to do an MMOPG cuz that's the latest thing.