Firing up the MMOG pack - Part 1
by David Dougher · in Torque Game Engine · 01/03/2005 (7:28 am) · 24 replies
OK. Things on Adellion have calmed for a bit and in the lull I thought I would turn my attention to the MMOG Kit that I was discussing before. I have at least a tentative go ahead from HonourBound Ltd. to share some of the tech we have done, but their request is that we build from a scratch 1.3 base to avoid any possible license complications. (Adellion uses just about every Pack GG, Bravetee, Synapse and Tim Aste made.)
So initially we will have to roll back to the original gorpe code and bring it up to 1.3 standards. We will also need to isolate features that some developers may or may not want. For example, Gorpe includes support for both python (Thanks Josh Ritter) and XML. It had an optional publisher and subscriber system as well to speed responses.
So, my thought would be to take each major piece and turn it into an add on that the developer can choose to combine with a basic 1.3 system or not.
This would include the terrain paging system, since it is possible to make a perfectly good zone based MMOG system without the terrain paging system. A hybrid zoned multiterrain system might also be desireable for many people.
The first thing to realize is that such a system will involve a lot more code than the current game has. MOst of this code will actually be repeated code - taken from the original engine and reused to make a server/service with a specific function.
Adellion was created by analyzing the way that information flows through an MMOG and then dividing it into those operations which must be realtime and those for which realtime response is not required.
For example, it is not critical when sorting through your inventory to make it a real-time operation. It must happen, it must be a reliable transaction, but it does not need to occur in real-time the same way that say, a jump across a chasm needs to work.
Some game operations need not even be part of the "game" at all. For example, character creation does not need to occur on the game server. The character can be created and stored to the database. When the player starts the game the newly created character is retrieved from the database and introduced to the game world.
Furthermore, the volume of transactions for some MMO operations change over time. When you first open your game world much more time will be spent by players in Character Creation and playing the game than in Guild Chat or Inventory. As the game world gets rolling people spend more time considering the objects they own (or would like to possess or make) and conversing with guild members. Also, consider the volume of data that moves when you announce your latest monthly update. There is a significant uptick on the load that would be placed on an update server.
So initially we will have to roll back to the original gorpe code and bring it up to 1.3 standards. We will also need to isolate features that some developers may or may not want. For example, Gorpe includes support for both python (Thanks Josh Ritter) and XML. It had an optional publisher and subscriber system as well to speed responses.
So, my thought would be to take each major piece and turn it into an add on that the developer can choose to combine with a basic 1.3 system or not.
This would include the terrain paging system, since it is possible to make a perfectly good zone based MMOG system without the terrain paging system. A hybrid zoned multiterrain system might also be desireable for many people.
The first thing to realize is that such a system will involve a lot more code than the current game has. MOst of this code will actually be repeated code - taken from the original engine and reused to make a server/service with a specific function.
Adellion was created by analyzing the way that information flows through an MMOG and then dividing it into those operations which must be realtime and those for which realtime response is not required.
For example, it is not critical when sorting through your inventory to make it a real-time operation. It must happen, it must be a reliable transaction, but it does not need to occur in real-time the same way that say, a jump across a chasm needs to work.
Some game operations need not even be part of the "game" at all. For example, character creation does not need to occur on the game server. The character can be created and stored to the database. When the player starts the game the newly created character is retrieved from the database and introduced to the game world.
Furthermore, the volume of transactions for some MMO operations change over time. When you first open your game world much more time will be spent by players in Character Creation and playing the game than in Guild Chat or Inventory. As the game world gets rolling people spend more time considering the objects they own (or would like to possess or make) and conversing with guild members. Also, consider the volume of data that moves when you announce your latest monthly update. There is a significant uptick on the load that would be placed on an update server.
About the author
Owner - Pariah Games, Adjunct Professor - Bristol Community College, Mentor - Game Design - Met School Newport, Mentor - Game Design - Met School Providence
#22
03/06/2006 (1:47 pm)
I'm pretty interrested in this. Can you give any dates and possible pricing info?
#23
03/06/2006 (2:20 pm)
As I stated above - at this point in time the project is not making any progress, although Adellion is. So, any timetable, or estimate of a selling price would be putting the cart WAY beore the horse. Sorry.
#24
www.mydreamrpg.com
07/15/2006 (9:31 pm)
Looks like there is a product that is something like what we're looking for... costs $100, tho.www.mydreamrpg.com
Torque Owner David Dougher
Pariah Games
Although progress on Adellion (and the underlying technology we planned to use for this project) continues at a pretty steady pace, I am sorry to say that almost nothing has been done specifically on this since it was first conceived. We had a lot of initial interest and more than a few offers from people who wanted to contribute but we stalled soon after starting.
The fact is, GarageGames has been far too busy to devote any of the promised resources to assisting us in this project, and the situation does not look like it will change any time soon. It might pull a phoenix after the next GDC, but barring that, I think the project should be considered dormant, if not extinguished. This is not to condemn GarageGames in any way. They have a very big plate and a lot of demands and this one simply does not merit their attention enough to divert resources from other projects they consider more critical.
Other paths have opened up that you might want to explore. Minions of Mirth and myDreamRPG both offer support for the MMO path through the torque engine and may fit the bill for what you are after. TSE remains the better solution for large terrain worlds.
And of course Adellion iself continues and is always looking for contributors in the programming and art space. Persons who join and are TGE/TSE owners get full access to the Adellion code and may use it in their own projects as long as they inidcate when they join that this is their intention and they do not directly compete in the narrow niche we are aiming for.