Dedicated Server?
Moderator: Moderators
Dedicated Server?
Has this been suggested before? 'All' it would need is a modified TASclient to handle, maybe, player voting and other such automation. And maybe a spring executable that didn't bother to do useless things like render. Just a thought I had last night.
Well here's a nice enticemant to make people want to make it more. If you make a dedicated server, Game Arena will be happy to host a server or two for us. If we get some servers going there, the we will definately get more players.
http://cvs.sourceforge.net/viewcvs.py/s ... iew=markup
Start there....
That is the netcode for the client/server relationship of the lobby.
Beyond that, the netcode for the game is subject to change right about now. For example, the current goal is to get the game so that it can recover from desync (this might require a major rewrite of the netcode, but it could make it faster also).
EDIT: Also, at the end of every command you send you must put a \n character! This took me a couple of days to figure out, and I was literally pulling my hair out by the end until I talked to Betalord about it.
Every command you recieve will also have a \n character at the end of it
Start there....
That is the netcode for the client/server relationship of the lobby.
Beyond that, the netcode for the game is subject to change right about now. For example, the current goal is to get the game so that it can recover from desync (this might require a major rewrite of the netcode, but it could make it faster also).
EDIT: Also, at the end of every command you send you must put a \n character! This took me a couple of days to figure out, and I was literally pulling my hair out by the end until I talked to Betalord about it.
Every command you recieve will also have a \n character at the end of it
While we're on netcode. The AI itnerface function to give a unti orders creates a mass of netcode and sends ti off tot he host.
This function doesnt always work reliably though. For example if I tell a unti to change tis move state ro start using high trajectory weapons, the command is ignored totally.
This function doesnt always work reliably though. For example if I tell a unti to change tis move state ro start using high trajectory weapons, the command is ignored totally.
by dedicated server, i was thinking for the actual battle, like you would use for counter strike, or bf2. The docs you linked to appear to be just for the lobby client/server netcode.
Starting with a simple proxy model you could start to add features like prediciton, and stats. It seems like there are quite a few players who are on dialup and like the sync nature of TA, but it would be nice for broadband optimized server option.
-jess
Starting with a simple proxy model you could start to add features like prediciton, and stats. It seems like there are quite a few players who are on dialup and like the sync nature of TA, but it would be nice for broadband optimized server option.
-jess
greenail wrote:by dedicated server, i was thinking for the actual battle, like you would use for counter strike, or bf2. The docs you linked to appear to be just for the lobby client/server netcode.
Starting with a simple proxy model you could start to add features like prediciton, and stats. It seems like there are quite a few players who are on dialup and like the sync nature of TA, but it would be nice for broadband optimized server option.
-jess
Well then, this is the code for you my friend. Download all of it and jump in! Try not to get too lost.

http://cvs.sourceforge.net/viewcvs.py/s ... pring/rts/
I don't think there is any netcode documentation for the current game protocol unfortunately.
http://developer.berlios.de/projects/taspring-linux/
The above funnily enough tends to eb ahead of the sourceforge cvs usually. In tiem widnwos and linux will both draw release versions from there and the sourceforge cvs will not be updated regularly.
That and the linux people work with that build aswell as zaphod whereas they ahve to wait for zaphod to move changes in the sourceforge CVS over.
The above funnily enough tends to eb ahead of the sourceforge cvs usually. In tiem widnwos and linux will both draw release versions from there and the sourceforge cvs will not be updated regularly.
That and the linux people work with that build aswell as zaphod whereas they ahve to wait for zaphod to move changes in the sourceforge CVS over.
Ok then ya:
http://svn.berlios.de/wsvn/taspring-linux/trunk/rts/
However...no documentation isn't really a problem atm. The netcode will probably change, and therefore it wouldn't be too efficient to write tons of documenation for obselete netcode.
When the netcode is finalized in a later version, then documentation for the entire project should be written.
http://svn.berlios.de/wsvn/taspring-linux/trunk/rts/
However...no documentation isn't really a problem atm. The netcode will probably change, and therefore it wouldn't be too efficient to write tons of documenation for obselete netcode.

When the netcode is finalized in a later version, then documentation for the entire project should be written.
- SwiftSpear
- Classic Community Lead
- Posts: 7287
- Joined: 12 Aug 2005, 09:29