More Players?

[shanu]

Any chance of making ufo2000 team more than just two players.. eg. two on each side? or even three independant teams battling it out together

[Brick-To-Face]

This was mentioned before, It's not anywhere near the top of the priority list:

1. Major changes to the code (though I'd like to see that part done, so we can have civvies!)

2. People have to wait three times as long to move.

3. People would disconnect twice as frequently.

 

I believe those were the three arguments presented. Maybe there are more.

[Sporb]

This was mentioned before, It's not anywhere near the top of the priority list:

1. Major changes to the code (though I'd like to see that part done, so we can have civvies!)

2. People have to wait three times as long to move.

3. People would disconnect twice as frequently.

 

I believe those were the three arguments presented. Maybe there are more.

 

Thats exactly right.

 

also to add to the reasons that it will probably not happen any time soon: Allowing multiple players to play in one game if done incorrectly would be hugely bad. The load being sent to and from the server would suddenly expant exponentially since all the computers need to know the details of what the other players are doing instead of just the two computers communicating with each other. This could also potentially boost the number of CRC errors by a huge degree since the game needs to keep track of a huge amount of data more than it would with 2 players.

[Brick-To-Face]

...This could also potentially boost the number of CRC errors by a huge degree since the game needs to keep track of a huge amount of data more than it would with 2 players...

Didn't know that, I'll try to get in contact with some network programmers for other games, I have a friend who's an awesome programmer.

[jedandjess]

This was mentioned before, It's not anywhere near the top of the priority list:

1. Major changes to the code (though I'd like to see that part done, so we can have civvies!)

2. People have to wait three times as long to move.

3. People would disconnect twice as frequently.

 

I believe those were the three arguments presented. Maybe there are more.

 

Thats exactly right.

 

also to add to the reasons that it will probably not happen any time soon: Allowing multiple players to play in one game if done incorrectly would be hugely bad. The load being sent to and from the server would suddenly expant exponentially since all the computers need to know the details of what the other players are doing instead of just the two computers communicating with each other. This could also potentially boost the number of CRC errors by a huge degree since the game needs to keep track of a huge amount of data more than it would with 2 players.

 

Surely in a turn based game, the load on the server would be next to nothing even with 10 players, as the players take turns at moving anyway. I can imagine in a 32 player / 64 player game of quake or battlefields what you say is true, but have 1 extra person on a TBS game ... come on! The only way i can see have 4 players enhancing the game is 2 teams of 2 players, each team moving simulataneously.

[Sporb]

i dont know the specifics of server/client interaction but i know this: we get enough CRC errors and such with just two players. More players = more possible things that can go wrong with the interaction = bad Edited September 19, 2006 by Sporb

[Guest lxnt]

CRC errors only exist due to the current engine architecture, where two copies of the game world

are kept in sync on both players' machines. This is called 'peer-to-peer' mode. In this mode, adding more players is quite hard and bandwidth-intensive in theory, and plain insane given current quality of code in the project.

 

Moving to the 'client-server' mode will both get completely rid of CRC errors, as there would be no

multiple copies to synchronize, and allow as many players, observers and AI-controlled entites as

desired. But this move requires complete rewrite of all game logic and internal representation of

the game world, and this also pulls a rewrite of the user interface because it is now tightly coupled

with the logic.

 

Seems like writing a new code base from scratch will be less work than rewriting current one.

 

Any volunteers?

[-Og-]

Ok surely only the network infrastructure of the game would have to be re-writen (not an easy task I grant you, especialy if the code is interweaved with other parts of the code)

 

Or as a quicker more hacky fix, could we not implement a Re-sync process where by if a CRC is detected the game will re-sync by Replacing all the variables on one machine with the ones of the other

 

I dont volunteer, but ill look though the code soon and see what looks to be the best, most viable course of action