server/client engine

if your pong game has only 2 players a peer to peer system would be just as easy, in fact almost the same

in my online Pong, i made the server a different program then the client. the server was just a windows console application, and the client had all the graphics and stuff. you can do it whatever way you want. i prefer to make the server seperate, because i feel putting both the client and server in the same program will kind of be overwealming, e.g., you have to look at code and think "hmm, is this for the server or the client?". to me it just confuses my logic when trying to keep them in the same spot.

also, writing it as a client / server model as opposed to p2p is probably a good idea. it will help you prepare for writing a bigger game which you would most likely want to use client / server for.

Quote:

Original post by graveyard filla in my online Pong, i made the server a different program then the client. the server was just a windows console application, and the client had all the graphics and stuff. you can do it whatever way you want. i prefer to make the server seperate, because i feel putting both the client and server in the same program will kind of be overwealming, e.g., you have to look at code and think "hmm, is this for the server or the client?". to me it just confuses my logic when trying to keep them in the same spot. also, writing it as a client / server model as opposed to p2p is probably a good idea. it will help you prepare for writing a bigger game which you would most likely want to use client / server for.

Although im sure that if you were a competant enough programmer and split the code up extensively into its client and server components it would be easily readable. Take the Quake 2 source it has the cpp files split accordingly.

ace

everytime I tried, it's always been a good idea to split the files, entities, systems into two separate client and a server objects. Also, it make it a lot easier if you decide to have your client and server into separate DLLs. keeping everything rolled up into one and separating the paths for both client/server is just adding to the confusion.

Two different engines is a bit far fetch, since there will be a lot of components shared by both clients and server. Just have a good naming convention, then it should be quite easy (Pong_Client_EntityPlayer, Pong_Server_EntityVehicle, Pong_Client_ConnectionState, Pong_Server_GameState...).