New answers tagged

0

Not really an expert on the subject but after reading your question it gave me an idea, why not use a static IP to host the server which you then set the limit to the maximum player size(16). Similar to how you would host a private Agar.io server(https://www.youtube.com/watch?v=2NSBKKj8kqg). Probably sounds stupid but just trying to help you see it from a ...


2

Is this a reasonable idea? Yes What are the possible drawbacks? Packet Loss, more code complexity, another connection to manage == more chance for disconnects, time outs, exceptions, whatever ... Are there better ways to handle this? Use an existing Reliable UDP Library. Two of the most popular are: Lidgren Network (C#), RakNet (C++). From ...


4

It results in packet loss for UDP due to contention between the two protocols - remember that UDP is not guaranteed delivery, while TCP is. More TCP packets will get through while UDP suffers - TCP induces UDP packet loss. There has also been the (historical) idea that router infrastructure favours TCP over UDP, though I doubt that is still true by this late ...


0

I do not how much do you know about Photon networking. Briefly, you have to put the unzipped Photon Server SDK folder to server. After that you have to create a server as a .dll application and but the dll file to deploy folder which is inside the SDK folder. Both server and client side should implement ClientPeer classes for communication. Using events, ...


0

There are for sure more that one way to achieve this. I am not familiar with smartfox server's setup to help you on that regard & there is no point on writing about all the possible routes you could take since this topic may result in a huge discussion. No matter what way of networking setup you do you should have to 'simulate' that destruction on the ...


0

On the most basic level, if your destruction algorithm is deterministic then the server can notify each client and the result should be the same. The more complicated manner is that this system is notify independent of your other systems. It will probably be best for you to handle destruction in the same way. E.g., if you are calculating ...


0

I am guessing that in your case, clients are sending all their shots to the server, and the server is broadcasting them back to clients. Basically, your clients should simulate the physics (animations and sounds as well) based on objective information coming from the server. This data can be differently structured: One approach is sending data about who ...


1

It depends on how your software architecture looks. When you have a clear separation between user input, ai input, game mechanics and graphic engine, then you should be able to simply switch out the ai input with a second user input. But when you have tight coupling between the AI code and the other components of your game, then it might be quite a lot of ...


0

It's a lot harder than you think (Judging by your words), plus being able to create a game doesn't mean you are able to handle multiplayer gameplay. I don't know about specific engines, some might help you with online gameplay, but talking just about code: You need a server (or servers) (Even if the game is based on 1v1 gameplay, you need a place to store ...


2

You probably do want to communicate to this information to the client so that they are able to view it. You can treat the client as a dummy terminal though with sparkly representation and have a "neutral" server as the authority. Lets consider League of Legends in this context. At the beginning of the game, each client connects to Riots servers. Every ...


3

I wouldn't store them on clients, or even ship that information with game. If security is your main concern, fetch all informations required at the beginning of the game (from a database), store them temporarily, and delete them after the game session ends. You can also cross-check local data with database during gameplay if you want an extra layer of ...



Top 50 recent answers are included