yep so tasks to do ->
hmm get code unreal engine online system plugin version. (of would be titled planet3954 online system plugin) which would use webrtc and it would have a counterpart node.js based webrtc server with some session setup mechanism. yep then such plugin would be compiled for unreal specific version.
so agenda is to use stun instead of relay. (since we dont want to add additional costs) and some geographically distributed nodejs instances with no db simple server with ephemeral cache of game players connected to relevant nodejs servers just only to have some signalling protocol of webrtc + some additional session layer logic. or maybe i might use some nondb having redis and some microservice and redis in server instances. (for session discovery and session establishment. e..g player with score 100 and battle pass level passed could join to some new game version session alike p2p specialization of game versions. so battle pass logic would be alike -> game is free and game would be freely downloadable but the updated new game versions would be able to play with such game versions. so this would enable in tandem not requiring dedicated servers. but rather still enabling p2p but where each p2p could have new game versions. so that new game versions could heterogeniusly distribute to players wihout effecting strict version updates but that rather a mixture of multiple game versions at the same time alike.
so i at last figured out a viable design for how to design battle pass concept for game.
hmm so would add some logics alike micro service for service discovery but with oauth jwt tokens to be verified. i mean e.g. user logins from game window to game platform. and receives jwt. then user communicates with jwt to microservice distributed (protocol that also enables webrtc's signalling establishment (e.g. nodejs based webrtc servers alike) and also stores the user in a centric distributed multiple servers and the web service would then both enable signalling protocol handshake and both session logic on over it. )
so the micro service first would verify the jwt received from the game platform from there. then would if correct register user game version/session info in either redis or cache of the server. (since its ephemeral thats only valid when user is logged on ) (e.g. webrtc stun server based discovery information which is ephemeral)
then wishfully i complete this design tasks.
then afterwards i would retest multiplayer with this design from also testing from vagon multiplayer instances alike.
then remaining tasks would be game additions. e.g. adding game environments. new textures/meshes to battle mars rovers. adding new turrets etc. so there we would have time for learning some more graphics design to better design style of battling mars rovers. e.g. more realistic vfx additions or more realistic pbr textures etc.
so after all initial game design challenges are finished then the following design challenge starts. where artistic design challenges would be tackled fixed.
so lets first finish the webrtc architecture in today and tomorrow. then test multiplayer on vagon. and then fix a remaining defect (There is one important defect issue I detected when testing on vagon (the player controller is floating pawn controller and not the gameplay controller strange. somewhat issue happens and user cant play afterwards so a very severe defect type in jira side) ) then after that is all completed of such tasks then to start implementing a battle pass level (where score 100 is required to be able to enter to) and then also some much more artistic mars rovers etc.
yep now time to code the plugin first. and then test with node.js.
yep.
Yorumlar
Yorum Gönder