Jake_28

Tbh I know nothing of programing.

unity is the most easy tool for those things

If i ever decide to do it, heres some ideas i came up with.

Players would be stored on the server host’s computer as (zip?) files (which have a maximum size to prevent exploitation). They would contain model data, probably Modelbench or Blockbench, and keyframe support. You would be able to script animations, variable changes, etc to happen with events like button presses, interations with scenes (I’ll get into that when I talk about scenery), etc. 

AFTER SOME MORE TYPING: I’ve realized it would be more efficient to store player script files on the client’s computer (but keep account, model, texture, and keyframe files on the host’s), because it would split up the processing power over multiple users rather than just the server host. I can keep sending rudimentary commands like for drawing models and updating animation frames to the server, but the code behind them can be kept on the client side. pls keep this in mind while reading. I have modified some specific parts relating to this in further texts.

So you’d script a character, animate it, texture it, and model it, and then and it would be able to receive inputs from the server and the client.

There would be a player editor, and a base player zip, that has movement, support for 1.8 skins, and some basic idle and walk animations. In the editor you can test your character scripts in a test environment which acts like a server, but offline and with real time editing. Scripts receive inputs, and update their data accordingly, for example a movement script would receive WASD and update its position variable, and then in the main script it would (kindly) tell the server “DRAW MODEL AT POSITION_x POSITION_Y WITH ANIMATION.WALK FRAME.7”.

Models will have a (file size/ part amount) limit, and a script can ask the server to draw (arbitrary number) of models on each frame. The amount can be adjusted in server settings to help counter lag with more or less people?

Account data is stored on the host’s computer. You type in the server adress (maybe somehow have the server host input an adress, encrypt it in some obscure way, and the decrypt the ip after it is input at the end user’s computer? But maybe it is secure enough already? Plus the user can probably just monitor the connection for the ip?) and then you input your username (check if the server already has someone with that name), and password (HUGE RED WARNING TO NOT USE THE SAME PASSWORD AS EVERYWHERE ELSE SINCE IT IS VISIBLE IN PLAIN TEXT ON THE HOST’S MACHINE? Or i could salt it, hash it, and store it like that but idk how) (maybe just randomly generated and store it on th user’s computer?) and then every time you want to log back in, you input the same stuff and you have your username.

There would be 2 chats: RP and OOC (out of character). (Or just make OOC the same but different color?) and you can type. Yep.

Maybe add recording and replaying with a moveable keyframable camera and then export images and use  FFMPEG to export video? And add chat bubbles for in character chat, so words appear above heads? Save chat as text for a backup?

Now time for: SCENERY!

Scenery can be submitted by users to the server host to add to the world map. Scenery is made in a similar editor to the character editor, and is composed of Objects and Structure. Structures are just structure block exports that you can position. Objects are able to be interacted with via scripting, and can activate those scenery events i mentioned earlier. For example you could have an object send a packet to a user saying “LINK.PLAYER TYPE:TYPE” here player can be replaced with any player submitted “player” (which can.. also be objects? You could model a carcand submit it as a “player”) and if the “player” can comprehend it (if it has handling code) it can do anything. For example you could make the player just display a simple holding animation with the model in their hands unless the TYPE is a.. for example gold sword, in which case they could make it levitate on their head. 

The world map is composed of scenery which can be travelled to “by foot” (by moving eg with WASD) or by using the minimap. Scenery fits together like a puzzle, but you have to make the transitions yourself. If there is a mountain and then a cut and a flat plains, you’d have to make a scenery to link them together yourself.

The world make would start as an ocean. Players would spawn on a designated spawn object. Eg a boat. And if it is destroyed, they would spawn somewhere else.

Speaking of objects, i should probably not limit them to scenes. Objects can be placed in the map and linked to players, or players linked to them. (The boat!). Players can edit variables by interaction via scripts (The boat might have a steering wheel!).

Basically rpg

Oh and it’s all top down like in minecraft dungeons.