## Features targeted for 20.2.0.0-alpha.3 * `[PlutoLib]` Completely redo the ModLoader system * The current implementation is a result of 5 years of feature creep * Large scale API changes, however the general idea should stay the same * Rethink the class loader system. * `[PlutoLib]` Redo the resource system ## Features targeted for 20.2.0.0-alpha.4 * The stage subsystem * A "stage", in this context, is a set of assets bound together by programming logic, not necessarily a game level. Stage switching and asset management are handled by the engine. * Upon stage switch 1. Unload unused assets 2. Load new assets * Provide multiple means of stage switching * Three modes with the initial release 1. Instant switch * Assets will be loaded and unloaded synchronously * The stage switch will happen in one frame 2. Deferred switch * The stage will continue running until all assets load * Assets will load synchronously, but at a slower pace to avoid frame stutter 3. Asynchronous switch * Assets will be loaded in asynchronously, where applicable * Falls back to deferred switching for synchronous loading, such as OpenGL texture upload * Automated asset loading * All asset management will eventually be handled by `PlutoCore` * This includes audio clips, textures, sprites * Add a common interface for all assets * Let the stage system handle audio playback * This API should be as neutral as possible and avoid steering towards game-only use * The stage manager should be relatively low-overhead and allow multiple instances * Allow stages to be inherited from, creating a stack-like structure * `[PlutoAudio]` Integrate the Audio API with the Stage API * Improve upon the support of thread-local Pluto instances * The long term goal is to allow an unlimited amount of Pluto applications at any given time ## Features targeted for 20.2.0.0-alpha.5 * The initial minimal release of `[PlutoCommandParser]`