Hi guys, first time posting here. I've been using Unity for 6 years, and daily in my work for about 4 years or so. This year I wanted to start a project with some friends, just for fun, nothing serious, and I've been struggling with this problem for a few weeks. I wanted to see your opinions on this. So, the project I'm working on is a survival horror with ps1 graphics (another one, yeah), and I want to make a working persistent system for objects. All interactable objects in my game are considered Entities, when I build the level, I can set an Entity ID (string) and add logic classes to it that will execute when certain things happens, like the player interacting with it. This logic classes are called Processes. I can create any kind of process, like if I interact with a box, one process can show the text "It is a simple box", and then another process can delete the object, or spawn a enemy, etc... In my game I have an invisible trigger that would inherit from Entity as a Trigger class, and it has a list of processes that will execute when the player touch it, like showing a text, disabling an object, spawn a enemy, etc... One of those processes can add this entity's ID to a serializable dictionary in my local data file, the keys of this diccionary are strings corresponding to the name of the current scene (each level will be in a scene), and the values are entities IDs. Then, if I save the game and I enter again in the scene, before the player could even see anything, my LevelController class would check the local data, trying to find if the dictionary has the scene's name that is loading right now as a key, and if it does, iterate throught the values to delete the objects with the same ID, deleting the trigger the user used before saving the game. With this I managed to create some kind of persistence, but problems began when I tried to add pickable and droppable items. See, I have a Database system to create ScriptableObjects that serve as a collection of anything that requires to be reference in runtime, an example is an Item Database where I reference all the Items in the game, so if I need to spawn pistol ammo, I can go to the database manager, search for item "pistol ammo" item in my Item Database, and spawn it (I will upgrade the spawn system with pools after solving the persistence problem). So, if I put an Item Entity on the level (Key01 for example), and the player interacts with it, various procceses start to execute in this order; adding the database ID to the player inventory and the quantity of 1, then adding this entity ID to the dictionary of persistent objects of this scene, and finally it delets itself. For the player, he picked the object from the floor and then it is in his inventory. So, my problem (finally), is that I can't see how to do a system where the player picks up an item and then drops it on the same scene, somewhere else, save, reload the scene and then see this item in the exact same place where he left it. Because until now, the system worked with Entities IDs, but adding the pistol ammo item to the inventory, I lose the Entity ID of this item, I only have the Database ID to reference the item to instantiate it when the player drops it. Dropping a bunch of pistol ammo items on the floor would not add any of them on the dictionary of persistent entities, so if I reload the scene I would lost all the ammo. Do you think I need to have two dictionaries instead of one? Like one for static entities (like the trigger, that will never appear again), and another one for dynamics entities (like the ammo), and use GUIDs when I "spawn" the item entity when I drop the item from the inventory? My idea is that the player can pick ammo from a room, and drop it somewhere else because he needs some inventory space, and maybe come back later to grab it again. I would love to see your opinions on this, thanks for reading all of this, it means a lot.