Alpvax

If you are really desperate to share your resource packs, I believe that using symbolic links should work.

These are 2 different tags ("minecraft:carnivore" and "mymod:carnivore" respectively). You should make 1 tag "mymod:carnivore" and add all the defaults to it (as in all the vanilla items, and your custom items if applicable), then if you want to add items in another mod, you also define the "mymod:carnivore" tag (not the "mysecondmod:carnivore", again that would be a different tag). As an extension on what draco said, the only time you should put tags in the minecraft domain is if you are overriding/adding to an existing vanilla tag.

That is not valid json. Currently you have an array with keys. Either it should be an array of objects "[ { ..." or it should be an object with keys of "one, two,...". I would suggest going for the first option (especially as that is what you support in your code):

Worked around the issue by double unchecked casting the type: public abstract class IOType<T> extends ForgeRegistryEntry<IOType<?>> { @SuppressWarnings("unchecked") public static final Class<IOType<?>> CLASS_GENERIC = (Class<IOType<?>>)((Class<?>)IOType.class); } //Can then be used as follows public static final DeferredRegister<IOType<?>> IO_TYPES = DeferredRegister.create(IOType.CLASS_GENERIC, MODID); By putting the constant in the API class, it can be used by others who are consuming the API.

DeferredRegister has been updated recently to make it work with custom registries, but it still doesn't work with custom generic RegistryObjects (similar to TileEntityType<T>). When trying to create the DeferredRegister, it errors with: //Registry public static final DeferredRegister<IOType<?>> IO_TYPES = DeferredRegister.create(IOType.class, MODID); //IOType.java (currently just a stub) public abstract class IOType<T> extends ForgeRegistryEntry<IOType<?>> { } Is there a workaround for this issue, or can I just not use DeferredRegister? This will be available in an API, so I would ideally be able to support the more popular DeferredRegister approach. Removing the parametised type (DeferredRegister<IOType> IO_TYPES = ...) then throws the error of IOType not implementing IForgeRegistryEntry<IOType>, so that isn't a fix. I am wondering if the only workaround is to make the registry in a static initialiser (and overriding it in the NewRegistryEvent), but that seems like a step backwards. Does anyone have any suggestions?

I believe that the last time I looked into doing something similar, I decided to make my HashMaps per-chunk to prevent this issue.

What are you actually trying to achieve for the players? If you want a slime block, register a new SlimeBlock.

When I was starting to use DataGenerators, I found that Lex's mod CobbleForDays helped me understand.

I don't agree with that logic. Yes, 1 bucket of a fluid is 1 block of the fluid (although it actually doesn't quite reach the top, so you could argue that is not true), but there is nothing saying 1 solid block is equivalent to 1 bucket of that material melted down. Liquid doesn't have ingots or nuggets, so those solid conversions do not need to apply to them. If we keep it as 1 block of a solid becomes a bucket-and-a-bit, all it means is that you don't fit an even number of blocks into your tanks. But I don't see why that is an issue, because once you have converted it to a fluid, whatever you're using it for (machines?) probably doesn't care about the solid amount. If you want to transfer exact amounts, use the blocks. They are probably easier to transport anyway (there is a vanilla mechanic for it, and you can usually carry 64 per inventory slot as opposed to the 1 in a bucket).

I don't see how that is true at all. Why does 1 block have to fit in 1 bucket? If we just accept that 1 block is 1296mb, everything works. I don't see why we have to add the restriction that 1 bucket is 1 (solid) block. We can just accept that melting it makes it less dense, therefore bigger.

Bear in mind that more players use metric systems than imperial (according to minecraft server statistics, and the fact that there are few countries which still use the imperial system). It is far easier for users to understand that half a bucket is 500mb, than it is for them to see 648 "atoms" (If that is the approach, I would prefer to use something like "voxellites" which has no real-world bearing than "atoms" which could cause huge confusion for users of chemical mods). And to the next argument that you would just display it as "3 1/2 buckets" or 3 1/1296 buckets", that is hugely unnatural for me (and presumably many others). I'm not sure what the issue with 144mb/ingot is. Why does a melted compound have to fit in the same space as a solid one? It doesn't in real life. And as far as cauldrons, I think it's fair enough to just lose 1mb in a cauldron. Who cares about 1/1000 of a bucket of an infinite resource? And for those mods which have finite supplies of water, they usually add their own mechanisms for collecting/storing it, so the cauldron issue is moot.

Do you just mean with spawn eggs? Or do commands not work either? Vanilla spawn eggs (I believe) have the entity set when they are created, so if you override the entity, I think you also need to override the spawn egg to make it work. Disclaimer: I've not actually done it myself, but I believe that is the case.

I don't believe that is the case. Look at the subclasses of JSONReloadListener (recipes, loot tables, advancements etc.). You don't care which files exist, just try to load all of them.

You create a new (empty) set, then loop through it. Twice. Neither of your loops will do anything, as they are both always going to be empty when the loop is run. I would suggest including your modid in the path, to stop there being conflicts. I would also not limit it to "supported" mods. That way support for third party mods can be added by datapacks (or even included in those mods if your mod becomes popular enough). I would recommend using a path such as data/<datapackid>/<your mod id>/entities/<modid>/<mobname>.json That would mean that your zombie definition would be at data/<your mod id>/<your mod id>/entities/minecraft/zombie.json It means your modid is duplicated, but there is no risk of another mod using the entities path and causing your mod to crash.

As the Optional annotation has now disappeared in favour of capabilities, there is now no way to add optional features to non-TileEntity blocks. An example use case would be proper fluid handling for vanilla cauldrons, or handling of different blocks' "heat" values. It would have to return a function (BlockState, IWorldReader) -> custom capability instance. And couldn't be cached without more work. The capability would also not be able to be serialisable to NBT (obviously non-TEs don't have NBT). It would effectively be a mapping of blockstate -> capability. If we restricted it to readonly capabilities there would be no issues with modders caching the capability and modifying something they shouldn't, but it would for example allow extraction from cauldrons, but not insertion. The optimum solution would be for the capabilities to be cached when first accessed (Probably by calling World/Chunk#getBlockCapability(Capability, BlockPos)) which would then be returned every time afterwards (until the position was unloaded). It would allow ModB to support ModA's X system, without having to inherit from an API which may not exist at runtime. This post is intended as the starting point of a discussion, if it is an idea that no-one is interested in, that is fine. The alternative is to just ship ModA's API with ModB, or at least the interfaces you are implementing in your block, but that has no way of adding capabilities to blocks which you don't own (e.g. vanilla cauldrons, torches).