DreadKyller

checks the 6 adjacent sides of the cable in question and finds any blocks that are of your cable type. If it's of your cable type then get the tile entity at that location via world.getTileEntity(x, y, z) and add it to a list.

The most common are NBT and damage values. Upon using your wand set it to 0 damage, then wit the onUpdate method increase the damage value by 1. The damage bar would be an indicator to the time. When the damage is maxed, the tool is ready again. If you must use the damage for something else, then you can store an integer in the NBT of the item stack.

Yes, this should work, however another option commonly used is as follows: 1: When a cable is placed, check if there are cabled next to it. 2: If no cables are next to it, create a custom TileEntity there. This will be the master tile entity for that system. Don't proceed farther down the logic. 3: If there were cables attached, get the tile entity at that location and find it's master. 4: Create a tile entity at the location of the placed cable, and set it's master to the master of the cables that are attached. 5: If the cables have different masters, we need to combine them into one system. Choose one of the systems and set all the When a cable is broken, you need to separate the cables into separate systems: 1: Iterate through the cables attached to each of the cabled next to the one that was broken, check if they are still connected to their master. If not, set them as the master of their own system, otherwise return and stop doing the next logic. 2: Iterate through the connected cables that were made into their own system and set their master to the new one. To check is a block is connected to the system, get the master of the cable next to the machine and see if that tile entity has certain data. The tile entity would store the following: TileEntity master boolean isMaster int powerInSystem etc... The code would be similar to the following: public void addCable(World w, int x, int y, int z) { // Create your tile entity here and add it to the world ... // Rest of code List<TileEntityPowerSystem> systems= getConnectedWires(w, x, y, z); if(systems.size() == 0) { tileEntity.setIsMaster(true); return; } TileEntityPowerSystem master = systems.get(0).getMaster();// Select the first system as the master system tileEntity.setMaster(master) for(TileEntityPowerSystem system : systems) { if(system != master) { system.getMaster().setMaster(master);// Make the master tile entities of the other systems no longer masters // And set their master to the new master. } } } public int getPower(World w, int x, int y, int z) { TileEntity te = w.getTileEntityAt(x, y, z); if(te instanceof TileEntityPowerSystem) { return ((TileEntityPowerSystem)te).getMaster().getPowerInSystem(); } return 0; } etc... This method requires using more, smaller, tile entities than one single one which holds all the information. In return it makes al operations faster, instead of having the cylce through or iterate through all the connected blocks, you can use stored data to access directly the tile entity that has been put in charge of storing the main data for the system. public TileEntityPowerSystem getMaster() { if(!this.master.isMaster) { this.master = this.master.getMaster();// If the blocks previous master was merged with another system, update it now } return this.master; } I don't have the development environment up atm, but this method probably isn't completely safe with chunk loading/unloading. You may want to store positions for the master instead of the actual tile entity to avoid issues, and maybe have the tile entities store a n instance of a object called System (or whatever) that stores the actual data instead of the tile entity, so that it can be access from anywhere, even if the master is unloaded.

I'm going to assume you wish to have this structure variable sized? If so here's a few steps of the process. Note that this is for explaination purposes, there are more efficient ways of doing parts of this, but as it's a task for a lesson you can try to figure some of that out Do something similar to these when your blocks are placed: 1: Identify the corners of your structure: 1a: Since this seems to be class project related, I won't give exact instructions 1b: determine if the block was placed on a wall of floor/ceiling (using blocks next to it) 1c: Depending, iterate through the blocks negative and positive to find the corner blocks. 1d: For easy later, it's easiest to make the corners the absolute negative corner and the absolute positive corner. 2: Now that you have the corners, check the border blocks: 2a: Start at the negative corner: 2b: For each axis iterate and check the blocks the ensure they're the proper type 2c: Do the same with the positive corner. 3: Check the walls of the cuboid for your type: 3a:you can get a wall by taking 2 of the axis from one point and iterating towards the values of the other point, but always keeping the remaining axis equal to the value of the starting point. 3b: You can also (although less efficient as it cycles through each block) cycle through each block, a block is located on the wall (or floor/ceiling) if any of the coordinates are equal on any axis to the values of either corner, all blocks in the middle will have 0 matching locations (the x will not be equal to either of the point's x location, same for y and z) 4: If these conditions are met, create the tile entity, store the two corners inside the tile entity. 5: On the tile entity update you can perform your operations within the two corners (be careful with how intensive your operation is or how often you set the tile to be ticked, you can cause a lot of lag). 6: You may have issues if your greenhouse crosses multiple chunks, if the chunk the tile entity is in isn't loaded, but the rest ofthe greenhouse is, the greenhouse will not function. If you need it to function regardless you can setup a master-slave system with the tile entities. Create a tile entity on each of the 4 corners on the bottom (or top if you wish) of your greenhouse. Set one as the master, another one as the secondary, another as the third, and the last as the fourth. When these tick they will first check if the tile entity one higher on the master-slave ladder is loaded, if so it will tick that instead, this would continue until it reaches the highest ranked loaded tile entity. On the highest ranked tile entity check if it has already ticked within the last while (however long you have between ticks) to prevent the entity ticking too often. At least one of the corners will be loaded any time any of the greenhouse is loaded (unless the greenhouse is like 32 or something chunks wide, which in that case that's a major problem). 7: You can also choose to, if you wish to make it easier at the cost of some slight memory, you can create a tile entity per chunk in that greenhouse that doesn't tick. Upon reaking a block of your type, you may iterate through the list of tile entites in that chunk, find that one and you may store on that tile entity the locationns of the 4 master-slave tile entities. Based on these locations you can determine with higher efficiency whether your block you broke was part of the greenouse (without having to do the stages in 1 again). If it was, destroy the tile entities, your greenhouse has been deconstructed.

Another thing to add is don't expect every modder to use the standards, even though you should extend ItemSword there are definitely modders out there that do not. You should use what Ernio posted above, but note that if you want your mod to be completely compatible with certain mods, you may need to add specific support for the mods that don't use standard practices.

Can you post the stacktrace so we can see what line is causing the NullPointerException? I'm going to assume however that this line is going to cause problems: for (int i = 0; i < 8; i++) { if (inv.getStackInSlot(i).getUnlocalizedName().contains("sword")) { ... } } inv.getStackInSlot(i) may return null there, and then you're calling null.getUnlocalizedName(). This may be causing your exception. you should add another null check there.

Those methods were added in 1.8 and are not available in 1.7. 1.8 introduced a new system of block models, for which those methods were created. The tutorial you're using (if any) is from 1.8. The RenderItem type you're trying to get the ModelMesher from is the physical renderer for Item Entities. The class that getRenderItem() returns in 1.8 doesn't exist either in 1.7.10. In 1.8 the block model and the textures are set inside the JSON model. pre 1.8 textures were registered via the methods setBlockTextureName and registerBlockIcons within the block class.

Ok, if you're checking the block in the isprovidingWeakPower, then you can just call the notifyBlocksOfNeighborChange within the onNeighborChange. public void onNeighborBlockChange(World world, int x, int y, int z, Block block) { world.notifyBlocksOfNeighborChange(x, y, z, this); }

I edited my last reply, but didn't finish before you viewed.

You should execute that whenever you change the signal that you are outputting, so wherever that is being changed, add that line. In the case of what you're doing, you'd put it in the onNeighborBlockChange. onNeighborBlockChange: 1) Check if there is a block connected to the front face 2) Set the metadata of the block depending on that result 3) call the notifyBlocksOfNeighborChange isProvidingWeakPower: 1) return a boolean based off the metadata that you set in the onNeighborBlockChange 2) You can get that metadata using the IBlockAccess.getBlockMetadata 3) The first 3 integers in the isProvidingWeakPower are the X, Y and Z coords. An example of the metadata could be: 0 ) North - No power 1 ) South - No power 2 ) East - No power 3 ) West - No power 4 ) Up - No power 5 ) Down - No power 6 ) North - Power 7 ) South - Power 8 ) East - Power 9 ) West - Power 10 ) Up - Power 11 ) Down - Power You can get direction (0-5) using meta % 6 You can get power using meta > 5 If you only want North, South, East and Wesk, just make it 0-7, use meta % 4 and meta > 3

world.notifyBlocksOfNeighborChange(int x, int y, int z, Block b) requires the location of the block that changed, it automatically calls the function world.nofityBlockOfNeighborChange(int x, int y, int z, Block b) for all 6 connected blocks. Btw, as of this message, congrats on 64 posts, you're now a full stack!

onNeighborBlockChange is for when a nearby block changes, it is not used to update other blocks when the block is placed. If you look at the redstone torch code you'll see: world.notifyBlocksOfNeighborChange You may need to execute this for the blocks that your redstone device is connecting to to cause those blocks to update and detect that the block is providing them power.

You appear to be registering the block to the GameRegistry inside the constructor for the block. One question I'd like to ask is in your main class are you creating a new instance of the block? If not then the call to the registry is never being executed.

What type of schematic are you referring to? If you're talking about the types of schematics created by MCEdit and WorldEdit and various other external programs? If so it's not really "converted". You can read up on the format here. You may have to create your own parser for the format as not many public tools appear to exist for it. However to aid you, you can find some code that could make it easier to understand here You may also be interested in checking out a related thread If you're referring to converting a schematic to hardcoded java, that's a bit more tricky, although there are some resources that seem to have done it. These are old, but the code should still work with slight modifications, and should at least give you something to work with,

I didn't say anything about the parsing of the JSON being the thing that slows it down, why are you putting words in my mouth? I stated a fact that the same exact model in two different versions of Minecraft, giving massive differences in performance, I didn't blame the JSON parsing, but something isn't as fast when dealing with complex models. I understand that the JSOn is only parsed at the creation of the object and then cached for use later when being rendered, and I havn't modded much with 1.8 so I don't understand exactly how the new system works, as in how the data is physically stored (my guess would be some form of list storing a primitive type that defines the corners texture and vertex UV's of each box) and how that data is then used when rendering (I'm guessing iterating through the list and calling the GL11.glVertex3f and GL11.glTexCoords2f with the data on the current cuboid). Those are only my guesses, having never actually viewed the code. In my experience, every time I've used the 1.8 object format the performance was far worse than the same model in 1.7, you can discuss the efficiency of the method as much as is possible, it won't change the fact that I personally have had issues with it. I said this conversation is getting off topic, I will not reply to this again because it's becoming a stupid argument over different peoples experience with a feature.

Then I'm extremely confused about how updating an old mod of mine from 1.7 to 1.8 using the JSON format the new models (which were fairly complex) would decrease my framerate to 15 fps when over 100 of them were rendered. In 1.7 they dropped my frame rate down by 20 (from 110 average). If they were just as efficient then something else is completely messed up. This is getting off the topic of this post, the person asked a question, I answered with what I know from my experience using both, if my experience with the JSON formatted blocks is atypical then I'm sorry, but this feels like it's becoming an argument, which is not what this topic is about.

Many complex blocks still use hard-coded renderers unless a model is found simply because it's always more efficient to have it code-level rather than interpreted. If the code in Minecraft for 1.8 really uses JSON for every block model by default then Mojang understands nothing about optimization. Even if blocks in 1.8 typically use JSON, if you want your block to look a specific way it's still more efficient to hard-code it. You can add support via a simple if statement if you wish that the model can be modified and have it use the JSON if it exists, but by default using the JSON for everything seems like a really inefficient way of making custom blocks.

Thank you for that explanation. I figured the ThrownPotion was just the registered name, just wasn't sure. As for the command then you for the explanation. I guess that makes sense, I was wondering how it managed to insert only some NBT with omitting other data and not change anything else of the entity, and that is the most obvious way of doing so (not that I'll be using that to achieve my goal)

What JSON are you referring to? there are several places in the code that use JSON. If you're referring to custom block models, I'd be interested to see why you're using JSON instead of a custom block renderer? Since the JSON needs to be parsed, it's be better optimized to do it with a custom renderer.

I tried that, it says it can't resolve 'player'. That means you haven't set the player variable yet. You have an instance of the player somewhere if you plan on adding something to the inventory. If you have experience with Java, the reason the variable can't be resolved should be obvious. you may have it called something else, you weren't being told to use specifically "player", but "player" represents whatever you call your actual player instance.

Ok, that makes more sense. I've never actually worked with a project that set that up, every time I usually click on those it goes to "No linked source could be found for this file", so I never gave it a thought to try it here. I see, thanks for that explanation then. That makes me curious how the entitydata command works, perhaps it calls the readfromNBT function again. Ok. this actually also means theoretically I could use reflection to change the ItemStack info (although I won't do that, I'll just set it at creation of the potion).

Ok, so I have a script I made in batch to create new workspaces for projects, and in that file I use gradlew setupDecompWorkspace --refresh-dependencies gradlew eclipse It ends up i was already using decomp workspace, not dev, and it did not set up a workspace that had the source code. The resulting eclipse workspace contains a src folder with only a simple example mod that prints out the unlocalized name of dirt, and none of the code. To compile against it contains a referenced library called forgeSrc which is a jar containing only empty methods to allow the IDE to compile. Where would I find the source code? Did you mean setupDevWorkspace would generate the source code instead, did you say them backwards? EDIT! apparently the source code is obtainable in "\build\tmp\recompSrc\", bit of a strange place to be, and not the most obvious place to find it, but it works at least. I find that very strange, considering you're supposed to be able to modify the NBT data of any entity, if that NBT data really can't be modified after creation it's a bit hard to work with. That's a bit convoluted. Considering the fact that I can with in-game commands assign NBT data to an entity that I summon, and can also be done with /entitydata, I find it really strange that I'm being told this is not possible to do within the code.

I am familiar with the custom potion effect format, I've studied that page and the Chunk_format page as well, the problem is even after applying the NBT to the potion it doesn't seem to have any effect. Is potionEntity.getEntitydata() not the right NBT to modify? From what I see it's the only method that returns NBT other than the readFromNBT and WriteToNBT which are only used when loading from the save file. The problem with just overwriting the attackEntityWithRangedattack (which I use already to do the lightning strikes) is that no where there can I change the potion types. The only way is to copy all of the code, so I'll end up doing that, I just am having a hard time getting the NBT to apply. I've checked that it's being structured the right way, I built a class that prints out the NBT of an entity in JSON and it prints out the correct NBT, but when the potion hits me it doesn't have the effect, while the one I summon in-game with commands using ThrownPotion works fine. I think I can figure the rest out from here now that I've been told how to get the decomp workspace, so thanks for that. I'll report back if I have other questions related to this.

Ok, thank you, that's very useful. I wonder why that's not noted in any tutorials I've ever found, all of say to use Dev. The question still remains how to make the potion give wither, since I know it's possible in game it's obviously possible with code, but I guess I can look at EntityPotion and figure that out.

With the changed to how forge decompiles in 1.7 (I don't know which version it changed) you're building against a placeholder jar. There is absolutely no source code that I can access besides the obfuscated code for the actual game. Because of this I can't just copy and paste the witch code over because I can't get it. The last time I used forge you built your mod with a full decompiled minecraft code, able to see the actual game code from within Eclipse. Even if I did summon the potion myself, I'd still have the same issue. I've done 2 days of research and can't find any links on the internet about how to set the CustomPotionEffects of an EntityPotion. Since I'm trying to use Wither, which is not obtainable via metadata this is the only way that's possible to do it short of making your own custom potion entity. Overwriting can only do so much. I want it to have completely the same behavior as a normal witch except throwing wither potions and summoning lightning bolts (which I already have working). Unless I can get the deobfuscated code for EntityWitch, I don't get how I'm supposed to simply overwrite the method, I still would need to set the potion data of the EntityPotion.