[1.70b] Bug: Certain items not stacking [SMP]

Looking into it

I can confirm the bug

So far it has only occurred while playing with crops, however I/we have not done much since the upgrade to 1.70

//MG0

This bug happens quite often to be we the RE-BATTERY when i autosort with Inventory Tweaks

better add exception handling for null nbt then creating it...

I can confirm this bug still exists in 1.71. Thanks for reporting it; now I know I don't need to hunt through forge items.

Observed effected entities:

Fertilizer (bonemeal + scrap)
Crop Yeilds
Stakes (er 4 sticks; forget the name offhand)

I also noticed inconsistency between client/server stacks (the client would stack but the server would say no).

From the description and observed behavior: I -think- the issue is opening the crop-analyzer shows the hotbar, and thus this code tags /all/ items with an empty NBT array, irrespective of the item type.

A test for confirming this with known clean parts (or dirty parts to see if they all become dirty); would be to place test items in the hotbar, open the scanner (might require a scan operation on an unknown seed sack) and see if the problem is reproducible.

I've confirmed that items in the hotbar which are clean are tainted when the crop scanner is used; it does not even have to process a crop; merely opening the scanner seems to be sufficient.

The check for 'bad' crops should probably only trigger when one is dropped in to the input of the scanner, and should require it to match the ID of a seed bag.

I can confirm that crafting RE-Batteries in a RP2 Project Table (SMP) can make them unable to stack. Charge a bit and uncharge all the batteries, fixed.

I have found that if you have Equivalent Exchange installed, throwing what you can into the Transmution Table's fire and then making it again form the EMC made also corrects it a bit.

The 1.2.4 version will have a hack to try to prevent that bug. Once it's released, I need you guys' feedback.

This bug is still present in 1.97 (installed only with forge 3.3.8 or 3.3.7). I added information to wiki's article about cropanalyzer, because it took me 3 hours to track it down.

Edit: I decompiled HandHeldCropnalyzer.class and I found out, that in "public void onGuiClosed" in a loop for every item in inventory until cropnalyzer is found "StackUtil.getOrCreateNbtData" is used.

for(int k = 0; k < entityhuman.inventory.getSize(); k++) 
 { 




 ItemStack itemstack = entityhuman.inventory.getItem(k); 
 if(itemstack == null)                        continue;                    NBTTagCompound nbttagcompound3 = StackUtil.getOrCreateNbtData(itemstack);
 if(nbttagcompound.getInt("uid") != nbttagcompound3.getInt("uid"))  continue;
 itemStack.count = 1; 
 entityhuman.inventory.setItem(k, itemStack); 
 break;                }

Possible fix would be to replace

NBTTagCompound nbttagcompound3 = StackUtil.getOrCreateNbtData(itemstack);  
if(nbttagcompound.getInt("uid") != nbttagcompound3.getInt("uid"))                        continue;

with

NBTTagCompound nbttagcompound3 = itemstack.getTag();  
if(nbttagcompound3 == null || nbttagcompound.getInt("uid") != nbttagcompound3.getInt("uid"))                        continue;

Something went wrong with code tags.

You shouldn't modify non cropanalyzer items in any way, but getOrCreateNbtData is doing it. Also I don't think that "entityhuman.inventory.setItem(k, itemStack)" is necessary, because you operate on references (ignoring this fact seems to be cause of this bug).

That's the second time I've used code fixes from the community on the main source, I'm feeling uneasy...

If you fixed this bug, could you post updated .class file, because I didn't go into whole recompiling thing yet.

Bug is still present in 1.103 and is EXTREMELY annoying.

I play on a lot of modded servers and while it seems rare, this has happened to me twice in total with many items such as red dye, or more recently water melons given from crops.

There isn't anything else I can do. All stray mentions of NBT are removed from the cropnalyzer.

Also, they broke in 1.3