[SPIGOT-6278] Persistent piglins count towards mob cap Created: 15/Dec/20  Updated: 02/Jan/22  Resolved: 02/Jan/22

Status: Resolved
Project: Spigot
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Bug Priority: Minor
Reporter: Bowserinator Assignee: Unassigned
Resolution: Fixed Votes: 4
Labels: mobcap, piglin, spawning, spawnreason
Environment:

Windows 10

Version: CraftBukkit git-Spigot-628435a-768d7fc (MC: 1.16.4) (Implementing API version 1.16.4-R0.1-SNAPSHOT)
Guidelines Read: Yes

 Description   

Piglins with the persistence tag set to true after they have spawned (for example, naming piglins, or piglins that have picked up golden shovels) count towards the mob cap, which is not vanilla behavior.

This has the effect that a lot of nametagged mobs, for example, can stop mobs from spawning when loaded by filling the mob cap. This affected my barter farm (which stops all other nether farms from working due to the piglins being loaded, which wouldn't happen in vanilla) and my wither skeleton farm (which uses nametagged piglins, so rates are much lower in spigot than vanilla).

 

Server config: spigot default

Tested with fresh server, no plugins

 

Steps to reproduce:

  1. Go to nether roof, make a pen and fill with 70+ piglins using spawn eggs. You can either name tag every one of them, or drop a bunch of golden shovels so they all pick them up. (Optional: verify they have the persistence tag with /data)
  2. Fly up 130 blocks so all mobs below nether roof despawn, as well as any piglins that aren't persistent
  3. Go below nether roof under where the piglins are, no mobs spawn

You can further verify by:

  1. Chunk load the piglin pen with either a portal chunk loader, or /forceload
  2. Teleport 1000 blocks away, no mobs should spawn
  3. Do `/kill @e[type=piglin]`
  4. Mobs will resume spawning normally

EDIT: this issue only affects piglins, thanks Airtheon for testing

 Comments   
Comment by Josh B [ 01/Oct/21 ]

Do we just edit the EntityPiglin.java to fix this? A friend runs our server and I'm not sure of the ins and outs of where to make this edit.

Comment by Seth Johnson [ 25/Sep/21 ]

This is a pretty significant bug. It prevents creation of a gold farm with a gold bartering farm, which pretty much are only useful in tandem.

 

I'm running Spigot 1.17.1 and would like to see this patched. Thank you.

Comment by jacob1 [ 04/Jul/21 ]

The fix for this is very simple. Bowserinator gave me a short patch, which I applied to my server, and persistent piglins no longer count towards the mob cap (restoring correct vanilla behavior).

 

The fix is in EntityPiglin.java , the isTypeNotPersistent should always "return true;". This is actually the default behavior from EntityInsentient.java. I looked at the spawning code and believe mojang may have implemented this method in EntityPiglin by mistake. It doesn't cause issues in vanilla mc, but because of spigot's differing logic, it is important that isTypeNotPersistent always returns true to keep hostile mob spawning working properly.

 

I've seen this same "return true" patch applied elsewhere, hope it can be fixed in spigot.

Comment by Bowserinator [ 17/Feb/21 ]

Someone brought this up again so I decided to investigate

It appears the issue is in EntityPiglin.java:

 

@Override
 public boolean isTypeNotPersistent(double d0) {
 return !this.isPersistent();
 }

This means that the check for not counting a mob to the mob cap in SpawnerCreature.java:

 

if (entityinsentient.isTypeNotPersistent(0) && entityinsentient.isPersistent()){
continue;
}

will always be false, thus piglins are always counted to the mob cap.

A solution would probably require a patch into EntityPiglin to match the patched mob cap check.

Comment by Airtheon Thesalion [ 04/Jan/21 ]

Update: to be honest, I can't find anything out of the ordinary, so I have no idea why it behaves like this.

Comment by Airtheon Thesalion [ 15/Dec/20 ]

EDIT: fyi you should remove the comma from the link, otherwise it won't work.

 

EDIT2: Spigot does have spawn reasons, I'm going to test if that matters somehow.

 

I see, well it is a CraftBukkit issue either way, as that is the server part that both Spigot and Paper base there stuff on.

On the paper issue it seems it has something to do with the SpawnReason.NATURAL, however I don't think Spigot has spawn reasons. So it might be something deeper that Paper fixes somehow when the spawn reason is not natural.

I believe it goes wrong in this part of the NMS code:
Lines 507-513 in net\minecraft\server\SpawnerCreature.java

// CraftBukkit start
private boolean a(EnumCreatureType enumcreaturetype, int limit) {
    int i = limit * this.a / SpawnerCreature.b;
    // CraftBukkit end

    return this.b.getInt(enumcreaturetype) < i;
}

This paper issue might also be useful https://github.com/PaperMC/Paper/issues/1922

Comment by Bowserinator [ 15/Dec/20 ]

It appears there's also an issue open for this on Paper: https://github.com/PaperMC/Paper/issues/4635 could it be a CraftBukkit issue?

Comment by Airtheon Thesalion [ 15/Dec/20 ]

To answer your edit, it does not affect all persistent mobs. I have tested your scenario with Piglins, Piglin Brutes, and Zombiefied Piglins. The latter 2 do not have the problem, so the issue seems to be specific to Piglins.

I am not sure what the cause is yet, but being able to compare how Piglins and Piglin Brutes are implemented, will help with finding it.

Generated at Wed Apr 02 10:50:50 UTC 2025 using Jira 10.3.3#10030003-sha1:d220e3fefc8dfc6d47f522d3b9a20c1455e12b7b.