I used to play on a 1.12.2 server, took a few months break and went back to play as 1.13.2.
All of the hostile mob farms (endermen, blaze, dark room mobs) seem to have massively lower spawn rates.
I googled for answers and I see many people complaining. But when I followed the links to the jira bug reports, I see all of the bug reports marked as "resolved" and comments from the devs claiming that spawn rates are higher in 1.13.2 than before... even though all evidence points to the opposite.
1) Does anybody understand what is really happening? If so, could you please explain it?
2) In order to spend my time building and not just gathering resources I painstakingly created a set of heavily enchanted tools with Mending and created/used XP farms to repair my tools.
This worked great in past versions, but since spawn rates are massively crippled, it now takes forever to mend a tool.
Being unable to quickly mend tools makes the game virtually unplayable for me. Well, it can be played, but it is no fun at all... which is the same.
Is there a new way to construct a fast XP farm in 1.13.2? I googled a lot and could not find one. All of the 1.13.2 tutorials I could find are based on spawn boxes which besides being quite rare are not very productive.
3) Does anybody know if Mojang plans to fix the hostile mob spawning in a future patch? Or are they still thinking that everything is great?
4) What is the most productive 1.13.2 XP farm design for mending?
IF the "Resolution" tag says "Fixed", this may be the case (although there are enough instances of comments made after putative fix dates to make this less than certain).
If the resolution tag says "WAI" or "Works As Intended", this is being considered as a 'feature'.
1)The core spawning system in 1.13 is identical to 1.12. It has the same mob cap (70) and the same number of attempts (1 attempt per loaded chunk per gametick). The placement-of-mobs condition set is still very similar, but 1.13 did add/change a few things:
y-level weighting has been equalized. In other words, mobs are just as likely to spawn in a cave at y=15 as they are in a dark room built at y=250. This weighting gets diluted by the number of spawnable blocks (dirt, cobble, stone, etc) in the spawn attempt column, so a column with just 5 blocks (ie, a 5-platform grinder in a void) will actually spawn mobs faster than a column with 100 blocks (ie, that same 5-platform grinder with the entire ground remaining beneath it). Dig out or replace with glass anything in the chunk that is not your grinder to speed it up (this is an old bit of advice that is applicable to 1.12, not something that's been added to 1.13, but in 1.12 basically the space below still counted even when dug away/replaced). Y-level still matters if you're going for extreme efficiency, but in 1.13 this change makes casual-grade grinders viable at any y-level.
pack size averages have been bumped upwards (with few exceptions). Most mobs kept their maximum pack sizes, getting their minimums raised. This means that more packs will have more spawns rather than "lucking out" with lots of small pack sizes. This equates to what appears to be more mobs spawning faster, but that's only a perceptual difference.
spawn attempts only happen on spawnable blocks (dirt, leaves, stone, gravel, clay, water, etc). In 1.12, pack centerpoints would only land on spawnable blocks but the resultant member spawns could then wander away from that spot and into mid-air (no solid block to spawn on) OR into the middle of a hill (no space to spawn). There was no preference for direction in this, so a pack could completely fail if it landed on the edge of a drop-off simply because all of the member attempts drifted off into space instead of onto the large flat space found in the opposite direction. In 1.13, this was changed so that the game only ever sees the large flat areas of spawnable blocks. This still means the pack could still fail by wandering inside the hill (this is why large we have large flat platforms in our grinders, so this isn't actually a new-to-1.13 idea...or new to 1.12, for that matter.)
mobs cannot spawn in spaces occupied by other mobs. This really only affects low-pack mobs like witches, but in general you will surely want to get the mobs off your spawn platforms faster. The more crowded it gets the slower the spawns will be because more spots will be occupied.
One thing that seems wrong to me is that all of the spawnable blocks are apparently globally applied to the entire mob group. Water counts as a spawnable block in every biome because in oceans and rivers Drowned can spawn. Leaf blocks count as spawnable blocks because in jungles parrots (which are totally irrelevant to this conversation) and ocelots (which are a part of the hostile mob cap) can spawn. Outside of those locations, these blocks probably shouldn't be counted as spawnable since nothing else can spawn on them but this is probably far too much effort to be worth changing.
These changes work together to fill the cap 10x faster (Ilmango visually demonstrates this in his video), and to reduce the number of failed spawns. They are NOT intended to actually increase the number of mobs from before.
2)Kelp farms with an attached autosmeltery setup can be used to store xp until the player is ready to collect it. This is because furnaces now keep track of their entire smelting history and award all of the experience generated when a player next grabs a result item from the output slot. Xisumavoid has a tutorial video about this that goes into the details like how much xp for a given tool and how many kelp smelts that translates into. I can't think of any other fully-automateable item that has this property (smeltable as well as being fuel), but a tree farm comes pretty close if you can keep up with it.
3)The point about the spawnable blocks above and the various movement-related bugs regarding spawning aside, there's no problem with mob spawning.
Thanks DuhDerp, that might be the technical explanation but itit do not explain in an easy to understand way why virtually every mob farm now produces a tiny fraction of what they used to or how to create a new XP farm for mending.
The YouTube links provided by ScotsMiser though do give some useful hints. I will have to create a new one based on that.
Try copying your world and then opening the copy in the latest 1.14 snapshot (don't open your world in 1.14, open a copy of your world because you can't go back to 1.13 after it's opened in 1.14). I created a world that put me on an island, and hardly anything spawned on it in 1.13.2, but spawned so many baddies in 1.14 that it was easy to get overwhelmed by all of the enemies.
But I think there's a couple of bugs still on the bug-stomping list: Enemies spawn at very high rates on chunks that are on the border of where they are about to despawn, and very low enemy spawn rates on servers where the chunk view limit has been set to a lower value than the default number.
1)The core spawning system in 1.13 is identical to 1.12. It has the same mob cap (70) and the same number of attempts (1 attempt per loaded chunk per gametick). The placement-of-mobs condition set is still very similar, but 1.13 did add/change a few things:
y-level weighting has been equalized. In other words, mobs are just as likely to spawn in a cave at y=15 as they are in a dark room built at y=250. This weighting gets diluted by the number of spawnable blocks (dirt, cobble, stone, etc) in the spawn attempt column, so a column with just 5 blocks (ie, a 5-platform grinder in a void) will actually spawn mobs faster than a column with 100 blocks (ie, that same 5-platform grinder with the entire ground remaining beneath it). Dig out or replace with glass anything in the chunk that is not your grinder to speed it up (this is an old bit of advice that is applicable to 1.12, not something that's been added to 1.13, but in 1.12 basically the space below still counted even when dug away/replaced). Y-level still matters if you're going for extreme efficiency, but in 1.13 this change makes casual-grade grinders viable at any y-level.
pack size averages have been bumped upwards (with few exceptions). Most mobs kept their maximum pack sizes, getting their minimums raised. This means that more packs will have more spawns rather than "lucking out" with lots of small pack sizes. This equates to what appears to be more mobs spawning faster, but that's only a perceptual difference.
spawn attempts only happen on spawnable blocks (dirt, leaves, stone, gravel, clay, water, etc). In 1.12, pack centerpoints would only land on spawnable blocks but the resultant member spawns could then wander away from that spot and into mid-air (no solid block to spawn on) OR into the middle of a hill (no space to spawn). There was no preference for direction in this, so a pack could completely fail if it landed on the edge of a drop-off simply because all of the member attempts drifted off into space instead of onto the large flat space found in the opposite direction. In 1.13, this was changed so that the game only ever sees the large flat areas of spawnable blocks. This still means the pack could still fail by wandering inside the hill (this is why large we have large flat platforms in our grinders, so this isn't actually a new-to-1.13 idea...or new to 1.12, for that matter.)
mobs cannot spawn in spaces occupied by other mobs. This really only affects low-pack mobs like witches, but in general you will surely want to get the mobs off your spawn platforms faster. The more crowded it gets the slower the spawns will be because more spots will be occupied.
One thing that seems wrong to me is that all of the spawnable blocks are apparently globally applied to the entire mob group. Water counts as a spawnable block in every biome because in oceans and rivers Drowned can spawn. Leaf blocks count as spawnable blocks because in jungles parrots (which are totally irrelevant to this conversation) and ocelots (which are a part of the hostile mob cap) can spawn. Outside of those locations, these blocks probably shouldn't be counted as spawnable since nothing else can spawn on them but this is probably far too much effort to be worth changing.
These changes work together to fill the cap 10x faster (Ilmango visually demonstrates this in his video), and to reduce the number of failed spawns. They are NOT intended to actually increase the number of mobs from before.
2)Kelp farms with an attached autosmeltery setup can be used to store xp until the player is ready to collect it. This is because furnaces now keep track of their entire smelting history and award all of the experience generated when a player next grabs a result item from the output slot. Xisumavoid has a tutorial video about this that goes into the details like how much xp for a given tool and how many kelp smelts that translates into. I can't think of any other fully-automateable item that has this property (smeltable as well as being fuel), but a tree farm comes pretty close if you can keep up with it.
3)The point about the spawnable blocks above and the various movement-related bugs regarding spawning aside, there's no problem with mob spawning.
Some of these have not been changed, even back to 1.6.4, where the altitude was evenly weighted over the entire range from y=0 to the highest solid block (or highest loaded chunk section, which was changed to the highest block in 1.8):
/**
* Given a chunk, find a random position in it.
*/
protected static ChunkPosition getRandomSpawningPointInChunk(World par0World, int par1, int par2)
{
Chunk var3 = par0World.getChunkFromChunkCoords(par1, par2);
int var4 = par1 * 16 + par0World.rand.nextInt(16);
int var5 = par2 * 16 + par0World.rand.nextInt(16);
int var6 = par0World.rand.nextInt(var3 == null ? par0World.getActualHeight() : var3.getTopFilledSegment() + 16 - 1);
return new ChunkPosition(var4, var6, var5);
}
In particular, "rand.nextint()" returns a uniformly distributed value so mobs are just as likely to spawn at y=1 as at y=200 or whatever, but the chance of spawning at a particular level does indeed depend on the available range (a flat surface at y=1 will have 256 times the span rate as at y=256, or prior to 1.8, 16 times since "getTopFilledSegment() returns the height of the highest non-empty chunk section).
In addition, mobs already couldn't spawn if they intersected other mobs:
/**
* Checks if the entity's current position is a valid location to spawn this entity.
*/
public boolean getCanSpawnHere()
{
return this.worldObj.checkNoEntityCollision(this.boundingBox) && this.worldObj.getCollidingBoundingBoxes(this, this.boundingBox).isEmpty() && !this.worldObj.isAnyLiquid(this.boundingBox);
}
Also, 1.6.4 does not appear to have a set min-max spawn count; it just makes 3 attempts each at spawning up to 4 mobs (12 total attempts), so anything less is simply due to the failure to spawn mobs and they always have a maximum pack size of 4 (even mobs set otherwise; only passive mob spawning during world generation actually uses the min/max pack sizes set in the spawn list. This may have been fixed in a later version but in 1.6.4 Endermen were the only mobs with a min that was less than 4, other than Ghasts, which have a "maxPerChunk" of 1, which is handled differently).
So really, the only change seems to be the "only spawn on spawnable blocks rule", which I'm not even sure how they would apply unless the game makes a map of the whole area around the pack center, and the reported issues with mob spawning have to be due to something else. Spawning "10x faster" also wouldn't make a difference since mobs spawn so fast that you wouldn't even notice (if I switch from Peaceful to Easy+ in 1.6.4 the entity count almost immediately rises by around 80 (the mob cap was actually higher before 1.8, 79 instead of 70) since with 225 attempts per tick of spawning a pack of 4 mobs the game can in theory spawn 18000 mobs per second; even a 99% failure rate is 180 mobs), and if anything, mob farms ought to be 10x more efficient (once all potential spawning places have been removed - failed spawn attempts outside of a farm have no effect on how well a farm works since it is only down to whether the mob cap has been reached; in fact, the cap isn't (at least not in older versions) even checked during a spawn cycle so you can get more mobs than the cap allows, as seen with squid in oceans).
Of course, 1.13+ is still affected by a render distance bug which has been around since forever, in addition to the issues with mob spawning on multiplayer servers (all players must spawnproof the areas around them):
MC-2536 View distance affects mob spawning (the view distance must be at least 10 or mobs can wander outside of entity processing chunks before they reach the despawn distance; 128 blocks = 8 chunks, plus 2 more for them to be processed = 10. The main symptoms are a progressive decrease in spawn rates as more mobs fill up the cap, occurring faster as the view distance is decreased)
MC-42053 Low mob spawn rates on low render distances (this also affects singleplayer since 1.7.4, and possibly before 1.3.1, when the internal server was implemented in singleplayer, which had a fixed view distance of 10 until 1.7.4. Either way, this is always the first thing I think of when I see somebody complain about poor spawn rates; newer versions are more demanding so it is natural that they lower the view distance)
But I think there's a couple of bugs still on the bug-stomping list: Enemies spawn at very high rates on chunks that are on the border of where they are about to despawn, and very low enemy spawn rates on servers where the chunk view limit has been set to a lower value than the default number.
I think you nailed it right there. I mainly play on a server and it would not surprise me if they adjusted the view limit for performance reasons.
Maybe I should take a vacation from Minecraft until 1.14
I've been playing the 1.14 snapshots pretty regularly and it looks like it's going to be awesome. They're still going through some bug fixes, but so far it looks like it's going to be a lot of fun. I started a server with vanilla Minecraft on the latest snapshot, and the enemy spawn rate is pretty insane (insane = much chaos = good fun). It also sounds like version 1.14 is going to be out soon. Woot!
I am puzzled by 1.13.2 hostile mob spawn rates:
1) Does anybody understand what is really happening? If so, could you please explain it?
2) In order to spend my time building and not just gathering resources I painstakingly created a set of heavily enchanted tools with Mending and created/used XP farms to repair my tools.
3) Does anybody know if Mojang plans to fix the hostile mob spawning in a future patch? Or are they still thinking that everything is great?
4) What is the most productive 1.13.2 XP farm design for mending?
MS/Mj follows a rather disingenuous practice with the "RESOLVED" label as it includes:
IF the "Resolution" tag says "Fixed", this may be the case (although there are enough instances of comments made after putative fix dates to make this less than certain).
If the resolution tag says "WAI" or "Works As Intended", this is being considered as a 'feature'.
1) You are in good (and numerous) company
2) See EnderMINI - Simple Enderman Farm [Tutorial], and 1.13 Mob Spawning EXPLAINED! [Fun Farms 23] from gnembon or Simple Enderman XP Farm for Minecraft 1.13.1 from docm77 (basically the same farm)
[gnembon also explains some of the changes to mob spawning in 1.13 (vs 1.12) that cause "lots of farm designs to need to be revised or thrown into the garbage" – 1.13 Mob Spawning Changes / Witch Farms Nerfed by ilmango is also recommended as helpful in explaining the changes]
3) Know? No, but this seems unlikely.
Yep, that seems to be the MS/Mj attitude in a nutshell… :facepalm:
4) The Ender Enders referenced above are probably fastest, with some of the upper level guardian farms (eg [Tutorial] Bubble Based Guardian Farm! (30 levels in 2 mins) from cubfan135) as the next tier.
1)The core spawning system in 1.13 is identical to 1.12. It has the same mob cap (70) and the same number of attempts (1 attempt per loaded chunk per gametick). The placement-of-mobs condition set is still very similar, but 1.13 did add/change a few things:
One thing that seems wrong to me is that all of the spawnable blocks are apparently globally applied to the entire mob group. Water counts as a spawnable block in every biome because in oceans and rivers Drowned can spawn. Leaf blocks count as spawnable blocks because in jungles parrots (which are totally irrelevant to this conversation) and ocelots (which are a part of the hostile mob cap) can spawn. Outside of those locations, these blocks probably shouldn't be counted as spawnable since nothing else can spawn on them but this is probably far too much effort to be worth changing.
These changes work together to fill the cap 10x faster (Ilmango visually demonstrates this in his video), and to reduce the number of failed spawns. They are NOT intended to actually increase the number of mobs from before.
2)Kelp farms with an attached autosmeltery setup can be used to store xp until the player is ready to collect it. This is because furnaces now keep track of their entire smelting history and award all of the experience generated when a player next grabs a result item from the output slot. Xisumavoid has a tutorial video about this that goes into the details like how much xp for a given tool and how many kelp smelts that translates into. I can't think of any other fully-automateable item that has this property (smeltable as well as being fuel), but a tree farm comes pretty close if you can keep up with it.
3)The point about the spawnable blocks above and the various movement-related bugs regarding spawning aside, there's no problem with mob spawning.
Thanks DuhDerp, that might be the technical explanation but itit do not explain in an easy to understand way why virtually every mob farm now produces a tiny fraction of what they used to or how to create a new XP farm for mending.
The YouTube links provided by ScotsMiser though do give some useful hints. I will have to create a new one based on that.
Try copying your world and then opening the copy in the latest 1.14 snapshot (don't open your world in 1.14, open a copy of your world because you can't go back to 1.13 after it's opened in 1.14). I created a world that put me on an island, and hardly anything spawned on it in 1.13.2, but spawned so many baddies in 1.14 that it was easy to get overwhelmed by all of the enemies.
But I think there's a couple of bugs still on the bug-stomping list: Enemies spawn at very high rates on chunks that are on the border of where they are about to despawn, and very low enemy spawn rates on servers where the chunk view limit has been set to a lower value than the default number.
Some of these have not been changed, even back to 1.6.4, where the altitude was evenly weighted over the entire range from y=0 to the highest solid block (or highest loaded chunk section, which was changed to the highest block in 1.8):
In particular, "rand.nextint()" returns a uniformly distributed value so mobs are just as likely to spawn at y=1 as at y=200 or whatever, but the chance of spawning at a particular level does indeed depend on the available range (a flat surface at y=1 will have 256 times the span rate as at y=256, or prior to 1.8, 16 times since "getTopFilledSegment() returns the height of the highest non-empty chunk section).
In addition, mobs already couldn't spawn if they intersected other mobs:
Also, 1.6.4 does not appear to have a set min-max spawn count; it just makes 3 attempts each at spawning up to 4 mobs (12 total attempts), so anything less is simply due to the failure to spawn mobs and they always have a maximum pack size of 4 (even mobs set otherwise; only passive mob spawning during world generation actually uses the min/max pack sizes set in the spawn list. This may have been fixed in a later version but in 1.6.4 Endermen were the only mobs with a min that was less than 4, other than Ghasts, which have a "maxPerChunk" of 1, which is handled differently).
So really, the only change seems to be the "only spawn on spawnable blocks rule", which I'm not even sure how they would apply unless the game makes a map of the whole area around the pack center, and the reported issues with mob spawning have to be due to something else. Spawning "10x faster" also wouldn't make a difference since mobs spawn so fast that you wouldn't even notice (if I switch from Peaceful to Easy+ in 1.6.4 the entity count almost immediately rises by around 80 (the mob cap was actually higher before 1.8, 79 instead of 70) since with 225 attempts per tick of spawning a pack of 4 mobs the game can in theory spawn 18000 mobs per second; even a 99% failure rate is 180 mobs), and if anything, mob farms ought to be 10x more efficient (once all potential spawning places have been removed - failed spawn attempts outside of a farm have no effect on how well a farm works since it is only down to whether the mob cap has been reached; in fact, the cap isn't (at least not in older versions) even checked during a spawn cycle so you can get more mobs than the cap allows, as seen with squid in oceans).
Of course, 1.13+ is still affected by a render distance bug which has been around since forever, in addition to the issues with mob spawning on multiplayer servers (all players must spawnproof the areas around them):
MC-2536 View distance affects mob spawning (the view distance must be at least 10 or mobs can wander outside of entity processing chunks before they reach the despawn distance; 128 blocks = 8 chunks, plus 2 more for them to be processed = 10. The main symptoms are a progressive decrease in spawn rates as more mobs fill up the cap, occurring faster as the view distance is decreased)
MC-42053 Low mob spawn rates on low render distances (this also affects singleplayer since 1.7.4, and possibly before 1.3.1, when the internal server was implemented in singleplayer, which had a fixed view distance of 10 until 1.7.4. Either way, this is always the first thing I think of when I see somebody complain about poor spawn rates; newer versions are more demanding so it is natural that they lower the view distance)
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?
I think you nailed it right there. I mainly play on a server and it would not surprise me if they adjusted the view limit for performance reasons.
Maybe I should take a vacation from Minecraft until 1.14
I've been playing the 1.14 snapshots pretty regularly and it looks like it's going to be awesome. They're still going through some bug fixes, but so far it looks like it's going to be a lot of fun. I started a server with vanilla Minecraft on the latest snapshot, and the enemy spawn rate is pretty insane (insane = much chaos = good fun). It also sounds like version 1.14 is going to be out soon. Woot!