You want to mean that my mob tower trap have a low item/hour because the monsters spawn in caves too?
if yes, what i can do?
edit: i have made a pillar of dirt that bring me to y:128 (max)
with that, mobs can't spawn in caves and my tower works like if it was a perfect world
You could always post in my thread rather than cluttering this thread with unnecessary discussion. To answer your question though, I suggest you use MCEDIT and fill in all air below sea level in a 272x272 square centered on your collection point. Then fill in all the other caves manually.
Im building the SMP version on the server i play on, built by myself....
its almost finished, except for the signs.
Ive completed approx 1 floor of signs, its whats taking the most time
its 31x31x128 (including collection shaft)
all legit except for the signs which i can get stacks of 3 from a kit (i chopped stacks of wood before i found this out :sad.gif: )
luckily it hasnt been griefed yet too
What exactly can block spider spawns when the 3x3 area is checked? Full blocks and half blocks obviously work, but what about something like mushrooms? Or fluid blockers like pressure plates and signs?
I apologize if this question has already been asked, but where would I find the design of the pit modules referred to in the first video? I'd like to build on to this whole trap, and incorporate it into my mega-city.
I think your theory is slightly flawed. Consider the pathing algorithm. It likely goes something like this:
1. Pick some place nearby.
2. Go there.
With 1., in your test without a sign bridge, the mob must pick exactly the sign as the destination in order to fall. The probability of that is very small because there are more blocks to not fall off of (or, in this case, the same number of blocks). When the sign bridge is created, the probability of picking a sign isn't much greater, but, in addition to sign blocks, the solid blocks across the bridge will also cause the mob to fall.
Your test 1: Pig never falls off the ledge. Creeper has ~50% chance of falling off the ledge each time it moves.
Your test 2: Both Pig and Creeper have a ~70% (21/30) chance of falling each time they move.
By this logic, we should conclude that the most efficient method for coercing a fall is a pattern where every spawnable block is surrounded on every side by a block which will make the monster fall. This would achieve 100% chance of falling each time they move.
Edit: I'm trying to think how this can be proven.... If you did a similar test, still using mob spawners and the same spawnable area, but isolated platforms with different sized sign borders, the larger virtual platform should have more deaths from it. This will show that it is the chance for moving into a sign block vs a not sign block, and nothing to do with the proximity to the wall.
My description of the tests was a bit simplified. I started with the two platforms, one with 12 signs around it and the other one completely without signs. The tests showed that the mobs from both platforms have virtually the same chance to fall, and adding the wall didn't change anything.
So, the pig spawner in your picture had mobs jump off that platform (at the same rate), even though there was no sign bordering it? How can that be true when the new mob intelligence prohibits them from jumping down more 4 or more blocks?
For all people saying "invedit all that stuff" gtfo this "Thing" here can be build legit, it was just research.
Naturally, given enough time. On a smp server a team can even managed to complete it fairly fast. Relatively speaking. The question is how much time you have and for most of us, building this trap is not feasible.
Built the minimal bioreactor with the help of WorldEdit (anybody got 2652 signs laying around? No? Okay...) on our server.
Unfortunately it overlaps 4 chunks (would be MY luck) instead of being entirely contained in one chunk. The resulting CPU load is not prohibitive, but is noticeable. The funny thing is, even when there are no mobs inside the trap, it continues to put a load. I suspect it's the sheer number of signs/transparent blocks. A similar-capacity conventional mob trap with NO signs does not lag this much even when it is cooking along with mobs on nearly every level.
So word to the wise: if you want to build this, it is possible to fit it into one single chunk. Whatever you can do to this end, do it. Use mods that let you swap chunks and just MCEdit it in. OR if you want a more vanilla-ish experience, try identifying chunks using any one of the many tools you can use in-game to delineate chunk borders. Survey a chunk with a low incidence of lava, excavate, etc.
And be prepared for a constant, low-to-moderate demand on your server CPU.
You are better off placing this a few minutes walk away from where most people are working. If they want to idle next to the trap they can go and do that. Set up a warp or minecart track there, so it's convenient, but far enough out of range so that the guy who just wants to hop on the server and harvest his wheat doesn't have to deal with lag. Encourage people who want to AFK at the trap, to do so when fewer people are on.
You don't need a lava block to destroy drops - it's not THAT productive. And that's not the source of the lag. It's the signs/fences, and the fluid flow mechanics (to a much smaller degree).