B R

Yes. On 1.8.1-pre4, if I create a new world in single player or on a server with the seed I gave and at the coordinates I gave, the spawners and chests appear. But if I take my existing world and delete those chunks or even the entire nether dimension folder, it regenerates with no spawners or chests.

After some more testing, I discovered that the problem will go away after deleting world/data/Fortress.dat. So I guess that solves the problem for me, but someone may still need to know where it originated.

I've tried a few tests now. On the same computer, with the same version of Java, using the same seed and visiting the same coordinates, on any of the pre-releases, and in single player or on a server, I cannot reproduce this bug in a newly generated world. However, on the server where I first experienced it, it keeps happening not matter how many times I delete and regenerate the chunks. I don't know what to make of that.

I am experiencing this problem in 1.8.1 pre-releases 2 and 3. I didn't test pre-release 1, but the problem was not occurring in 1.8.0. No monster spawners or chests generate in the fortresses. Example: seed=-3320457680109975269, coordinates=(351x, 55y, 187z); (394x, 55y, 174z).

I don't know why this is marked as fixed. I'm still seeing the problems (maybe slightly different than this but similar) in the issue I posted that got marked a duplicate, here: MC-31492

I just experienced this bug in 1.6.2. I was pulling a cow by a lead while in a boat and had to turn around. After colliding with the cow, I died instantly, being at full health and armor, with the message saying I "fell from a high place."

It seems 1.6 totally revamped this part of the code, so it's no longer and issue.

My report was based on the code in EntityLiving.java, the same code that Kumasasa posted. When it comes to empirically testing this, I urge you to be careful with how you interpret your observations. 2% and 5% are relatively small percentages that are close together. To confidently confirm those numbers through in game tests, you would need a large sample size and rigorously recorded results.

I personally ran a test spawning 100 zombies on both Normal and Hard. On Normal, I observed 4 zombies with armor, and on Hard, I observed 2. 100 is hardly a large enough sample to definitively make a conclusion either way, but what you can safely conclude from that is that the odds are close, and not a "clear increase".

Again, just to make sure you're watching for the right thing, I'm not talking about whether they're wielding a weapon or not, or how much armor a particular mob is wearing. They could be wearing 1 piece or all 4, the random percentage in question is how many are wearing any armor vs. none.

So CubeTheThird and Joss Eden, I encourage you to test it again, in larger numbers, and to write down what you see.

Add sponges to the list too. As long as adventure mode is going to be breaking blocks only with the right tools, then everything should be breakable with some tool for it to be a fully usable game mode. There should be a no-build/break mode too, but that's another issue.

Also, stone half-slabs and double-slabs can be targeted with an axe, which seems incorrect. And axes should probably be able to target leaves and melons, but currently they do not.

I would propose the following changes:

TNT: either shears or anything
Monster Egg: Pickaxe
Pistons: Pickaxe or anything
Cactus: Axe
Bed: Axe
Fire: Anything
Sponge: Anything
Cake: Anything

The first 1.5 snapshot seems to have slightly affected this behavior, but still not fixed it. Torches still don't respond to short pulses in an expected way.