This is the intended use for activator rails - ejecting entities. The shaking indicates that. I see nothing wrong with this. Did you confuse activator rails with powered rails?
This doesn't happen in spectator mode as far as I can tell, but I get particles - related issue?
Can confirm for 20w10a but the beacon effect particles are very transparent, but with the shell on they are solid.
Anvils are random. From the wiki: With each use, an anvil has a 12% chance to become damaged – degrading one stage at a time, first becoming chipped, then damaged, then eventually destroyed. An anvil typically survives for 25 uses on average, or approximately one use per 1.24 iron ingots used in crafting the anvil.
Mobs are not supposed to, but the player can respawn into lava, for example from a bed. Remotely related design issue or reason to mark it WAI?
In acacia trap doors you get the metal looking part always away from the "actual hinge". Is this related?
This doesn't make any sense, when I bonemeal grass it generates different results depending on biome and location, it's the same item, used on the same block. Why should it be different with trees. Clearly double standards at play.
The duplicate is just plain unrelated, is this a joke? You can place beds perfectly dry and make them still respawn you in lava, which is the core issue, lava is not obstructing the respawn.
@Mega_Spud Not to be rude or overly offensive, but this is the worst way to handle this. Get a tag "inactive" or something, but not like this. I updated the issue, as soon as I saw this - no warning even. As if this does any good or bad, open or not, this exact issue exists on java for three years. This is purely trust breaking, I trusted this platform to ignore my issues in perpetuity.
Please reopen this.
I placed black dots in the diagonals and white lines through the positions that are not altered. This is an extrapolation and cleanup from my previous findings. The south west corner is correct btw, it looked weird and I tested it.
[media]@Auldrick The requested tests with different facing did not deviate from my working theory of how it is behaving. Aligning myself with a /tp command (previously I was tired and forgot it) allowed me to reliably be placed towards the same direction, the east-west direction is prefered over north-south. My facing is saved only half correctly, the declination (up down) is always reset to some neutral.
Very important details: old world type, also in 1.8. I didn't have a working hypothesis for the expected behavior yet. I went back to a spot (survival world, original issue) where I got placed on the foot of a hill and it perfectly fit my model, although I still admit the lack of data for that version.
I made an image out of my tests, the player is moved from the color in the middle to the same color on the outer edges(straigth lines, recycling of colors), the upper right red spot is (63,63), the lower left red spot is (47,47). While making this graphic I could not recreate my own findings of the anti diagonal being safe.
This is an upgrade to the previous model, all quadrants are tested, I checked chunk borders via proper Minecraft and the east-west preference is left only on the main diagonal. This edge case only applies with positioning with tp command. Conclusion 2 from above still holds up. 12,1% (31/256) of the world is naturally safe! Everything is predictable now. I noticed no indications for any y level issues.
The target block being blocked, so one would suffocate or placed somewhere else potentially.
Different y levels, in my survival world I was placed only once only one block deeper. Round errors, sneaking? I don't know.
Nothing of this was done in survival, I just assumed this not to be related to the issue.
[media]
Being locally on the highest spot is related to the death issue. E.g. jumping and exiting quickly, standing ontop of a hill or a single block sticking out of a flat plane.
Standing ontop the eastern edge of a steep hill I could get myself to be placed into the world 5 blocks to the west (east, with a different place). I was immediately on the foot of the hill, had already taken damage, when I could see anything ingame.
Standing at any of the invisible walls sets you ontop of that wall. All tests I conducted (y64 and y83) did that.
Standing ontop of a pillar at (z,x)=(63,63) didn't reposition me. Standing at (62,62) sent me to death, where there are no coordinates shown... So only creative now: I am placed at (62,63). Next I tried on blocks on the diagonal to (56,56), I got placed on (56,63), consistent with the (62,63) result. While testing the diagonal I noticed that the precise position determines wether one is moved from (m,m) to (m,63) or (63,m). Further mapping: (62,62)→(62,63), (61,61)→(61,63), (58,58)→(58,63), (57,57)→(63,57)[precise positioning...], (50,56)→(47,56) [random position for good measure and a little surprise again], [symmetric to previous] (54,60)→(54,63), (50,60)→(50,60), (52,58)→(52,58), (57,53)→(57,53)
I could repeat [some] of the stuff above at (47,47) as starting point [, but couldn't bother to do it all again].
I can't be bothered to reload my game as often as the mapping with this blackbox would require. Also exact placement of the player is important and can lead to false results due to rounding errors.
one coordinate is kept, the other is set to the closest coordinate a =16n-1; n ∈ ℕ, it appears to be that one is placed to the closest a and the other coordinate is kept.
Relogging on chunk diagonals south west to noth east doesn't move the player.
We really need the F3 features, like chunk borders, precise coordinates and "facing north/.." (I used the sun)
Duplicate of MC-128257
Also: why is this so surprising? Jumping up and down in water was derpy anyway, this at least looks in first person somewhat good and is not gamechaning in any reasonable way.
It is still in 1.13-pre1 and I want to add one thing: This bug makes breaking sugar canes really weird, it feels unnatural to break 4 sugar canes infront of you, then having to going out and turn 180 degrees, to break the first one, of course you could harvest differently, but sometimes people run sloppyly ❓ into their fields...
commandblock_minecart should be command_block_minecart and
ender_crystal should be end_crystal in the new namesystem.
Affected version/s: new snapshots (e.g. 16w35a)
Can we agree on: "It is a bug and cannot be a feature."?
Shears should do that, since they are the only tool to break vines.
Dogs just teleport themselves back to you, when far enough away. What you are seeing is your dog falling that distance downwards and it teleporting back next to you. It should take the fall damage for the distance it fell, before teleporting.
I would believe this to be intended behaviour, at least if the damage calculation is working as I described.