blablubbabc

Can confirm for 1.16.5

Tested on 20w22a with two items with names A and B and trading removed the correct items in correct quantities. So this seems to be fixed.

Can confirm the weird behavior when interacting with a villager while holding a signed book in hand. MC 1.15.1. Does not affect villagers without profession / trades.

I just rebuild the reproduction contraption I outlined above and am still able to reproduce this on 1.14.4. I haven't tested this on MC 1.15 yet.

Since my description of the issue / how to reproduce this with the outlined command block contraption might be confusing, I made a quick video of this and uploaded it here: https://www.youtube.com/watch?v=AUhcTnryQds (or if that link doesn't work, alternatively here: https://oc.h-in-t.net/nocloud/index.php/s/voXdVOtKBcyBtCn)

This contraption actually moves the villager 21 blocks this time, and directly teleports the villager back to the original height at the start of the next cycle. The video also shows the commands updated for 1.14.4. The first demonstration at 1:40 already perfectly shows how the villager appears to be inside the orange block, even though the output of the last teleport command reports that the villager got correctly teleported to y=6. (Though, with the video quaIity on youtube it might be harder to see than it actually is when playing the game; but the ones at 2:21, 2:25, and 2:42 should be visible nevertheless)

Can confirm for 1.14.3: Villagers will still look at nearby invisible villagers.

Confirmed for snapshot 17w50a. Hope this finds its way into 1.13, with all those nbt-related command changes.

I am not sure I understand this ticket correctly, but the following bug can be reproduced easily on MC 1.12.1, both on a server and client world, and I assume it relates to this issue:

• Give yourself a mob_spawner item and place it: Initially it will have SpawnData for 'Pig' and an empty list of SpawnPotentials internally. (Note: If you run /blockdata ~ ~-1 ~ {} it will report that the SpawnPotentials list contains an entry for 'Pig', however, internally, it actually doesn't. Reason for that mentioned below.)

• Go far away in order to trigger saving of the BlockEntity. Or stop/start the server/world. When saved and having an empty list of SpawnPotentials, it will add an entry corresponding to its SpawnData to that SpawnPotentials list in its saving output. (That's also the reason why the /blockdata ~ ~-1 ~ {} command shows that entry: The command shows the saving output of the BlockEntity, not the actual data currently in use, at least not for the mob_spawner BlockEntity..)

• Come back to the mob-spawner / load the world again: This will load the mob-spawner from the saved data. It now actually contains a 'Pig' entry in its internal SpawnPotentials list.

• Get yourself a spawn_egg, for example of type zombie from the creative menu.

• Make sure to get in survival mode. Not sure if creative mode has this issue as well.

• Right-click the mob-spawner with the spawn_egg. This will set the entity type of the SpawnData.

Expected result: The mob-spawner will now spawn zombies instead of pigs.

Actual result: The mob-spawner will spawn one (wave of) zombie(s), and then switch back to spawning pigs. Reason: After each spawn wave the mob-spawner checks its SpawnPotentials (if it isn't empty) for the next entity to spawn, and replaces the SpawnData with the picked entity. And this list only contains 'Pig' after the save-load-cycle and even after changing the SpawnData with the spawn_egg.

Possible fix: Maybe clear the internal SpawnPotentials of the mob-spawner after the mob type was changed via a spawn_egg.

And / Or: Don't include the SpawnsData in the SpawnPotentials when being saved and having an empty SpawnPotentials list. So the blockdata-command will report correct (the actual) data, and the data doesn't change after a save-load-cycle.