I created a machine for the sole purpose for testing if/when pistons bug out and how they do so. Two weeks ago I posted two results, the first I added to [MCPE-16314], which I deemed most relevant to the issue, and is still marked as "Resolved: Fixed". The second I added to MCPE-28260 at a moderator's request, which I questioned due to the lack of any signs of crashing when I found the bug; I have yet to receive a reply.
Thus, I have decided to repost these here. The first bug is that the piston deletes the block it moves into. Upon closer inspection, it seems the piston extends, begins to retract but instantly begins moving, gets stuck in extended position with the arm inside of T2, then deletes the block when it decides to retract. This occurs when T1 is air and T2 is a normal block.
The second bug is that the piston carries a block sideways then (the piston) turns into an item drop. Again, upon closer inspection the piston spits out T1, begins extending right as it is moved by the machine, gets its arm stuck in T2, proceeds to bring T2 to T1's starting location as the piston is moved back, then the piston drops itself as an item. This occurs when both T1 and T2 are normal blocks.
Both of these glitches appear to stem from extending/ed pistons remaining movable if powered in a special way.
World download:
[media]
(!)Warning: Using Note Blocks creates a chance for the game to crash. This seems to be limited to just Note Blocks for now.
A schematic (with original coordinates) of the original machine is below:
Facing positive z, x (left): 165, x (right): 161
z: 10
RB | SB | RB | ||
[RB] | P> | PA | SB | <P |
z: 9
Ov | ||||
NB | S* | NB |
z: 8
T1 | T2 |
Key:
> | Left |
< | Right |
v | Down |
* | Towards the screen |
NB | Normal block |
RB | Redstone block |
[RB] | Remove this block to start the machine, replace to stop |
SB | Slime block |
P | Regular Piston |
S | Sticky Piston |
O | Observer |
T1, T2 | Testing blocks, known to change the end result of the machine |
P.S.: Sorry mods, I just didn't feel like this was getting the right attention in what I consider to be a dead ticket and a separate and/or blanket issue.
Linked issues
is duplicated by 4
relates to 2
Attachments
Comments 14
I was making a dedicated world and found that adding signs changed its behavior, so an instructional book will have to do.
Further Testing
Can break: (assume "no drop" unless stated otherwise)
Bedrock
Extended piston arms (piston base will turn into drop)
Extended pistons
Diamonds
Trapdoors
Doors (upper half: drop item, lower half: no drop)
2-tall flowers
Beds
Item frames w/ items
Jukebox (drops music disk only, stops playing)
Chest (pushes forward, then drops only its contents)
Mob heads (drops as item)
Monster eggs (no silverfish, I did change the difficulty)
Dragon egg
Torches
Hoppers (drops contents only)
Flower pots (drops flower only)
Growing Netherwart
Growing crops
Signs
Banners (lower half, ground mounted only)
Pressure plates
Slime blocks
Rails
Glass panes
Ladders
Shulker boxes
Droppers (drops contents only)
Redstone blocks
Terracotta
Glazed terracotta
Note blocks
Mob Spawners
Stone walls
Packed ice
Moves:
Chests (piston does not break)
Slime blocks with or without other blocks
Redstone blocks
Terracotta
Glazed terracotta
Note blocks?
Stone walls
Packed ice
Deals damage to:
Shulkers
Only breaks piston:
Observers
Crash the game:
Note blocks (I give up trying to figure out what's going on)
Pistons
Transmute:
Pistons
⚠️ Revision to the book: Glitch usually occurs within the first few cycles, causes of variation are still unknown. Piston seems to prefer moving the block when piston and/or block is on the left.
Edit: It seems even the crash caused by the note block is inconsistent as there is the only one of 5 crashes with a certain exception code and different crashes list different failing modules and only happens most of the time, when it feels like it.
This is starting to feel like a mega-report but I have no idea how to break it up. In a recent setup with sand, I created a copy of the testing world, made two sand pillars to near build limit in front of the machine and a hole to the void for the sand to fall into after being pushed and then saved. I created a copy of that and ran the test in the copy to compare file sizes afterwards and killed all entities before saving. In game it said there was an additional 1.2MB in the copy world and after exporting there was a 25kB difference between the two worlds with no difference between them other than running the machine to remove a sand pillar and removing all entities afterwards.
This is going to be my last post for a while (I'm done with all of these stupid PE bugs). The in game file size in my previous comment might be caused by MCPE-23462, MCPE-25613, MCPE-7749, etc. Still no explanation for exported file size differences
I built a snow farm with pistons quickly destroying the snoe a snoe golem places. The pistons are connected to a fast redstone clock. When a full blocks is attached to the piston, the full block can be destroyed.
Cleaning up old tickets: This ticket has not been updated recently (~1 year+) so is being closed as Cannot Reproduce. If you feel this is still a valid issue then please comment, or create a new ticket following the Issue Guidelines.
Quick Links:
📓 Issue Guidelines – 💬 Mojang Support – 📧 Suggestions – 📖 Minecraft Wiki
Thanks for the extra information. If you wanted to include a test world showing these scenarios it would be helpful.