I left it in from another branch, didn't want to force push.
This reverts commit 58d2f59192.
Revert "Walls use `source` parameter for default hardness and blast resistance"
I left it in from another branch, didn't want to force push.
This reverts commit e8944cc145.
Revert "Fix prismarine bricks and dark variant blast res."
I left it in from another branch, didn't want to force push.
This reverts commit 6125d625bc.
Revert "Un-hardcode blast resistance, hardness for walls,"
I left it in from another branch, didn't want to force push.
This reverts commit 26e8737031.
Revert "All wood-type and nether-type fences now match material's blast resistance"
I left it in from another branch, didn't want to force push.
This reverts commit a1e20f2916.
- Notably, smooth sandstone (and red) is much more durable than other sandstone variants
- Ender chest isn't actually more explosion resistant than obsidian, hmm
Change blast resistance to 0.75, add an optional hardness override parameter, because cobble is unique here
(Still instant-minable, this is a separate issue)
The old code takes the first return val of `minetest.item_place_node'
which is `itemstack'. Therefore, the variable `success' in the old code
is always true. The new code takes the second val which will be nil if
an invalid node placement occured.
This check is necessary since the ladder may be placed in the front of
pointed block while there is a node with hole (slabs, fences etc.) at
the same place resulting an invalid placement and sound played when
it shouldn't be played.
LUAs `math.random(a,b)` expects a and b to be integers. `size` was only randomized once.
Reviewed-on: VoxeLibre/VoxeLibre#4583
Reviewed-by: Mikita Wiśniewski <rudzik8@protonmail.com>
Co-authored-by: kno10 <erich.schubert@gmail.com>
Co-committed-by: kno10 <erich.schubert@gmail.com>
Items are instead written as a serialized string into ItemStackMetaRef, and read from there as well. Old itemstacks get converted to the new format automatically.
Now you can define custom groups for the chests, as shown in the trapped
chests example (now they show up under "Mechanisms" tab in creative!).
I'm a bit suspicious of the new return-wrapped functions, as in *they
might break under some circumstances I didn't observe*. It will require
some heavy testing to make sure nothing crashes in the future.
Also, `on_rightclick` for double chests is something I really want to
return-wrap as well, but failed to do so. 🤷
Pass chest entity initialization data to on_activate as staticdata so
initialization is atomic, preventing premature deletion of chest entity
by concurrent server steps.