| Commit message (Collapse) | Author | Age |
| |
|
|
|
|
| |
'y_min' and 'y_max' are still accepted for compatibility.
|
|
|
|
|
|
|
|
| |
* ObjectRef::set_local_animation: fix wrong lua return (should push a boolean, currently returns nil)
* ObjectRef::set_eye_offset: fix wrong lua return (should push a boolean, currently returns nil)
* Fix various Server functions which depends on RemotePlayer objet and return true/false when player object is nil whereas it's a caller implementation error. Change those bool functions to void and add sanitize_check call instead. Current callers are always checking player object validity
* Optimize Server::setClouds : use CloudParams object ref instead of attribute deserialization from structure & perform RemotePlayer::setCloudParams directly in server class like many other calls
* Optimize Server::SendCloudParams: use CloudParams object ref instead of deserialized attributes
|
|
|
|
| |
Call directly accessible RemotePlayer::getHotbarSelectedImage() from server api
|
|
|
|
| |
Call directly accessible RemotePlayer::getHotbarImage() from server api & make it const ref
|
|
|
|
| |
Call directly accessible RemotePlayer::getHotbarItemcount() from server api
|
|
|
|
|
|
|
| |
(#7108)
* mods.cpp/h: little performance improvement in getModsInPath
|
|
|
|
|
|
|
| |
Use the --color command line parameter instead of a setting for coloured logs
This fixes the missing warningstream bug, g_settings->get mustn't be used there.
Also, the decision about en- or disabling log colours fits better to the command line parameters than minetest settings.
|
|
|
|
|
|
|
|
|
| |
Change to --worldlist instead of --world list.
Gets rid of --worldpath parameter added as part of this pull request,
instead moving the listing function to a command --worldlist that
accepts either name, path, or both and prints out the corresponding
information.
|
|
|
|
|
|
| |
Returns the decoration ID for the provided decoration name string.
For use with gennotify, to know the decoration IDs for use in
'minetest.set_gen_notify'.
|
| |
|
| |
|
| |
|
|
|
|
| |
Use already existing collision results for the nearest colliding node
Fix slippery effect in free_move mode
|
| |
|
| |
|
|
|
|
|
|
| |
For 'place schematic' and 'place schematic on vmanip' APIs.
Fix 'place center' code to properly centre schematics.
Fix some comments.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
generating Visual Studio projects.
The TREE argument of source_group appears in CMake 3.8 therefore check for version.
|
|
|
|
| |
targets separately from common_SRCS
|
|
|
| |
Center text, text area doubled now
|
| |
|
| |
|
|
|
|
| |
Preserve the upper limit used in mgvalleys.
|
|
|
|
|
| |
user by the server. (#6878)
This doesn't check the fields in anyway whatsoever so it should only be seen as a way to mitigate exploits, a last line of defense to make it harder to exploit bugs in mods, not as a reason to not do all the usually checks.
|
| |
|
| |
|
| |
|
|
|
| |
Called when a client fails to supply the correct password for the account it's attempting to login as.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously only 'mapgen water source' and 'mapgen river water source'
were checked for. Games can use multiple liquid nodes defined for biomes,
many of which will not be aliased to those 2 mapgen aliases, causing
floating dungeons to generate in some liquids.
Now we check for liquid drawtype instead, so can remove liquid nodes
from dungeonparams.
Also check for 'airlike' drawtype instead of 'CONTENT_AIR' to avoid
generation in 'airlike' nodes in some rare situations. This will also be
needed for when we add definable biome air nodes.
|
|
|
|
|
|
|
| |
Cleanup:
* Drop unused Map::transforming_liquid_size()
* NodeNeighbor must use const ref for v3s16
* Add a missing default in a switch case
|
|
|
|
|
|
|
|
|
| |
* Rename IWritableNodeDefManager to NodeDefManager
* Make INodeDefManager functions const
* Use "const *NodeDefManager" instead of "*INodeDefManager"
* Remove unused INodeDefManager class
* Merge NodeDefManager and CNodeDefManager
* Document NodeDefManager
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
* Make hud_get return aligment and offset.
* Return size aswell.
|
|
|
| |
Also remove the unit test that tests the removed algorithms.
|
|
|
|
| |
* Account for walking speed in vertical dir
* Avoid undefined behaviour due to division-by-zero
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* Add minetest.bulk_set_node call + experimental mod unittest
* Optimize set_node function to prevent triple lookup on contentfeatures
Do only one lookup for old, and try to merge old and new lookup if node is same than previous node
* Add benchmark function + optimize vector population to have real results
|
| |
|
| |
|
|
|
| |
The CSM HUD PR caused some strange behavior including aborts due to parts of it using some slightly hacky code, the event refactor changing how events are processed and a minor oversight.
|
| |
|
|
|
|
|
| |
Reverted from commit 19960e26c672c6337f8c6ffbe27f2c6bca49750c
(* [CSM] add screenshot api lua)
|