It uses the (`const T &&... p_args`) forward reference, to avoid copying the
memory in case it's an rvalue, or pass a reference in case it's an lvalue.
This is an example:
```c++
PagedAllocator<btShapeBox> box_allocator;
btShapeBox* box = box_allocator.alloc( btVector3(1.0, 1.0, 1.0) );
```
Input buffering is implicitly used by event accumulation, but this commit makes it more generic so it can be enabled for other uses.
For desktop OSs it's currently not feasible given main and UI threads are the same).
- API has been simplified: all events now go through `parse_input_event()`. Whether they are accumulated or not depends on the `use_accumulated_input` flag.
- Event accumulation is now thread-safe (it was not needed so far, but it prepares the ground for the following changes).
- Touch drag events now support accumulation.
With this commit the macro `memnew_placement` uses the standard memory
placement syntax: `new (mem) TheClass()`, and removes the outdated and
not used syntax:
```
_ALWAYS_INLINE_ void *operator new(size_t p_size, void *p_pointer, size_t check, const char *p_description) {
```
Thanks to this change, the function `memnew_placement` call is compatible with
any class, and can also initialize classes with non-empty constructor:
```
// This is valid, like before.
memnew_placement(mem, Variant);
// This works too:
memnew_placement(mem, Variant(123));
```
This makes it clearer that this property is only about physics FPS,
not rendering FPS.
The `physics_fps` project setting was also renamed to
`physics_ticks_per_second` for consistency.
`PackedScene`s can be configured to be spawnable via a new
`MultiplayerAPI.spawnable_config` method.
They can be configured either to be spawned automatically when coming
from the server or to always require verification.
Another method, `MultiplayerAPI.send_spawn` lets you request a spawn on
the remote peers.
When a peer receive a spawn request:
- If it comes from the server and the scene is configured as
`SPAWN_MODE_SERVER`:
- Spawn the scene (instantiate it, add it to tree).
- Emit signal `network_spawn`.
- Else:
- Emit signal `network_spawn_request`.
In a similar way, `despawn`s are handled automatically in
`SPAWN_MODE_SERVER`.
In `SPAWN_MODE_SERVER`, when a new client connects it will also receive,
from the server all the spawned (and not yet despawned) instances.
The Transform::xform and xform_inv are made safe for Planes when using non-uniform scaling.
Optimization of calling sites to prevent loss of performance from the changes to xform(Plane).
These methods were broken by 22419082d9
5 years ago and nobody complained, so maybe they're not so useful...
But at least this should restore them to a working state.
- Fix library loading and initialization.
- Add extra methods/parameters in the interface needed by extenstions.
- Add Variant destructors and functions for extracting values and
creating Variants from values.
This sends the Godot object instance back to the extension so they can
keep a pointer for function calls.
Incidentally fix argument order on instance bindings callback for free()
- Fix type size information.
- Validate sizes at compile time (for the current build configuration
only).
- Normalize type names.
- Add extra information.
Adds a few more complex edge cases which are supported.
Also adds some documentation, simplifies the code a bit and forbids using
double quotes as a delimiter.
This seems to be a pretty old bug, older then originally reported (at
least under certain circumstances).
The IP singleton uses a resolve queue so developers can queue hostnames
for resolution in a separate while keeping the main thread unlocked
(address-resolution OS functions are blocking, and could block for a long
time in case of network disruption).
In most places though, the address resolution function was called with
the mutex locked, causing other functions (querying status, queueing
another hostname, ecc) to block until that resolution ended.
This commit ensures that all calls to OS address resolution are done
with the mutex unlocked.
Implemented "output" event
Refactored "seq" field generation
Prevent debugging when editor and client are in different projects
Removed unneeded references to peer on the parser
Refactored way to detect project path
Implemented "setBreakpoints" request
Fix double events when terminating from client
Refactored "stopped" event
Implemented "stopped" with breakpoint event
Implemented "stackTrace", "scopes" and "variables" request
Report incoming number of stack dump variables
Implemented proper reporting of scopes and variables from stack frames
Prevent editor from grabbing focus when a DAP session is active
Implemented "next" and "stepIn" requests
Implemented "Source" checksum computing
Switched expected errors from macros to silent guards
Refactored message_id
Respect client settings regarding lines/columns behavior
Refactored nested DAP fields
Implement reporting of "Members" and "Globals" scopes as well
Fix error messages not being shown, and improved wrong path message
Variants like dictionaries and arrays can have cyclic references, which
caused `encode_variant` to run an infinite recursion.
Instead of keeping a stack and looking for cyclic references which would
make serialization slower, this commit adds a `MAX_RECURSION_DEPTH`
constant to Variant, and have `encode_variant` keep track of the current
recursion depth, bailing when it's too high since this likely means a
cyclic reference has been encountered.
Used by ENetMultiplayerPeer and WebSocketServer to generate network IDs,
and exposed to the user for p2p networks (e.g. WebRTCMultiplayerPeer)
and custom MultiplayerPeer implementations.
This should fix various issues where retrieving enum values from
scripting languages would result in corrupted values (where 32 bits
were valid, and the other 32 random data).
Found via `codespell -q 3 -S ./thirdparty,*.po,./DONORS.md -L ackward,ang,ans,ba,beng,cas,childs,childrens,dof,doubleclick,fave,findn,hist,inout,leapyear,lod,nd,numer,ois,ony,paket,seeked,sinc,switchs,te,uint`
* Most resource types now have unique identifiers.
* Applies to text, binary and imported resources.
* File formats reference both by text and UID (when available). UID always has priority.
* Resource UIDs are 64 bits for better compatibility with the engine.
* Can be represented and used textually, example `uuid://dapwmgsmnl28u`.
* A special binary cache file is used and exported, containing the mappings.
Example of how it looks:
```GDScript
[gd_scene load_steps=2 format=3 uid="uid://dw86wq31afig2"]
[ext_resource type="PackedScene" uid="uid://bt36ojelx8q6c" path="res://subscene.scn" id="1_t56hs"]
```
GDScript, shaders and other special resource files can't currently provide UIDs, but this should be doable with special keywords on the files.
This will be reserved for future PRs.
* Clean-up of node_3d_editor_plugin.{h,cpp}: removed unused code, fixed some bugs.
* Moved node_3d_editor_gizmos.{h,cpp} to editor/plugins.
* Added support for multiple gizmos per node. This means custom gizmos will no longer override the built-in ones and that multiple gizmos can be used in more complex nodes.
* Added support for handle IDs. When adding handles to a gizmo, an ID can be specified for each one, making it easier to work with gizmos that have a variable number of handles.
* Added support for subgizmos, selectable elements that can be transformed without needing a node of their own. By overriding _subgizmo_intersect_frustum() and/or _subgizmo_intersect_ray() gizmos can define which subgizmos should be selected on a region or click selection. Subgizmo transformations are applied using get/set/commit virtual methods, similar to how handles work.
The various get_meta, set_meta, has_meta, get_meta_list, remove_meta
functions now uses StringName, allowing further optimizations via the
SNAME macro when used from C++ (this PR does not change the various
usage though).
* Friendlier with version control.
* Generates pseudo unique IDs, to minimize conflicts when merging, but still
user readable (so, not UUID).
* Eventually will also allow to have more precisely named sub-resources in
imported files.
* This will allow better reloading on changes (including resources already
loaded) as well as better keeping track of changes on the DCC.
* Keeps backward compatibility with the old formats.
* Binary and text format version incremented to mark breakage in forward
compatibility.
The `Math_INF` and `Math_NAN` defines were just aliases for those
constants, so we might as well use them directly.
Some portions of the code were already using `INFINITY` directly.
* Added explicit return type to the SNAME macro.
* Add some extra SNAME usages.
* Change some ClassDB methods to use const StringName & arguments.
* Cache the Window parent in Control because it's used in
is_layout_rtl(), which is called often.
* Only enable internal processing for viewports that need it.
* Change CanvasItem::group to be a StringName because it's only used as
that.
- Move the "sync" property for RPCs to RPCConfig.
- Unify GDScript annotations into a single one:
- `@rpc(master)` # default
- `@rpc(puppet)`
- `@rpc(any)` # former `@remote`
- Implement three additional `@rpc` options:
- The second parameter is the "sync" option (which also calls the
function locally when RPCing). One of "sync", "nosync".
- The third parameter is the transfer mode (reliable, unreliable,
ordered).
- The third parameter is the channel (unused for now).
* Added a new macro SNAME() that constructs and caches a local stringname.
* Subsequent usages use the cached version.
* Since these use a global static variable, a second refcounter of static usages need to be kept for cleanup time.
* Replaced all theme usages by this new macro.
* Replace all signal emission usages by this new macro.
* Replace all call_deferred usages by this new macro.
This is part of ongoing work to optimize GUI and the editor.
* This PR adds the ability to disable classes when building.
* For now it's only possible to do this via command like:
`scons disable_classes=RayCast2D,Area3D`
* Eventually, a proper UI will be implemented to create a build config file to do this at large scale, as well as detect what is used in the project.
Now the process uses a Map to lookup node pointers instead of iterating
over all modified node paths in a list and comparing them for each
property to check.
The process also avoids checking properties with empty node paths and
does an early exit on deleted nodes to avoid checking the node and its
descendants.
Also made a minor change in NodePath::rel_path_to() to avoid resizing a
Vector many times for long paths (with copy-on-write each time). Now
it's down to 2 resize calls in any case.
This PR implements range iterators in the base containers (Vector, Map, List, Pair Set).
Given several of these data structures will be replaced by more efficient versions, having a common iterator API will make this simpler.
Iterating can be done as follows (examples):
```C++
//Vector<String>
for(const String& I: vector) {
}
//List<String>
for(const String& I: list) {
}
//Map<String,int>
for(const KeyValue<String,int>&I : map) {
print_line("key: "+I.key+" value: "+itos(I.value));
}
//if intending to write the elements, reference can be used
//Map<String,int>
for(KeyValue<String,int>& I: map) {
I.value = 25;
//this will fail because key is always const
//I.key = "hello"
}
```
The containers are (for now) not STL compatible, since this would mean changing how they work internally (STL uses a special head/tail allocation for end(), while Godot Map/Set/List do not).
The idea is to change the Godot versions to be more compatible with STL, but this will happen after conversion to new iterators have taken place.
* The harcoded 8 slots are no more and impose limits in the new extension system.
* New system is limitless, although it will impose small performance hit with a mutex.
* Use a token to request the instance binding.
**Warning**: Mono will most likely break as a result of this, will need to be modified to use the new system.
* Lua table syntax uses named indexing: `{ mykey = myvalue }`
* Python style syntax uses string indexing: `{ "mykey" : myvalue }`
* Both are incompatible since a StringName key wont fetch a String key, hence confusing.
* This PR proposes always using String for indexing at a very minimal performance cost. Always indexing with StringNames will be faster, but they are considerably more expensive to create.
