Age | Commit message (Collapse) | Author |
|
SCons: Re-allow upcoming GCC 8.4, fixes C++17 copy elision
|
|
Follow-up to #36484.
The patches for https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86521 have now
landed in the `releases/gcc-8` branch and will be in GCC 8.4.
|
|
Added missing destructor for Navigation2D
|
|
Improve LSP completion using scene owner
|
|
Improved docs wording and added links
|
|
- Added missing links to the Control node in BoxContainer
- Added Oxford commas in BoxContainer and CanvasItem
- Clarified ambiguous boolean wording in BoxContainer
- Improved paragraphing in ScrollContainer's description
- Simplified ControlPicker description
|
|
Fixes: #36473
|
|
Although destructor call was missing, it still doesn't heal #36537 memory leaks. Further description how that might be overcome - on GitHub
Partialy covers #36537
|
|
doc: Sync classref for Packed{Int,Float}{32,64}Array additions
|
|
|
|
Fixed remove region method.
|
|
Update PackedInt/FloatArray icons for the new types
|
|
|
|
Update docs and bindings for new integer vector types
|
|
Fix debugger crash inspecting freed object.
|
|
|
|
Variant: Added 64-bit packed arrays, renamed Variant::REAL to FLOAT.
|
|
This seems to be the correct way to validate a reference.
Why is cast_to failing? Is this the correct way of checking if the
object is valid?
|
|
Expression: Fix parsing integers as 32-bit
|
|
Fixed crash when connecting a signal in GDScript
|
|
|
|
- Renames PackedIntArray to PackedInt32Array.
- Renames PackedFloatArray to PackedFloat32Array.
- Adds PackedInt64Array and PackedFloat64Array.
- Renames Variant::REAL to Variant::FLOAT for consistency.
Packed arrays are for storing large amount of data and creating stuff like
meshes, buffers. textures, etc. Forcing them to be 64 is a huge waste of
memory. That said, many users requested the ability to have 64 bits packed
arrays for their games, so this is just an optional added type.
For Variant, the float datatype is always 64 bits, and exposed as `float`.
We still have `real_t` which is the datatype that can change from 32 to 64
bits depending on a compile flag (not entirely working right now, but that's
the idea). It affects math related datatypes and code only.
Neither Variant nor PackedArray make use of real_t, which is only intended
for math precision, so the term is removed from there to keep only float.
|
|
Use PascalCase file names for editor icons
|
|
Travis: Use Python 3.8 alias instead of specific version
|
|
Otherwise it breaks when they update the container to a new version,
like they did today with 3.7.6.
|
|
Fix project.godot for projects with class_name
|
|
Re-Remove this signal call that was mistakenly added in #36244
|
|
Fixes #36438
|
|
Using PascalCase file names without any conversion step is
less confusing to new contributors.
|
|
|
|
Core: Change _Marshall class from inherit Reference to Object
|
|
|
|
Scons: fixed build for vanilla clang in mac os x
|
|
|
|
The original change was in #36340
|
|
Fix some signals and non-debug branch for callable_mp
|
|
Add icons for some new variants
|
|
Changed default capsule axis to vertical
|
|
These bugs existed since those lines were added, so I assume that
their intended use is no longer relevant.
|
|
Was missed in #36393 because no `callable_mp()` calls were actually
compiled with `tools=no` in that PR.
Also work around GCC warning that also affects the
`call_with_variant_args_ret_helper` variant.
|
|
|
|
Store arrays inside of Variant as shared.
|
|
Arrays inside of Variant are unique and use reference counting.
When you assign a variant containing a packed array to another, or
when you call non const functions to arrays, this will work even
if the array is inside a dictionary, so they will from now pass
as reference.
The difference with regular variant arrays is that, once passed
to a function in the C++ API, they are no longer shared. This is
required for security and thread safety, as those arrays are
mainly used to pass data back and forth even between threads.
|
|
Fix: auto brace complete for quoted strings
|
|
Fix crash after closing a GDScript LSP session
|
|
SCons: Add GCC/Clang minimum version check
|
|
Co-authored-by: Hugo Locurcio <https://hugo.pro>
|
|
Prevent using GCC 8 as it does not properly support C++17's
guaranteed copy elision which we now need.
(Upstream bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86521)
Follow-up to #36457 and #36436.
|
|
|
|
|