Add `PackedVector4Array` Variant type #85474

fire · 2023-11-28T14:56:23Z

Was curious how hard it was.

Bugsquad edit: Keywords for easy searching: PackedVector4Array, Vector<Vector4>

core/variant/variant_internal.h

AThousandShips · 2023-11-28T14:59:55Z

Unfortunately swapping current TypedArray<Vector4> to PackedVector4Array breaks compat it'd be nice to know the scope of where this can be applied to see the possible improvements from adding it 🙂

modules/lightmapper_rd/lightmapper_rd.cpp

core/extension/gdextension_interface.h

core/io/marshalls.cpp

fire · 2023-11-28T19:29:36Z

We had an argument about PackedVector4Array Type in Godot Engine and here's the ai summary.

Key Points from Lyuma:

Suggests using PackedColorArray instead of creating a new PackedVector4Array type due to potential casting capabilities between Color, Quaternion, and Vector4.
Acknowledges that Color is float32 and Vector4 could be float64 or float32, but sees no real need for Vector4 to be double precision.
Notes the convenience of being able to cast directly between color arrays, quaternion arrays, and shader arrays.
Points out that while individual floats in GDScript are doubles, PackedFloat32Array used for shader inputs is not.

Key Points from iFire:

Argues that Color and Vector4 are not the same; Color can be different numerically (e.g., srgb color vs. color).
States that Vector2 and Vector3 are dual precision and wishes to maintain parity by having Vector4 with similar precision.
Believes he needs a compelling argument to remove double precision from Vector4, as it matches the current precision types in real_t, Vector2, and Vector3.
Mentions that shader input has its own system for conversion to vec4[], and that the problem of precision also occurs with vec3[] and vec2[], hence viewing the double precision issue as not unique to Vector4.

Conclusion:

Both participants highlight various aspects of type compatibility, precision, and practical use cases within Godot's engine. Lyuma seems more inclined toward simplifying things without necessarily enforcing double precision for Vector4, whereas iFire emphasizes consistency across vector types and existing systems that handle precision in shaders.

The discussion centers around the implications of introducing a separate PackedVector4Array type versus adapting current arrays and their interactions with other data types, along with shader input considerations.

fire

There was an unresolved comment about asking reduz about bumping the version, but I am ok either way.

akien-mga · 2024-04-15T22:25:49Z

Ah yes thanks for the reminder, I'll check that with reduz.

dalexeev · 2024-05-02T16:38:59Z

Missed here, was recently added in #82952:

godot/modules/gdscript/gdscript_parser.cpp

Lines 4126 to 4133 in 4e9543d

    
           case Variant::VECTOR3: 
        
           	types.push_back("PackedVector3Array"); 
        
           	break; 
        
           case Variant::COLOR: 
        
           	types.push_back("PackedColorArray"); 
        
           	break; 
        
           default: 
        
           	break;

AThousandShips · 2024-05-02T16:43:15Z

Also here:

diff --git a/core/io/marshalls.cpp b/core/io/marshalls.cpp
index 03677c6649..a9ffa15130 100644
--- a/core/io/marshalls.cpp
+++ b/core/io/marshalls.cpp
@@ -1327,6 +1327,7 @@ Error encode_variant(const Variant &p_variant, uint8_t *r_buffer, int &r_len, bo
                case Variant::VECTOR4:
                case Variant::PACKED_VECTOR2_ARRAY:
                case Variant::PACKED_VECTOR3_ARRAY:
+               case Variant::PACKED_VECTOR4_ARRAY:
                case Variant::TRANSFORM2D:
                case Variant::TRANSFORM3D:
                case Variant::PROJECTION:

And possibly here:

diff --git a/modules/mono/editor/bindings_generator.cpp b/modules/mono/editor/bindings_generator.cpp
index 9aef338891..ac34f74474 100644
--- a/modules/mono/editor/bindings_generator.cpp
+++ b/modules/mono/editor/bindings_generator.cpp
@@ -330,6 +330,8 @@ String BindingsGenerator::bbcode_to_text(const String &p_bbcode, const TypeInter
                                output.append("'" BINDINGS_NAMESPACE ".Vector3[]'");
                        } else if (tag == "PackedColorArray") {
                                output.append("'" BINDINGS_NAMESPACE ".Color[]'");
+                       } else if (tag == "PackedVector4Array") {
+                               output.append("'" BINDINGS_NAMESPACE ".Vector4[]'");
                        } else {
                                const TypeInterface *target_itype = _get_type_or_null(TypeReference(tag));

@@ -646,6 +648,8 @@ String BindingsGenerator::bbcode_to_xml(const String &p_bbcode, const TypeInterf
                                xml_output.append("<see cref=\"" BINDINGS_NAMESPACE ".Vector3\"/>[]");
                        } else if (tag == "PackedColorArray") {
                                xml_output.append("<see cref=\"" BINDINGS_NAMESPACE ".Color\"/>[]");
+                       } else if (tag == "PackedVector4Array") {
+                               xml_output.append("<see cref=\"" BINDINGS_NAMESPACE ".Vector4\"/>[]");
                        } else {
                                const TypeInterface *target_itype = _get_type_or_null(TypeReference(tag));

diff --git a/modules/mono/editor/bindings_generator.h b/modules/mono/editor/bindings_generator.h
index bb0ba0cb00..fce530a832 100644
--- a/modules/mono/editor/bindings_generator.h
+++ b/modules/mono/editor/bindings_generator.h
@@ -689,7 +689,7 @@ class BindingsGenerator {
                StringName type_Vector4i = StaticCString::create("Vector4i");

                // Object not included as it must be checked for all derived classes
-               static constexpr int nullable_types_count = 18;
+               static constexpr int nullable_types_count = 19;
                StringName nullable_types[nullable_types_count] = {
                        type_String,
                        type_StringName,
@@ -711,6 +711,7 @@ class BindingsGenerator {
                        StaticCString::create(_STR(PackedVector2Array)),
                        StaticCString::create(_STR(PackedVector3Array)),
                        StaticCString::create(_STR(PackedColorArray)),
+                       StaticCString::create(_STR(PackedVector4Array)),
                };

                bool is_nullable_type(const StringName &p_type) const {

akien-mga · 2024-05-02T17:02:40Z

Indeed, thanks both. I had found some of these while doing another thorough review after rebase, but missed the XML stuff in bindings_generator.cpp.

I also found a handful more places where we were missing the PackedVector4Array equivalent (comparing with all PackedVector3Array usage/handling).

And I fixed a bug where type hinted PackedVector4Arrays in the Inspector were not properly editable as arrays of Vector4. We were missing to pass the hint string along to EditorPropertyArray.

Still pending: Evaluating the compatibility breakage in the text and binary resource formats, and see how to handle it gracefully (like we did for PackedByteArray serialization changes).

fire · 2024-05-02T18:34:16Z

With all this work, I was reminded of this quote:

The Programmers’ Credo: we do these things not because they are easy, but because we thought they were going to be easy

akien-mga · 2024-05-02T22:38:22Z

Still pending: Evaluating the compatibility breakage in the text and binary resource formats, and see how to handle it gracefully (like we did for PackedByteArray serialization changes).

I checked and indeed the changes would break forward compat in 4.2.

I added this diff here to bump the format version for the binary format too (like we did already for the text format when adding the PackedByteArray base64 serialization), while still saving by default in the previous format if possible.

diff --git a/core/io/resource_format_binary.cpp b/core/io/resource_format_binary.cpp
index 99cffde082..ab460c5f4c 100644
--- a/core/io/resource_format_binary.cpp
+++ b/core/io/resource_format_binary.cpp
@@ -90,11 +90,12 @@ enum {
 	OBJECT_EXTERNAL_RESOURCE = 1,
 	OBJECT_INTERNAL_RESOURCE = 2,
 	OBJECT_EXTERNAL_RESOURCE_INDEX = 3,
-	// Version 2: added 64 bits support for float and int.
-	// Version 3: changed nodepath encoding.
-	// Version 4: new string ID for ext/subresources, breaks forward compat.
+	// Version 2: Added 64-bit support for float and int.
+	// Version 3: Changed NodePath encoding.
+	// Version 4: New string ID for ext/subresources, breaks forward compat.
 	// Version 5: Ability to store script class in the header.
-	FORMAT_VERSION = 5,
+	// Version 6: Added PackedVector4Array Variant type.
+	FORMAT_VERSION = 6,
 	FORMAT_VERSION_CAN_RENAME_DEPS = 1,
 	FORMAT_VERSION_NO_NODEPATH_PROPERTY = 3,
 };
diff --git a/scene/resources/resource_format_text.cpp b/scene/resources/resource_format_text.cpp
index f70a00a9f6..2e27ac9198 100644
--- a/scene/resources/resource_format_text.cpp
+++ b/scene/resources/resource_format_text.cpp
@@ -37,11 +37,12 @@
 #include "core/object/script_language.h"
 #include "core/version.h"
 
-// Version 2: changed names for Basis, AABB, Vectors, etc.
-// Version 3: new string ID for ext/subresources, breaks forward compat.
-// Version 4: PackedByteArray is now stored as base64 encoded.
-#define FORMAT_VERSION_COMPAT 3
+// Version 2: Changed names for Basis, AABB, Vectors, etc.
+// Version 3: New string ID for ext/subresources, breaks forward compat.
+// Version 4: PackedByteArray can be base64 encoded, and PackedVector4Array was added.
 #define FORMAT_VERSION 4
+// For compat, save as version 3 if not using PackedVector4Array or no big PackedByteArray.
+#define FORMAT_VERSION_COMPAT 3
 
 #define BINARY_FORMAT_VERSION 4
 
@@ -1979,6 +1980,9 @@ void ResourceFormatSaverTextInstance::_find_resources(const Variant &p_variant,
 				use_compat = false;
 			}
 		} break;
+		case Variant::PACKED_VECTOR4_ARRAY: {
+			use_compat = false;
+		} break;
 		default: {
 		}
 	}

PR #91486 adds compat code to 4.2.3 (and 4.1.5 once cherry-picked) so that those versions can properly parse scenes and resources using PackedVector4Array.

Properties of that type are converted to plain Arrays, and thus on the trip back to 4.3 there's a risk of losing data.
Here's the diff in my test scene going from 4.3 with this PR to 4.2 with #91486:

diff --git a/node_3d.tscn b/node_3d.tscn
index 64be59a..72aa829 100644
--- a/node_3d.tscn
+++ b/node_3d.tscn
@@ -1,4 +1,4 @@
-[gd_scene load_steps=4 format=4 uid="uid://bs0waftols4bk"]
+[gd_scene load_steps=4 format=3 uid="uid://bs0waftols4bk"]
 
 [ext_resource type="Texture2D" uid="uid://4siemblwxumr" path="res://icon.svg" id="1_r5ffe"]
 [ext_resource type="Script" path="res://Sprite3D.gd" id="2_u0x6a"]
@@ -15,13 +15,13 @@ mesh = SubResource("BoxMesh_ku6f7")
 transform = Transform3D(2, 0, 0, 0, 2, 0, 0, 0, 2, 0, 2, 0)
 texture = ExtResource("1_r5ffe")
 script = ExtResource("2_u0x6a")
-pba1 = PackedByteArray("MnsQF/8slhgAAAAAAAB7ewADAIsAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA")
+pba1 = PackedByteArray(50, 123, 16, 23, 255, 44, 150, 24, 0, 0, 0, 0, 0, 0, 123, 123, 0, 3, 0, 139, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
 pba2 = [51, 123, 15]
 pv3a2 = [Vector3(1, 1, 1), Vector3(2, 2, 2), Vector3(3, 3, 3), null, Vector3(5, 5, 5)]
-pv4a1 = PackedVector4Array(11, 22, 33, 44, 19, 29, 39, 49, -8, -123, 14, 129)
+pv4a1 = [Vector4(11, 22, 33, 44), Vector4(19, 29, 39, 49), Vector4(-8, -123, 14, 129)]
 pv4a2 = [Vector4(9, 82, 7, 7), Vector4(1, 1, 1, 1), Vector4(0, 2, 2, 2), Vector4(3, 3, 3, 3)]
 pv4a3 = [Vector4(10, 20, 30, 40), Vector4(110, 120, 130, 140), null, Vector4(2, 2.22, 2, 2), Vector4(141, 3123.12, 21390.1, 1.1), Vector4(0, 0, 0, 0), Vector4(3, 3, 3, 3)]
-metadata/meta_pv4a = PackedVector4Array(0, 0, 0, 0, 1, 2, 3, 4, 10, 20, 30, 40, 0, 0, 0, 0, 100, 200.5, 300.8, 400.9)
+metadata/meta_pv4a = [Vector4(0, 0, 0, 0), Vector4(1, 2, 3, 4), Vector4(10, 20, 30, 40), Vector4(0, 0, 0, 0), Vector4(100, 200.5, 300.8, 400.9)]
 
 [node name="Sprite3D2" type="Sprite3D" parent="."]
 transform = Transform3D(-8.74228e-08, 0, 2, 0, 2, 0, -2, 0, -8.74228e-08, -3.32284, 2, 0)

So PackedVector4Array(...) gets converted to [Vector4(...), ...], so here the values aren't lost, just the main container format.

On the way back however, those properties will stay Arrays, and some of them can be lost:

diff --git a/node_3d.tscn b/node_3d.tscn
index 72aa829..d28b7f4 100644
--- a/node_3d.tscn
+++ b/node_3d.tscn
@@ -1,4 +1,4 @@
-[gd_scene load_steps=4 format=3 uid="uid://bs0waftols4bk"]
+[gd_scene load_steps=4 format=4 uid="uid://bs0waftols4bk"]
 
 [ext_resource type="Texture2D" uid="uid://4siemblwxumr" path="res://icon.svg" id="1_r5ffe"]
 [ext_resource type="Script" path="res://Sprite3D.gd" id="2_u0x6a"]
@@ -15,10 +15,9 @@ mesh = SubResource("BoxMesh_ku6f7")
 transform = Transform3D(2, 0, 0, 0, 2, 0, 0, 0, 2, 0, 2, 0)
 texture = ExtResource("1_r5ffe")
 script = ExtResource("2_u0x6a")
-pba1 = PackedByteArray(50, 123, 16, 23, 255, 44, 150, 24, 0, 0, 0, 0, 0, 0, 123, 123, 0, 3, 0, 139, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
+pba1 = PackedByteArray("MnsQF/8slhgAAAAAAAB7ewADAIsAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA")
 pba2 = [51, 123, 15]
 pv3a2 = [Vector3(1, 1, 1), Vector3(2, 2, 2), Vector3(3, 3, 3), null, Vector3(5, 5, 5)]
-pv4a1 = [Vector4(11, 22, 33, 44), Vector4(19, 29, 39, 49), Vector4(-8, -123, 14, 129)]
 pv4a2 = [Vector4(9, 82, 7, 7), Vector4(1, 1, 1, 1), Vector4(0, 2, 2, 2), Vector4(3, 3, 3, 3)]
 pv4a3 = [Vector4(10, 20, 30, 40), Vector4(110, 120, 130, 140), null, Vector4(2, 2.22, 2, 2), Vector4(141, 3123.12, 21390.1, 1.1), Vector4(0, 0, 0, 0), Vector4(3, 3, 3, 3)]
 metadata/meta_pv4a = [Vector4(0, 0, 0, 0), Vector4(1, 2, 3, 4), Vector4(10, 20, 30, 40), Vector4(0, 0, 0, 0), Vector4(100, 200.5, 300.8, 400.9)]

Here's how those properties are defined:

@export var pba1 : PackedByteArray
@export var pba2 : PackedByteArray = []
@export var pba3 : PackedByteArray = [0, 2, 13, 8, 12, 151, 90, 42]

@export var pv3a2 : PackedVector3Array = []

@export var pv4a1 : PackedVector4Array
@export var pv4a2 : PackedVector4Array = []
@export var pv4a3 : PackedVector4Array = [Vector4(10, 20, 30, 40), Vector4(110, 120, 130, 140)]

I think this is a GDScript issue outside the scope of this PR.

There might still be areas where we would need to start handling PACKED_VECTOR4_ARRAY. I did a quick comparison with files where PACKED_VECTOR3_ARRAY is included but the VECTOR4 one wasn't, and found this which is likely worth looking into:
diff --git a/servers/rendering/shader_language.cpp b/servers/rendering/shader_language.cpp
index 8fbad346a4..90b1492d2f 100644
--- a/servers/rendering/shader_language.cpp
+++ b/servers/rendering/shader_language.cpp
@@ -4183,7 +4183,7 @@ PropertyInfo ShaderLanguage::uniform_to_property_info(const ShaderNode::Uniform
 				if (p_uniform.hint == ShaderLanguage::ShaderNode::Uniform::HINT_SOURCE_COLOR) {
 					pi.type = Variant::PACKED_COLOR_ARRAY;
 				} else {
-					pi.type = Variant::PACKED_FLOAT32_ARRAY;
+					pi.type = Variant::PACKED_VECTOR4_ARRAY;
 				}
 			} else {
 				if (p_uniform.hint == ShaderLanguage::ShaderNode::Uniform::HINT_SOURCE_COLOR) {
But I didn't commit it, I think further work will be needed to review all places where either PACKED_COLOR_ARRAY or PACKED_FLOAT32_ARRAY were used for vec4 in shader code, and could be replaced by PACKED_VECTOR4_ARRAY, provided we properly handle compatibility with existing data.

I actually went ahead and committed this change earlier today. Needs testing of shader uniforms to confirm it's still working, if not I can roll it back.

Co-authored-by: A Thousand Ships <96648715+AThousandShips@users.noreply.github.com> Co-authored-by: Rémi Verschelde <rverschelde@gmail.com>

akien-mga · 2024-05-02T23:02:04Z

I actually went ahead and committed this change earlier today. Needs testing of shader uniforms to confirm it's still working, if not I can roll it back.

Ok that didn't work, so I reverted this part of the change. vec4 uniform arrays still (de)serialize to PackedFloat32Array for now.

Here's a start for a follow-up if someone wants to tackle changing it PackedVector4Array:

diff --git a/servers/rendering/shader_language.cpp b/servers/rendering/shader_language.cpp
index 99b3f54379..9231a91732 100644
--- a/servers/rendering/shader_language.cpp
+++ b/servers/rendering/shader_language.cpp
@@ -3962,12 +3962,9 @@ Variant ShaderLanguage::constant_value_to_variant(const Vector<ShaderLanguage::C
 						}
 						value = Variant(array);
 					} else {
-						PackedFloat32Array array;
+						PackedVector4Array array;
 						for (int i = 0; i < array_size; i += 4) {
-							array.push_back(p_value[i].real);
-							array.push_back(p_value[i + 1].real);
-							array.push_back(p_value[i + 2].real);
-							array.push_back(p_value[i + 3].real);
+							array.push_back(Vector4(p_value[i].real, p_value[i + 1].real, p_value[i + 2].real, p_value[i + 3].real));
 						}
 						value = Variant(array);
 					}
@@ -4183,7 +4180,7 @@ PropertyInfo ShaderLanguage::uniform_to_property_info(const ShaderNode::Uniform
 				if (p_uniform.hint == ShaderLanguage::ShaderNode::Uniform::HINT_SOURCE_COLOR) {
 					pi.type = Variant::PACKED_COLOR_ARRAY;
 				} else {
-					pi.type = Variant::PACKED_FLOAT32_ARRAY;
+					pi.type = Variant::PACKED_VECTOR4_ARRAY;
 				}
 			} else {
 				if (p_uniform.hint == ShaderLanguage::ShaderNode::Uniform::HINT_SOURCE_COLOR) {

The implementation should make sure to preserve compatibility for pre-existing shader params.

akien-mga

I think this is good to go now.

AThousandShips · 2024-05-03T07:56:54Z

Should we get a @godotengine/dotnet look at the added cases above to check if we've missed anything? I just did a search when I found those cases and I'm not familiar with the mono side

AThousandShips

Might pick up the shader code task, will look once we merge

Edit: Won't be focusing on it at the moment

akien-mga · 2024-05-03T10:17:00Z

Did a quick C# test and base functionality seems to work:

akien-mga · 2024-05-03T10:41:41Z

Thanks to everyone involved!

AThousandShips added enhancement topic:core labels Nov 28, 2023

AThousandShips added this to the 4.x milestone Nov 28, 2023

fire force-pushed the packedvector4array branch from b6359dc to 32ec12f Compare November 28, 2023 14:57

AThousandShips reviewed Nov 28, 2023

View reviewed changes

core/variant/variant_internal.h Outdated Show resolved Hide resolved

core/variant/variant_internal.h Outdated Show resolved Hide resolved

fire force-pushed the packedvector4array branch from 32ec12f to 26ac16c Compare November 28, 2023 14:58

fire force-pushed the packedvector4array branch 11 times, most recently from c4a697b to d2590f5 Compare November 28, 2023 16:57

AThousandShips reviewed Nov 28, 2023

View reviewed changes

modules/lightmapper_rd/lightmapper_rd.cpp Outdated Show resolved Hide resolved

fire force-pushed the packedvector4array branch 5 times, most recently from 9ac6de8 to 45e3191 Compare November 28, 2023 18:01

fire commented Nov 28, 2023

View reviewed changes

core/extension/gdextension_interface.h Outdated Show resolved Hide resolved

fire force-pushed the packedvector4array branch from 1adbb54 to b989023 Compare November 28, 2023 18:43

fire commented Nov 28, 2023

View reviewed changes

core/io/marshalls.cpp Outdated Show resolved Hide resolved

fire force-pushed the packedvector4array branch from b989023 to 60f90ed Compare November 28, 2023 18:52

fire marked this pull request as ready for review November 28, 2023 19:33

akien-mga added this to the 4.3 milestone Apr 15, 2024

akien-mga removed the needs work label Apr 15, 2024

fire commented Apr 15, 2024

View reviewed changes

akien-mga mentioned this pull request May 2, 2024

DAP: Fix typo in parsing of PackedVector3Array #91468

Merged

akien-mga force-pushed the packedvector4array branch 2 times, most recently from ef96abe to d929f27 Compare May 2, 2024 17:00

akien-mga mentioned this pull request May 2, 2024

[4.2] Add forward compat parsing of PackedVector4Array #91486

Merged

akien-mga force-pushed the packedvector4array branch from d929f27 to 4608dc3 Compare May 2, 2024 22:30

Add PackedVector4Array Variant type

f9b4885

Co-authored-by: A Thousand Ships <96648715+AThousandShips@users.noreply.github.com> Co-authored-by: Rémi Verschelde <rverschelde@gmail.com>

akien-mga force-pushed the packedvector4array branch from 4608dc3 to f9b4885 Compare May 2, 2024 23:00

akien-mga approved these changes May 2, 2024

View reviewed changes

akien-mga mentioned this pull request May 2, 2024

Remove unused code to convert text resource format to binary #91488

Merged

AThousandShips approved these changes May 3, 2024

View reviewed changes

akien-mga mentioned this pull request May 3, 2024

Ability to convert native engine types to JSON and back. #91503

Closed

akien-mga merged commit 03e6fbb into godotengine:master May 3, 2024
16 checks passed

dsnopek mentioned this pull request May 6, 2024

Add support for PackedVector4Array godotengine/godot-cpp#1456

Merged

fire deleted the packedvector4array branch May 11, 2024 01:00

caimantilla mentioned this pull request May 23, 2024

Add an AABBi type godotengine/godot-proposals#3679

Open

CedNaru mentioned this pull request Jun 1, 2024

Add new core type PackedVector4Array utopia-rise/godot-kotlin-jvm#646

Closed

akien-mga mentioned this pull request Jun 26, 2024

Use PackedVector4Array instead of float array for vec4 array uniform #93635

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add `PackedVector4Array` Variant type #85474

Add `PackedVector4Array` Variant type #85474

fire commented Nov 28, 2023 •

edited

Loading

AThousandShips commented Nov 28, 2023

fire commented Nov 28, 2023 •

edited

Loading

fire left a comment •

edited

Loading

akien-mga commented Apr 15, 2024

dalexeev commented May 2, 2024

AThousandShips commented May 2, 2024 •

edited

Loading

akien-mga commented May 2, 2024 •

edited

Loading

fire commented May 2, 2024

akien-mga commented May 2, 2024

akien-mga commented May 2, 2024

akien-mga left a comment

AThousandShips commented May 3, 2024

AThousandShips left a comment •

edited

Loading

akien-mga commented May 3, 2024

akien-mga commented May 3, 2024

Add PackedVector4Array Variant type #85474

Add PackedVector4Array Variant type #85474

Conversation

fire commented Nov 28, 2023 • edited Loading

AThousandShips commented Nov 28, 2023

fire commented Nov 28, 2023 • edited Loading

Key Points from Lyuma:

Key Points from iFire:

Conclusion:

fire left a comment • edited Loading

Choose a reason for hiding this comment

akien-mga commented Apr 15, 2024

dalexeev commented May 2, 2024

AThousandShips commented May 2, 2024 • edited Loading

akien-mga commented May 2, 2024 • edited Loading

fire commented May 2, 2024

akien-mga commented May 2, 2024

akien-mga commented May 2, 2024

akien-mga left a comment

Choose a reason for hiding this comment

AThousandShips commented May 3, 2024

AThousandShips left a comment • edited Loading

Choose a reason for hiding this comment

akien-mga commented May 3, 2024

akien-mga commented May 3, 2024

Add `PackedVector4Array` Variant type #85474

Add `PackedVector4Array` Variant type #85474

fire commented Nov 28, 2023 •

edited

Loading

fire commented Nov 28, 2023 •

edited

Loading

fire left a comment •

edited

Loading

AThousandShips commented May 2, 2024 •

edited

Loading

akien-mga commented May 2, 2024 •

edited

Loading

AThousandShips left a comment •

edited

Loading