Moving objects being added during instance_moving_create() were incorrectly not forcing a full check to find which room they were within. This could result in moving objects being re-added not correctly identifying their current room, and thus culling incorrectly. This PR forces a full check on calling instance_moving_create.
* Replace case-by-case extraction with PNAME & GNAME
* Fix group handling when group hint begins with property name
* Exclude properties that are PROPERTY_USAGE_NOEDITOR
Utility functions for NavigationServer2D/3D to find missing RID information when working with Server API directly. e.g. from map to regions and agents, from agent or region to map, from region to map and agents and so on ....
Requirement to work with NavigationServer API exklusive without SceneTree nodes and when juggling agents and regions between multiple navigation maps.
(cherry picked from commit 371054e3e5)
In order to prevent glitches when unhiding nodes, set_transform() is still called to the VisualServer even for hidden nodes when the node is interpolated. This activates a lot of logic which is not necessary just to keep the previous transform updated.
This PR adds an early out which misses out on the unnecessary logic when instances are invisible.
It has been pointed out to me that it is far more useful to display the NodePath in the warning than the name of the node, as there may be lots of nodes sharing the same name in a project. This PR fixes this.
When physics interpolation is active on a node, it is essential that transforms are updated during "_physics_process()" rather than "_process()" calls, for the interpolation to give the correct result.
This PR adds optional warnings for instances, cameras and multimeshes which can flag updates being incorrectly called, and thus make these problems much easier to fix.
Move VisualServer interpolation data out of Scenario and into VisualServerScene, so the interpolation data and enabled status is now common to all Scenarios.
Fix physics interpolation in multithreaded mode by ensuring tick and pre-draw are called.
In order to properly support the resource sharing paradigm, Occluders are split into Instances and Resources in the VisualServer. Instances are owned by a Scenario, and Resources are global. OccluderShape resources can now correctly be shared by multiple OccluderInstances.
Adds fixed timestep interpolation to the visual server.
Switchable on and off with project setting.
This version does not add new API for set_transform etc, when nodes have the interpolated flag set they will always use interpolation.
Using codespell 2.2-dev from current git.
Added `misc/scripts/codespell.sh` to make it easier to run it once in a
while and update the skip and ignore lists.
(cherry picked from commit 1bdb82c64e)
- Prevent debug enforced use of ubershader on shaders not supporting it
- Use unsigned integer for ubershader flags
- Add project setting for disabling async shader compilation on mobile
- Stop sampling some textures through different kinds of samplers at the same time
- Add a range hint to the Viewport `shadow_atlas_size` property.
- Add range steps to the shadow resolution project settings that
match the lowest allowed values (since lower increments don't
make sense).
When editor continuous redraws is switched off, the editor only redraws when a redraw_request was issued by an element in the scene. This works well in most situations, but when scenes have dynamic content they will continuously issue redraw_requests.
This can be fine on high power desktops but can be an annoyance on lower power machines.
This PR splits redraw requests into high and low priority requests, defaulting to high priority. Requests due to e.g. shaders using TIME are assigned low priority.
An extra editor setting is used to record the user preference and an extra option is added to the editor spinner menu, to allow the user to select between 3 modes:
* Continuous
* Update all changes
* Update vital changes
Due to an optimization to prevent processing except when camera rooms changed, the ticking synchronization and updating of previous and current lists could get out of sync for affected objects, leading to missing gameplay notifications.
This PR adds new paths to properly support and synchronize objects in this "room based" path.
Refactors the BVH to make it more generic and customizable. Instead of hard coding the system of pairable_mask and pairable_type into the BVH, this information is no longer stored internally, and instead the BVH uses callbacks both for determining whether pairs of objects can pair with each other, and for filtering cull / intersection tests.
In addition, instead of hard coding the number of trees, the BVH now supports up to 32 trees, and each object can supply a tree collision mask to determine which trees it can collide against.
This enables the BVH to scale to either the two or 3 trees needed in physics, and the single tree used without pairing in Godot 4 render tree.
The gameplay monitor wasn't being unloaded correctly in between levels. This meant that exit signals were not being sent, and entered signals for the new level were being missed.
This PR sends appropriate exit signals on unloading, and clear the data.
Change the entire navigation system.
Remove editor prefix from nav mesh generator class. It is now used for baking
at runtime as well.
Navigation supports obstacle avoidance now with the RVO2 library.
Nav system will also automatically link all nav meshes together to form one
overall complete nav map.
The line width of thick lines was being applied on both sides of the line, resulting in a line that was twice as thick as requested.
This PR fixes this embarrassing oversight.
In rare circumstances, changing the geometry data attached to an instance, there was the opporunity for the lighting_dirty flag to get out of sync, which could lead to access to a stale light RID, and warnings or worse.
This PR fixes the problem by ensuring the lighting is always updated on the instance when first adding GeometryData.
Margin needs to have a high enough value for test body motion to work
properly (separate using the margin, move without then gather rest info
with the margin again).
Fixes issues with test motion returning no collision in some cases with
margin equal to 0.
(cherry picked from commit 0c354047e1)
This PR adds a define BVH_EXPAND_LEAF_AABBS which is set, which stores expanded AABBs in the tree instead of exact AABBs.
This makes the logic less error prone when considering reciprocal collisions in the pairing, as all collision detect is now taking place between expanded AABB against expanded AABB, rather than expanded AABB against exact AABB.
The flip side of this is that the intersection tests will now be less exact when expanded margins are set.
All margins are now user customizable via project settings, and take account of collision pairing density to adjust the margin dynamically.
Updating the broadphase to find new collision pairs was done after
checking for collision islands, so it was working in most cases due to
the pairing margin used in the BVH, but in case of teleported objects
the narrowphase collision could be skipped.
Now it's done before checking for collision islands, so we can ensure
that broadphase pairing has been done at the same time as objects are
marked as moved so their collision can be checked properly.
This issue didn't happen in the Octree/HashGrid because they do nothing
on update and trigger pairs directly when objects move instead.
Applying overlay materials into multi-surface meshes currently
requires adding a next pass material to all the surfaces, which
might be cumbersome when the material is to be applied to a range
of different geometries. This also makes it not trivial to use
AnimationPlayer to control the material in case of visual effects.
The material_override property is not an option as it works
replacing the active material for the surfaces, not adding a new pass.
This commit adds the material_overlay property to GeometryInstance
(and therefore MeshInstance), having the same reach as
material_override (that is, all surfaces) but adding a new material
pass on top of the active materials, instead of replacing them.
Implemented in rasterizer of both GLES2 and GLES3.
Correct transformation of normals that works with a Basis containing non-uniform scale is difficult to get correct for those not familiar with the maths, it is also rather verbose and hard to read in calling code. This PR adds helper functions which both standardize the approach and make it clearer in calling code what is being done and why.
Previously a crude metric was used to decide on the roaming expansion margin, but it created unexpected results in some scenarios. Instead this setting is exposed to the user via the RoomManager, allowing them to tailor it to the world size, room sizes, roaming objects sizes and the speeds of movement.
In all physics servers, body_get_direct_state() now silently returns
nullptr when the body has been already freed or is removed from space,
so the client code can detect this state and invalidate the body rid.
In 2D, there is no change in behavior (just no more errors).
In 3D, the Bullet server returned a valid direct body state when the
body was removed from the physics space, but in this case it didn't
make sense to use the information from the body state.
Async. compilation via ubershader is currently available in the scene and particles shaders only.
Bonus:
- Use `#if defined()` syntax for not true conditionals, so they don't unnecessarily take a bit in the version flagset.
- Remove unused `ENABLE_CLIP_ALPHA` from scene shader.
- Remove unused `PARTICLES_COPY` from the particles shader.
- Remove unused uniform related code.
- Shader language/compiler: use ordered hash maps for deterministic code generation (needed for caching).
Changes the Path2D drawing to use POLYLINE instead of thick lines.
Add a path to translate thick lines (that are not using anti-aliasing) to draw as polygons instead. This should be faster because polygons can be batched.
Sets `AlignOperands` to `DontAlign`.
`clang-format` developers seem to mostly care about space-based indentation and
every other version of clang-format breaks the bad mismatch of tabs and spaces
that it seems to use for operand alignment. So it's better without, so that it
respects our two-tabs `ContinuationIndentWidth`.
