Add Procedural Layer support for Weightmaps, including Brush support that affect Weightmap and/or Heightmap.
#jira UE-61718:
#rb Jack.Porter
[FYI] Jeremy.Moore, Richard.Malo
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: michael.dupuis
#ROBOMERGE-SOURCE: CL 5433816 via CL 5437700 via CL 5437807
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5440742 by michael dupuis in Dev-Networking branch]
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: nick.darnell
#ROBOMERGE-SOURCE: CL 5430467 via CL 5430470 via CL 5431124 via CL 5437649 via CL 5437709
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5440524 by nick darnell in Dev-Networking branch]
Change summary:
- Modified UBlueprintGeneratedClass::GetDefaultObjectPreloadDependencies() to include non-inherited SCS node objects as a preload dependency. This ensures that the reference they hold to the associated component template object will be serialized prior to the point at which we need to use it to generate the runtime data that's needed to support the fast path.
- Modified UBlueprintGeneratedClass::SerializeDefaultObject() to better handle the state in which we cannot generate runtime data to support the fast path when cooked data is present (i.e. emit a warning and fall back to the slow path). Previously we'd end up in a state where an ensure() was later triggered at instancing time when cooked data is present (which I've been using to catch unexpected cook results). Leaving the ensure() in place for now, but it should no longer be triggered.
- Renamed FBlueprintCookedComponentInstancingData::bIsValid for clarity (per code review).
#rb Dan.OConnor
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: phillip.kavan
#ROBOMERGE-SOURCE: CL 5430354 via CL 5430391 via CL 5433546 via CL 5433610
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5436713 by phillip kavan in Dev-Networking branch]
Reasons for the flush:
In Editor realtime rendering changing a material connection (or via undo/redo) changes what is rendered on the RT (specifically in FMeshDecalMeshProcessor::AddMeshBatch() Emmissive connection test)
before the Shader Map has been updated in PostEditChangeProperty() - Viewport render command enqueue has already happenned so we have a data race and we need to flush that before the material change.
#jira UE-70565
#rb mark.satterthwaite
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: richard.wallis
#ROBOMERGE-SOURCE: CL 5418937 in //UE4/Release-4.22/... via CL 5418940
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5421514 by richard wallis in Dev-Networking branch]
This CL also add an internal delay per frame, allowing to wait a certain amount of time before capturing a movie frame.
#jira none
#rb juan.canada
#rb max.chen
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: charles.derousiers
#ROBOMERGE-SOURCE: CL 5405810 in //UE4/Release-4.22/... via CL 5405811
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5418598 by charles derousiers in Dev-Networking branch]
* Removed code to invalidate makefiles when adding new source files. UBT should be reliable enough to make this determination itself nowadays, and ignored -invalidatemakefilesonly argument was causing modules to be recompiled.
* Fixed incorrect config section name when determining whether to allow hot reload from IDE. Now prevents hot reload from IDE when live coding is enabled.
* Added error message when trying to add a new class with Live Coding enabled.
* Added error messages when trying to start Live Coding after a hot reload has taken place.
* Added error messages when trying to hot reload with Live Coding enabled.
#jira UE-71253
#rb none
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: ben.marsh
#ROBOMERGE-SOURCE: CL 5403464 in //UE4/Release-4.22/... via CL 5403485
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5417897 by ben marsh in Dev-Networking branch]
#ROBOMERGE-AUTHOR: robert.osborne
#ROBOMERGE-SOURCE: CL 5399925 via CL 5401157 via CL 5402217
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5417746 by robert osborne in Dev-Networking branch]
This issue was introduced by the fix that allowed MotionControllerComponents to correctly replicate in networked games. Previously the local motion controllers overrode the remote controller's positions. This was fixed by only updating if HasLocalNetOwner was true. But that was returning false for VREditorAvatarActor, as it's Owner was null.
So for now, VREditorAvatarActor's HasLocalNetOwner now always returns true. When using Multi-User with two VRMode editors, we might have to do something else here for when this is called by MotionControllerComponent::PollControllerState()
#rb Lauren.Ridge
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: joe.conley
#ROBOMERGE-SOURCE: CL 5391928 in //UE4/Release-4.22/... via CL 5391930
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5415909 by joe conley in Dev-Networking branch]
Automation - Adding more options for forcing and waiting for loading and streaming to finish and more utilities for waiting for that to happen entirely for tests.
Lauren.Ridge
#ROBOMERGE-OWNER: ryan.gerleve
#ROBOMERGE-AUTHOR: nick.darnell
#ROBOMERGE-SOURCE: CL 5382989 via CL 5383481 via CL 5383540
#ROBOMERGE-BOT: ENGINE (Main -> Dev-Networking)
[CL 5415617 by nick darnell in Dev-Networking branch]
