#rb ryan.gerleve
[FYI] david.harvey
#ROBOMERGE-SOURCE: CL 7477716 via CL 7493666 via CL 7545632
#ROBOMERGE-BOT: (v372-7473910)
[CL 7545879 by bart hawthorne in Main branch]
This should help reduce the number of crashes being reported as abnormal shutdowns because the analytics are never written out.
#jira UETOOL-1719
#rb chris.gagnon
#ROBOMERGE-SOURCE: CL 7473700 in //UE4/Release-4.23/...
#ROBOMERGE-BOT: RELEASE (Release-4.23 -> Main) (v371-7306989)
[CL 7473701 by sebastian nordgren in Main branch]
[FYI] Justin.Marcus
#ROBOMERGE-SOURCE: CL 7445561 via CL 7446165 via CL 7450457
#ROBOMERGE-BOT: (v371-7306989)
[CL 7450607 by bob tellez in Main branch]
-Add FileExistsInPlatformPackage() to platform misc
-Add HasBuildMetaData() to bundle manager interface
[at]Thomas.Ross [at]Daniel.Lamb [at]Hongyi.Yu
#ROBOMERGE-SOURCE: CL 7437649 via CL 7437657 via CL 7438904
#ROBOMERGE-BOT: (v371-7306989)
[CL 7438929 by justin marcus in Main branch]
Move previous implementation into its own GetPathLeaf function which should allow getting the leaf of the path regardless if its a file or dir
#jira none
#rb none
[FYI] Fred.Kimberley
#ROBOMERGE-SOURCE: CL 7437816 in //UE4/Release-4.23/...
#ROBOMERGE-BOT: RELEASE (Release-4.23 -> Main) (v371-7306989)
[CL 7437820 by brandon schaefer in Main branch]
We've seen the compiled-in engine versions get out-of-sync when dealing with pre-built binaries vs binaries built locally by a programmer, so we've changed the validation to use the current engine version, and to override the changelist with the Build.version data (if available). The changelist in the Build.version data will always be the latest CL# synced in UGS, regardless of whether you're using a pre-built or locally compiled editor.
This change makes the older BaseRevision setting redundant, but we've kept it around for now (set to zero) so that we can hopefully change it to store the real content base revision in the future (in a way that works for projects not using UGS, eg) using pre-built editors from the public launcher).
Since the base revision is now always zero, it's been removed from the UI, and the UI has been updated to show you the engine version of a session instead.
#jira UE-77465
#rb Francis.Hurteau
#rnx
#ROBOMERGE-SOURCE: CL 7429459 in //UE4/Release-4.23/...
#ROBOMERGE-BOT: RELEASE (Release-4.23 -> Main) (v371-7306989)
[CL 7429461 by jamie dale in Main branch]
- Added better defaults for depth and width
- Hid the static mesh from the water actor. This is defined by the water type and should in general not be changed. An override is still around as an advanced property as needed
- Tied depth and width properties to spline scale x and y. They should always be in lockstep now.
- Added fixup code to fix up exising tests
#ROBOMERGE-SOURCE: CL 7325640 via CL 7329846
#ROBOMERGE-BOT: (v371-7306989)
[CL 7330063 by matt kuhlenschmidt in Main branch]
Merging using Fortnite_Main<->Fortnite_ControlRig
#rb none
#ROBOMERGE-OWNER: helge.mathee
#ROBOMERGE-AUTHOR: helge.mathee
#ROBOMERGE-SOURCE: CL 7321580 via CL 7323175
#ROBOMERGE-BOT: (v371-7306989)
[CL 7323351 by helge mathee in Main branch]
- This was due to ThreadArraySection and the underlying allocators Mutex (MallocBinned2)
#rb none
[FYI] josh.adams
#ROBOMERGE-SOURCE: CL 7314184 via CL 7317993
#ROBOMERGE-BOT: (v371-7306989)
[CL 7319774 by stu mckenna in Main branch]
#rb none
#ROBOMERGE-SOURCE: CL 7298882 via CL 7298885 via CL 7298886 via CL 7298888 via CL 7309214
#ROBOMERGE-BOT: (v371-7306989)
[CL 7316865 by peter sauerbrei in Main branch]
Submitting in Main since GenericOctree.h got moved recently and the change hadn't make it to FN yet
#jira none
#review-7308607 @Steve.Robb
#rb Steve.Robb
#fyi Yoan.StAmant
[CL 7308751 by Mieszko Zielinski in Main branch]
-------------------
New "Object" parameter type for Niagara.
These allow you to pass in Raw UObject pointers for use in data interfaces or renderers.
Current use case is to pass in Skeletal Mesh references from the outside.
This avoids having to create a whole new data interface when passing in external object references.
Other use cases could be passing in;
- material overrides
- Static mesh references to override the mesh a renderer uses.
- Basically any UObject a DI or renderer may use.
These are weakly typed, meaning that they have to be UObjects but beyond that, they can be anything.
So in theory you could pass in a Material to these Skeletal mesh references.
It's on the using code to validate and act appropriately.
This weak typing does allow a gread deal of flexibility though.
E.g. you can now pass in a Skeletal Mesh Actor OR a SkeletalMeshComponent to Niagara via the same parameter and the DI will interpret it accordingly.
The ability to pass in direct sub component refs is something cascade could never do.
Deterministic Randoms for Data Interfaces
-------------------------------------------
To support functional tests for the object parameters, this CL also includes deterministic randoms for DIs.
Random seeds for deterministic randoms are now packaged up into a NiagaraRandomInfo that can be passed to DI functions.
There is a helper class that will handle the parameter and let you generate randoms that are deterministic or not based on the passed NiagaraRandomInfo.
Non deterministic is indicated using a special case value of 0xFFFFFFFF for Seed3.
Misc
-------------
Some utility additions to int vector class.
Slight refactor of mesh helper for the skel mesh DI.
Fixed issue with incorrect delta time on first tick of skel mesh DI.
Work Needed
------------------------
UI customization for these parameters in the stack and the component override panel.
As these are typeless probably the best thing here is to have the stack UI be an asset picker and the component UI be an actor picker?
Is there such a thing as a combined asset/actor picker? IDK.
Deterministic Randoms need to be utilized in other DIs/Functions.
#rb Frank.Fella, Shaun.Kime
#ROBOMERGE-SOURCE: CL 7288634 via CL 7301760
#ROBOMERGE-BOT: (v370-7290619)
[CL 7302160 by simon tovey in Main branch]
Note: Post process volume saved between 5993259 and this CL will loose their focus distance. But that is OK, because configuring a focus distance in a post process volume where camera came move arround doesn't make any sens anyway.
#rb marcus.wassmer
#jira UE-76709
#ROBOMERGE-SOURCE: CL 7283824 in //UE4/Release-4.23/...
#ROBOMERGE-BOT: RELEASE (Release-4.23 -> Main) (v369-7254125)
[CL 7283826 by guillaume abadie in Main branch]
- CreativeAssetRegistry that will become the interface through which creative assets are accessed. Enabled with CreativeAssetRegistry.Enabled console variable; false by default.
- Temporary debug (only when CreativeAssetRegistry.Enabled is set): Can spawn and cycle through playset props with the creative move tool by pressing the comma key (previous prop) and the period key (next prop).
#rb Mike.Fricker
#ROBOMERGE-SOURCE: CL 7273653 via CL 7276311
#ROBOMERGE-BOT: (v369-7254125)
[CL 7276325 by dave belanger in Main branch]
