- When a remote user saves a map, Concert will hot reload the map on all other clients
- Hot reload creates a temporary map called "Untitled"
- During hot reload, ignore the delegates for changing maps
#jira UE-228018
#rb jason.walter
[CL 37258527 by dominik peacock in 5.5 branch]
- Reverted back to old logic (since that worked) and made the condition for exclusion stronger (AllOf instead of AnyOf).
- Added check for empty property list (AllOf returns true if empty)
- Also added CVar MultiUser.AllowTransactionsOnReplicatedproperties to disallow this feature altogether
#jira UE-227473
#rb jason.walter
[CL 37063424 by dominik peacock in 5.5 branch]
During testing, I found a crash on exit while in Multi-user that was missing a nullptr check in STakeRecorderCockpit.cpp
#jira UE-226665
#rb dominik.peacock
#lockdown alejandro.arango
[CL 37035463 by jason walter in 5.5 branch]
Now, transactions are blocked only if all properties being changed by the transaction are replicated. If at least one property is not replicated, the entire transaction is allowed through.
#jira UE-225658
#rb jason.walter
[CL 37022760 by dominik peacock in 5.5 branch]
Remove the package update table in favor of an activatable query that sweeps through all actors in the Scene to update their overlays
Activate said query when the SCC status of an object changes
TEDS:
Fix a crash with activatable queries where you could have duplicates in the pending activatable queries, but unregistering a query only removed one instance
PrimitiveComponent:
Add functionality to persist OverlayColor in UPrimitiveComponent using the ComponentInstanceDataCache
#rb brooke.hubert
[CL 36828353 by aditya ravichandran in 5.5 branch]
Original CL Desc
-----------------------------------------------------------------
[Backout] - CL36477705
[FYI] brooke.hubert
Original CL Desc
-----------------------------------------------------------------
[TEDS] Deprecate accessor functions for TEDS features inside the TypedElementRegistry class
#jira UE-224434
#rb logan.buchy, ronald.koppers
[CL 36749089 by brooke hubert in 5.5 branch]
[FYI] brooke.hubert
Original CL Desc
-----------------------------------------------------------------
[TEDS] Deprecate accessor functions for TEDS features inside the TypedElementRegistry class
#jira UE-224434
#rb logan.buchy, ronald.koppers
[CL 36747703 by bob tellez in 5.5 branch]
#rb juan.legaz
#jira UE-224100
- Calls to FPerforceSourceControlProvider::GetStatus on the GameThread will end up using the persistent connection to determine the current workspace root. Before this change the persistent connection was not getting updated when the user changed the workspace via the UX, so we'd return the root of the original workspace when the editor started up, if one had been set.
- Added a call to reset the persistent connection when the FConnect command is run, we can then let the persistent connection be recreated when it is next requested.
- Changed FPerforceSourceControlProvider::PersistentConnection from a raw pointer to TUniquePtr to make management of it a bit easier.
[CL 36258840 by paul chipchase in 5.5 branch]
[FYI] paul.chipchase
Original CL Desc
-----------------------------------------------------------------
Perforce will now return the correct workspace root after a user has changed the workspace via the UX.
#rb juan.legaz
#jira UE-224100
- Calls to FPerforceSourceControlProvider::GetStatus on the GameThread will end up using the persistent connection to determine the current workspace root. Before this change the persistent connection was not getting updated when the user changed the workspace via the UX, so we'd return the root of the original workspace when the editor started up, if one had been set.
- Added a call to reset the persistent connection when the FConnect command is run, we can then let the persistent connection be recreated when it is next requested.
- Changed FPerforceSourceControlProvider::PersistentConnection from a raw pointer to TUniquePtr to make management of it a bit easier.
[CL 36230464 by paul chipchase in 5.5 branch]
#rb juan.legaz
#jira UE-224100
- Calls to FPerforceSourceControlProvider::GetStatus on the GameThread will end up using the persistent connection to determine the current workspace root. Before this change the persistent connection was not getting updated when the user changed the workspace via the UX, so we'd return the root of the original workspace when the editor started up, if one had been set.
- Added a call to reset the persistent connection when the FConnect command is run, we can then let the persistent connection be recreated when it is next requested.
- Changed FPerforceSourceControlProvider::PersistentConnection from a raw pointer to TUniquePtr to make management of it a bit easier.
[CL 36228097 by paul chipchase in 5.5 branch]
#rb juan.legaz, wouter.burgers
#jira UE-223965
### Problem
- Previously when storing the workspace root we assumed it would be a directory path and so normalized it to make sure that there was no trailing slash so that our file path parsing code in ::ParseOpenedResults would be consistent. Note that this was changed back in CL 23945151, before that the root could have a trailing slash or not which could cause issues.
- This assumption didn't work if the user decided to use a volume root (i.e D:\) as the workspace root, as in our path utilities normalizing code assumes that volume roots should end with a trailing slash. This in turn would result in local file paths being generated by ::ParseOpenedResults that contained duplicate slashes in the path.
- It is also worth noting that other source control implementations were specifically adding trailing slashes to their equivalents of the workspace root, so forcing perforce roots to not have the trailing slash was inconsistent with other implementations too.
### Solution
- We now enforce that the workspace root ends with a trailing slash unless the workspace root is null, and parse the results accordingly.
- We also removed duplicate slashes from the client root as perforce seems to technically support that even if p4v has issues with workspaces set up like this. At the very least this will move any issues from that to be pure perforce issues and not in our own parsing code.
[CL 36220660 by paul chipchase in 5.5 branch]
The bug was that objects that had been trashed previously were being added instead of the live instances.
#jira UE-207192
[CL 36035402 by dominik peacock in ue5-main branch]
[FYI] eric.renaudhoude
Original CL Desc
-----------------------------------------------------------------
MultiUser: Added new media plate resource type to include object class filters.
#jira UE-212269
#rb jason.walter
#rnx
[CL 36010777 by jason walter in ue5-main branch]
