I came across a lecture/training video that talked about dividing up your persistent level into sub levels. One level for your landscape features, one to have all your game assets in.
Does anyone do this for non-streaming reasons? If so, do you see much benefit in that effort?
One thing we do is to create game type levels using the same map for all game types. Capture The Flag for example will have the content for that game type in a sub level along with a level for team based spawn points.
Another ability is to be able to make a single lighting solution and make use of it in all environments.
Another ability is having more than one person working on an environment broken down into blocks, with always loaded turned on, which is similar as to how xreferencing is done using 3ds Max
There a number of benefits that may or may not be appropriate to you.
One of the biggest benefits we’ve found in keeping lighting separate to the geo is that we can fix a number things with out having to worry about the lighting which, besides save time, means much small files to transfer to/from local/remote source control system.
Using this method, we support multiple game modes, lighting scenarios and different versions of the same map.
e.g. TE-Depot & TE-Depot_Compound
Both support the Team Elimination game mode.
Both use Depot_Geo
Both support Depot_Day_Scenario & Depot_Night_Scenario
Both include Depot_Audio
TE-Depot uses the entire map - no restrictions.
TE-Depot_Compound uses blocking volumes and various static meshes to restrict people to the “Compound” section of the map.