viewer.git/indra/llxml, branch main

Merge remote-tracking branch 'secondlife/release/2026.01' into 2026.01

2025-12-04T08:48:50Z

#5078 Replace boost::function with std::function

2025-12-01T10:46:57Z

* Replace boost::function usage with std::function for easier debugging and reduced compiler warnings * Remove a few remaining instances of boost::noncopyable that were missed in tests Signed-off-by: Rye

Merge tag 'Second_Life_Release#a6d4c1d3-2025.07' into 2025.07

2025-10-01T08:56:56Z

Merge branch 'develop' of github.com:secondlife/viewer into rye/infinitemac

2025-08-20T22:04:55Z

Merge tag 'Second_Life_Release#922d2324-2025.06' into 2025.06

2025-08-14T22:14:30Z

#4524 LLControlVariable error

2025-08-12T17:31:00Z

Have unencapsulated Expat headers in any condition

2025-05-12T12:14:23Z

The only condition where Expat headers would be encapsulated is when using LL's Autobuild-based prebuilt libraries, and we're never using any of LL's prebuilt binary for Expat on desktop, since Expat is practically available on any supported desktop platform. The system Expat headers are never encapsulated in any of those platforms. This is the beginning of not relying on the LL_USESYSTEMLIBS macro any more (eventually not relying on the custom USESYSTEMLIBS CMake setting either). Keeping the build system still flexible to have the option to use LL's prebuilt libraries *fully* (we still use some of them in cases where the platform may not have the related system library or for convenience, so the term USESYSTEMLIBS may not always be consistent), is getting harder to maintain. The way it's done is using #if 1, in order to minimise difference from upstream.

Merge tag 'Second_Life_Release#895a6739-2025.03' into 2025.03

2025-03-21T11:36:44Z

Merge branch 'release/2025.03' into rye/forevermac

2025-03-12T02:44:49Z

Merge branch 'marchcat/maint-b-cherrypicks' into marchcat/2025.03-maint-b

2025-03-11T03:10:11Z