DolceTriade

I think this flag needs to be set correctly for M1 Macs? https://github.com/DaemonEngine/Daemon/blob/17ba767bf0b128368af1f8e7d6a75f36385a1702/cmake/DaemonFlags.cmake#L541

``` Detected architecture: amd64 Enabling amd64 architecture intrinsics Enabling i686 architecture intrinsics CMake generator: Unix Makefiles Detected C compiler: AppleClang_16.0.0/Clang_16.0.0:cc Detected C++ compiler: AppleClang_16.0.0/Clang_16.0.0:c++ ``` Seems like its just not...

Sure, but that does mean that Arm M1 Mac isn't being handled correctly, or it needs some special settings that are undocumented.

I can confirm that https://github.com/bazelbuild/rules_rust/commit/8de8f2b89f71819c1d226ff0b87a62bdb3f13c32 is the breaking change. Can this behavior be toggleable?

I debugged this a little more... I probably should have included this but I'm still using `WORKSPACE` on Bazel 7.4.1 (behavior is the same on OSX and Linux tho). I'm...

I guess the "fix" here then is to make sure the manifest and directory based approaches are consistent by making sure the directory based approach cannot return a reference to...

If bazel chooses to be inconsistent, why can't we force the runfiles library to pick the underlying implementation we prefer? Would we be open to accepting a PR to set...

That sounds good! Let me know if I can help.

I think this is a good proposal. We can use the `id` field to store this. The ability of having a stable way of addressing entities is surely a win....

The issue is with layouts... Maybe there is a way to re-order things such that given a particular layout, we can have deterministic starting entity numbers, but even then, it...