🚀 Feature
While inspecting rs_loader, it was noticed that the same source input can be compiled multiple times within a single MetaCall process under certain flows. This is not a bug, but a potential optimization opportunity.
Instead of introducing a loader specific cache, this could be handled at the MetaCall core level so that caching is shared across all loaders and languages thus, avoiding duplicated logic.
A possible approach could be a simple process local cache mapping source input to the compiled artifact, with no persistence. Opening this issue to discuss feasibility and design before any implementation.
🚀 Feature
While inspecting rs_loader, it was noticed that the same source input can be compiled multiple times within a single MetaCall process under certain flows. This is not a bug, but a potential optimization opportunity.
Instead of introducing a loader specific cache, this could be handled at the MetaCall core level so that caching is shared across all loaders and languages thus, avoiding duplicated logic.
A possible approach could be a simple process local cache mapping source input to the compiled artifact, with no persistence. Opening this issue to discuss feasibility and design before any implementation.