Decision service optimization into interpreted/compiled runtime vs incremental query #636
Description
Currently
Everything is re-parsed every time and the EVAL is evaluated.
In Future
At compile time, an L4 project comprising multiple files get parsed and loaded into RAM.
That project then remains resident, and can answer various queries that are submitted incrementally.
From parent issue
Performance must be improved a lot. I think ignoring # keywords will go a long way and also not compiling for debug, but it really needs to run super quick and with low resource usage even for thousands of loaded files.
Load and unload L4 files without service restart