Today I added a review of the current iniload's control flow. Specifically, this is for the lDOS entrypoint and doesn't trace into subfunctions such as error, query_geometry, read_sector, check_clust, clust_next, clust_to_first_sector, nor the multi-sector loader.