ParentFull threadp_l·Isn't PER or OER more compact? especially for the per-bit charging thingView on HN