Re:TreeFlat: Building a (possible) faster tree for Rust, ins...
by mamcx on Thursday March 03, 2022 @11:44AM
(#30543329)
Yeah, that is why I say upfront that is "nice one-trick pony" and tied to a specific (if very useful) data pattern.
How update after the fact is something I haven't implemented and one of the reasons I deviate from the way is explained in the APL talk (where it materialize the full path per node) and just keep parent/level is with the hopes it later make updates easier...
How update after the fact is something I haven't implemented and one of the reasons I deviate from the way is explained in the APL talk (where it materialize the full path per node) and just keep parent/level is with the hopes it later make updates easier...